IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

Maya Man is an artist focused on contemporary identity culture on the internet. Her websites, generative series, and installations examine dominant narratives around femininity, authenticity, and the performance of self online.

She is the creator/co-creator of several NFT collections including Art Blocks curated NFT collection, FAKE IT TILL YOU MAKE IT, as well as Ugly Bitches and Little Darlings, both with Ann Hirsch. She is also creator of the browser extension Glance Back.

She has exhibited internationally at bitforms, NYC; SOOT, Tokyo; Verse, London; HEK, Basel; and Feral File, online. She has been invited to speak on her work at The New Museum, NYC; The V&A, London; and MOCA, Los Angeles.

Over a video call, Man and I chatted about cultivating physical spaces, Chinese finger traps, the role of software in art, labels for girlhood, and more.

The following interview has been edited and condensed for length and clarity.

Katie Chiou: For those who may not be familiar with your work, can you share more about your background and journey as an artist?

Maya Man: I'm an artist interested in identity, culture on the internet, and ideas around the self––specifically self-presentation and curation––and how those things interface with being on screens like phones and computers and being online. That curiosity has been a driving force for me for a long time. My background is in computer science and media studies. I studied software during college, as well as media studies, which is more focused on theories around how technology affects society at large.

My primary medium is software. I often write code, specifically web-based projects that run in the browser, using JavaScript to create my work. But I also do installation and physical textile work as well.

KC: Right before the call, we were talking about how you got an MFA, but didn’t study art in college. At what point did you know that you wanted to form an art practice, and specifically an art practice related to software? In college?

MM: During college, I didn’t have the language or structured thought process to articulate that I wanted to be an artist. I knew that I was really invested in exploring certain themes and ideas, but it wasn’t until after I graduated and worked at the Google Creative Lab in New York that I realized I felt so driven by my investment in exploring ideas around the self and the internet. I wanted to create things that were more strange or critical than I could at a big tech job. I’m very conceptually driven, and being an artist has allowed me to follow my desired lines of exploration in the world. It’s the best framework for life I’ve found. As an artist, everything you do operates under a coherent practice. There’s a lot of flexibility in how you craft your practice, even more than in other artistic fields like filmmaking or music. Being an artist gives me a lot of freedom.

KC: You’ve created a lot of artwork in web3. You’ve launched an NFT collection with Art Blocks and multiple other NFT collections on Solana since then. What was your initial foray into crypto like?

MM: I had been making software-based artwork for years, since like 2014, when I was first introduced to p5.js via Lauren Lee McCarthy. I was in this community of artists who were making not only digital work, but specifically code-based artwork. The way I heard about crypto and NFTs was through being involved with a lot of people who were already really interested in technology. At first, it came in whispers, and then it kind of came out all at once, making it difficult to ignore.

It was so divisive in the different scenes that I was in. I felt really scared thinking about working with NFTs because I didn't want to engage with something a lot of my peers were so critical of. But at the same time, I'm very excited about new technologies and always curious to understand how they could operate in a way that might be beneficial to digital artists.

My first NFT was the piece I did in the first Feral File show. I was invited by Casey Reas, who is an artist I've admired for a long time and has been a really amazing mentor to me. From there, I was pretty available in response, is what I often say. For example, the artist Dmitri Cherniak suggested that I do an Art Blocks project. I had seen Art Blocks a bit and wasn't super familiar, but then I started thinking about doing one, and thinking about what I could make that felt true to my practice but also would fit into this very specific format of generative art releases that they were doing. And that's how I came up with FAKE IT TILL YOU MAKE IT.

It was really exciting because there was a lot of energy and attention being put into means like software-based artwork, digital artwork, that were very much—and I would argue, still are—seen as, Michael Connor once said in a panel he was on, "the basement of the art world.” At the same time, it also felt really complicated because there were a lot of aspects of crypto and NFT culture that I did not align with. I felt like, “Wow, I believe so deeply in art on the internet, and it’s so sad that so much of the NFTs that are getting sensationalized and popular are such terrible representations of, I think, the culture of artists making this type of work at large.” There’s been a lot of frustration alongside the joy.

KC: Given all your frustration and misalignment with crypto culture, what motivates you to engage with web3? What does working with web3 add to your practice—experimentation, monetary value, something else?

MM: There are multiple vectors that I feel are additive for me as an artist and potentially for other artists. For me personally, platforms that allow for releases of generative artwork have been central for me. You upload the JavaScript file and when someone mints or collects a piece, the code runs live and produces that piece within the system you've designed. That, to me, is a major part of my practice. The process of introducing randomness into a system that I'm writing and building in code, conceptually and in every way, plays a major role in how I think about my artwork. But it’s very difficult to showcase or convey that to people when you're not showing it on a platform built specifically for this type of work, or when you're showing it in a gallery. In a gallery, maybe you've written an algorithm that's live, running via projection or on a screen, or you've written an algorithm that produced a number of outputs, and you've curated a selection of those to make physical in some way and show. But I think, especially for people who are less familiar with code as a medium, it's difficult to understand the level of chance that's embedded in that process and system, which I think is a really important part of anyone’s practice who works with software in a generative way. I sometimes have a challenging time emphasizing that code-as-a-medium aspect of the work when the end result is just someone viewing an image on Instagram or a printed physical piece. They can’t see the process. That was one of the amazing aspects of NFTs—I think they helped popularize and increase literacy around what it means to work with code as an artist.

NFTs also brought money to artists for their work. They created a contemporary art market for digital artwork that hadn’t really existed in a strong way before. I have a real belief in the inherent value of digital objects and for the first time, NFTs offered a technical solution that, in large part, allowed for mass engagement. There was a critical mass of people who were open to using this mechanism to participate in a digital art market, and that was really valuable in a lot of ways.

The last thing is that it brought energy to the community. I've gone to events all over the world in the past couple of years where I've met other people who are excited about digital art and serious about it. Like I mentioned before, a lot of web3 work that gets the most publicized is not necessarily a reflection of the people who are most passionate, the people who are making the strongest artwork in this space. But I feel really lucky that I’ve had the chance to meet people who share the depth of love and attention that I want to give to digital art practices.

KC: In your experience, how has the role of software in art changed over time? How much have new, buzzy technologies like crypto or AI affected the field?

MM: I’ve been thinking about this general question a lot especially because the conversation around AI has intensified over the past couple of years, particularly in the past year. I guess I’m a bit bored with the larger conversation about AI and art, questions like—”Who's making the artwork? Is AI displacing artist labor?” There’s this sensationalized fear around AI’s involvement in the arts because people often imply that it’s offloading the artist's labor to a new technology. But I believe the labor of the artist is all in their mind—it's more about how they execute their ideas and how they use these new technologies. Approaching the question from that perspective is more interesting to me. I’m not that interested in using AI personally, but I’m curious about why other artists are. So when we're talking about an artist using AI, I’m more curious about why they want to use it, how it serves the concept of their work, and why it’s essential to them. That becomes the really exciting question for me.

KC: We just spent a lot of time talking about technology and internet culture, but you also just opened a physical space in SoHo called HEART. Can you tell us more about that?

I wasn’t really seeking to open a space, so it felt almost fateful because I was also craving to do more curatorial work and be with people in person. Coming off a couple of years deeply embedded in the web3 art scene, I realized that the moments I valued most were when I could show work IRL and hang with people in physical spaces. The intensity of the online scene really drained me. So, once I had access to this space, I felt a strong desire to serve the community I felt part of in New York.

I was frustrated with the lack of shows focusing on art on the internet, pop culture, and work that uses digital mediums. For obvious reasons, galleries are still mostly showing work made in traditional mediums, since they’re businesses. But I wanted this space to become something that I, as an artist, would be excited to see exist. I decided to call the space HEART because in my MFA thesis, I talk about the concept of the heart on the internet as an icon. Amidst all these tech platforms born out of the masculine Silicon Valley "tech bro" archetype, they use the heart icon almost as currency. The heart emoji has the most versions in the emoji set, and it’s been a continuous symbol of emotion in our digital communication. It’s a very feminine symbol by nature, so I decided to call the space HEART in the spirit of all those ideas and feelings.

It’s been really exciting to have a space to bring people together. Right now, there’s a show up called Sacred Screenshots, showcasing 26 different artists. Each artist submitted one unedited screenshot from their phone that they felt was sacred in some way. I worked with Jason Isolini, another artist, who helped me print them four feet tall, and they’re hung all around the space. These are 26 of my favorite New York City-based artists who all work with the internet or new technologies in some way. I wanted everyone to be present in the same room in some way, and it’s very cool to have the space to do that.

KC: I’m curious if you’ve encountered any difficulties, challenges, or complexities in trying to display digital, internet-native art in a physical space. What has that process or the thinking around it been like?

MM: I have a really tenuous relationship with the idea of the "white cube" gallery. In many ways, I want it, because to me, it represents ultimate validation for a working artist to show their work in this sanitized, bright space. But I’ve also been thinking a lot about attention. It’s very important to me to make work that operates online, that people can view on their phones or laptops at home. To me, if I make a website as an art object, that’s my work. If you’re viewing the website on your phone, you’re viewing the work in its truest form. But also, if I’m showing it in a gallery and the website is running live on a screen or in an installation, that’s also the work.

Digital work, by nature, has this very fluid quality where you can view it in many different ways, and those are all representative of the experience I want my audience to have. But what I’ve found is that people’s attention spans on their phones or computers are really short—mine is too. I wanted to show the work here and offer experiences at HEART in a way that allows people to spend time with what artists are making, in a way that you simply can’t do on the internet. So the choice to print these screenshots out and have them physically hanging around the room versus doing a slideshow of screenshots or something like that, was very much about attention and also about absurdity. It’s about taking something you’re so used to seeing on your phone and transforming it into something absurd by printing it out four feet tall next to many other artists’ screenshots.

KC: I went to the opening night of Sacred Screenshots, it did feel super absurd to see the screenshots blown up. My favorite screenshot was the Chinese finger trap.

MM: That was Jared Hoffman’s screenshot. A lot of the screenshots have crazy lore that you just can’t know by looking at them, so it’s been really special to talk to all the artists about the stories behind their screenshots. Jared Hoffman talked about how he thought of a Chinese finger trap as an analogy for when you’re trying to get out of a situation that you know is bad for you. You’re really deep in something, but you know you want to get out, and it’s really hard to escape. That was the Chinese finger trap.

KC: That’s so funny because that analogy is so relevant to our conversation… Shifting a little bit to online identity and the internet which is at the center of a lot of your work—how do you think the internet has changed our collective process of meaning-making and identity and community?

MM: We're in a moment where, at least in my life and among the people I interact with—caveat city—I feel like we're very meaning-poor. People are struggling to find meaning in life, and I find that people, myself included, are turning to the internet to try to find meaning and instructions on how to live and be a person in the "right" way. I've always been the type of person who wants to do the right thing, and sometimes I couldn’t figure out what that was, which really stressed me out. So, I spent a lot of time online trying to understand how to build a life that was valid in the ways I wanted it to be.

But with people being less and less religious these days, as religion has faded from the role it used to play in culture, I think that has deeply contributed to this meaning-poor moment. People are looking online for a sense of direction, to figure out who they think they are, and to ascribe an identity to themselves. I think that’s been a really challenging shift for people because the way the internet operates is extremely intense.

KC: An interesting tension I feel is that people used to turn to the internet to seek answers to their deepest, most vulnerable questions about identity—that’s the basis of most early 2000s coming-of-age movies. But now, in 2024, it almost feels like the narrative has flipped. You spend so much time on the internet that it almost becomes the driver of the lack of meaning because your eyes are glued to the screen. You still turn to the internet for meaning, but if you spend too much time there, you end up devoid of meaning. Maybe that’s even related to why you opened a physical space? I don’t know if you have thoughts on that.

MM: My thoughts are that it’s really complicated. The way the internet has entered and infiltrated society is complex, and that’s why I’m so invested in thinking about it. But I’m also hesitant to make any blanket statements about it because it’s so multifaceted. There’s a tendency to want to label the internet as either net good or net bad, but I’ve said this before: I feel like, at this point, the internet is like air, like the universe. How can something so vast be purely good or bad? It’s just everything—it’s all there. It’s a question I struggle with a lot because the internet has made my life beautiful in so many ways, and I wouldn’t want to live without it. Yet, at the same time, I recognize and criticize the ways it has also destroyed parts of people and the way we structure relationships of all kinds.

I’ve noticed in myself that my relationship with the internet has evolved to be more and more one of escapism. I find myself turning to it in small, mundane moments during my day, and that’s something I feel very conflicted about.

KC: It’s a Chinese finger trap.

MM: Totally Chinese finger trap.

KC: Your work specifically often explores femininity and girlhood. Can you share more about how you think modern notions of femininity may have been uniquely defined by internet culture?

MM:  People sometimes ask how I decided to make work about the subjects I focus on like girlhood, and honestly I don’t think I ever definitely decided. These interests have been so innate in me since I was a child. If I go back and read old journals, I can see that I was grappling with concepts around femininity and being online long before I knew I wanted to be an artist. In the early days of being online, I felt a lot of guilt and shame about how I used the internet—it felt extremely egotistical and self-centered and that was really uncomfortable for me. At the core, I’m interested in identity. And as an artist, I’m a vehicle for my own practice, so I use my identity as a case study and a tool to explore how I’ve formed my concept of self, especially as I’ve grown up in the daily process of consuming and producing content online.

I’m really curious about the extremes of girl culture online because I grew up fascinated and awed by it. My work explores the ideas of femininity that I consumed in magazines, on Tumblr, Instagram, and now, to some extent, on TikTok. The way femininity is performed in these very classic ways has always been a fascinating line of inquiry for me.

For example, the artist Ann Hirsch and I did a project together called Ugly Bitches, and we’ve talked about this a lot. In 2024, the popular understanding of gender and how it operates in culture has evolved to recognize gender as a spectrum, allowing people to identify in ways that feel more aligned with their true selves. But still, the extremes of binary gender performance on the internet operate in these really intense ways that I find fascinating.

KC: Can we take “Brat Summer” as a case study for thinking about this? Brat by Charli XCX is an album all about the dark, ugly, complex parts of girlhood, yet it’s somehow garnered extreme mass appeal and gone viral, which is so interesting.

MM: Oh my god, I’m so scared to comment on Brat Summer. What you said made me think about how there have been many phases where a genre of femininity becomes popularized in a way that people really want to attach to. I think of things like "nasty woman," which was tied to Hillary Clinton. There’s this constant push and pull, this desire to label and then escape the labels we make for ourselves. It reflects an underlying desire to feel like you fit into something, to belong to something that aligns with how you want to see yourself.

All these different labels—like "that girl," "girl boss," "nasty woman," and now "Brat Summer"—are very aspirational. They offer a way to talk about a part of yourself that feels glamorous, even if the label is meant to embrace the unglamorous aspects of your life or self. By labeling a piece of yourself, it sort of glamorizes it. I think people want to feel like they belong to something in a glamorous, fun way, and Brat Summer is just the latest iteration of that.

But it’s something we’ve seen in a really cyclical way, again and again. Even recently, with the way people are using the word "demure.” I think because media is driven by narrative—the media, films, TV shows, TikToks we consume—they’re all driven by these narratives that we can label. It feels good to narrativize your life and label it in a way that’s cute and fun.

KC: That reminds me, you actually make TikTok dances to promote HEART. When you’re making them, is there some sort of calculation? Like, “If I do this TikTok dance, people will watch it, they’ll see I have this space, and they’ll come to the space”? Or does it feel more like, “This is just fun, and I want to do it”?

MM: I don’t believe in the idea of just doing something for fun and not for an audience. There’s always an aspect of being on display in some way through the video. But honestly, the true psychology of it is that I find a song and think, “Oh, this is fun.” Sometimes I do an actual TikTok dance, sometimes I’m just flailing around in freestyle mode.

When I think about TikTok, I don’t expect anyone I know to see it because I don’t have a huge following there, and I don’t expect this style of video to go viral. It’s just me dancing with a bunch of filters layered on top. But when I post the videos on the HEART Instagram, part of me thinks, “Oh, this might be fun content that catches people’s attention and makes them want to come see the screenshot show I’ve put a lot of energy into.” I think the screenshots the artists have contributed are really special, and it’s a temporary opportunity for people to see them in person, so I want people to come see it.

I guess that’s the full psychology of it. But the videos themselves are pretty non-calculated. It’s just like, I make the video, post it on TikTok, have it on my phone, and then when I’m playing on my phone, I think, “Oh, I’ll just share this on HEART.”

KC: It’s important that we’re talking deeply about these topics, but also to some extent—they’re so deeply silly and funny. It’s hard to reconcile the two parts of my brain.

MM: I totally agree. My life feels like an ongoing attempt to put these two pieces of my brain in conversation. It’s something I’ve often struggled with in my work because I’m very indulgent in my relationship with the internet. There are people who make work critical of being online or writers who are very offline and critical—they don’t really engage. But I find a lot of fun and joy in being online, and I want to follow that, while also engaging with the part of me that’s hyper-critical of the many detrimental aspects of online culture.

To me, the conversation between these two perspectives is where the most interesting things emerge. This also relates to femininity because young girls generally rule social media platforms—they’re the best at them and know how to wield them in a powerful way. But there’s also a lot of societal violence and disrespect towards young girls, and I think that influences how people view those who engage deeply with the internet.

KC: What’s been inspiring you in digital art lately? What do you hope the space looks like 5 years from now?

MM: I recently went to see the Brian Eno documentary at Film Forum in New York twice. Seeing it twice was fascinating because it’s a generative documentary, so it’s different each time. But both times I saw it, toward the end—I'm pretty sure it was toward the end both times—Eno emphasizes that it took him a long time to realize this, but to him, art is about feeling.

I’ve been thinking a lot about feelings lately because I’m also someone who likes to be very heady. I enjoy theory, and I like to operate up here in my mind. Lately, I’ve been focusing more on feelings and how to let them guide me and reveal my true desires. I feel that sometimes the way we talk about art, think about art, and the way the art world operates at large gets pulled away from the idea of feeling. And I’ve felt very inspired lately to be more consciously driven by feelings.

I hope that in five years, there’s widespread respect for digital art and software as a medium for creating art, as well as continued excitement. There was a lot of excitement over the past couple of years with the rise of NFTs, but I feel that energy has dissipated a bit over the past year. I’ve never been particularly excited about NFTs specifically—I’ve always been really passionate about digital art and art on the internet. I hope that excitement for these forms of art sustains over the next five years. I also hope there are many opportunities for artists working in this way to show their work in physical spaces.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

Nic Hamilton is a multifaceted visual artist based in Melbourne, whose work spans across mediums and themes. His practice-led work draws inspiration from software, nature, online and underground music cultures, and his formal background in architecture. Hamilton's art frequently explores themes of decay, the visual perception of environments, and interpretations of digital spaces, creating interpretations of the natural and digital worlds.

Throughout his career, Hamilton has collaborated with numerous artists from the electronic music scene, showcased his work in galleries across the globe, and lent his creative expertise to consumer brands. Hamilton is a founding partner of ONE, a boutique creative agency. He also worked as creative director at JPG, a decentralized digital art curation platform. Nic owns and operates a digital art co-working space, NUX.

In 2022, Hamilton released NUXUI, a series of 333 unique generative and hand-finished digital artwork NFTs. This collection was followed by NUX2UI, a Web GL / CGI hybrid and SVG collection. In 2024, Hamilton released Rez Tabs, a collection of  ultra-high-resolution degraded digital artifacts. Alongside this, he released Xookt, a WebGL / CGI hybrid series that transforms and degrades imagery in a web browser into a field of liquid-like pixels.

Over a video call, Nic and I talked about launching NFT collections, balancing personal and commercial art practices, AI music visualizers, and more.

The following interview has been edited and condensed for length and clarity.

Katie Chiou: For those who may be unfamiliar with your work, can you share more about your background and journey as an artist?

Nic Hamilton: I describe myself now as a visual artist, but across my career, I've done a whole lot of different things––always mainly based in image making. I was originally trained as an architect, but I’ve always just wanted to create things, which didn’t really jive with architecture. Originally, I was mainly focused on documentation and designing townhouses and offices, but I actually really enjoyed making the images of the buildings. I started pursuing architectural image-making more specifically as a career, and got a job focused on architectural storytelling and communication and making films, and I loved it. I still worked in the property industry, but I loved working on the cultural and conceptual projects and creating content for architecture pitches and things like that.

Around that time, I also started making music videos in my spare time because I’ve always loved techno and dance music. I would send the video to the artist saying, "Hey, I made you something," and then that kind of caught on. One thing led to another, and I started making bigger music videos and becoming more interested in leveraging technology in my work. I was really interested in the intersection of technology and image making, and I saw a real commercial niche for using these new technologies as soon as they came out.

Eventually, I got tired of the property and architecture world. I started independently doing brand creative, creative direction, content and strategy, and interacting with big Nike-style companies, which I loved for a while. And then along the way, we had COVID. During lockdown, I started making digital art for myself again, which was about the same time that NFTs came about. I thought, finally, that there might be a place for me to publish, share, and maybe sell my artwork.

Socializing online through Discord became a huge part of my life. I reconnected with friends overseas in similar positions, made new friends in small Discords, and saw artists I had respected move into NFTs. From that point, I started making a lot more of my own art and a lot less commercial work. I found that really compelling and rewarding, especially the fact that I might be able to sell artwork and interact with an audience.

Now, I’ve dialed back a lot of my commercial work and am much more concentrated on building my personal art career.  I'll always do commercial work in the background to pay my rent and living expenses, but COVID and Discord and NFTs were the real catalysts for me to take my own personal art seriously.

KC: In terms of the decision to do less commercial work, was that influenced by gaining an audience, the opportunity to monetize with NFTs, a general preference to work independently, or something else?

NH: Being able to explore the intersection of visual art and online culture through my own practice was certainly the biggest factor. But also just being able to get recognition and potentially payment for my work certainly helped make concrete in my mind that I could legitimately leverage NFTs. I haven't really sold personal artwork outside of NFTs apart from, I suppose, commissions and installations. And in those works, I don’t really have any ties to them. I didn't really feel that proud of them because people don't really know they’re mine. The only record of those works exists on my website in a small little blurb, and I just don't feel like they have a continuation or a real connection to me, unlike the kind of digital works I've published.

As far as commercial work, commercial budgets really seemed to plummet during COVID. I think on jobs for larger clients, we saw upwards of 50% budget cuts. In terms of digital content, people were spending a lot on digital marketing during COVID, so that was quite lucrative, but it was just work to me at that time. I didn't find it creatively rewarding. The only goal was to run those jobs as efficiently as possible, get a good margin, and to make sure the client was happy. I just wanted to get back to my own personal work.

There’s a tension in commercial work between being in control of my own art and sending work to a client. It's not a compelling way to spend eight weeks for me anymore. I still take on smaller jobs, but I’ve stopped taking larger jobs, the ones that require a bigger team, just because they are no longer worth the time to me.

KC: Something I’ve found in similar conversations with artists who work on institutional scales, is that in those types of projects, you have no idea who is engaging with your work on the audience side, which I can understand to be disappointing. But also, when working with institutions, you don’t have to worry as much about marketing and distribution and personal brand because the institution bears a lot of that burden for you––so there’s tradeoffs. I would love to hear if that resonates with you at all.

NH: Yeah, certainly. Not having to promote your own work is a big upside of commercial work. The business already has the audience in mind, they know exactly what they want, they pay you. It may or may not get credited. The dynamic never really bothered me, but it wasn’t particularly fulfilling either. If it was a job that I liked, I would share it on my social media or send it to other people I knew and say, "Hey, I worked on this. I like this. Maybe we could do this with your brand."

The social side of personal artwork for me is really interesting. I talk to people all day and night about everything––art, digital work, and technology. I genuinely love those subjects, and making those relationships is really rewarding. It means you can go anywhere in the world, to any city, and you've got someone you know there to chat with or catch up with.

I feel like the majority of all my commercial work is actually word-of-mouth referrals. I don't advertise; I don't do cold calls. Developing those social, kind of invisible networks, whether it's commercial or personal artwork, is absolutely key. If you don't share stuff, people don't know you're around. You might not get new work unless you get lucky.

KC: You launched the NFT collection NUXUI in 2022, what was the motivation for launching the collection and what was your experience like?

NH: For NUXUI, I was seeing other people I admired releasing work in the NFT space and I wanted to try it myself. In the past, I had really only made singular artworks, I had never thought about working generatively on a larger scale, so I was very intrigued by watching people make 1000 pieces of artwork that were essentially the same. Working that way allows you to put at the forefront all your experimentation on a work in progress and to show the entire gamut of any idea. For NUXUI, it was an artwork in the form of a collection of around 300 pieces. All the software tools I was already working with at the time actually suited this kind of algorithmic, generative process, and automated a lot of stuff anyway. I also had two friends locally, two developers, who worked on it with me.

At the time, there were no platforms for launching your own collection. There was no Highlight, I think Zora was around and I think there Art Blocks was as well, but we had to apply for that. I had no idea how to make a smart contract. My friends and I got together and learned how to do it all ourselves, which was really cool. The whole process was fascinating. I enjoyed writing about the collection, explaining it, and making a nice website for it. I wanted to create the whole context around everything. I didn't quite get as far as I wanted. I wanted to do videos and stuff, but in the end, we ran out of time. It was like, crap, let's call time on this and release it, otherwise, we would have worked on it forever. I think that was all done in 3D software with procedural methods. I think I made probably thousands of those things, but in the end, whittled it down to around 200. At the time, I was spending a lot of time with Friends With Benefits, and people there were very supportive of it. They were like, this is really cool, release it. I was hesitant, but they encouraged me. That support and reinforcement from the online community I knew was really helpful. I don't think any of my real-life friends at the time really wanted to know about NFTs; they didn't care or weren't interested. They were like, Nic's making his digital artwork again.

KC: Do you recall anything about the process of launching the collection being particularly difficult or unique from your typical processes?

NH: I think the whole process of actually using a terminal to deploy smart contracts was kind of scary. I think there's so much money at stake. It cost about $800 to launch that smart contract at that time. I thought, "Man, if I get this wrong, this is an expensive mistake." On the technical side, getting the wallet to make you feel good and informed while connected and making sure the server was strong enough to meet the minting demand involved a lot of technical overhead that I didn’t expect. I was super interested to learn about it, but I found it fairly scary.

KC: You’re still actively creating NFTs and a lot has changed since you initially launched NUXUI. How has the process changed, and how have various platforms and tools made the process easier?

NH: It's totally different now. I feel like pretty much anyone could deploy an NFT collection. Platform-wise, I think what Nat and Modi have built at Highlight is probably the best user experience out there for launching a collection. They've got a lot of sophisticated tools and are very product and user-focused. On the other side, you have Zora, which is even easier to use for open editions and very fast minting. Both platforms serve different purposes, and they're both extremely easy to use.

I'm not sure if I'll use either of them again in the future for a large collection. I'd probably go back to making my own contract because I like the ownership aspect. I like that it can be completely self-owned and operated, rather than linked to another brand or platform, as good as the platforms out there are. They serve a great purpose, but for a larger, more deliberate body of work, it might be worth the effort of doing your own contract. Adding some novel mechanics can differentiate it and make it totally independent with no one else involved.

The self-publishing platforms are great for providing access for everyone, but it also makes it difficult to stand out. This forces people to up their game to get noticed, which I think is worthwhile. The improvements in the past two years have been tenfold; it's changed completely. Everyone gets their artwork now, even on social apps like Lens. It's very easy to mint stuff there quickly.

KC: You make art that isn't NFTs and also make art that you mint as NFTs. Do you have a particular framework for deciding what's better suited to be onchain? For example, NFTs often involve high volume and many editions, so maybe algorithmic generative art is better suited for that.

NH: At the moment, I have no clear framework for what should be an NFT and what shouldn’t.  If I'm publishing work in progress or tests, I'll often chuck stuff on Zora without tweeting about it. I'll just publish my work in progress, and people might see it or they might not. I don't shout about it. Currently, I'm working on physical editions backed up with digital ones, trying to create a hierarchy of work. I'm doing giant light boxes of AI-generated imagery with RFID chips to link them to digital versions. Beneath that, there will be a series of high resolution, longer format videos and maybe 200 still images. I'm trying to create a real-life exhibition so local people can see the work and engage with it at different levels.

KC: For any artists today looking to engage with NFTs or crypto, what advice would you give them?

NH: If I was just starting out, I'd say that you've got to get involved. You have to reach out to people also in the space, meet the people building the products, making the artwork, buying the artwork, and selling the artwork. The social aspect is huge. You need to engage with it on a genuine level. Don't expect any financial reward at the moment. If you've been making art, you're probably going to continue making it regardless. Be deliberate about what you mint, how you mint it, and how you price it. Make it accessible to as many people as you can and then go from there.

I've always had day jobs and other sources of income. Even with a full-time job, I still tried to find time to work on my art. There's something about getting home from work and thinking, "That was such a drag of a day, let's do something for myself." I've always managed to fit it in.

KC: How do you draw distinction between your commercial work and personal work? Do you think of them separately?

NH: I think it really depends on the purpose. For many people, it's hard to delineate between art and design because the context is so important. For example, graphic designers work in a commercial sphere, but their work often exists as art. I feel that Eric [Hu], who I consider an artist, is also a very skilled creative director and designer. His work bridges both art and design, and there's no real endpoint. His work is art, design, and a commercial project simultaneously. That's something I've struggled with. I've always felt a delineation where my commercial work could never be considered artwork. It probably reflects my feelings about having others influence my work, making me think it's not really art. I've struggled with calling myself an artist because it didn't feel genuine. I'm still working it out.

There's a lot of discourse around the idea that real artists don't sell out and that monetizing art makes it commercial and therefore not art. Crypto for artists allows for easier monetization, changing the discussion around commercial versus fine art. I have no issues with people making money from their artwork. It's about the manner, context, and intention. Some art projects are commercially focused from the start, and that's fine. Some interesting projects begin as commercial and gain a cult following, like some PFP projects and Solana edge-case art. They bridge that gap well.

KC: How do you think about the role of emerging technologies in your work and how it affects commercialization potential?

NH: Everything I do is seeded by new technology or techniques. I closely study open-source AI, 3D, and VR. For my whole career, these have been jumping-off points for recontextualizing ideas or revisiting projects through a new lens. In 20 years, it will be interesting to see the influence of major software releases and innovations on people's artwork. Early AI art, for example, has a certain character marking it as of a certain time.

Technology as a marker of time is interesting, but as a conceptual driver, it’s less so. I'm more interested in the aesthetics these tools provide for building on larger themes and concepts. There are some genuinely cool conceptual blockchain-native artworks, like Terra0, which is particularly cool. The early projects by Hito Steyerl, for example, include the garden cedar where the blockchain updates parts of the work algorithmically. That's strong, both aesthetically and culturally.

However, generative art using Python or JSON to produce thousands of iterations isn't particularly interesting technically. It's more a reflection of the tooling and marks a time when processing was huge. It’s great for exposure but doesn’t always result in strong artwork. It can lead to a kind of zombie formalism, which may look fine but lacks longevity.

KC: What topics and themes are you thinking most about exploring nowadays?

NH: The themes I keep coming back to are entropy, decay, and the passage of time, particularly through a digital lens. And particularly, I think about landscapes and how people perceive them over time. These ideas have always fascinated me, especially the way people inhabit and view landscapes and how these perceptions evolve.

When I was younger, it was just through a camera and I was always trying to manipulate that kind of footage to make it look like it was more messed up or looked like it was decaying or rotting. Now, with open-source AI, there are new ways to interpret landscapes, people, and culture. It always feels like everything is constantly melting and decaying, which ties into a sense of mortality and beauty.

I love the idea of reflecting the passage of time over a career, showing snapshots that melt and degrade through the lens of contemporary technology. Imagine having a body of work in 50 years that shows a gradual growth or decay, a natural cycle viewed through the tools of the time.

I've never been particularly articulate about the conceptual grounding behind my work. I consider myself someone who has a feeling and then makes stuff. I talk about my work in abstract or metaphorical terms related to a time, place, or experience. I'm not a conceptually driven artist; my work comes from my own experiences and reflects what I see in landscapes and people.

I think that’s why I like music so much. Whenever I'm making artwork, I test it with music, putting it into Premiere Pro and adding different tracks to see what kind of vibe it gives off. Sometimes the artwork needs to be more messed up, sometimes more ambient and beautiful. Music helps me check my work. I love going out to dance parties and listening to loud music; the social aspect is invigorating. The otherworldly experience of strobe lights and smoke, disconnected from reality, somehow connects to how I see nature.

KC: Using music is interesting. Do you usually have a track in mind for what you want an artwork to evoke or is the song choice more so informed by the art?

NH: Sometimes I will. I’ll find a piece of music and make something specifically about that song. Whether I tell people about it or not is another thing. Recently, I’ve been making these large-scale AI creations, like waterfalls and giant slow-motion waves. That inspiration came from listening back to my friend Kane Ikin’s unreleased music, which I've had on my hard drive for ages. As I was really listening to it, I thought, "This stuff is so romantically melancholic and euphorically epic. It has this giant grinding sound that reminds me of an upside-down waterfall or something." So, I decided to create something for it. Those tracks ended up being the conceptual side that actually pushed me to act on making it. There’s no way I would have just sat down and said, "I’m gonna make a giant waterfall today." That was really cool. I like the symbiotic relationship with that—it gave me a jumping-off point.

KC: On that note, how do you feel about generative music visualizers? I could see having access to that sort of tooling eliminating the kind of collaborative practice you’re describing with Kane.

NH: From an artist's perspective, I love all the AI tooling. I think it’s really exciting. It’s the most excited I’ve been about a tool or technique, or something that can change the way you look at stuff, since probably the release of Unreal Engine, which made real-time graphics accessible to everyone.

In the early days of new tools, people, including myself, make all sorts of stuff—terrible stuff, good stuff—so I’m kind of used to that. In the first wave of 3D, when Blender became good, people were making orcs, fairies, women with swords, superheroes, and aliens. We’re not there yet with AI tools and it’s fine—people are just messing around. It’s how they learn. A lot of those people end up making cool stuff as they learn the techniques.

I think AI visualizers for music are very cool. It’s just a new tool, and people are making really good, interesting stuff. It’ll be a fun moment to look back on, like, "Wow, 2024, when everyone got involved in creating and started making these image interpolation, morphing graphics—wasn’t that cool." It’s the people who grab those tools, subvert them, and use them in the "wrong" or interesting way—that always happens—that create a cool niche. They find stuff that people think is stupid at the time and say, "Actually, no, it’s not. It’s really cool, and we’re going to use this."

It’s the same as NFTs. Mainstream people thought NFTs were a stupid scam, but underpinning all that, it’s just people messing around. Ultimately, over time, all that critique and criticism will wash away, and the interesting activity will continue. When I started making 3D art, I made some pretty awful, cringeworthy stuff. But that’s just how it happens. You learn through making, through sharing, and eventually, things get better.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

Mike Pollard is the CEO and Co-founder of Nina Protocol, alongside Eric Farber and Jack Callahan. Mike is also a musician and has released his own music under the projects Treetops and Pale Blue Sky. Before starting Nina in 2021, Mike started and ran the music label Arbor in 2004 which has released music from artists including Oneohtrix Point Never, Horse Head, Emeralds, and more. Mike is also a member of the art group FPBJPC and former Director at Svetlana Gallery.

Nina Protocol is a platform and toolkit designed to give artists total freedom and ownership of their work. Nina’s vision is to empower musicians, labels and archives to connect directly with their fans, establish new communities, and create context-rich experiences with unparalleled depth.

Over a video call, I asked Pollard a series of questions ranging from his experience starting and running a music label, music journalism and media, tapping into the music community in New York, and more.

The following interview has been edited and condensed for length and clarity.

Katie Chiou: Long before you started Nina, you started and ran a music label called Arbor. Can you tell us more about the inspiration for starting a label?

Mike Pollard: I started Arbor in 2004 which was 20 years ago, so starting a label probably means something different today than what it meant back then. But at the time I was 14, and I was getting deep into music because of the internet. I had an older cousin who showed me Napster and Internet Relay Chat (IRC), and it was this interest in peer-to-peer internet technologies that got me into music. I started by trying to download music from skateboard videos I’d find on IRC and got into all these other IRC music sharing communities. I would burn CDs of my own collected versions of soundtracks to different skating or snowboarding videos. That was my first journey into searching for music, finding all these different niche music communities online.

I then found my way into following these smaller music labels that had websites where you could email the people who ran them, or you could even email the artists directly to buy a tape or CD that they had self-produced. I started buying CDs and tapes from these small labels and artists and then I realized that I could just ask them if I could release their next album. And in many cases, they said yes. I don’t think anybody realized I was 14 or 15, I would just be on these different message boards online like the Sonic Youth message board or the Animal Collective message board and labels would also gather there to advertise stuff. I would lurk and post and advertise Arbor releases.

KC: From a business perspective, what did running a label at that time look like?

MP: I know most about DIY labels of the previous decade. I’d reach out to an artist, and tell them that I wanted to do a tape with them. For bigger artists we would do 200 copies, for a middle artist we’d do 100, and for a small artist maybe we’d do 50. We’d go back and forth about the details of the release, and eventually I’d get a CDR mailed to me with the tracks, and I’d reproduce them. That could mean dubbing the tapes or burning the CDs at home, and I’d make all the packaging. For compensation, the artist would get 20% of the run. So, for example, if there were 50 tapes in a run, they would get 10. They’d usually sell them through their own website or while on tour. I would sell the remaining 80%. I would sell to distributors or directly to people. I had an email list where I would add everyone who ever bought something from Arbor, and I’d email them about new releases—there were usually 2-5 new releases a month. There were people who would buy every single release, and others that would pick and choose.

I think something interesting about labels is that a label is a kind of curator where if you can trust the label, you can probably guess that everything they release will be interesting to you. We’re also talking about the era of around 2004-2012, which is sort of pre-SoundCloud and pre-easy access to MP3s and sound samples. At that time, you basically had to make a bet based on the label or the press release or the cover.

KC: How do you think music discovery and curation have changed as sharing has gone from word of mouth to small internet forums to bigger platforms like SoundCloud?

MP: The pre-internet version would be things like fanzines and newsletters and magazines basically. If you look at music magazines from the 90s, like Forced Exposure, which was run by a distributor, it would be full of "send a self-addressed envelope to this address to get stickers" and "this label's available catalog." It was like a network about knowing the trusted nodes and who were the good sellers or labels. It was probably just a lot slower than it is now.

The Web 1.0 era was putting those newsletters and catalogs online and then having some forum-based conversation. I think what always matters is taste. There are probably good labels that make bad business decisions. People are sad when they shut down, but the mark a label leaves is whether they can make something coherent or intriguing by bringing together a body of artists. You see that through all eras of independent music. It's a little different when talking about a giant label, but with independent record labels, it's all about taste and trust. If you work with artists who then escalate in their career and move beyond the independent label, then that's probably a big sign that something was chosen correctly early on.

KC: It’s interesting that you emphasize taste because doesn’t everyone think they have good taste? So it more so boils down to correctly identifying and connecting with the people who have the same taste as you. When tastes are local, it’s a lot easier because you have these geographical constraints. Even with social networks like MySpace, it's still based on a social graph. Now things have completely scaled where the likelihood of you finding someone who has the same taste as you is probably higher, but also the means of connecting with that person is much lower. That’s not really a concise question, but I want to poke at this.

MP: If you look at SoundCloud right now, I would say that the artists that have the most potential have probably somewhere between 1,000 and 10,000 followers. You can look through these smaller accounts, see who they follow, who's following them, and see that there's usually a tight network of shared interest or collaboration there.

That can get lost when an artist blows up. There's almost this sense where once you get a certain number of followers, then the game changes; your replies are less interesting or things get more combative. So, I think there's something interesting about the current era where people aren’t trying to blow up their networks. They want them to remain small, vital, and sustainable.

KC: How have all your observations from running a label and existing in niche music communities influenced your role as a Co-founder of a music tech platform?

MP: For the first year, Nina was invite-only, but we gave access to most people who asked for it. The original artists on Nina were basically seeded from people who were already friends and family of me, Eric, and Jack. It just naturally spread through our networks.

We still constantly get this question, "Is Nina a label? Can I put my music out on your label?" People say they like the curation of Nina, but Nina is actually a fully open thing that anyone can go onto and put their music on.

We’re always focused on "narrative capture," as we call it. We didn't try at first to get as many artists or tracks on the platform as possible because we wanted to make sure that there was a solid foundation of people who understand how independent music works. This means understanding how we do things, the importance of context, and that artists don't exist in a vacuum. So, we moved slowly at first to make sure that there wouldn't be something that would come onto the platform that would change the narrative. We wanted Nina to be about intentional context-building, supportive, real-life music scenes, rather than just like a new song website with an interesting financial model or something like that.

KC: You started Nina in 2021 and launched V2 in Fall 2023. What learnings did you take from V1 to V2?

MP: The biggest learning from V1 to V2 was that crypto UX is not ready for a mainstream audience. With V2, we added fiat on and off-ramps, email wallets, and the ability to publish music without needing any crypto to start with. It confirmed a lot of our assumptions that things like browser extension wallets will not be what bring the normal person into this world of peer-to-peer financial interactions and provenance.

With V2, we also introduced multi-track releases, which I would say account for like 90 to 95% of releases on Nina—EPs and albums. I'm a huge fan of albums and having an artist be able to tell a story across a few tracks. It was a feature that we always wanted, but the hardline crypto UX of V1 made album releases extremely difficult. Musicians think across multiple tracks, and I think the prevailing thoughts in the crypto music space are very track-based or a Rube Goldberg machine-like style method of thinking of ways to connect tracks across multiple tokens or something like that. I think that only appeals to a very initiated group of artists.

KC: Nina is quickly becoming a pillar of the underground music scene, both in media and IRL. You have cult-following music blogs like No Bells and First Floor writing columns for Nina Editorial, throw Nina Nights events at beloved venues like Nowadays, and have the 400 Floor podcast. How do these efforts tie into Nina’s core vision and strategy?

MP: On Nina Nights in particular, Jack, Eric and I—who all started Nina together—and other people on the Nina team have deep experience putting on shows, running labels, and being in bands, so it comes naturally to us. We thought live shows could be a place where people can come and we can talk to them about Nina. We can also elevate people who are friends of Nina or Nina-adjacent. The podcast is kind of similar. They're both about the wider context of music. Music isn't just a song online. The strongest interactions occur when you can meet someone and talk to them. It's definitely not scalable, but something nice about the Nina Nights is that they've all been different scenes. There's maybe a core group of 10 or 15 people who come to every event, but there’s also always new faces.

The goal of Nina Editorial is to highlight trends in different kinds of online music discourse and give them a place to hangout. It's not really about the columns in particular, First Floor and No Bells aren’t related to crypto. Both columns have premiered tracks and onboarded artists on the platform, but it's really about supporting their work.

KC: Do you have thoughts about all the recent events in music media, things like Bandcamp editorial shutting down to Pitchfork folding into GQ?

MP: It's definitely scary. Before Pitchfork was bought by Condé Nast, there were all these blogs in the MySpace era and then you had a sort of consolidation and corporatization of all of those bloggers. They were all hired by different companies—whether it was Vice or Pitchfork or somewhere else—and then all those companies said, "Okay, this no longer works." Then all those people who had been corporatized got pushed back out at no fault of their own. It’s brutal.

But you need vital writing, you need scene reporting. Pitchfork and publications like it are good for distribution and having this shared idea as to what the most important stuff happening in independent music is. But does that match the experience being on the ground with up-and-coming bands? I think that people just need more direct ways of supporting artists and critics and scenes. There are very dedicated people, and I think that the dedicated people will continue to be dedicated.

KC: How do you think about Nina being a bridge from crypto to the mainstream? Does that cross your mind at all?

MP: Spotify, SoundCloud, and Bandcamp all started over a decade ago, and I think they've all shown signs of weakness in recent years. Bandcamp has been sold twice in the last few years. Spotify has demonetized 85% of artists on the platform with less than 1,000 streams a year. SoundCloud is constantly being put up for sale.

There is value in permissionless cultural tools. That's what we are dedicated to thinking about at Nina: a platform that could exist independent of a solo operator. A platform where you can avoid things like MySpace's servers going down and 15 years of music getting deleted, where you can have peer-to-peer artist interactions, whether it's financial or communicative. That's the thing that we're trying to build with Nina.

I think that the mainstream crypto discourse currently is very speculation and degenerate-oriented, and crypto culture can be off-putting to anyone not involved in it. So, I definitely understand where people's apprehensions come from because we largely share them. But I also think that there are benefits to thinking about how culture could exist on the internet. We’re thinking about peer-to-peer music online from the perspective of the experiences that we've received from the past decade of music tech. People are exhausted with change and the ever-increasing intensity of the firehose of content and roles that you have to play.

KC: You’re a big music collector and think about archiving and provenance a lot. Can you tell us more about that personal philosophy?

MP: At the speed at which culture currently moves online, it's super easy for things that don't have the right rollout plan or PR strategy or just engine behind it to launch and then immediately fade out forever. But I think that these forgotten things always come back and end up being the inspirations for future generations. There's a real value to not letting the things that are loudest in their infancy be the only things we remember. There should be an archive to dig through for people who are interested—to follow the traces back from the things that they know to the things that they don't yet know. This is how we keep culture lively, rather than letting whatever an algorithm wants to show you push you based on whatever reasons it has.

KC: In the context of Nina, what does building a “context-rich” platform or experience look like?

MP: I would say that it's a North Star, we don't have it yet. Discogs is a good example of cataloging the cultural history of recorded music. Think of archive.org. The important thing is having open access to the interconnectedness of things. So, if one label puts out music by 10 artists and those 10 artists are also put out by 4 labels each or something, then you have this network of nodes that you travel through and follow your interests. There’s value to archiving the past in order to preserve the future. If you don't have a past, then you just have a perpetual present.

KC: What would you like Nina to look like 1, 2, 5 years from now?

MP: Something I really like about SoundCloud is how well it facilitates the move from being someone who listens to music to becoming someone who makes music, which I think is a pretty natural progression for a larger subset of people than you would think.

In general, I don't think that much about Spotify. I don't think that Spotify is where music is born. I think a lot more about where music is born, I guess. So I'd like to see Nina facilitating that role. I'd like to see us find more ways to make context-rich experiences happen there where people can actually dig through the history of a scene or a style. I think that we're at the beginning of a Nina-specific sound developing. I won't say more, but there are groups of artists from all around the world who share something that brings them to Nina.

I'd also like to see us make better tools to be the backend for artists' personal sites. Having a personal page is a good way to say, "This is where I live, this is my context." In light of platform fatigue and streaming fatigue, I would like to see Nina be a sort of Discogs or dictionary of everything that's happening in independent music. Whether it’s Nina or not, I’d like to see a place for artists to have their own place to live. Things like Linktree are a start and maybe a sign that things could go one step further––I'd like to see that step beyond.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Earlier this year, my colleague Benji Funk wrote a piece called “Crypto-Powered Information Games” detailing the design space for coordinating, aggregating, and rewarding quality information across decentralized networks. Benji’s post begins with the following thesis: “Humans have an inherent desire to create, search for, and speculate on information.”

Since then, Benji and I have spent a lot of time talking to teams building at the intersection of financialized information games and social platforms—this category is currently referred to as “Social Finance - SocialFi.” SocialFi teams vary widely across mediums (text, image, video, audio) and audiences (pro crypto traders, artists, content creators, fans, and influencers), but the overall strategy is to create a fun social network based on high-skill financial information games or to financialize platforms that were initially primarily social.

The central goals of information games are to gain critical insights, improve individual or collective decision-making, and to accrue capital. Social networks and newer forms of SocialFi tend to magnify the relationship between information and capital accumulation. To understand the holistic value of information, however, it is crucial to consider all forms of capital, not just financial capital.

In this post, I will explore the interplay between social capital and financial capital in different types of social platforms, as well as where crypto may be a useful tool in these different contexts. By completing this exercise, we can better identify what goals/features matter for different goals and user personas and then build better-aligned platforms for those objectives.

The Relationship between Social and Financial Capital

Economic theory tends to focus on financial and transactional exchange, thereby classifying other forms of exchange, like cultural and social exchange, as non-economic and insignificant. However, this narrow view ignores the interplay between different types of capital and power, creating a false distinction between profit-maximizing economic activities and seemingly frivolous cultural activities.

One of the most desirable types of information, in fact, is social information. Social information gives us crucial cues as to what is considered morally right or wrong and how to behave in certain situations—fundamentally how to “belong” in specific groups and societies. Having accurate and relevant social information allows humans to earn capital, both social and financial. Synonymous with the concept of social capital is the term “status.” To quote Eugene Wei quoting Jane Austen, "It is a truth universally acknowledged, that a person in possession of little fortune, must be in want of more social capital."

Status/social capital refers to the resources available to an individual through their relationships within a network. These relationships form in a variety of ways including familial ties, institutional recognition, and other types of acknowledgment or respect based on some situational rubric.

Whereas financial capital can be utilized for immediate exchange, social capital is accrued through the establishment and maintenance of relationships over some period of time. These two forms of capital are not necessarily mutually exclusive. Converting financial capital into social capital might look like using financial means to invest into social relationships to improve status, transforming economic debt into social gratitude, or purchasing goods and symbols to gain social recognition or respect. This form of capital conversion is often faster, but less stable as there is a risk of ingratitude or failure to reciprocate in social exchanges or the risk of the gestures being perceived as inauthentic or reeking of ulterior motives. Conversion of social capital into financial capital is often more concealed so as to not disturb the ongoing maintenance of longstanding relationships or perception of meritocracy, but its investment can yield financial benefits through economic opportunities. For the sake of simplicity, we’ll refer to this dynamic as financial capital being more liquid, while social capital is generally more illiquid.

Designing a Modern Social Platform

In its early days, Instagram was largely used for friends to share moments with each other. Throughout the 2010s, the term “Instagram influencer” came into being as users moved beyond sharing photos with first and second-degree connections to promoting products and content for upwards of 7-figure sponsorship deals with brands. The influencer marketing economy (including platforms beyond Instagram) is now valued around $20B+. Instagram has dived headfirst into commerce, launching Instagram Shopping in 2020 (though it has retracted some of its more explicit shopping features such as live shopping). Instagram’s evolution over the past decade is a prime example of seamlessly converting social capital into more liquid financial capital.

The shift of social platforms more broadly has been one from “networking” to “media.” Platforms are increasingly designed to be consumed at-scale (broadcast media, think TikTok), rather than to foster interpersonal connection (P2P networking, think WhatsApp).

Between these two ends exists a wide, fluid spectrum. As a general rule of thumb, more broadcast-oriented platforms rely much more upon discovery mechanisms and speculative symbols of status and reputation (likes, follows) with the ultimate goal of discovering and scaling a community/audience. In P2P-oriented platforms, target users tend to already be in pre-existing social/network groups and the goal is to deepen relationships within these groups rather than to necessarily scale them to an infinite degree.

When we look at existing social platforms, we can identify two central dynamics by which to categorize the types of status games they play. There are platforms that lean into liquid, financial games, and there are platforms that lean into more illiquid, social status games. Similarly, platforms can make these games and status mechanics explicitly core to the experience of the platform or more implicit and subconscious in the background. With these lenses, two high-level axes emerge:

• Social capital <-> Financial capital

• Implicit capital <-> Explicit capital

We can use these axes and subsequent quadrants to broadly categorize social platforms and their ideal users. Again, these axes/quadrants do not mean mutual exclusivity but there are natural tradeoffs to consider when building social platforms according to specific goals—breadth vs depth, scale vs intimacy, instant exit liquidity vs long-term retention, etc.

SOCIAL/IMPLICIT - FRIENDS

In 2019, Venkatesh Rao coined the term “cozyweb” to refer to places people can hang out online away from the hubbub of the public internet.

“Unlike the main public internet, which runs on the (human) protocol of ‘users’ clicking on links on public pages/apps maintained by ‘publishers,’ the cozyweb works on the (human) protocol of everybody cutting-and-pasting bits of text, images, URLs, and screenshots across live streams. Much of this content is poorly addressable, poorly searchable, and very vulnerable to bitrot.”

In their purest forms, cozyweb platforms look like email, text message, and related messaging apps used primarily between friends where information remains relatively secluded. Similarly, private online groups like Discord servers and Slack workspaces have some semblance of openness and discoverability (less cozy) but are still highly gated.

As a step further, there are social media platforms such as BeReal or Snapchat that give new utility/entertainment to existing social groups. However, these apps have little-to-no focus on growing your social network, status, or user discovery; they’re focused on giving you tools to engage your existing communities.

Lastly, and perhaps the hardest type of platform to replicate in this category, are platforms on which users have public presence and discovery mechanisms exist, but “status” has little utility. Examples of these platforms include Letterboxd, Are.na, and Spotify to some extent. Each of these platforms has their own unique combination of features that limits the proliferation of status games.

For example, Spotify has very few social features—no direct messages, limited UGC/post capabilities; engaging with a friend with Spotify is relatively high friction (sending links off-platform).

Letterboxd is based on user-generated content (film reviews), but a user’s status level changes very little of the experience on the platform. Having more friends or likes on your reviews does not necessarily amplify your content in any particularly desirable way or unlock any power features or utility in the way that, say, Yelp power users (also based on user reviews) may receive invitations to review new restaurants. That being said, there’s little reason why this couldn’t become a dynamic for Letterboxd users as well in the future.

In the case of Are.na, the team has always been a major proponent of “cozyweb” spaces, creating a highly aligned user base and culture that generally discourages status games. Instead, social features on Are.na encourage fundamentally unscalable IRL community meetups rather than highly scalable digital audiences.

To summarize, cozy platforms tend to limit scalable status games through a few key features or lack thereof: limited UGC or P2P social features, highly aligned user culture, and/or lack of algorithmic discovery or amplification. Instead, these platforms focus on providing experiences that can help deepen existing relationships and connections with people already in your social network including direct messages, coordination tools for IRL community, and focus around niche interests.

SOCIAL/EXPLICIT - INFLUENCERS

Whether a platform makes social status games explicit or implicit is as much a decision made by its user culture as it is a function of actual platform buildout and design. For example, because of the widespread acceptance of Twitter KOLs and Instagram influencers, it’s natural that any new platform similar in UI to Twitter or Instagram will replicate similar results, hence making the goal of accruing social capital almost inherently explicit for these types of platforms.

As a baseline, a fundamental core question to ask when evaluating the positioning of a social platform is whether the goal of the platform is to aggregate attention and build audiences (one-to-many) or to promote smaller, high-touch networks and relationships (one-to-one). However, in many cases, aggregating as much attention as possible at scale can be an effective way to then filter or organize suitable candidates for a smaller, intimate social group. More structural features that support social capital games include easily legible/quantitative metrics such as follower/following/likes. Prominent discovery features on a platform also usually point to more explicit social status games, as users typically want to be featured prominently in the discovery feed/algorithm to increase their audience and reach. It’s worth noting that quantitative status markers (following, followers) on social platforms fundamentally improve content recommendations and user discovery, so these features often go hand-in-hand.

A prime example of an influencer platform is TikTok, where the goal is for short-form content to reach the most people as possible (go viral) and to accrue likes and followers, but more intimate 1:1 communication is extremely uncommon (though TikTok does have DMs). A more recent example that illustrates this dynamic is the privatizing of “likes” on Twitter, meaning it’s now impossible to view any user’s likes or the likes on any given post. Though a small feature change, the change dramatically transforms the social dynamics on the platform where social graphs/networks become less legible, instead shifting to more quantitative measures. By obfuscating certain social details, the goal becomes less so to discover relevant people in your social network, and more so to pump numbers and reach. An interesting side consideration: How much more data can Twitter collect from users now given that likes are private/users may like posts more freely and often now? The opposite may also be true––users may like posts less often because likes are private.

A nuanced factor related to influencer platforms is that, as mentioned earlier, social capital generally relies upon one’s ability to build strong relationships over time. However, if said relationship-building is perceived as inauthentic or in explicit search of capital (“clout-chasing”), such activity can actually overall reduce one’s social status as people can be less open and willing to engage in relationship-building that is perceived as not genuine.

A platform that thrives on influencers and social status games must maintain a small number of influencers, such that 1) there is a level of perceived desirability or rarity in each influencer and 2) there are remaining users that can serve as an authentically engaged audience and that can aspire to higher influencer status. The obvious questions then become how do users sort themselves into these tiers and what platform features best facilitate this organization?

FINANCIAL/IMPLICIT - PROFESSIONAL

It is generally known that social capital can be converted into financial capital, but there are also explicitly professional social platforms. These platforms typically focus on career-building, whereas more traditional social platforms rely on parasocial attachment for later capital conversion (i.e., an Instagram influencer building a social audience, and then later launching merchandise). These professionalized platforms often take the form of “patron” platforms, subscription businesses for gated content/insights, or professional networking platforms for hiring and finding new roles.

In a simple example, the goal of LinkedIn is to find professional roles that lead to earning financial capital. This goal is rather explicit and socially accepted. Professional platforms can become more nuanced, however. Take Zora, where in its base form, artists post work to be minted/purchased for a fee so that the creator may earn income. While the exchange of selling goods for a fee is rather direct, Zora is becoming an increasingly more social-oriented platform rather than traditionally professional.

The line between “influencer” and “professional creator” continues to blur where influencer is a much more common professional career, and professional creators must increasingly build audiences on social platforms to be successful. For example, while a vast majority of Substack newsletters are unpaid/written by hobbyist writers, there are many professional writers who rely on Substack as a principal source of income. Many of these writers start a Substack newsletter after already accumulating professional audiences, but also use social platforms—Twitter in particular—as a key distribution channel to develop more social relationships with their audiences. As a side note, Twitter throttles Substack links, making it more difficult for Substack writers to build audiences and distribution on Twitter.

I’ve written about this dynamic for music artists, in particular. For example, from my post on the music industry “where does music go from here”:

“With the death of the capital-a Album, you optimistically have about 200 seconds to capture a listener’s attention with your music. In an era in which artists are discovered on “social” platforms, the focal points become the individuals themselves. Music artists who capitalize on this feature can experience major boosts in popularity. [...] Another way to frame this shift is that art and media have become much more identity-centric.”

A core question to ask when considering/designing these dynamics on a given platform is: “How directly is a user seeking financial capital on this platform (by selling goods and services, earning income, etc.), as opposed to social capital (building an engaged audience through more personal, social content)?” The former results in more immediate financial gain, though perhaps is more difficult to build social engagement from, while the latter often is a slower process as you have to socially engage an audience and develop an authentic relationship with them before monetizing the relationship.

FINANCIAL/EXPLICIT - EARNERS

The last quadrant of our framework includes platforms that place financial games at the center. These platforms are often pay-to-play and/or rely on some sort of speculative price action or prospect of earning money to rally users. Casino games or traditional betting/prediction market platforms are classic examples of “earner” platforms.

Earner platforms are particularly effective at quickly aggregating users and attention, as financial capital is much more liquid and therefore much more exciting and active. These platforms gain momentum quickly and generate content rather self-sufficiently, as prices fluctuate in real-time and user actions are usually rather straightforward (buy, trade, sell). These platforms are usually well-suited for short-form, viral content that complements live price action. Similarly, the ROI from engaging on these platforms is much more immediate and clear than attempting to painstakingly build reputation on a new social platform. It’s also worth noting that in these platforms, users rotate in and out much more frequently than a social platform, where the goal is generally to deepen engagement with a specific community of users.

There has been recent exploration in overlaying these financial games explicitly on top of more elusive social dynamics. These platforms seek to place a dollar amount on usually more illiquid individual social status or on the strength of a community. While these platforms are still social to varying extents, the prospect of earning is at the forefront. Earnings can be for individual gain or for the goal of investing these funds back into the social group/community, depending on the dynamics of the platform itself.

The extent to which and methods by which financial layers are made explicit in these platforms varies widely. For example, Patreon is a subscription platform by which users pledge a subscription in order to receive premium content, but primary content usually exists on a separate platform—therefore making the financial layer largely the sole function of the platform, but not necessarily financializing core social content that may exist elsewhere. In the case of Twitch, financial tips during streams have become extremely common, though a majority of Twitch streams are still free.

A few high-level questions to ask to determine the level of financial games on a platform are:

1. Is creating financial markets a core function of the platform?

2. Do money/price actions affect the UX of the platform in real-time? (Unlonely, Twitch)

3. Are there prices/price charts prominent in the UI? (Friend.tech, pump.fun)

4. Is financial gating a core functionality of the platform? (Patreon, PartyDAO, Friend.tech)

“CRYPTO SOLVES THIS” OR SOMETHING

As attention and social experiences move online and content and culture move at the speed of light, managing the information we have about ourselves and each other becomes nearly impossible. The cultural cues we use, financial and social capital, and the relationship between them become even more difficult to decipher and manage. Crypto infrastructure helps us manage, contextualize, and connect indicators of social and financial capital in a hyperspeed, hybrid online/offline world.

In the case of financial capital, the role of crypto is perhaps more straightforward.

Real-time price discovery

The ability to permissionlessly and programmatically create new markets allows us to discover the price and underlying sentiment/demand for practically anything under the sun in the blink of an eye. In this way, blockchain rails allow us to most explicitly and purely express and act upon the will of groups of users (both people and bots) by programmatically generating quantitative information in real-time.

Capital and social coordination and deployment

Quantitative signals generate important social signals in two ways: they create binary, legible outcomes (number go up, number go down) that allow us to measure sentiment and direct latent social energy through extrinsic rewards and incentives (perks, streaks, and rewards). The value unlocked from gaining and leveraging real-time social information in this way is extremely powerful, allowing decentralized networks of people to aggregate and deploy capital in new, innovative ways that were previously impossible.

In the case of social capital, I’ve expressed before that social reputation is incredibly broken in a few different ways. The user journey across online and offline is incredibly fragmented, as online platforms mine and hoard social information and data, and as users move back and forth between IRL and URL. Platforms then channel this information into centralized algorithms that they can use to silo users into personalized feeds that amplify whatever content the platform so chooses, meaning creators must play by the platform’s rules in order to participate and culture submits to power-law like distribution. Kyle Chayka shares an anecdote about an artist, Hallie, in his book Filterworld that illustrates this point:

“Hallie also realized that the Instagram feed rewarded specific qualities. She had always combined visual art and writing, but posts with clear written messages got the most engagement. [...] It was a meme-like assembly-line process perfectly suited for Instagram: the bright colors and simple text added a little spice to her followers’ feeds along with simple moral messages. Followers came to rely on her account for those pieces alone. [...] The pressure that Hallie felt to make the rest of her artwork similarly bright, clear and simple is much like the pressure that a musician feels to frontload the hook of a song so it succeeds on TikTok or a writer feels to have a take so hot it lights up the Twitter feed.”

Small algorithms and local curation

Algorithmic feeds have granted users significant power through advanced discovery mechanisms, yet these same feeds have transformed into a new type of monolithic platform, centralizing curation around scalable metrics and uniform cultural tastes. Blockchain rails offer tools for rethinking curation, promoting structures where individuals are encouraged to co-create and curate their own digital community spaces, moving away from centralizing platform mechanics that lean towards large scale and uniformity. Tools like token-curated registries and markets could allow each user to actively build a community and elevate their own "digitally local" curators, instead of passively consuming a single mega-feed tailored for broad appeal—consider Facebook group pages, Subreddits, NTS channels, Discord servers, and similar platforms.

Holistic and verifiable user identity

We identified a few key challenges in user identity already, namely that identity is fragmented both IRL/URL and cross-platforms. Leveraging blockchains means leveraging open data, where users can move freely between platforms and bring their data with them. Not only does crypto enable open data sharing, but it also enables trustless verifiability as people can attest to information as correct/incorrect, true/false, relying on the wisdom of the crowd, rather than on platforms with ulterior motives. This becomes increasingly important as more content becomes generative and as social platforms move to become more decentralized.  Identity primitives must exist to help us better identify, incentivize, and match people across social platforms much more efficiently and effectively.

A key way we generate information and its subsequent verifiability is through social information games that allow users to attest to qualities and behaviors that others can similarly react and respond to. With open data sharing and open data verifiability of these information games, we are then able to build robust, decentralized reputation that is able to draw verifiable information holistically across platforms, improving both its quantity and quality.

While many SocialFi platforms to date focus on speculative financial games, there is an incredibly vast design space for platforms that create status games across both financial and social capital. A key to building a powerful, engaging social platform is to identify a target user persona and their unique preferences for financial or social status games. Appealing to different types of users and leveraging different tools necessitates certain tradeoffs in all cases, but is an important exercise for builders.

Across both axes of financial capital and social capital, blockchain rails offer exciting new opportunities for the next generation of social applications. Real-time, decentralized markets allow everyday users to discover, speculate, and trade with strangers anywhere in the world on major global events like elections or play ephemeral social games with friends like betting on the winner of a hotdog eating contest. Communities can rally around cultural objects and launch life-changing crowdfunds to support creators and projects in less than a day. Decentralized algorithms and curation give networks the ability to create subcultures at the most granular levels, discovering and engaging the most relevant people in the digital underground, away from monolithic broadcast feeds.

The delineation between the four types of social platforms is not to declare a hierarchy between them, but to illustrate that building for different audiences with different goals necessarily comes with different tradeoffs. Rather than attempting to build for all quadrants all at once, it’s helpful to identify which features work best for which audiences and goals.

I hope this framework is helpful, and we cannot wait to see what you build.

Thank you to Yash Bora, Jihad Esmail, baz, jtgi, Brian Kim, Lyron Co Ting Keh, Holyn Kanake and Archetype colleagues Benjamin Funk, Tyler Gehringer, Ash Egan, Danny Sursock, and Dmitriy Berenzon for thoughtful review and feedback on drafts of this post.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

Adam Ho is a designer and artist based in Queens, NY. He currently runs Adam Ho, I.S.P., his own independent studio practice. Adam’s primary focus is on creating large-scale visual brand identities and interaction design for emerging technology companies.

Adam is a prolific brand designer in crypto, having worked with companies like Eigen Labs, Axelar, Coinbase, MoonPay, Mona, Valhalla, and more. Adam also released his own collection of NFTs in 2022 on Foundation called Liquidities, a series of NFT abstract motion expressions. Apart from his own studio practice, Adam also actively works with VectorDAO as a Design Lead.

Before forming Adam Ho, I.S.P., Adam worked in-house on design teams at Cash App and Zendesk. He has also collaborated with other design-forward brands such as Nike, Square, Medium, and Airbnb.

Over a video call, Adam and I talked about launching an NFT collection, creating the brand for EigenLayer, general design principles, and more. Adam even designed a new NFT to accompany this article, which you can collect for free on Foundation!

R1Z0 | Foundation

The start of something new

https://foundation.app

The following interview has been edited and condensed for length and clarity.

Katie Chiou: What was your initial foray into crypto like?

Adam Ho: My first time interacting with crypto was hearing about the coins they sold on Coinbase back in 2016-2017, like Bitcoin, Litecoin, Ethereum. I talked about it with my friends from college at the time. I didn’t really think about crypto again until around 2020 when I started seeing a lot of people talking about crypto in very cryptic ways I didn’t understand—like what does the acronym “NFT” mean?

Foundation actually reached out to me before they started using the term NFT, and I thought it was just kind of a way to securitize art. I didn’t really put 2 and 2 together until after Foundation launched, where I could see people uploading their art for other people to buy. This to me felt like a shift, a second phase of crypto based on art, whereas the first phase was based on money and speculation. Now, I feel like we’re in a third infrastructure phase right now.

I tend to have an obsessive personality, so when Loot came out in 2021, I was really into it. That’s actually how I first met Jon Yan at VectorDAO. He was working on a Loot project at the time, and we started talking about other NFT collections and communities like Solvency and Azuki. I thought the whole space was really fun and new. So that was my foray into crypto, it started off mostly as money speculation and treating it as stock and then later figuring out utility for crypto through NFTs.

KC: You launched your own NFT collection on Foundation called Liquidities in 2022. What made you want to launch a collection, and how was that process?

AH: I was watching an interview recently with Ezra Koening from Vampire Weekend and Zane Lowe, and Ezra talked about how musicians get into music through different ways like live music or the gear, but he’s always been interested in discographies and the idea of entire bodies of work. To some extent, that’s similar to how I approached Liquidities. My first few NFT uploads were a bit random, and then I thought to myself, “Wait a minute, this is on the blockchain, there’s going to be a history to this.” I didn’t want to continue to upload and mint just random things. For about a month or two after that, I really sat down and thought about the kind of work that I would want to produce that would be constrained and have a specific point of view—both in terms of the aesthetic but also the narrative.

Liquidities is a play on the word solvency, and it also sounds similar to Solidity. Once I had a name and concept I liked, I made the art feel almost liquid, like they were on loop in this liquid pattern. The art looks generative, but it’s not. I did everything in a lo-fi way frame by frame in Photoshop, similar to what you would do if you were making a GIF. When I exported the files, I constrained the amount of colors which is why they look both random and specific. That’s on the art side. When I decided to launch the collection, I wanted it to feel like an album. That's why actually there's only around 10 pieces. I haven’t touched the collection in a while because I’m focused more on client work nowadays.

KC: When you uploaded NFTs for the first time, do you remember it being particularly confusing or difficult or different from other platforms you had used before?

AH: It was way different. I think I lost around $600 the first time I tried to mint because the gas fee was like $300 and then everyone told me, “Yeah, that's just how it goes.” It actually felt like a massive barrier to entry. To some extent, there’s a charm to engaging with things that have a bit of friction behind it. After the first or the second time, it also gets a bit better and more familiar. There was definitely friction and a learning curve, and to be honest, I'm still learning more about crypto every day. But the more things are out in the market and the more people are working on tooling, the more processes become easier.

KC: You’re known to focus on working with emerging technology companies, specifically in crypto. Are there any considerations you take into account when you work with a crypto company versus our average tech startup?

AH: There’s still a lot I don’t know about crypto, I still ask you and Jon questions all the time. But that’s also something I enjoy, I like learning new things and facing new challenges. There’s a challenge to developing a net-new visual metaphor for an object that no one else has had to think about before—like an asset bridge for example. Or, how do you make visual metaphors for zero knowledge?

I also have a very specific point of view in my work, it’s very graphic design-driven. A lot of inspiration I pull from is from the 1960s and 1970s, so part of the fun is developing a visual metaphor for something entirely new, while also pulling from older references. This is a tangential point, but I love the design inspiration I can take from someone who may not have had the same tools as I do now, but still did incredible work that I couldn’t do today.

In terms of consideration for working with teams, I try to just make sure we are as aligned as possible. By alignment I mean that I’m very upfront about how I work and my point of view. What you see on my website is what you get. Once we’re aligned on process and method and there’s implicit trust, we can work together.

KC: There’s sometimes criticism about how crypto website copy and design tend to use a lot of very technical jargon that isn’t accessible to most people. Curious if you have thoughts or reactions to that discourse?

AH: I tend to not participate in these types of discussions publicly, but I have had private conversations about this. I think my work was even mentioned in a thread about this, but I don’t take offense to it. I understand where the perspective comes from, especially since the industry can seem convoluted to outsiders. However, many of these products are designed for insiders, they’re not meant for mass-market—particularly infrastructure companies aiming to improve the world through partnerships with other companies that later are consumer-facing. They need to be able to talk to developers in a certain way that makes sense to them, and then those companies can more easily build for consumers, be it Coinbase or MoonPay. It’s an evolution, I guess.

While we can always strive to make things more approachable and do a better job of it, my main goal isn't necessarily about accessibility. Rather, I aim to help differentiate the company I work with from its peers. This may involve making things more approachable as a byproduct, but it's not my primary focus.

KC: Walk me through the process of how you work with teams as clients.

AH: One of the challenges of working with emerging technology companies is that oftentimes, they’re proposing something entirely new to the market. Only, maybe 1,000 people actually understand what they’re building. Even the clients sometimes have trouble trying to explain to me what they’re building in simple language. They’ll apologize and say something like, “Sorry, this is really complicated, we just need a new brand.” I totally understand that, and it’s my job to be honest when I don’t understand something and to do my best.

For example, with EigenLayer, restaking just sounded like some kind of buzzword to me at first, but when I dug deeper into it, it actually was pretty profound—the thesis of expanding resources and non-zero-sum games around data availability and economic security. Even those terms are their own separate mouthfuls.

I feel more aligned with a company when they’re able to have a longer discussion with me about what they’re building. In the past, I would turn away clients because I didn’t understand what they were building. Now, I take on more clients than I would in the past because the learning curve has gotten easier. As time goes on, I build a better foundation of knowledge that allows me to better understand future clients. Especially working with foundational infrastructure companies in the past, it helps me better understand other companies both vertically and horizontally.

That all being said, I really don’t need to know that much about a company in order to work with them. With EigenLayer, I had to understand the baseline concept of restaking, they told me about their culture and how the team is very academia-driven. The branding reflects that, with my own twist on it, incorporating typefaces typically used in direct-to-consumer contexts.

KC: How do you think about the importance of a personal brand as a designer?

AH: I'm probably a bit of a purist in some way in that I want to be known for my work and for my work to speak for itself. That's the most important part. I try to do a lot of work so that I can show a lot of work and therefore get more work. Some people try to fill the gaps between work with more personal brand-oriented posting, but I just feel really weird when I try to act cool on the internet. I did a lot of that early in my career where I would respond to posts, and sometimes I would get hundreds of followers from that, so it definitely works.

But I think these days, I'm a little bit more focused on showing my work. Hopefully it’s good work that people resonate with. When I meet people that know of my work nowadays, it’s usually through my website. That’s also a part of personal brand, I think it’s important for designers to update their portfolios. You learn so much in a year, it’s like a year of studying or a year of school. I really enjoy having annual portfolios that reflect that growth, it’s meditative for me. So I think it’s a shame when people learn so much but don’t update their portfolio.

KC: What/who are your biggest design inspirations nowadays?

AH: There's a Japanese designer that I think passed away 4 or 5 years ago, Mitsuo Katsui. I found his work somewhere and he had such a deep body of work that I felt resonated with what I'm trying to do with my own work. I’m kind of speechless at how he was able to accomplish some of his work without the tools that exist today. Besides that, it’s no secret that I’m a massive fan of Karel Martens. I have books from both of them.

KC: Do you have any general design principles that you live by?

AH: The first is always maintaining a healthy balance of art and design. By that I mean, in the corporate branding world, art is usually reference and design is what you actually do. I think those two things can be much closer than people tend to think. I try to merge those worlds as much as possible.

It also doesn’t have to be profound, it’s just graphic imagery—like the Baldessari dots, for example. Brands need to be instinctively visually captivating, people crave unexpected experiences, even if they don’t think that they do.

Thirdly, there should be a method to the madness. Whenever something feels chaotic, make sure there’s a reason to it, so that you can scale that reasoning.

Next is similar to the first one, but conflict brings cohesion. I don’t mean interpersonal conflict, I mean playing with tension between concepts and understanding them.

Then lastly is paying homage. Especially as we’re talking about emerging technologies that can be extremely referential, I think it’s important to understand the past and why people made the decisions that they did.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

Jared Madere is an artist and curator based in New York, as well as the Co-founder of Galerie Yeche Lange. The Yeche Lange team also includes Milo Conroy, Wretched Worm, Miles Peyton, Supermetal Bosch, Anastasios Karnazes, and Hikkimourning. Yeche Lange has collaborated with crypto-native communities including little swag world, Tojiba CPU Corp, and more. Yeche Lange is opening a physical gallery in New York on May 17, 2024.

Prior to founding Galerie Yeche Lange, Madere co-ran other galleries in Los Angeles and New York. His own works have been displayed at institutions such as the Whitney Museum of American Art, Le Magasin, La Panacée, The Watermill Center, and others.

Over a video call, Madere and I talked about the relationship between digital art and physical space, bridging traditional art and degen culture, the next generation of art collectors, and more.

The following interview has been edited and condensed for length and clarity.

Katie Chiou: For those unfamiliar with your work, can you share more about your background as an artist and curator? And specifically about your foray into digital art?

Jared Madere: I've always made digital work, but my baptism into the traditional art world was mostly large-scale, collaborative installations. The scale of this work requires institutional budgets and bureaucracies, so the process can be less than magical. You plan everything out, follow some sort of script, and the thing materializes exactly how it is in your mind. At a certain point, I became really bored with this direct 1/1 relationship, it lacked the sort of material improvisation and self reflective reaction associated with painting.

I became interested in this idea of working with other people to introduce an element of something that can't be anticipated into the process. I would come up with an initial script or prompt that was intentionally worded in a way to create ambiguity in its possible interpretation. So there would be material instructions like “Create xyz image out of bisected squid and sand” or “Create xyz image out of shredded beaded curtains and raspberries.” I was working a lot with tattoo artists and graffiti artists who were used to taking descriptions and very quickly interpreting them in a utilitarian way versus someone who comes from a conceptual art background that tries to be clever with their translation. I got interested in the way that something physically more reductive, i.e., the written word or a still image, could do that in a more portable way. The thought of compressing an idea into something that would fit into someone else's head and then could be unpacked according to their own idiosyncrasies. This happens commonly in literature, where there's this kind of transference between the author and the reader, where the vision gets compressed and committed to a textual format and then gets re-inflated in the mind of the reader. I became very hypnotized by how elegant that way of interacting with an audience was.

Also I think, with showing other people's art, there was always this interest in the democratization of ideas. I feel like that word often gets used to mean something specific related to class and accessibility. I don't necessarily mean it in that way. I guess to a certain extent I do, but I think about things like what Kanye and Miami Basel did for exporting ideas that were traditionally relegated to a less accessible field. But there’s also a perverse thing that happens when you round off an idea's edges and prepare it for mass export. I think about it in terms of the way fashion discourse transformed around 2011-2012, where labels like Margiela and Comme des Garçons participated. They weren't just luxury brands, but had commentary—like Comme des Garçons on deconstruction at a material level, or Margiela on deconstruction of the meaning of normative clothes with a wink. Around the birth of Instagram, these meanings flattened, becoming luxury goods like any other designer label. Despite this flattening the original seed is somehow still transmitted the same way that if you understand Skims you understand Judd at least on a subconscious level.

I think this idea that in that reduction, somehow the ghost of the idea remains in there if you can read between the lines or listen carefully enough. That was always an interest of mine—how to get these ideas, typically relegated to contemporary art, into the heads of more people.

KC: What about your interest in crypto?

JM: I had never been interested in crypto before NFTs. I didn't pay close attention to crypto beyond knowing it existed. But when NFTs appeared, it was immediately clear how they could be a very powerful vehicle for distributing art. I had always been envious of how music could penetrate people’s lives, while art felt less agile—you don't really bring a painting or sculpture on a road trip. When NFTs became a possibility, you had audiences curious about art that hadn’t explored it before. One example of this is Roman Signer, a Swiss sculptor with works like a kinetic sculpture of rain boots launching water or a white dress shirt mounted on an automatic zipline through a forest. By early 2021, I was in group chats where Apes were being discussed, and after 8 months of self-educating, the same group chats were talking about Roman Signer—this relatively obscure sculptor wasn't even on the radar of many trad art friends who live and breathe art.

NFTs provided a way for art to exist in people's lives differently, trafficking these ideas to new audiences. People who had been flipping coins were now interested in digging deeper into ideas previously relegated to other spheres. I started thinking about building Yeche Lange as a vehicle to bring my art friends into this new context with the idea of bridging them to the possibilities there.

KC: Earlier in our conversation, you referred a lot to the flattening of ideas especially due to speed of idea dissemination. In many ways, crypto culture is a manifestation of this type of speed.

JM: Yes.

KC: I’m curious if, by participating in crypto, you’re hoping to make some sort of commentary on that observation? Or if it’s generally intentional in any way?

JM: I never really think of it in terms of commentary. I guess what I'm hoping to do is bring ideas that have traditionally been relegated to less accessible corners of the contemporary art world to a new group of people so that they can unpack them from a different vantage point.

I think this is especially true with the new physical space we're doing with Yeche Lange in New York. It's a way of introducing traditional audiences to something that they haven't really had a way of viewing. With my traditional art friends, NFTs were just some sort of obscure esoteric interest that you really had to pour yourself into in order to understand. A lot of them didn't have the patience or didn't know where to begin. At the time, around 2021, NFTs also all looked the same—some pixelated animal with a rhyming name with different baseball hats and different 3D glasses on it. To me, this seemed so bizarrely reductive, it's like it's closer to stamp collecting if stamps had more limited imagery.

In the gallery, we're going to have a massive DriFella painting. To me, DriFellas represent some of the most radical image-making I've seen in the last 15 to 20 years. They stand out compared to what's been happening in the art world in recent decades, where the most advanced stances orbited around this sort of cynical celebration of the limpness of painting as an uninventive space, maybe seen most directly with artists like Michael Krebber or Merlin Carpenter.

DriFellas pick up from a very different place. The artists behind it aren't even engaging in those conversations, or if they are, it's through a completely different lens. The way they approach image-making, and the entire process, contrasts sharply. For instance, telling an artist to create a 3K or 10K piece collection is unfathomable to them. They may say, “I don't know if I'll ever make 3,000 works in my life. That seems psychotic.” DriFella’s approach involves creating a few thousand image components that are pseudo-randomly collaged together according to predefined weights. The resulting images feel like violent onslaughts of stickers piled on top of each other, creating complex foreground-background relationships. These are some of the most exciting images made in decades.

This crypto world has been opaque to the traditional art world—unless you're a degen or on crypto Twitter, you don't know where to look. You and I know how to browse collections on Tensor or OpenSea, but even then, you might miss something. Compared to the traditional art world, where you can visit a gallery website or do a Google search, this world is opaque to non-participants. Making the fruits of these cultures visible to audiences beyond collectors has been a huge driver for us to bring Yeche Lange into the form of a physical gallery as a bridge. We’re also launching this Yeche Lange archive platform, cataloging every collection these artists have done, including their IRL work to make all of this less opaque regardless.

KC: People have complained for years about how to display NFTs in physical spaces. How do you think about the relationship between digital art and physical space?

JM: I also think that’s a funny thing about the NFT space. Our gallery will also eventually have screens, as there’s motion work we want to show, but in our first show, everything is painting, sculpture, and prints. We're probably going to be the first NFT gallery to open a show without a single screen. You know how it goes—90% of NFT exhibitions are just flat screens with LED rimmed backs casting arbitrary colors onto the walls. I personally dislike being in dark rooms; I love sunlight, natural light, windows. It's important to me that when we present digital media, we don't confine it to caves. There's an added dimension when you put these things in a room with a window, where a breeze can come in, and you see leaves vibrating outside.

I've always been less interested in structuralism, meaning the belief that digital media should only be displayed digitally. This notion echoes through contemporary art and modernism in slogans like "the medium is the message," and similar ideas advocating that paintings should remain flat because that explores their full potential. In this modernist belief system introducing three-dimensional elements to a traditionally flat medium is a perversion of its essence. I’m more interested in them as substrates that carry communications. For me, these are historical conversations that demonstrated a point but should ultimately be moved beyond. Ultimately, the printing press wasn't just about the mechanics of printing but about distributing the Bible. I think of digital displays not just as screens but as another medium through which ideas are transported—like the literary form.

Many of the “sophisticated” NFTs in recent years have structurally engaged with blockchain, but I see those as inevitable conversations. Like in cinema, I'm not particularly fascinated by artworks like Tony Conrad’s sculptures where celluloid is put in pickle jars and this triple underlining this point that these too count as films. I'm more interested in Godzilla vs. Kong which engages with the core components of cinema from a less form and structural underpinning-obsessed place.

KC: Yeche Lange launched an NFT collection in 2022 “Galerie Yeche Lange Pie Keys” meant to act as a premium membership key of sorts. Any reflections or takeaways from that experience?

JM: Well, that was an interesting collection because it was released for free, serving as a way to introduce people to the gallery. I never expected it to become particularly popular. I imagined some would know about it, but many might remain unminted, with my friends being the primary audience. However, it minted out almost instantly, generating insane volume in a couple hours that basically funded the gallery treasury for two years.

There were a lot of interesting takeaways. Initially, I wasn't interested in making a 10K collection, I romanticized the idea of a single painting that takes 10 years, and we wanted to create a vehicle that allowed artists to work that way. I think a lot of us were thinking about this at the time. We wanted to have a vehicle that could allow artists to work within this medium that also had a way of bringing traditional artists onboard. For them, it might be unfathomable to create a 10K collection, but it was very easily conceivable to create a very labored single image or 3D object or whatever art object it is. However, between the Pie Keys and the gallery launch was the crash, and the meta changed drastically. Artists who had been market darlings in 2021, whose collections always minted instantly, suddenly saw 10 1-of-1 collections sitting for weeks. Our approach was to bridge worlds, aiming for price points comparable to what artists would get from traditional galleries. But post-crash, these prices seemed steep to even many degen collectors, and they were less interested in flipping single works. Our initial idea of bridging these worlds was shaken and has since remade its face many times.

KC: Launching an NFT collection at that time would’ve looked very different than launching one the year before or the year after, I think.

JM: A huge change we’ve seen as we’ve watched art scenes, particularly the scene that comes out of little swag world, is that there used to be a general degenship where people would buy Pepes and also generative art and also Murakami NFTs and also Petra Cortright because “someone in the trad art told me she’s in museums, so I should get that.”

There’s become much more focused collectorship recently, people who are really diehard. When I look at the traditional art world, I see a huge gulf generationally. You see the values of many traditional art collectors rooted in wanting to see their names on the steps of the opera house or have museum wings with their names above the entrance, but their kids aren’t really stepping up to fill those shoes in the same way. The majority of collectors that are interacting with projects like DriFella are participating at a very different price point but they are infinitely more engaged—eating, breathing, sleeping art from these scenes and spending huge swaths of their life in group chats focused on them.

I think collecting art is going to look different. Right now, a huge number of traditional collectors define the market by buying $200K paintings, but there are fewer degens willing to spend $200K on a 1-of-1. Comparatively, there's another economy where 500 people buy 100 things each for $10 to $100 a pop.

When I look at these two worlds, I see one slowly declining and in dire need of reinvention, and another burgeoning with extremely genuine engagement and consistent dedication that will grow as the scene matures. Many of these hyper-engaged younger collectors are under 30, which is a rarity in the traditional art world. These people are going to own homes and want art on their walls, but not the same decorative paintings you see in Chelsea, etc.

KC: You’ve mentioned in past interviews your belief that we shouldn’t necessarily laud the complexity of technology. How do you think people should engage with technology in the context of art?

JM: For myself, personally, when I first got out of school, it was very in vogue to transpose a structural conversation onto painting. There was a cliché from the late 80s into the 90s of "painting is dead," and a whole generation of painters reacted to the 90s with exhibitions like Sensation, which Damien Hirst, Tracey Emin, and others emerged from. These shows were sensationalistic, featuring things like a rotting cow carcass in a glass box being gnawed on by flies or a tent covered with the names of hundreds of people the artist slept with. You had this group of painters that were interested in structural conversations that looked up to conceptual artists like Tony Conrad. They aimed to show that painting didn't have to be a "dumb thing," it was like “I'm making paintings, but I'm doing it in this very self reflexive way.” You had these projects that had very indexical relationships to the labor that went into them. 20-foot canvases were being painted 24/7 by studio assistants in specific colors demarcating the length of their shifts and other metrics of their labor being visually charted out. All this felt like a very heavy justification, kind of like math class, where it was this kind of “show your work” approach.

It’s similar to why I'm more interested in DriFella than in code-based projects generated entirely onchain, which embody hardcore blockchain structuralism. DriFella engages with NFT cultural history in a very direct way, akin to how Godzilla reboots engage with cinema, not in a Derek Jarman’s Blue-like manner, where you stare at a blue screen for two hours to investigate materiality or mechanics.

KC: What do you hope Yeche Lange looks like 1, 2, 5 years from now?

JM: I guess I'd like to see these conversations just engaging with a wider audience, meaning, I think it's two things I'd like to see.

One is creating a bridge to make these conversations legible to a broader audience. We're working on a book, similar to the archive, cataloging these projects. These efforts speak to the diehard audience, degens or whatever you want to call them, but also serve to bridge to others not participating as directly.

I saw this at a party we did with Club Chess and Zora. We had Bosch's game set up due to its similarity to chess. A woman who had recently moved from Texas introduced herself saying she liked my work naming projects I had only really ever shown on Twitter. I asked if she collected NFTs, and she said, "No, I don't even have a wallet, but I love the art." She mentioned Wretched Worm and Bosch as favorites. This was exciting to hear because it showed that these projects have legs to reach outside the NFT sphere.

On the other side of this, many projects within the NFT sphere are still in their infancy. Most projects remain vaguely PFPs, representing something with eyes and a mouth, or engage in a structural way that can seem thin and lacks the ability to speak to a wider audience. They don't evoke the same emotional response as traditional art. I had a teacher in school who said their criteria for art was that their perception of the street was changed when they left the theater. I’m looking forward to seeing more of that onchain.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Verifiable compute is poised to become as important of a technological revolution as blockchain itself. It encompasses the commercialization of various cryptographic techniques that can be collectively as impactful for blockchains as HTTPS has been for the internet. While many technological hurdles and barriers to adoption exist, the density of sheer brain power pushing its limits points to an optimistic (or rather verifiable) future.

Verifiable compute and blockchains aren’t only groundbreaking technologies in their own right—they’re also uniquely complimentary. Where blockchains provide economically-guaranteed integrity by having multiple untrusted machines re-execute the same programs and agree on the end state, verifiable compute provides mathematically-guaranteed integrity by proving that some end state is the result of a set of inputs into some program.

Together, these two emerging technologies are poised to expand the design space of what is possible with blockchains, solve the issues present in today’s blockchain designs, and address broader societal concerns that arise from the proliferation of generative AI.

In this piece I will:

• Provide an overview and history of verifiable compute.

• Outline the use cases for verifiable compute.

• Introduce a framework for evaluating verifiable compute projects.

Note to the reader: there will be several abbreviations used throughout—if you are unfamiliar with these terms then you can reference the definitions at the end of the piece.

A Technology 90 Years in the Making

While verifiable compute is becoming an increasingly important tool for scaling blockchains, it is far from a new concept. We can trace its roots all the way back to the commercialization of the internet in the early 90s and, in fact, mathematical concepts even further back to the 1930s.

Kurt Gödel can be considered the grandfather of arithmetization, a key component of verifiable compute protocols. He figured out in 1931, before computers really existed, that you can efficiently encode complex mathematical expressions, logical proofs, or even entire programs into a numerical form. In other words, that’s right, you can turn code into math. While Gödel numbering played a foundational role in the development of mathematical concepts related to verifiable compute, the term as we know it today emerged because of two technological developments following the commercialization of the internet.

The first was distributed computation. Several projects emerged in the late 1990s and early 2000s that experimented with this idea, namely the Mersenne prime search in 1996, SETI@Home in 1999, and Folding@Home in 2000, all of which distributed computations to millions of internet clients to take advantage of their idle cycles. While the idea was novel, an issue that all of those projects faced was dishonest clients—end users could modify their client software to return plausible results without performing any actual work.

The second development was cloud computing. As cloud computing became more mainstream in the late 2000s, applications began outsourcing computation to commercial cloud computing services and increasingly wanted guarantees that the cloud performed the computation correctly to protect against faulty or malicious providers.

Before we continue, here is a quick refresher (or introduction) on computational complexity. Ideally we want constant or logarithmic complexity because that is when it becomes commercially viable at scale.

The initial cryptographic techniques for addressing these issues were introduced around the late 80s and early 90s, namely Interactive Proofs (IPs) in 1985 and (non-interactive) Probabilistically Checkable Proofs (PCPs) in 1992. In particular, Shamir’s IP=PSPACE paper in 1992 paved the way for people’s understanding around the potential of proof systems by demonstrating that any computation (which requires a generally practical amount of memory to solve) can be efficiently checked for correctness through the use of interactive proofs. These breakthroughs, however, were simply not commercially viable. IPs were exponential-time for the prover and PCPs were so long (superlinear-size) that it would have taken thousands of years to generate and check them.

Innovation began accelerating over a decade later in 2007 with two key advancements. One was the introduction of more performant Argument systems (IKO07) which were converted from PCPs and offered a succinct prover and more efficient verifier while maintaining computational soundness (i.e., making the assumption that a malicious actor does not have unlimited computing power). Another, GKR08, introduced an interactive proof protocol with a polynomial-time prover (prior protocols had superpolynomial-time provers) and a quasilinear-time verifier. That said, both were still completely impractical if implemented as-is.

It took five more years for researchers to propose notable improvements to these protocols in 2012, with Pepper refining IKO07 to achieve better costs (but still had a quadratic-time prover so did not scale to large computations) and CMT12 reducing the prover time in the GKR protocol from quartic-time to quasilinear-time. 2012 also saw an introduction of new “underlying cryptographic machinery”, which served as the theoretical foundation of the first SNARK that remains popular today-Groth16 (introduced in 2016).

Since 2016, advancements in proof systems have only continued to accelerate, with recursive SNARKs, STARKs in 2018, and folding schemes in 2021 making generating and verifying proofs orders of magnitude cheaper and less computationally intensive. Today, proofs for many common computational tasks can be generated in a matter of seconds using commodity hardware. While the performance still isn’t ideal for many blockchain applications, it is generally believed to be commercially viable at scale.

Verifiable Compute 101

The easiest way to verify a computation is to simply repeat the computation, but that’s clearly inefficient and results in wasted time and energy. Techniques for verifiable compute aim to run computations on your behalf while providing assurances around the validity of the computation’s inputs, outputs, and methods used.

While people often think of zero-knowledge proofs (ZKPs) when they think of verifiable compute, it is important to note that these concepts are related but different. Verifiable compute is a broader concept that encompasses various techniques for ensuring computational integrity. It doesn't necessarily involve zero knowledge and the goal is moreso to prove the correctness of computation rather than to preserve privacy. Moreover, zero knowledge is an additional property that verifiable compute protocols can obtain.

That said, there are similarities between the concepts. Both ZKPs and verifiable compute utilize cryptographic techniques designed to establish trust and verify information. In addition, the primary constraint for both concepts is that the verification of the computation should require substantially less computational effort than performing the computation from scratch.

The traditional framing of zero-knowledge proofs is that they provide a way to cryptographically prove knowledge of a particular set of information without revealing what that information is. While this framing focused on proving simple statements, recent advances have unlocked breakthroughs in "general-purpose" ZKPs that allow you to prove arithmetic circuits, which can be encodings of real programs. These circuits are fundamentally a series of mathematical statements that are used as the basis for the proof. The circuit verifies details about the program’s execution, such as:

• Who called the program’s function?

• Did the caller pass correct inputs to the program?

• Did the execution use the correct program?

• Was the final output the result of correctly executing the program with the provided inputs?

In other words, you are effectively modeling the physical logic gates and circuits that make up a program as arithmetic gates and equations that give you an equivalent output for that program. The broader technique involves a programmer specifying a computation in a high-level language and a compiler transforming the computation to the formalism that the verification machinery uses, ultimately outputting executables that implement the verifier and prover. In short, ZKPs turn code into math.

There are other tangential technologies that are often associated with verifiable compute and it is important to clarify the differences. Fully homomorphic encryption (FHE) is a technique that allows computations to be performed on encrypted data without revealing the plaintext. While FHE is also used for privacy, and in fact provides stronger privacy properties because it ensures full-state privacy, the technology does not ensure the inherent correctness of the computation itself. Like FHE, trusted execution environments (TEEs) are also primarily used for confidential compute, but do not provide verifiable compute. TEEs ensure computational integrity with an attestation-based trust model that provides verification for three things: the application’s identity, its intactness (that it has not been tampered with), and that it is running securely within an enclave.

Issues

While researchers and developers have made tremendous strides in commercializing the technology over the last decade, we still have a long way to go.

Costs

Verifiability does not come for free. There is an inherent overhead for proving computation, and it is high for cryptography-based systems. From one source, it is currently approximately 1,000x to 100,000x more expensive to generate a ZKP for a computation compared to natively performing the computation. From another source, the Jolt prover, which is currently 6x faster than other currently deployed zkVMs, it is 500,000x slower relative to native execution of a RISC-V program. In other words, proving one step of the RISC-V CPU requires ~500,000 cycles of the RISC-V CPU.

The good news is that the efficiency of chips (power usage per unit of compute) improves exponentially over time, so we can expect costs to gradually converge to the cost of native compute. Even still, without additional advances in hardware (e.g. ASICs for ZKPs) and algorithms (i.e. even more performant proof systems), it is hard to imagine the overhead being less than 100x over the next decade. And for mainstream adoption, the power required to produce a proof likely needs to be a fraction of the percentage of power required to do the computation, rather than orders of magnitude more.

Societal Value

That said, I believe that there will be increased societal value placed on verifiability, which will translate into a willingness to pay for commercial use cases for verifiable compute technology, especially in low-trust environments like blockchains.

We already see this today with ZK rollups (ZKRUs). If it costs a ZKRU ~$200K/month to run a prover but they earn ~$1M/month in sequencer revenue, the economics clearly make sense. Naively, one could argue that the economics already make sense from a cryptonetwork perspective. If Ethereum has 7,500 nodes (at the time of this writing), each of which re-executes every transaction, one could argue that a ZKRU with 5,000x overhead is still cheaper for the system as a whole.

We will see more data points here over time. For example, if consumers are willing to pay more for a device (e.g. cameras, microphones) with an attested sensor (a TEE that stores a secret key for digital signatures), then we can say that GenAI is likely driving real concern over the authenticity of content.

Tooling

Despite the power of verifiable compute, it remains largely inaccessible to the vast majority of developers. Domain specific languages (DSL) like Circom, Noir, Leo, Lurk, and Cairo are effective but require large communities and comprehensive libraries to be adopted at scale. Transpilers are slow, and virtual machines that provide abstraction come at the cost of an additional 100x to 1000x overhead.

Ideally developers can avoid thinking about circuits, constraints, and arithmetization altogether, and the developer experience is as close to that of the native blockchain on which the developer is building their application. The performance trade-off is glaring here, but I would argue that the experience as-is caters to too niche of a developer base to generate enough experimentation with new use cases (for if the use case is compelling enough, there will likely be a willingness to pay from users). Furthermore, developers need simpler deployment workflows that incorporate verifiable compute for their applications, including the underlying infrastructure and integration with onchain components.

An Overview of Use Cases

The blockchain industry often frames use cases through an introspective lens—L1s, EVMs, defi, bridges, etc. But because the use cases for verifiable compute can extend beyond blockchains themselves, they deserve a broader framing that extends beyond those we see in production today.

Privacy

The first production use case for verifiable compute has been privacy, starting with Zcash in 2014 and applied to projects like Aztec in 2017, TornadoCash in 2019, and Nocturne in 2022. These use cases were particularly focused on payments and aimed to make private the sender, receiver, and/or amount of a given transaction.

To provide a more tangible intuition of what exactly is being “proven” for privacy-related use cases, the zk SNARK for Zcash proves the following:

• Knowledge of Spending Key: The transaction creator knows the spending key corresponding to the notes they intend to spend without revealing the keys themselves.

• Unspent Note and No Double-Spending: Each note being spent has not been spent before, managed through nullifiers, which are cryptographic values derived uniquely and deterministically from the spending key and the note. The nullifiers for spent notes are revealed and checked against a list of all nullifiers from previously spent notes.

• Conservation of Value: The total value of inputs equals the total value of outputs.

• Merkle Tree Validity: The spent notes exist in the blockchain's current Merkle tree of all notes (spent and unspent).

The general issue with mainstream adoption of these solutions has been difficulty with regulatory compliance. In response, new projects have been leveraging verifiable compute to provide certain guarantees around this. For example, a user can generate a proof saying that “I didn’t interact with anyone on the OFAC list.” It is also possible to make the privacy properties adjustable and “pull back the privacy curtain” if necessary. For example, one can reveal the connections between addresses but keep the values and amounts hidden. This paper has a good overview of what can be done and what fundamentally can’t. This use case actually does demand the zero-knowledge property that ZKPs provide, but most other use cases rely on the succinctness properties of ZKPs.

Compression

The second production use case has been around succinctness, which can be reframed as state and compute compression. ZKPs can be used to post a proof onchain that says “here is some batch of valid transactions whose execution leads to some valid new global state.” There are several blockchain-native use cases here, notably ZKRUs and coprocessors. ZKRUs scale the execution throughput of blockchains by maintaining a partitioned persistent state that is periodically posted and finalized at the settlement layer. Coprocessors allow blockchain applications to use offchain compute while accessing the full state of the underlying chain without adding any trust assumptions to the application itself. The primary difference between ZKRUs and coprocessors is that the former is stateful while the latter is not.

Compute compression can also improve security for blockchain interoperability. Instead of relying on economic or game theoretic security, projects can efficiently verify proofs that the data posted on the destination chain has been signed by the validator set on the source chain, thus implementing “proofs of consensus.” In other words, instead of verifying a bunch of signatures onchain, which is generally cost-prohibitive, you verify a proof of a circuit that verifies a bunch of signatures. This is the basis for the “ZK bridges” we see coming to market today.

To provide a more tangible intuition of what exactly is being “proven” for compression-related use cases, a proof for a ZKRU proves the following:

• Validity of Transactions: Each transaction in the batch has been authorized by the owner of the funds.

• Correct State Transitions: The application of the state transition function is correct, including whether balances are updated correctly and ensuring that inputs, outputs, and fees are accounted for correctly.

• Conservation of Value: The sum of inputs equals the sum of outputs plus any transaction fees.

• Accurate State Root: Confirms that the transition from the old state root to the new state root is valid.

• No Double-Spending: Each account transaction must increment the nonce correctly.

Another example of an application using this property is onchain games. Instead of verifying individual actions of a game onchain, users can simulate the game locally and provide a proof of their game’s end state which can then be trustlessly verified. This also computationally enforces the game logic because the proof will be invalid if someone submits a forbidden action.

Data Integrity

Verifiable compute can be used to prove that some piece of data from either onchain or offchain sources, as well as any additional computation on that data, is accurate and has not been tampered with. This enables a class of oracles that can be framed as verifiable oracles, which can be used in a wide range of applications, such as price feeds for decentralized exchanges and ingesting arbitrary web2 data and identity onchain. Verifiable compute can also be used to scale oracles in general by performing price updates and signature verification offchain in a circuit and posting proofs containing the signed updates onchain. It can also be used for building verifiable databases, which ensure data provenance by proving correctness of both data injection and retrieval.

To provide a more tangible intuition of what exactly is being “proven” for data integrity-related use cases, the proof for a zkTLS project like DECO proves the following:

• Data Possession: The user (prover) possesses a value that corresponds to specific data obtained via a secure TLS session.

• Data Origin: The data originated from a specific server that is verified through the server's TLS certificate.

• Data Integrity and Freshness: The data is unchanged from what the server sent and was retrieved in a timely manner.

A note on oracles: one can consider something an oracle if it breaks the chain of custody of the data that is being provided. Any data ingested onchain from offchain like ML models or exchange feeds is certainly considered an oracle, but what about offchain computation on L1 state or execution of L2 state? In an attempt to further categorize this, I suggest that all coprocessors are also oracles, specifically verifiable oracles, but not all oracles are coprocessors. In addition, verifiable oracles can be considered coprocessors if they perform compute on the data being provided along with a proof for the output of that computation.

A related use case is storage proofs, which proves the value in a storage slot for an address at a specific block. The “proof” here is not in the context of a ZKP but rather a light client proof which consists of the block header, state root, and Merkle-Patricia inclusion proofs for the key-value pairs of account data in the state trie and storage slot data in the storage trie. Verifying this light client proof requires checking that the block header, state trie proof, and storage trie proof is correct, which involves many Keccak hashes and is very expensive to do onchain. Instead, one can do each of these checks directly in a circuit and only need to verify ZK proof onchain, which is much cheaper.

Machine Learning

Verifiable compute also has applications for machine learning (often referred to as “zkML”) by proving one or more steps along the “ML pipeline”. zkML can be used to delegate the execution of machine learning models to service providers and obtain proofs that a particular trained model produces the correct output given some inputs (i.e. verifiable inference), which ensures transparency and prevents model substitution. There are different flavors of verifiable inference: public model/private (but signed) input like Worldcoin, private model/public input like credit scores, and private model/private input like KYC.

It can also be used for verifiable training by generating proofs that the correct dataset and learning algorithm were used in the creation of a model (e.g. proving that no data poisoning occurred in a dataset of non-copyrighted works). zkML can also be used for proving specific properties of a model, such as fairness (e.g. the model is not biased), which can be important for certain applications, such as granting university admissions, assessing loan applications, and distributing medical resources. In a blockchain context, zkML effectively acts as an oracle that ensures the integrity of external models and their outputs that can be used to update mission-critical blockchain state.

In terms of the proofs themselves, the components being “proved” are generally:

• Correct Application: A specific model and set of parameters (e.g. weights of a neural network) were used to compute the output from given inputs.

• Integrity of Parameters: The parameters used are the ones claimed (i.e. they haven't been tampered with or altered).

• Correct Execution: Each step in the computation (e.g. for each layer in a model) was executed correctly.

To dive deeper into the intuition behind zkML, it is helpful to revisit ZKPs. In general, ZKPs enforce consistency between the circuit, committed inputs/outputs, public inputs/outputs, and intermediate values within the circuit (these are generally either committed to and/or not revealed at all to the verifier). Note that the only difference between a committed vs. public input/output is that the former is private (i.e. the prover says "I know some x such that...") while the latter is public (i.e. the prover says "here is x - you can check it for yourself!"). In other words, a ZKP shows that a circuit which takes in some committed/public inputs indeed produces the claimed committed/public outputs, with consistency between those and some set of intermediates which the verifier doesn't care about. Applying this framework to zkML, we have:

• Circuit ≈ model architecture.

• Committed inputs ≈ model weights, which are hidden but consistent between runs (enforcing the same parameters to be used each time), and potentially also private/hidden inputs.

• Committed outputs: not commonly used in zkML unless you want the outputs to be private and perhaps further used in another ZK computation (e.g. proving some property about the private output of a model).

• Public inputs ≈ model inputs, although this can also be model weights for public/open-source models too.

• Public outputs ≈ model outputs.

• Intermediate values within the circuit: typically not used in zkML.

Proof Generation and Aggregation

One of the magical capabilities of certain verifiable compute systems, namely SNARKs and incrementally verifiable Computation (IVC) schemes, is recursion, which is the ability to generate “proofs-of-proofs” by using the output of one proof as the input for another proof. This can decrease the cost of verification by aggregating proofs from different sources and recursively proving them into a singular proof, amortizing the cost for all participants involved.

Proof aggregation can generally be applied to scalability (e.g. a ZKRU ecosystem aggregating proofs from its L3s and L2s to settle onto the L1) and interoperability (e.g. composing proofs from different ZKRU ecosystems to enable more secure and seamless value transfer across them).

It can also be looked at from the perspective of vertical scaling (i.e. rolling up lots of disconnected proofs for blockchain-level scaling) and horizontal scaling (i.e. providing more expressive and interconnected programs).

Conclusion

Verifiable compute expands the design space of blockchains by enabling them to interact with external systems and data in a secure and mathematically verifiable way. In addition, it scales blockchains by removing the need to re-execute the same computations while guaranteeing that the execution was done correctly and effectively, enabling infinitely complex smart contracts that can bypass a blockchain’s gas limits.

On top of all of this, recent advancements in generative AI like OpenAI’s Sora have placed a significant societal pressure on ensuring verifiability and further commercializing the underlying cryptographic techniques. The drive is a fundamentally human one; in a world where we do not know what is real and what is AI-generated, verifiability impacts the identity of both the species and the self.

Despite the issues with performance and cost, we are close to achieving the dream envisioned 33 years ago, where: “...a single reliable PC can monitor the operation of a herd of supercomputers working with possibly extremely powerful but unreliable software and untested hardware.” -Babai, Fortnow, Levin, Szegedy, 1991

If you are building infrastructure or applications around verifiable compute, please reach out!

You can find me on Twitter/X and Farcaster.

Many thanks to Srinath Setty, Thor Kampefner, David Wong, Shumo Chu, JD Kanani, Ben Livshits, Tom Walton-Pocock, Kobi Gurkan, and Ryan Cao for conversations and feedback on this piece.

Definitions

• Verifiable Computation: Enables, through a variety of cryptographic techniques, a computationally weak client to “outsource” computation to a worker with more computational power, who then returns the result along with a proof that the computation was carried out correctly. Verifying the proof should require less effort than performing the computation from scratch.

• Arithmetization: The process of turning a generic statement or question into a set of equations to be verified or solved.

• Proof: a logical demonstration or evidence that establishes the validity of a statement.

• Interactive Proof (IP): An iterative exchange that permits the verifier to pose random queries and the prover to demonstrate the validity of a statement.

• Probabilistically Checkable Proof (PCP): A proof system where the proof's validity can be efficiently verified probabilistically and with high confidence by examining only a small, randomly chosen portion of the proof.

• Interactive Oracle Proof (IOP): A generalization of PCPs and IPs, where a prover interacts with a verifier by making queries to an oracle, aiming to convince the verifier of the correctness of a statement.

• Argument of Knowledge (the ARK in SNARK): A protocol where a prover can convince a verifier of the truth of a statement, while additionally demonstrating knowledge of some secret information related to that statement. An “argument” of knowledge, as opposed to a “proof” of knowledge, relaxes the inflexibility of proofs by stipulating that certain computational tasks are infeasible for the prover.

• Succinct Non-Interactive Argument of Knowledge (SNARK): A protocol where a prover can convince a verifier of the validity of a statement without revealing any underlying information, while maintaining a small and efficiently verifiable proof size. IPs, PCPs, and IOPs can all be compiled to a SNARK by utilizing polynomial commitments (for succinctness) and applying Fiat-Shamir (for non-interactivity).

• Incrementally Verifiable Computation (IVC): An ARK for incremental computations; enables producing proofs of correct execution of “long running” computations in a way that can be efficiently verified in an incremental manner.

• Folding scheme: A protocol used for IVC that is weaker, simpler, and more efficient compared to direct proof composition approaches where a verifier is implemented in a circuit. IVC proofs are usually constructed using a folding scheme and a SNARK at the end to finalize the proof.

Sources

• An incomplete guide to Folding

• Delegating Computation: Interactive Proofs for Muggles

• Do You Need a Zero Knowledge Proof?

• Efficient Arguments without Short PCPs

• Intro to Zero-Knowledge: A Guided Tour of the ZK Landscape, Part 1

• Incrementally Verifiable Computation or Proofs of Knowledge Imply Time/Space Efficiency

• KZG in Practice: Polynomial Commitment Schemes and Their Usage in Scaling Ethereum

• Making Argument Systems for Outsourced Computation Practical (Sometimes)

• Modern Verifiable Computation

• Non-Interactive Verifiable Computing: Outsourcing Computation to Untrusted Workers

• Polynomial Commitments

• Practical Verified Computation with Streaming Interactive Proofs

• Privacy-Protecting Regulatory Solutions Using Zero-Knowledge Proofs

• The Anatomy of Proof Generation

• The knowledge complexity of interactive proof-systems

• The Lifecycle of a ZK Proof

• Verifying Computations Without Reexecuting Them

• Zero knowledge proofs for gaming: how it works & why do we need them

• Zero-knowledge rollups | ethereum.org

• Zero Knowledge Virtual Machine Step by Step

• zkVMs are cool, but have you heard of zkCPUs?

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

In the nine years since Ethereum launched the first decentralized, programmable blockchain, crypto has faced multiple roadblocks in the quest to scale decentralized applications to billions of users. And in order to develop scaling solutions to address this, the crypto industry has continuously funded and developed entirely new types of blockchains to solve the “performance problem.”

However, the “performance problem” has been poorly defined and quantified. Synthetic memes such as “transactions per second” have neatly packaged what are really apples-to-oranges comparisons between transactions that do not require equivalent computational work. The lack of nuance in these metrics also shrouds our ability to assess the independent impacts of a blockchain’s components on performance, distracting us from a principled approach to identifying the sets of optimizations we can make to solving highly interdependent problems.

Despite this fog, we have seen credible, sustained improvements to blockchain scalability play out over the past few years. As Ethereum powers through its rollup-centric roadmap, a new wave of rollups, coprocessors, data availability (DA) layers, and competing L1s are emerging, each with unique design choices to provide developers with more performant environments for building scalable, user-friendly dapps.

Today, the introduction of EIP4844 and alternative DA layers have alleviated the critical DA bottleneck. Despite this critical milestone, evidence suggests other important bottlenecks must be solved. Last month, Base collected $1.57M in transaction fees in a single day while paying only $5K in data availability costs to Ethereum. This suggests that the computational work required to validate and process state updates remains a critical bottleneck and an opportunity for improvement.

This piece will evaluate the design choices made by both integrated and modular execution environments in their path to solving for higher performance and expanding the scope of applications that can live onchain.

Today’s Challenges

The performance of an execution layer can be benchmarked according to the computational work that executing nodes achieve relative to their chains block time, or  "gas computed per second.”

With this in mind, we can narrow execution layer bottlenecks to two interconnected factors: inefficient state access and inefficient computation.

Inefficient state access refers to the overhead of retrieving and updating the blockchain's state, which can slow down transaction processing. On the other hand, inefficient computation is a function of the overhead incurred by the algorithms executing operations and state transitions, which can include everything from simple transfers to complex smart contracts and signature verifications.

These bottlenecks are mutually reinforcing—delays in state access can prolong the time to computation, while inefficient computational practices can strain state management. Moreover, proposed improvements to addressing these issues often require systemic improvements such as sharding or adopting stateless architectures, which enhance state access and computation efficiency to improve execution performance.

Bottleneck #1: Inefficient State Access

The cost and speed required to access the state of a blockchain are critical bottlenecks to performant execution environments and can be distilled to the issue of state bloat.

In blockchains, the state of the world is managed and updated through specific data structures called trees. Trees are integral to blockchains, providing a secure and efficient way to give parties external to the executing node guarantees around the blockchain's correct state. Each update within a trie generates a new root hash, which light clients can reference to verify transactions and account balances without maintaining the entire chain.

Ethereum specifically relies on a data structure known as the Merkle Patricia trie (MPT), which comprises four sub-tries.

As Ethereum adds more smart contracts and tokens to its state, its state trie becomes larger and more complex. As the state grows, it requires more storage space, more computational resources to process, and more bandwidth to transmit. At the same time, the node’s hardware constraints stay roughly the same.

This state growth directly impacts Ethereum's performance because the state is stored in disk, and disk operations incur a high overhead. While accessing data from a CPU register can take 0.1 nanoseconds, it can take between 10 and 100 microseconds (100x–1000x slower) to access data from a disk, roughly translating to 200,000 CPU instructions that could have been executed in that time. That equates to a conservative estimate of 36 ERC-20 transfers that could have been!

Exacerbating this issue, blockchains have many inefficient access patterns for reading and writing to state. For example, the non-sequential structure of the Merkle Patricia trie inherently leads to these disk input/output (I/O) operations reading from and writing to various unpredictable locations on the disk. The random nature of transaction inputs and the subsequent state changes they trigger lead to a scattered data access pattern which significantly slows the process of verifying and updating state and only utilizes a portion of a hardware device’s capacity.

All in all, the state management primitives for blockchains are far from achieving their absolute potential, and numerous advancements can be made to improve computational efficiency.

Bottleneck #2: Inefficient Computation

Execution layers also face the bottleneck of inefficient computation, which manifests in various ways.

For one, many process transactions sequentially, underutilizing modern multi-core processors capable of handling multiple operations simultaneously. This sequential execution leads to inevitable CPU idle times between transactions, wasting valuable computational resources.

Additionally, using virtual machines involves translating high-level smart contract operations into bytecode—a lower-level, platform-independent code—which is then executed instruction by instruction. This translation and execution process introduces significant overhead, especially for applications with complex and frequently repeated application-specific tasks.

These inefficiencies lead to suboptimal utilization of computational resources and hinder the performance of execution layers.

Solutions: Inefficient State Access

There are a few distinct ways that teams are improving the rate at which state can be retrieved and updated from an executing node’s hardware, including simplifying complex data structures and finding ways to reduce the costly disk I/O operations that lead to state bloat.

Statelessness & In-Memory Computing

Some execution layers address state bloat by simply accepting it in the short run. They shift state data storage from slower disk-based systems to faster random access memory (RAM). Accessing state information in RAM significantly reduces the overhead associated with disk operations, which are slower and more resource-intensive.

However, this approach challenges the core principle of decentralization. Storing increasingly large amounts of state data in RAM necessitates more advanced and expensive hardware, which could limit the ability of individuals to participate as node operators. Consequently, as the hardware requirements escalate, fewer entities can afford to run these nodes

To balance the attractiveness of computing in memory with trust minimization, both L1s (such as Ethereum) and L2s are pursuing a scalability roadmap that relies on unbundling the role of a validator into separate, centralized executing nodes with many verifying nodes. In this model, highly performant block producers with the hardware requirements to compute in memory are responsible for generating blocks, and cryptographic proofs (fraud and validity proofs) are leveraged by verifying nodes to keep block producers accountable.

As a result, these systems should allow block producers to maximize their speed because they can be expected to compute in memory, eliminating disk I/Os completely during execution. Because the latency of RAM is typically under 100 nanoseconds, the latency of state access is reduced by up to 1000x relative to disk-based implementations.

In parallel, fraud and validity proofs are used in place of decentralized consensus to scale the trust-minimization properties of the system along with its throughput. As a result, powerful, centralized block-producing nodes are counterbalanced by verifying nodes that can be run on much less expensive hardware. These nodes perform the critical function of independently verifying proofs of state transitions (or invalid state transitions) to maintain an accurate view of the state without the burden of storing the entire blockchain state.

To facilitate this process in a trust-minimized fashion, execution layers must implement a degree of statelessness, the most popular being the concept of  "weak statelessness.” Weak statelessness is achieved by mandating that block producers provide a cryptographic attestation known as a witness to a verifying node. This witness encapsulates all the proposed state changes by the new block, enabling validators to verify these changes without additional historical data.

Although this concept can be applied using various tree structures, Verkle trees are often preferred to Merkle trees for their efficiency. Merkle trees require the inclusion of all sibling node hashes along the path from a data point (leaf) to the tree’s root to prove data integrity. This requirement means that the size of the witness (the proof of integrity) grows with the tree’s height, as each level necessitates additional hashes. Consequently, verifying data integrity in Merkle trees becomes computationally intensive and costly, especially for large datasets. In contrast, Verkle trees streamline this process, reducing the overhead associated with computation and storage in generating and verifying new blocks.

Verkle trees enhance the structure of traditional Merkle trees by streamlining the connections between leaves and the root and eliminating the need to include sibling nodes in the verification process. In a Verkle tree, verifying a proof involves only the value at the leaf node, a commitment to the root node, and a single vector commitment based on polynomial commitments, which replaces the multiple hash-based commitments found in Merkle trees. This shift allows Verkle trees to maintain a fixed-size witness, which does not increase with the tree’s height or the number of leaves verified, significantly improving the efficiency of storage and computation during data verification.

Over the coming years, we'll see implementations of statelessness happen at the L1 and L2 levels with varying configurations. According to the latest Ethereum roadmap, validators can rely on block builders to provide Verkle proofs regarding the state of certain blocks and verify these lightweight proofs instead of maintaining Ethereum's state directly.

At the L2 level, teams like MegaETH are actively applying the concept of statelessness to the design of optimistic rollups. In their design, the sequencer node generates a witness for each block containing the necessary state values and intermediate hashes while emitting a state delta representing the changes in the state. Verifier nodes can then re-execute any block by retrieving the witness from the DA layer or a peer-to-peer network without storing the entire state. In parallel, full nodes update their state by applying the state deltas disseminated through the network, allowing them to stay synchronized without re-executing transactions or storing the entire state history.

However, it’s also worth pointing out that the benefits of statelessness and the resulting ability to compute in memory isn’t a silver bullet for the performance of the execution layer.

As co-founder of MegaETH, Yilong Li, identifies in the following research presentation on Ethereum execution, there are other inefficiencies to the data structures and access patterns onchain that remain optimized.

Improving the Databases

Teams working on execution layers are finding ways to improve the structure of these databases themselves to eliminate some of the bottlenecks experienced by Ethereum and other EVM-compatible blockchains in dealing with inefficient state access, which has a domino effect on computational efficiency.

In fact, the limitations of existing database designs found in the EVM informed Monad's* decision to go beyond purely optimizing for computational efficiency to achieve parallelization. Monad found that even after implementing parallel execution, they only saw a small speedup in performance because multithreaded read and write requests to the database blocked each other. As a result, Monad implemented a database compatible with asynchronous IO (AIO), or parallel access, as a critical part of the solution.

Async I/O

I/O operations—such as reading from or writing to storage devices—often create bottlenecks, particularly with mechanical hard disk drives (HDDs). These drives require the physical movement of a read/write head to access data, which can significantly slow down data processing.

AIO addresses this challenge by allowing programs to perform I/O operations concurrently with other processes. Essentially, a program can initiate an I/O operation and move on without waiting for it to complete. It does this by registering a callback function or a promise that the operating system or an I/O library will fulfill once the I/O operation finishes. This asynchronous approach allows the main program to continue executing other tasks, improving overall efficiency by not stalling for I/O tasks to complete.

Asynchronous I/O can be implemented with both traditional HDDs and solid-state drives (SSDs), although the benefits are more pronounced with SSDs. HDDs can perform AIO, but their mechanical nature means they're inherently slower than SSDs which store data on flash memory and have no moving parts, resulting in faster access times.

For instance, Monad utilizes a custom state backend optimized for SSD storage, which supports high levels of parallel data processing and reduces I/O latency. This setup is more efficient than systems relying solely on traditional disk-based storage or those using in-memory databases, which may still face delays from frequent data writes to and reads from slower storage media.

Similarly, Reth employs a method that separates database operations from the core EVM execution engine. This setup allows the EVM bytecode to execute sequentially on a single thread to maintain consistency while database I/O tasks are offloaded to parallel processes. Reth uses the actor model—a software architecture pattern—to manage these parallel processes effectively, ensuring that I/O operations do not interrupt the EVM interpreter.

State Merklization Frequency

Another vector for optimization is the state merklization frequency. Ethereum's current model of merklizing state after every block introduces significant overhead, requiring frequent writes to and reads from the disk and continuous trie traversals. Merkle trees usually work by grouping intermediate hashes into sets of 16 (called a node) and storing them in a key-value store database where the key is the node hash and the value is the node itself.

Traversing this tree to find and update data requires one random disk access for each layer of the tree to be traversed, and traversing a naive Merkle tree will require roughly eight sequential database queries per entry.

Solana’s approach of updating the commitment to the state delta only at the end of each epoch allows for the amortization of write costs over many transactions within that period. If a state entry is modified multiple times within the same epoch, each write does not require an immediate update to the Merkle root. This reduces the overall computational overhead associated with state updates during the epoch. Consequently, the cost associated with reading from the state remains constant, or O(1), because the state can be read directly without needing to traverse a Merkle path each time.

Reducing the frequency of merklization in Ethereum could decrease the overhead from state reads and writes, enhancing performance. However, light clients would need to replay block changes to track state between epochs or submit onchain transactions for state verification, and such a change is not currently compatible with Ethereum.

Efficient & Specialized Data Structures

Moreover, layered tree structures within existing Ethereum clients generally cause inefficient state access patterns, further contributing to state bloat. While Ethereum's state is structured as an MPT, it is also then stored in Ethereum client databases such as LevelDB, PebbleDB (utilized by go-ethereum), or MDBX (employed by Erigon) that store data in Merkle trees such as a B-Tree or LSM-Tree.

In this setup, a data structure is rooted into another data structure of a separate type, creating "read amplification" from navigating internal tree structures atop clients that operate under another Merkle tree-based system. Read amplification can be understood as the result of the multiple steps for accessing or updating information contained within a state, which requires navigating the outer tree to find the entry point into the MPT before executing the required operation. As a result, the number of disk accesses for a random read is multiplied by a log(n) factor.

To solve this, Monad natively harnesses a Patricia trie data structure on disk and in memory. From a technical perspective, Patricia tries are often superior to other Merkle tree structures due to their unique combination of space efficiency, efficient prefix matching, and minimal node traversal. The trie's design collapses nodes with single children and streamlines lookups, inserts, and deletions, reducing the number of disks or network I/O operations required. Moreover, a Patricia trie's adeptness at handling prefix matching enhances performance in applications needing rapid partial key searches.

Another bottleneck specific to tree-based structures is that accessing or updating data requires traversing multiple layers, leading to numerous sequential disk accesses. Sovereign Labs addresses this inefficiency by advocating for a binary Merkle tree configuration. This pivotal shift to a binary structure drastically reduces the number of potential paths during tree traversal, directly reducing the hash computations needed for updates, insertions, and cryptographic proofs.

An additional example in this category is the Reth team configuring Reth to pre-fetch intermediate trie nodes from the disk during execution by notifying the state root service about storage slots and accounts touched.

State Expiry

State expiry is a mechanism to manage and reduce the size of the blockchain state by removing data that hasn't been accessed for a set period of time. While expiry is often bucketed under the “statelessness” category, it’s critical to distinguish these concepts in the context of execution.

Statelessness improves execution by increasing an executing node's ability to compute in memory, but the improvements to execution stem from the beefier hardware requirements across fewer nodes that execute transactions. In contrast, state expiry can be applied to blockchains with both few and many executing nodes.

There are a couple of methods commonly discussed for implementing state expiry:

• Expiry by Rent: This method involves charging a maintenance fee, or "rent," to keep accounts active within the state database. If the rent isn't paid, the accounts are archived until a fee is paid to restore them.

• Expiry by Time: Here, accounts are deemed inactive if they haven't been accessed—meaning no transactions or interactions—for a specified duration.

Both methods aim to maintain only actively used data in the immediate, accessible state while pushing out the older, less frequently accessed data to an archived state that does not burden the main system.

By maintaining a smaller and more manageable state, state expiry reduces the "state bloat" that can severely hinder blockchain performance. A smaller state size allows nodes to quickly navigate and update the state, translating to faster execution because nodes spend less time scanning and more time processing.

Execution Sharding

Sharding optimizes resource utilization and performance by distributing tasks and data across a limited number of specialized nodes (not every node executes a global state).

In a sharded blockchain architecture, the global state is divided into distinct partitions called shards. Each shard maintains its portion of the state and is responsible for processing a subset of the network's transactions. Transactions are assigned to specific shards based on a deterministic sharding function, which considers various factors such as the sender's address, the recipient's address, and the hash of the transaction data. This minimizes the need for cross-shard communication and enables more efficient transaction execution.

This becomes evident when exploring NEAR Protocol’s sharding design, Nightshade, which achieves statelessness to scale sharding without compromising trust minimization.

In Nightshade, the blockchain is structured as a single logical chain, with each block composed of multiple "chunks" and one chunk being allocated per shard. These chunks contain the transactions and state transitions specific to each shard. Including chunks from all shards within a single block allows for a unified view of the entire blockchain state and simplifies the process of cross-shard communication.

Similarly to proper-builder separation (PBS) on Ethereum, Nightshade explicitly delineates the roles of stateful and stateless nodes. On NEAR, stateful validators are assigned to specific shards and are responsible for collecting transactions, executing them, and producing shard-specific chunks. They maintain the full state of their assigned shard and generate state witnesses for validators to use during the validation process.

Meanwhile, stateless validators are randomly assigned to validate specific shards on a per-block basis. They do not need to maintain the full-sharded state and rely on state witnesses provided by the block producers from other shards to validate the state transitions and transactions within a chunk. The random assignment of validators to shards helps ensure the network's security and integrity, as it makes it more difficult for malicious actors to collude and control a specific shard.

Since each node in the network only needs to handle the data for its respective shard rather than the entire network's data, the storage and computational burden on individual nodes is reduced.

Solutions: Inefficient Computation

Parallelizing Execution

Time to address the elephant in the room: parallelization. Parallelizing transaction execution enables processing multiple transactions by utilizing multiple computing resources concurrently. This allows for increased throughput as hardware resources are scaled up during periods of high demand.

However, it's important to consider that multiple execution components can be parallelized, many of which are implemented by coprocessors such as Lagrange* and alternative blockchain clients such as Firedancer to improve blockchains' performance significantly. Specifically, parallelization can involve:

1. Parallelizing State Access

2. Parallelizing Specific Operations

3. Parallelizing Consensus and Execution

Parallelizing State Access

Parallelizing state access brings two critical benefits:

1. Parallel EVMs distribute transaction processing across several CPU cores. This setup allows multiple transactions to be handled concurrently rather than forcing them to queue up for a single resource.

2. When a transaction waits for data from storage—which can introduce significant latency—the system doesn’t remain idle. Instead, it can switch to another transaction that is ready to execute. This is possible because multiple cores can handle different tasks independently and simultaneously.

The primary challenge in parallelizing transaction execution stems from managing concurrent access to the shared global state without violating the ACID rules for updating distributed systems. If a blockchain has a bunch of transactions executing in parallel, some of them will conflict. As a result, the two primary methodologies for parallelizing state access differ on when they dedicate resources to resolving conflicts: the pessimistic execution (or memory lock) model and the optimistic execution model.

Pessimistic Execution

The pessimistic execution model is a transaction processing approach requiring transactions to declare the state variables they will access (read or write) during execution. This information is included in the transaction's metadata, allowing the runtime to analyze the access patterns before execution.

By examining the read-write access patterns, the runtime can identify transactions with non-overlapping access sets, enabling parallel execution of non-overlapping and read-only transactions and improving throughput. The runtime creates parallel transaction queues for each CPU thread on a validator node, ensuring that transactions with non-conflicting access patterns are processed concurrently.

As a result of this design choice, the pessimistic execution model benefits from fine-grained control over resource allocation, allowing for segmenting or partitioning of a blockchain's state space.

Parallelization effectively creates multiple, synchronously composable independent execution shards underpinned by a unified security model. It helps address network congestion and optimize gas costs through precise resource management and dynamic fee markets. By identifying state-access "hotspots" (areas of high transactional demand), the system can implement targeted optimizations like differentiated fee pricing, rate limiting, or allocating additional resources to high-contention states. It's important to note that Solana's current implementation of parallelization does not fully realize the potential of localized fee markets.

To ensure data consistency in concurrent access, the pessimistic execution model utilizes a locking mechanism. Before a transaction can access a specific state variable, it must acquire a lock on that variable. The lock provides the transaction with exclusive access to the variable, preventing other transactions from modifying it simultaneously. The lock is released once the transaction is executed, allowing other transactions to access the variable.

In Solana's Sealevel runtime, which implements this pessimistic execution model, transactions specify the accounts they will read or write during execution. Sealevel analyzes the access patterns and constructs parallel transaction queues for each CPU thread on a validator node. If an account is accessed multiple times, it is listed sequentially in a single queue to prevent conflicts. Transactions not processed within the leader node's block time are bundled and forwarded to the next scheduled leader for processing.

Unspent transaction output (UTXO)-based systems improve computational efficiency similarly. UTXOs involve specific units of currency—UTXOs—associated with an individual’s wallet. For each of said wallet’s transactions, UTXOs are expended and replaced with new ones; one or more UTXOs are created for the receiver, representing the payment, and another is typically created for the initiator, representing any change due back.

By defining which contracts will be touched, transactions that touch disjoint sets of contracts can be executed in parallel by executing nodes (which can be accomplished in the “accounts”: data model with strict access lists). However, to gain compatibility with Ethereum-style smart contracts, UTXO schemes such as Fuel's constrain block-producing nodes to execute transactions with overlapping access lists sequentially.

Nevertheless, the pessimistic execution model has limitations. Transactions must accurately declare their access patterns upfront, which can be challenging for complex or dynamic transactions where the access patterns may depend on input data or conditional logic. Inaccurate or incomplete access pattern declarations can cause suboptimal performance and potential runtime errors. Additionally, the locking mechanism can introduce latency and reduce concurrency when many transactions compete for the same state variables. This contention can form performance bottlenecks, as transactions may spend a significant portion of their execution time waiting to acquire locks on high-demand state variables.

More importantly, this model places a considerable burden on developers, who must have a deep understanding of their contracts' data dependencies to specify necessary state accesses beforehand accurately. This complexity can introduce challenges, especially in designing applications with dynamic and complex state interactions, such as decentralized exchanges or automated market makers.

Optimistic Execution

In contrast, the optimistic execution model adopts a “speculative” approach to transaction execution, allowing transactions to execute in parallel without needing upfront state access declarations.

Instead of preventing conflicts before they happen, transactions are optimistically executed in parallel, assuming they are independent. The runtime employs techniques like multi-version concurrency control (MVCC) and software transactional memory (STM) to track read and write sets during execution. After execution, the runtime detects any conflicts or dependencies. It takes corrective measures, such as aborting and re-executing conflicting transactions, but can do so by reading from memory instead of disk to identify conflicting transactions.

The optimistic execution model simplifies the development process, allowing programmers to focus on writing contract logic without worrying about declaring state access patterns. Because transactions do not need to declare their state interactions upfront, developers are afforded more freedom in designing their smart contracts, allowing for more complex and dynamic interactions with the blockchain's state. The optimistic execution model is particularly well suited for platforms that support a high volume of transactions and complex dapps, as it can offer higher throughput and scalability than the pessimistic model.

One notable implementation of this model is found in Aptos and the MoveVM of Movement Labs*, which employs a technique known as Block-STM. In Block-STM, transactions are first executed in parallel; then, conflicting transactions are identified and scheduled for re-execution based on the detected dependencies. This approach ensures processing resources are continuously utilized, improving throughput while maintaining the integrity of the transactional workflow.

Despite its advantages, the optimistic execution model also comes with challenges. The need for runtime conflict detection and the possibility of transaction aborts and retries introduce computational overhead and complexity. In addition, maintaining multiple versions of the state and managing the overhead associated with conflict resolution requires sophisticated system design and robust concurrency control mechanisms to ensure the blockchain's integrity and performance.

Block-STM leverages MVCC to effectively manage concurrent writes and maintain multiple versions of data, thereby preventing conflicts between simultaneous write operations. It incorporates a collaborative scheduler to coordinate the execution and validation tasks across multiple threads, ensuring that transactions are committed in the order they were started. This setup minimizes transaction aborts by using dynamic dependency estimation, which allows transactions with dependencies to efficiently wait and resolve these dependencies before proceeding.

Additionally, the account model used by MoveVM differs from that of Ethereum’s EVM, which leads to fewer collisions. In Ethereum, a token is typically managed by a single smart contract, potentially causing multiple token transactions to interact through the same contract address, increasing the likelihood of conflicts. In contrast, MoveVM assigns tokens to individual user accounts, reducing the chance of such conflicts as each transaction usually interacts with different account addresses.

In Monad, the initial set of transactions executed in parallel can be framed as an I/O phase, which may produce immediately-committable results, and the following “retry” phase, which requires a small amount of work to clear conflicting remaining transactions. These conflicting transitions are surfaced and pulled into cache, allowing for execution overhead to be reduced because they live in memory. While most state lives on disk, conflicting transactions are accessed quickly at execution time.

The pessimistic and optimistic execution models offer distinct approaches to handling transaction execution and state management in blockchains. The choice between these models involves tradeoffs between upfront complexity in state access specification and the computational overhead associated with dynamic conflict resolution.

Data & Task Parallelism

Data and task parallelism focus on optimizing performance by distributing computational loads across multiple processors: data parallelism segments a dataset for simultaneous processing, while task parallelism assigns different tasks to various processors to operate concurrently.

These optimizations are distinct but interdependent with state access parallelism, which manages and synchronizes access to shared resources like memory or databases to prevent conflicts and ensure data integrity when multiple processes or threads operate simultaneously.

Data Parallelism

Data parallelism involves parallelizing specific operations across multiple data elements simultaneously. This approach is particularly beneficial when the same operation needs to be applied to a large dataset or when performing computationally intensive operations on multiple input values. The key unlock comes from distributing the data across multiple processing units and executing the same operation concurrently on different data elements.

One common technique for data parallelism is single-instruction, multiple data (SIMD), which allows a single instruction to be executed simultaneously on multiple data elements. Modern CPUs often have built-in SIMD capabilities, enabling them to perform parallel operations on multiple data points. By leveraging SIMD instructions, developers can achieve significant speedups for certain types of operations, such as mathematical computations, data transformations, or signal processing.

For example, consider a scenario where you must apply a complex mathematical function to a large array of numbers. Instead of processing each number sequentially, SIMD can operate on multiple numbers simultaneously. This simultaneous processing is achieved by loading a subset of the numbers into the CPU's SIMD registers, executing the mathematical function on all the loaded numbers in parallel, and then storing the results back into memory. By processing multiple numbers at once, SIMD can greatly reduce the overall execution time.

Firedancer’s work on ED25519 signature verification demonstrates the power of SIMD for optimizing complex computations. The signature verification process involves arithmetic operations within Galois Fields, which can be computationally intensive. By leveraging SIMD instructions, Firedancer can perform these operations on multiple data elements concurrently, resulting in significant performance improvements. These optimizations will be critical in improving the performance of Solana, which has already implemented parallelization of state access.

Task Parallelism

Task parallelism involves parallelizing different tasks or operations within a program across multiple processing units. This approach is useful when a program consists of multiple independent tasks that can be performed concurrently. By assigning each task to a separate processing unit, such as a CPU core or a GPU, the overall execution time can be reduced.

Task parallelism is commonly used in scenarios where a program needs to perform multiple complex operations simultaneously. For instance, consider a video processing application that needs to apply different filters and effects to a video stream in real-time. Instead of using every compute unit to collectively apply each filter sequentially, task parallelism can distribute the workload across the multiple processing units. One processing unit can be responsible for applying a blur filter while another unit applies a color correction filter, and so on. By executing these tasks in parallel, the application can achieve faster processing and maintain a smooth user experience.

Lagrange's ZK MapReduce (ZKMR) leverages data and task parallelism to efficiently parallelize and generate proofs of distributed computations on large datasets. In the map phase, the input dataset is partitioned into smaller chunks, and each chunk is processed independently by a separate mapper worker or machine in parallel (task parallelism). The "map" operation can be parallelized within each mapper task across multiple cores or processors (data parallelism). Similarly, in the reduce phase, the "reduce" operation on the values associated with each key can be parallelized within each reducer task (data parallelism). In contrast, the reducer tasks are executed parallel across multiple workers (task parallelism).

By combining data parallelism and task parallelism, ZKMR can achieve efficient scaling and performance for complex computations on massive datasets while maintaining zero-knowledge guarantees through recursive proof composition.

Lagrange’s ability to generate storage proofs for SQL computations over 888,888 storage slots in 1 minute and 20 seconds demonstrates the power of ZKMR, as well as the task and data parallelism that underpin it. Moreover, Lagrange’s recent Reckle Trees paper underscores the need for parallelism in that it ensures that batch proofs of onchain data are also computable in 𝑂(log𝑛), independent of the batch size.

Parallelizing Consensus & Execution

While this piece doesn't address consensus, blockchains can also parallelize the process of consensus and execution. Traditional blockchains often process transactions sequentially, reaching a consensus on a block's transactions (block N) before executing them. Parallel processing of consensus and execution phases significantly boosts execution efficiency and is a technique exemplified by systems like Monad. As the network reaches consensus for block N, it concurrently executes transactions for the previous block (N-1).

This strategy ensures continuous, efficient use of computational resources, effectively reducing idle times and enhancing the network's ability to process transactions quickly. These improvements increase the system's throughput and the cost of capital required to spam the network.

Interpreters & Reducing Overhead

When smart contracts are written in languages like Solidity, they are first compiled into the lower-level bytecode. Then, the EVM uses an interpreter to execute this bytecode. The interpreter reads and executes each instruction sequentially, akin to translating a foreign language in real-time as it is being spoken. Paradigm’s latest piece on Reth points out that this leads to overhead since each instruction must be processed individually and converted from bytecode to machine instructions during runtime.

Reth is addressing EVM inefficiencies by incorporating a just-in-time (JIT) compiler. This compiler translates bytecode into native machine code shortly before execution, circumventing the resource-intensive interpretation process typically required during runtime.

The Reth article mentions that 50% of EVM execution time under an interpreter-based system is dedicated to processes that JIT could theoretically optimize, suggesting the possibility of doubling execution speed with JIT implementation. However, as Yilong points out in this presentation, while JIT can significantly decrease the time needed for processing specific opcodes, it may not drastically impact overall execution. This is because a substantial portion of the 50% of EVM execution times that JIT takes up involves "host" and "system" operations (Slide 13), which are not amenable to JIT optimizations like "arithmetic" or "control" due to their non-computational nature.

While interpreters may limit performance, they do create the opportunity for “translation,” which increases the scope of code that can leverage new virtual machines, lowering the overhead for developers to use designer blockspace. For example, Movement Labs has developed Fractal, enabling developers to deploy their Solidity-based contracts on MoveVM. Fractal works by compiling Solidity into an Intermediate Language containing instructions articulated in EVM opcodes, which are then mapped to their MoveVM bytecode counterparts, allowing Solidity contracts to run in the MoveVM environment.

Specialized & Customized State Machines

Customizing the execution layer involves designing specialized state machines optimized for specific applications. Not only does this mean that an execution environment can forgo the need for a virtual machine entirely, but it also enables applications to tailor the instruction set architecture (ISA), data structures, and execution model to their specific needs. The key performance benefit of tailoring an ISA to a specific application comes from reducing the overhead of translating the application's computational patterns into the general-purpose instructions of a traditional ISA. General-purpose CPUs use basic instruction sets (i.e., add, load, branch) to support running different types of software. However, when applications frequently repeat the same multistep operations, implementing these patterns using sequences of simple instructions becomes inefficient.

For example, database applications may need to constantly traverse tree data structures, look up entries, update values, and rebalance trees. On a normal CPU, mapping these higher-level operations requires breaking them into long sequences of low-level micro-ops, such as loads, stores, branches, and arithmetic, executing individually on the general hardware. In contrast, an ISA customized for databases can fuse these recurring patterns into optimized wider instructions that leverage specialized hardware. A "TraverseTree" instruction could calculate memory addresses, load relevant nodes, and compare keys using parallel comparison circuits designed for that operation. "UpdateEntry" could directly gather the entry from the optimized database storage layout, modify it, and commit the new state all in a single instruction.

This eliminates redundant overhead from translating high-level operations down to simple instructions. It also allows the hardware to optimally execute the application using fewer but wider, explicitly parallel instructions tailored precisely to its needs.

LayerN’s Nord demonstrates the performance benefits of specializing execution environments and data structures through their specific use case of a verifiable order book. LayerN's approach focuses on optimizing the placement of trades into the order book data structure, while their pipelining mechanism is designed to efficiently insert new orders into the appropriate position within the order book's data tree. By tailoring the data structure and insertion algorithm to the specific requirements of an order book, LayerN achieves low-latency order placement and high throughput.

Alternatively, it's possible to lean into general-purpose execution environments that enable arbitrarily programmable modules that apps can plug into to optimize their performance. This approach prioritizes the developer experience over raw performance.

Fluent and CWD utilize a strategy that balances the tradeoffs between optimizing for raw computational performance and enhancing the developer experience and compatibility of the ecosystem. This approach centers on using WebAssembly (Wasm) as the VM to execute code. Wasm has become a preferred choice in web development due to its broad language support and the wide degree to which it has been adopted.

A developer's decision to use Wasm rather than native client execution reflects a strategic preference for the versatility and broad accessibility of a general-purpose execution environment. Although native execution, which runs code directly on hardware without a virtual machine, can offer better performance, it restricts cross-platform compatibility and is less accessible to developers. In contrast, Wasm ensures a uniform and secure execution environment across different platforms despite not achieving the same raw speed as native execution. This tradeoff aligns with Fluent and CWD's design philosophies, prioritizing developer productivity and broader ecosystem integration over maximum performance efficiency.

CosmWasm deployment (CWD), in particular, exemplifies this approach by not just employing Wasm for smart contract execution but also incorporating it into a more extensive framework designed to support the intricacies of blockchain operations. Enriched with "periphery logic," this framework offers advanced account management, a customizable gas mechanism, and optimized transaction ordering. These features contribute to a flexible, efficient, secure development environment that empowers developers to build scalable and complex dapps relatively easily.

Stackr* takes a different approach by combining the benefits of customized execution environments with the flexibility of traditional smart contract platforms. Stackr allows developers to code applications as rollups, enabling them to define their own rules for transaction ordering, execution, and configuration. In the Stackr model, developers can choose the ISA, data structures, and execution model that best suit their application's requirements.

With Stackr, developers can apply state transition rules directly in the application's runtime rather than being constrained by the rules of a general-purpose VM, giving them the ability to streamline their instruction set to be more efficient and redefine the set of things that can be done in a runtime environment.

This results in more lightweight and efficient execution, as the business logic is implemented at the client level, eliminating the need for costly smart contract invocations and validation. As a result, the possibilities around how an application is configured expand in terms of the different types of languages, data structures, and signatures developers can use for a single app without sacrificing performance.

Conclusion

There are multiple paths to optimal execution layer performance.

No singular optimization to state access or parallelization stands out as a proprietary point of technical differentiation between execution layers when attempting to capture dapps. As we went through, the benefits of resource-based parallelization on Solana can be equally applied to Fuel’s UTXO model. Anyone can use Amazon’s insightful solutions to improve horizontal scalability through sharding and improve the execution layer performance.

While execution layer performance is a critical vector for winning over builders of decentralized applications, new L1s and L2s centered around improving execution must compete on other variables, including security, interoperability, and compatibility with existing tooling. For this reason, the proliferation of new interoperability layers—from Nebra to Statenet to Polygon’s AggLayer—will be critical to developers buying designer blockspace, as they can build or buy specialized blockspace without sacrificing the synchronous composability and shared liquidity of traditional, general-purpose L1s.

Improvements to state management and computational efficiency are interdependent.

Across the communities designing new execution layers, the parallelization of state access has become a defining meme for the performance improvements they promise to bring. While this is for a good reason, as it could lead to a 5x improvement in the execution of the EVM, evidence from Monad’s early experimentation with parallelization demonstrates that its role is overemphasized if other improvements, such as async I/O, aren’t developed in tandem.

Based on this, we can conclude that computational efficiency is often only achieved when we improve how state is accessed and stored. Efficient state management reduces the time and resources needed to access and manipulate data, which speeds up processing and reduces computational load.

Taking this a step further, incumbents may be making path-dependent choices that hinder their ability to compete with new blockchain designs that re-architect how state is managed and updated, given the inertia that a hard fork entails. As a result, specialized, modular execution layers and alternative L1s may be able to create defensibility around design choices for more efficient state storage and the protocols for reading from and writing to it. These design decisions offer a competitive advantage, as incumbents may encounter inertia in updating their database structures without a hard fork.

At the end of the day, a blockspace’s values impact the design space for execution layers.

In understanding how we can improve execution layers, we can now delineate that the classes of optimizations differ according to two critical design choices—who is executing transactions, and how many nodes need to be involved? The techniques available to developers for solving execution bottlenecks differ significantly depending on a team’s initial answers to these questions.

On one hand, monolithic L1s like Solana and Monad don't accept separating the validator role into heterogeneous powerful and weak nodes to accelerate performance. “Accepting” state bloat in the short-term isn't a viable solution, so they lean on improvements at the database layer and other components of the block production engine, such as consensus, to make up for the broader number of executing nodes deemed as a critical component and core value of the network. Because the security models of these L1s rely on the consensus of a more distributed set of validators with weaker hardware requirements, their data needs to be written to a database that lives on a disk, which is necessarily cheaper for a permissionless and maximally decentralized blockchain.

On the other hand, projects like Ethereum and its L2s are pursuing a roadmap that leans into centralization across their executing nodes through centralized block builders held accountable to weaker verifying proposer nodes through fraud or validity proofs.

Suppose centralized "executors" of transactions and state transitions are considered acceptable in pursuing a decentralized future. In that case, the law of physics states that systems that can 1) add blocks to a chain without requiring multiple actors to re-execute transactions, 2) increase validator requirements to maximize in-memory computation (and ignore the state bloat problem), and 3) reduce latency and consensus bottlenecks clearly win out compared to systems relying on extensive decentralization and consensus among nodes.

In seeking a balance between scalability and trust minimization, it's becoming apparent that the objective for execution layers should not be to optimize for decentralization blindly, nor must execution always be completely permissionless.

As we develop and implement a broader array of cryptographic tools, such as validity and fraud proofs, we effectively reduce the number of nodes necessary to resist censorship and maintain safety and liveness. This approach, however, involves tradeoffs, potentially impacting censorship resistance, ordering integrity, and liveness guarantees due to the possible centralization of executors.

As noted by Sreeram, the “minimum viable decentralization” does not mean that “validation should be permissionless” but that it should “just be rightly incentivized.” This implies that a well-monitored system, where validators face significant repercussions for misconduct, can maintain safety and liveness without the need for excessive decentralization (h/t Sreeram).

Such governance models are already being tested in practical applications. For instance, rollups like Arbitrum are exploring governance or committee-based systems to enforce transaction ordering and leader selection rules, and they are considering mechanisms where sequencers use onchain data to uphold transaction ordering policies.

Despite these advancements, there is no definitive "pareto optimal frontier" for balancing decentralization with performance.

Ideological and technical considerations continue to favor decentralizing executing nodes to validate the state. While centralizing nodes reduces consensus overhead and upgrading hardware can significantly enhance performance, it remains to be seen whether these optimizations will attract developers focused on creating censorship-resistant applications and to what extent censorship resistance remains a core value in the industry.

*denotes an Archetype portfolio company

Special thanks to all the people in the arena for the thoughtful conversations and feedback that went into putting this piece together: Ash Egan, Dmitriy Berenzon, Dima Romanov, Ismael Hishon-Rezaizadeh, Kautuk Kundan, Katie Chiou, Keone Hon, Larry0x, Movement Research Labs, Rushi Manche, Soumya Basu, Tyler Gehringer, Will Kantaros, and Yilong Li.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

This post is divided into two sections. First, I describe my belief that chain abstraction infrastructure is vital for consumer adoption of crypto and that intent-based architecture is the best way to engineer it. Second, I describe the main obstacle toward broader adoption of intents: solver network liveness.

I end the piece with a solution and introduce the standard that Across and Uniswap have collaborated on, using feedback derived from the CAKE working group. This standard is designed to optimize for the solver UX, lower barriers to entry to a universal solver network that most intents can be routed to, and ultimately enable a larger, more competitive solver network to flourish.

Agenda

The Problem

0. Defining the end state: what makes crypto applications "usable"

1. Why "chain abstraction" is a solution to a UX problem that arises out of the fundamental topology of modular blockchains

2. Why usable crypto applications must be built on top of chain abstraction infrastructure

The Solution Space

3. How intent-based architecture will give rise to chain abstraction

4. Understanding that intent marketplaces perform best when the solver network is large and competitive

5. Bootstrapping the intent solver network requires onboarding more applications that will produce intents

The Proposal

6. Why we need a cross-chain intents standard that prioritizes "solver UX" to grow the solver and intent marketplace to large enough scale to achieve network effects

Usable Crypto Applications Cannot Be Built Without Chain Abstraction

Are our best and brightest building redundant infra?

Many have lamented that the best crypto engineers and most based thinkers are overallocating attention and energy toward offering more blockspace to end users. This criticism has merit; there are too many L2s available for end users relative to the demand for them.

However, I reject the notion that there aren’t any useful crypto applications in existence.

Decentralized finance offers individuals the ability to self-custody digital assets, allowing them to work around draconian service providers and use their digital assets to buy things valued in the real world. The promise of self-custodied data also offers a utopian alternative for individuals that are becoming increasingly wary of trusting FAANG monopolies to keep their data secure.

The real problem in my opinion is not lack of useful crypto applications but friction for end users trying to access them. End users should be able to experience the following when interacting with crypto applications:

• Speed: Applications should feel as fast as web2 applications.

• Cost: Unlike web2, all web3 interactions must incur some cost, but the "cost per click" should be negligible.

• Censorship resistance ("permissionless-ness"): Anyone with a wallet should be able to interact with the application if they can afford to click.

• Security: Clicks should do what users expect them to do and not revert—all web3 updates should be permanent.

These are the properties of "usable" crypto applications.

We've been trying to build usable crypto for a long time

Today’s modular blockchain solutions offer consumers all of these properties, but they are not all available in the same place.

In 2020, blockchains were monolithic, offering two of three properties to end users: speed, cost, or security. We then envisioned a rollup-centric or modular future that would unlock all three properties simultaneously.

Today, we have built the groundwork for this rollup-centric infrastructure. L2s offer cheap and fast blockspace, and most of them offer permissionless blockspace. Conversely, L1 offers WW3-resistant, secure blockspace. (You can read more about the security-UX tradeoff offered by L1s and L2s in my short survey article). These L2s connect securely to L1 via canonical message paths, laying the foundation for a modular yet interoperable network. In the past four years, we've built the optical fiber between blockchains that will one day support useful crypto applications. But why are modular blockchains so unusable?

The inevitability of modular blockchain networks is that capital assets will accrue at the most secure layers while user clicks will accrue at the faster and cheaper layers.

Modular blockchain topology encourages secure blockspace to be offered on a different layer than cheap and fast blockspace. Users will naturally prefer to store their value on the most secure networks, but they will demand to interact most frequently with the cheap and fast ones. By design, the canonical paths between L2s and L1 are slow and/or expensive. These phenomena explain why users must traverse these canonical pathways to pay for L2 interactions using L1 assets. This results in "unusable" crypto UX.

The goal of chain abstraction is to reduce the friction of sending value across these in-protocol paths away from the user. Chain abstractors assume that users prefer to specify their desired end state to dapps as "intents" and it is the responsibility of the dapp to fulfill their intents. Users should not have to compromise secure custody of their assets in order to access low fees and low latency.

Therefore, chain abstraction depends critically on users being able to transfer value across networks securely, cheaply, and quickly. A common user flow today is that a user with a USDC balance on a "secure" chain like Ethereum wants to mint an NFT or swap for new tokens on a newer chain like Blast or Base. The way to do this in as few steps as possible is to sequentially execute a Bridge→Swap→Mint series of transactions (or Swap→Bridge→Mint).

In this example, the user's intent is to use their USDC on the secure chain to mint an NFT on another chain. The user will be satisfied as long as they receive the NFT and their USDC balance is charged wherever they choose to custody.

Intent-Based Architecture is the Only Way to Build Chain Abstraction

Chain abstraction relies on cross-chain value transfer, but sending value via canonical messaging paths is either expensive or slow. "Fast bridges" offer cheap and fast alternatives for users to send value across networks, but they introduce new trust assumptions. Message passing is the most intuitive way to build a fast bridge because it is modeled off of the TCP/IP architecture; it relies on a bridge protocol acting as the TCP Router to connect two chains.

Value transfer via message passing involves the bridge protocol sending messages between its contracts on the origin and destination chains. This message is triggered on the origin side by a user transaction and relayed to the destination side once the message's "validity" is verified.

A message can only be verified after the origin chain transaction initiating the message has finalized, meaning the transaction is permanently included in the origin chain’s canonical blockchain. This verification can be completed as a validity proof proving consensus inclusion of the transaction on the origin chain, as an optimistic proposal, or after a threshold of witness signatures attesting to its inclusion have accumulated on the origin side. Once the message is relayed to the bridge contract on the destination chain, tokens are released to the user.

There are several fundamental issues with this architecture:

• The verification mechanism must wait for full finality before sending the message to the destination chain protocol contract. This can take up to seven days for L2s with optimistic finality periods.

• One cross-chain message is sent per bridge transaction OR messages are batched together, but the batch can only be sent after the last message in the batch is finalized.

• The bridge has limited ability to source external liquidity to give user price improvements because it must be declarative about the fulfillment path of the user intent.

Message-passing fast bridges are going to be either unsecure, slow, or expensive depending on the verification mechanism. Intent marketplaces are an alternative architecture for fast bridging that arise from a key insight:

Value is fungible and it shouldn't matter to the recipient how the value is transferred as long as they receive the funds

Can a bridge outsource the value transfer to a sophisticated agent to gain speed and lower cost? Liquidity is dynamic on and offchain and price improvement can be realized if the bridge mechanism has flexibility to choose an optimal execution path at the time of the bridge transfer.

Intent mechanisms allow users to specify precise conditions or covenants under which their value transfer transaction can be executed.

A minimal viable intent is an order to pay X token from chain A to receive Y token on chain B.

The bridge protocol does not need to send a message between domains per bridge transaction to fulfill a user's cross-domain intent. Instead, the protocol outsources the value transfer to an agent drawn from a permissionless solver network, and the individual solver will seek repayment later from the bridge protocol. In comparison, message-passing mechanisms specify exactly how their transactions should be executed and do not need to rely on the availability of an agent.

Intent settlement protocols

Intent-based bridge protocols can be labeled more precisely as intent settlement protocols that are responsible for ensuring that solvers do not violate the user-specified conditions. Intent settlement protocols offer solvers security that they will be repaid and rewarded for fulfilling user intents. To do so, intent settlement protocols need to appeal to an oracle to verify the authenticity of the intent fulfillment. The oracle's security can be grounded in an optimistic challenge period, a witness threshold, or be ZK validity proof based, for example.

Intent settlement protocols offer fast and cheap value transfer because individual solvers can take on finality risk and identify optimal execution paths

Message-passing bridges can only communicate as quickly as finality is reached by the originating chain. Finality times are seven days on optimistic rollups and one hour on ZK rollups today. Even though these finality times should trend down following broader adoption of ZK light client technology and advancements in shared-sequencing pre-confirmation technology, it is unlikely that finality times for all blockchains will ever feel "instant" for users, suggesting a persistent need for fast bridging solutions. It is impossible to relay a message faster than the finality period without assuming finality risk—which is outside the scope of a message-passing bridge—unless the bridge wants to add an additional trusted agent to the relay path who will backstop losses due to chain reorganizations.

The speedup offered by intent-based architecture arises because individual solvers within a heterogenous solver network can assume more finality risk than a message-passing protocol can and fill a user's intent before chain-reorg risk completely disappears. Solvers will subsequently charge users for this finality risk they assume in exchange for faster fill times.

Outsourcing cross-chain intent fulfillment to an agent also leads to price improvement on average for users. In intent-based bridges, solvers who front user orders on the desired destination chain are repaid later by the system after their fulfillment is validated. These intent settlements can be batched together to amortize cost. Fillers, unlike users, do not demand instant repayment and will charge users accordingly for fronting them capital. Batch settlement is not unique to intent-based architecture, but the architecture is more synergistic with batch settlement because it separates the repayment step from the intent fulfillment step.

The larger source of price improvement comes from the intuition that value is fungible, and finding the best path just-in-time will usually outperform value transfer. (However, some paths will be impossible to beat on cost just-in-time, like when bridging USDC over CCTP.)

Message-passing bridges must encode how they will transfer value to the user. Some elect to send tokens out of a liquidity pool at a predetermined exchange rate, while others mint representative tokens to recipients who need to then swap for the desired canonical token asset.

When fulfilling a user's intent, an agent can source liquidity from a combination of onchain and offchain liquidity venues. Competitive solver networks offer users unlimited sources of liquidity in theory (but even these sources of liquidity can be depleted quickly when volume trends in one direction during high volatility onchain events like popular NFT mints, airdrops, and rug pulls).

Submitting a cross-chain order as an intent allows solvers to internalize the order's generated MEV as price-improvement.

Intent-based architecture is fundamentally designed to be secure

Intent-based bridges can be built securely because they separate the urgent demands of the user from the complex demands of the settlement network. Solvers can wait for repayment, unlike users, and they will charge users for the amount of time that the settlement protocol makes them wait for repayment. Therefore, intent settlements can be validated using very robust mechanisms without a strict time constraint. This is preferable from a security standpoint because verifying an intent fulfillment is intuitively complex.

As an example of intent verification in production, Across validates and repays fillers in batches following a 90-minute optimistic challenge period. Of course, settlement networks should strive to repay fillers as quickly as possible to reduce end user fees. An improvement on the optimistic challenge mechanism would be a ZK validity proof mechanism, which would require encoding the intent validation logic into a ZK circuit. In my opinion, it’s an inevitability that validity proving mechanisms will replace optimistic challenge mechanisms and enable intent settlement networks to repay users faster.

So, how does chain abstraction arise out of intent-based architecture?

Recall that chain abstraction requires fast and cheap cross-chain value transfer. It also shouldn't require the user to submit an onchain transaction on the network where their assets are stored.

A user's intent does not need to be submitted onchain by the user if it includes a Permit2 or EIP3074 signature. This is true for both message-passing and intent-based bridges. Both architectures can take advantage of the Permit2 pattern to allow the user to sign offchain the amount of tokens that they are willing to pay from their origin chain wallet.

Intent marketplaces best support chain abstraction because they offer cheap and fast cross-chain value transfer. Imagine a world where a user could request a solver to give them a quote to enter into a WETH staking position on Arbitrum, using their USDC on Optimism as payment. The user could send this intent offchain to an RFQ auction where solvers could bid on it. The winning solver of the auction could then receive the user’s signed intent, containing an allowance to spend their USDC on Optimism, their desired amount of WETH to receive on Arbitrum, and the calldata needed to deposit this WETH into a staking position on Arbitrum. The solver could subsequently submit this transaction on Optimism (on behalf of the user) to initiate the cross-chain intent and pull USDC from the user’s Optimism wallet. Finally, the solver could fill the user’s intent on Arbitrum by sending them WETH and forwarding the calldata to enter the user into the onchain staking position.

Building chain abstraction infrastructure means making this user flow feel instantaneous and cheap without requiring them to submit an onchain transaction. Let's conclude this article by discussing obstacles to broader adoption of intents.

For the best case user experience to materialize out of intent-based chain abstraction, we need a competitive network of solvers

Bridging with intents depends on solver network effects to perform better than message-passing variants. This is the core tradeoff of intent versus message-passing architectures. Realistically, not all applications producing intents will need access to a perfectly competitive set of solvers, and some might prefer routing their intents to oligopolistic solver networks. However, the current state of solver networks is immature and isn't close to fulfilling the assumptions of solver network liveness that intent marketplaces depend on.

We don't want a world where each dapp is routing intents to isolated solver networks. The best case for UX is that many dapps communicate with the same solver pools, and all dapps have the freedom to change which solver pools they send their intents to.

How do we bootstrap the solver network?

We must prioritize solver UX.

Running an intent solver is complicated and requires expertise in building highly performant software as well as managing cross-chain inventory risk. Naturally, there will be limited parties interested in paying the startup cost to run this code. In the best case scenario, a solver written for one dapp, like a UniswapX solver, could be reused to solve for other intent-producing dapps like Across and CowSwap.

We really need to increase the aggregate capital efficiency of the solver network for all intent-based dapps. This will require addressing the barriers to running a solver.

For this, we'll need dapps producing intents to be visible to any solver and ensure all solvers have access to many differentiated and competitive intent settlement networks. This would give confidence to solvers that they could choose to route their intent fulfillments to a settlement network that they trust. The competition between settlement networks would also lower costs for solvers.

The value proposition of intent settlement networks is offering security to solvers as well as other features that would affect the solver's ability to fill an intent.

The solver's choice of intent settlement network will affect their ability to offer fees and execution-time guarantees to users. Some settlement networks might offer solver exclusivity periods, which would support the development of offchain auctions where solvers and users could negotiate and commit to relay fees. (These intent auctions might moreover offer economically bonded pre-confirmations, further enhancing UX. To learn more about a user flow featuring intent discovery via auctions and pre-confirmations I recommend this talk by Karthik of Sorella.)

Some settlement networks might offer intent expiry (i.e., sending value back to users after some fulfillment deadline is passed), intent backstopping (i.e., the settlement network using its own balance sheet to fulfill a user's intent if no solver does), or flexible repayment chain (i.e., allowing the solver to get repaid on their chain of choice).

In the end, settlement networks will compete ferociously to repay solvers quickly and cheaply without compromising on security. In turn, solvers will send their orderflow to the settlement networks that allow them to offer the cheapest fees to users so that they can win the dapp orderflow. Competition in settlement and solver networks depends on all parties in the intent supply chain coordinating to speak the same language, and competition will lead to the best UX for cross-chain value transfer.

It is clear that we need a standard for cross-chain intents

If solvers can assume that intents will share common elements, then they can reuse their code to solve intents originated by different dapps and subsequently lower their setup costs. If different dapps create intents that conform to the same standard, then they can all route their intents to the same solver pools. This would help onboard the next generation of dapps by giving them the ability to plug their cross-chain intents directly into an existing and mature solver pool. New dapps would not have to individually onboard solvers and, instead, would get access to cheap, fast, secure, and permissionless value transfers.

Third party tracking software would also be more easily able to track intent statuses for any new dapp if they conform to a standard.

This intent standard should allow the intent principal or the solver to specify which settlement network they wish to settle their intent on.

I envision competing settlement protocols like SUAVE, Across, Anoma, and Khalani offering differentiated features to intent solvers. Depending on which settlement network is repaying the solver, the solver can offer different price and time guarantees to the owner of the intent. The dapp and solver could agree to route a user's intent to a settlement network they trust to avoid censorship, maintain data privacy, and also be secure enough to be trusted to repay the solver.

By enshrining the choice of settlement network into the intent order itself, the solver could bake this certainty into the quote that they would show to the user. The solver and user would eliminate upfront uncertainty about bridge pricing before submitting the intent onchain, reducing costs.

In collaboration with Uniswap and based on feedback from the CAKE working group, Across and I propose the following cross-chain intent standard prioritizing solver UX:

/// @title CrossChainOrder type
/// @notice Standard order struct to be signed by swappers, disseminated to fillers, and submitted to settlement contracts
struct CrossChainOrder {
    /// @dev The contract address that the order is meant to be settled by.
    /// Fillers send this order to this contract address on the origin chain
    address settlementContract;
    /// @dev The address of the user who is initiating the swap,
    /// whose input tokens will be taken and escrowed
    address swapper;
    /// @dev Nonce to be used as replay protection for the order
    uint256 nonce;
    /// @dev The chainId of the origin chain
    uint32 originChainId;
    /// @dev The timestamp by which the order must be initiated
    uint32 initiateDeadline;
    /// @dev The timestamp by which the order must be filled on the destination chain
    uint32 fillDeadline;
    /// @dev Arbitrary implementation-specific data
    /// Can be used to define tokens, amounts, destination chains, fees, settlement parameters,
    /// or any other order-type specific information
    bytes orderData;
}

This standard is designed to make a solver’s job easier. One opinionated choice it makes is supporting Permit2/EIP3074 natively with the nonce and initiateDeadline and it gives fillers some guarantees, such as the amount they will be refunded from the settlement network and the format of the user intent that they can track. Moreover, an initiate function is defined in the standard that crucially allows the filler, the one who will bring the order onchain, to specify additional “fillerData” onchain that the user would not have known about at the time that they signed the CrossChainOrder. This allows the filler to make sure that they are rewarded by the settlement contract for submitting the user’s meta-transaction and also set repayment specific information such as repayment chain.

This standard is also designed to make it easier for dapps to track intent fulfillment status throughout its lifecycle. Any settlement contract implementing this standard should create a custom sub-type ResolvedCrossChainOrder that can be parsed from the arbitrary orderData field. This may include information such as the tokens involved in the swap, the destination chain(s), and other fulfillment constraints. A resolve function is included in the standard to enable dapps to understand how to display intent statuses to users and for solvers to know the exact intent order structure they are working with.

/// @title ResolvedCrossChainOrder type
/// @notice An implementation-generic representation of an order
/// @dev Defines all requirements for filling an order by unbundling the implementation-specific orderData.
/// @dev Intended to improve integration generalization by allowing fillers to compute the exact input and output information of any order
struct ResolvedCrossChainOrder {
    /// @dev The contract address that the order is meant to be settled by.
    address settlementContract;
    /// @dev The address of the user who is initiating the swap
    address swapper;
    /// @dev Nonce to be used as replay protection for the order
    uint256 nonce;
    /// @dev The chainId of the origin chain
    uint32 originChainId;
    /// @dev The timestamp by which the order must be initiated
    uint32 initiateDeadline;
    /// @dev The timestamp by which the order must be filled on the destination chain(s)
    uint32 fillDeadline;
    /// @dev The inputs to be taken from the swapper as part of order initiation
    Input[] swapperInputs;
    /// @dev The outputs to be given to the swapper as part of order fulfillment
    Output[] swapperOutputs;
    /// @dev The outputs to be given to the filler as part of order settlement
    Output[] fillerOutputs;
}
/// @notice Tokens sent by the swapper as inputs to the order
struct Input {
    /// @dev The address of the ERC20 token on the origin chain
    address token;
    /// @dev The amount of the token to be sent
    uint256 amount;
}
/// @notice Tokens that must be receive for a valid order fulfillment
struct Output {
    /// @dev The address of the ERC20 token on the destination chain
    /// @dev address(0) used as a sentinel for the native token
    address token;
    /// @dev The amount of the token to be sent
    uint256 amount;
    /// @dev The address to receive the output tokens
    address recipient;
    /// @dev The destination chain for this output
    uint32 chainId;
}

A compliant settlement contract implementation MUST implement the ISettlementContract interface:

/// @title ISettlementContract
/// @notice Standard interface for settlement contracts
interface ISettlementContract {
    /// @notice Initiates the settlement of a cross-chain order
    /// @dev To be called by the filler
    /// @param order The CrossChainOrder definition
    /// @param signature The swapper's signature over the order
    /// @param fillerData Any filler-defined data required by the settler
    function initiate(CrossChainOrder order, bytes signature, bytes fillerData) external;
    /// @notice Resolves a specific CrossChainOrder into a generic ResolvedCrossChainOrder
    /// @dev Intended to improve standardized integration of various order types and settlement contracts
    /// @param order The CrossChainOrder definition
    /// @param fillerData Any filler-defined data required by the settler
    /// @returns ResolvedCrossChainOrder hydrated order data including the inputs and outputs of the order
    function resolve(CrossChainOrder order, bytes fillerData) external view returns (ResolvedCrossChainOrder);
}

The design goals of this standard were to enhance solver UX, make it easier for them to support multiple settlement networks, and deterministically compute their rewards. I believe that this will allow them to give more accurate and tighter quotes to users. You can read more details about this standard, code-named ERC7683, in this X/Twitter post and the discussion surrounding it on the Ethereum Magicians forum.

Closing Thoughts

"Intents" are confusing because they are not defined, and this lack of a definition is creating real UX defects.

Everyone wants everyone else to use their standard definition of an intent, so I fully acknowledge that standards are practically impossible to establish. I don’t think that defining an intent settlement system first and trying to attract orderflow second is the right approach to establishing an industry-wide standard.

In my opinion, the more tractable approach is for dapps that already own a lot of user flow and originate many user intents will agree to conform to some minimal standard that their existing solvers will adopt. This will form a new and larger solver pool. By getting access to merged orderflow from already prominent venues, this new solver pool will earn more profits and be able quote better prices to end users. Eventually, newer dapps will also demand to route their intents to this solver pool and will support its intent standard.

To get us kickstarted, Across and Uniswap are jointly proposing a standard for all intent supply chain parties to use when handling user orders to send X tokens from chain A and receive Y tokens on chain B. The orderflow running through UniswapX (having a comparative advantage in auction design and originating intents) and Across (having a comparative advantage in settling intent fulfillments) can merge and kick off the process of nurturing a larger, more competitive solver network.

This article was inspired by many IRL and online discussions with the following folks building actively within the intents supply chain:

Ankit Chiplunkar from frontier.tech, Mark Toda, Alice Henshaw, and Emily Williams from Uniswap, Arjun Chand from Li.Fi, Karthik Srinivasan from Sorella, Kevin Wang from Khalani, tumilet from The Spartan Group, Matt Rice, Ryan Carmen and Hart Lambur from Across, Benjamin Funk and Dmitriy Berenzon from Archetype, Will Kantaros from Decent, and Francesco from Flood.

Finally, the ideas in this article are summarized in the following ETH Denver talks that Hart Lambur and I gave:

Where to learn more about intents and chain abstraction

• UniswapX

• CAKE Framework by frontier.tech

• Anoma blog on Intents

• 
  

• Intent-based Architectures and their Risks

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

IN CONVERSATION WITH is a series from Archetype where we interview artists in/at the edges of crypto across music, visual art, design, curation, and more.

André Allen Anjos, professionally known as RAC, is a Grammy award-winning musician, record producer, and DJ. RAC is most notably known for pushing the boundaries of remixing, expanding the historically dance-forward genre to include indie and rock. RAC has released 200+ remixes over the course of his career, which you can listen to on SoundCloud. Artists RAC has collaborated with include The Shins, the Yeah Yeah Yeahs, Tegan and Sara, Phoenix, Kings of Leon, Lana Del Rey, and many more. RAC has also released three of his own solo albums, STRANGERS(2014), EGO (2017), and BOY (2020). Anjos has also been incredibly active in the crypto ecosystem, releasing an onchain, tokenized cassette tape known as the $TAPE token in 2020, as well as the $RAC community token launched the same year. He is currently building a new company at the intersection of music and crypto called Oscillator.

Over a video call, I asked Anjos a series of questions ranging from his experience as a major label musician, remixing and technology, personal brand, the opportunity to improve the music industry by leveraging crypto, and more.

The following interview has been edited and condensed for length and clarity.

Katie Chiou: When you think about the music business, the first thing that comes to mind is streaming because that’s where people interface the most with artists. Can you tell us a bit about the dynamics of music streaming?

André Anjos: Let’s break it down from recorded music, which includes streaming. You can cut it a few different ways depending on what type of recorded music it is. There’s physical sales and the costs of creating those artifacts—vinyls, CDs, cassette tapes, whatever. There’s a real cost to producing and distributing those artifacts, and we’ve taken that model and applied it to something digital which virtually has no distribution cost.

Spotify dominates digital distribution. For every play of a song, you receive a fraction of a cent, which is tiny. To give some concrete numbers, about a million plays is worth probably about $4,000. That’s assuming you don’t have a label, a manager, or a distributor that’s taking a cut. Now, let’s say you’re signed to a major label. They’ll typically take about 80% and give 20% to artists. Without getting super in the weeds, the artist is usually the last one to get paid after everyone else gets paid. If you’re on an indie label, there’s typically a 50/50 split. You get less visibility, usually in exchange for a higher percentage of ownership.Then, there’s the full independent route where perhaps you own everything or you use a digital distributor like DistroKid, but you get no support so you have to promote it yourself.

KC: Streaming isn’t usually an artist’s biggest revenue stream, there’s other things like touring and partnerships. Can you break some of these down?

AA: In general, I would say the three broad main categories are recorded music, publishing, and touring, and every artist has a slightly different version of that. I separate publishing from recorded music because that’s essentially like writing. You can be a writer on something, but not own the recording, so I treat them differently.

I talked about streaming, but on the publishing side, it's kind of its own industry. It involves a lot of government-mandated royalties and is a complex, convoluted industry. They are very litigious, they love to sue people. So, writing and publishing can actually be one of the more profitable sides of music, but you have to play their game and you can’t really be independent. You have no sway.

To talk about touring for a bit, I’ve done almost every type of touring. It generally depends, but I will say that touring can be profitable once you get to a certain level, but for most people it’s actually a loss or breakeven at best. There are a ton of hidden costs that people don't realize until they're too deep in it. You can get caught in this cycle where you're constantly touring just to make ends meet, and it can really spiral. I would say, right now, the only real way to be profitable is to be a solo artist who DJs and is a somewhat well-known person and tours non-stop.

I also don’t see enough people talk about how touring is just fundamentally different after COVID. There used to be a balance to it before where there was enough ebb and flow—people take a break and write an album, come back. The issue that happened was that everybody tried to come back all at once after not touring for 2 or 3 years. With limited supply of venues, that inevitably pushed everybody's fees down and the cost of touring is suddenly 3x more expensive. You basically have a supply shock.

KC: Artists today are increasingly personal brand-driven and are expected to form more social relationships/communities with their fans. Frankly, it feels unfair to me. Maybe some people just aren’t built to do that. Curious if you feel the same/if you have thoughts around that?

AA: I think that’s definitely become more true over time. I had this realization years ago that I'm effectively competing with Netflix for attention. Everybody's competing for the same thing. And with attention comes money, so it's like we're all chasing the same thing. So naturally, it doesn't surprise me that artists suddenly have to play this other game that maybe they didn't sign up for or that maybe they were promised some version of being an artist that now looks very different.

I think change is inevitable. You kind of have to adapt. There are aspects of that I’m not personally comfortable with, but you have to make those choices for yourself and do it in your own way. If your choice is to not play that game, then maybe that's cool too. But I definitely still empathize with newer artists today. If you asked me what an artist just starting out right now should do to succeed, I have zero clue. I think everybody's kind of in the same boat. You try different things, see what works and what doesn’t work, and trends emerge from there.

I remember when I first started doing remixes, I wasn't really touring for the first year or two. I was perfectly happy doing studio work. I love being in the studio; that's what I like doing. Then suddenly, I started getting offers to play shows. It wasn't really my thing, but I thought, "I guess I can learn." It's funny because I learned how to DJ while in the club, even though I never used to go to clubs.

KC: Yeah, I read in an early interview of yours that clubs aren’t your scene.

AA: Yeah, not at all. But I get booked for shows all the time, and maybe that's a good thing because I bring my own energy to it. I remember in 2009, there were all these bedroom producers suddenly being thrown into clubs and DJing. Performing is such a different environment, it’s like learning a whole other art form. Even to this day, it's not the most comfortable thing to do for me, but everybody has different comfort levels, and you have to find your own path. I empathize with people that have to do a lot more social media now than before, but it feels like it’s just the world that we live in now. I kind of just accept it.

KC: You livestreamed on Twitch a lot during COVID after you had to cancel your tour. What was that like?

AA: I was pretty aware of the culture on Twitch, so I think that gave me a little bit of a leg up when I started streaming. I knew that I had to solve a few things like filling a lot of space and being loose and improvisational. There were experimental musical things I did that sometimes worked and sometimes didn’t, but I got to have fun with it. But the music was just there to fill the dead space. The real powerful thing about Twitch is talking to people. During COVID, everybody rushed to do Twitch streams and “live” performances that were actually pre-recorded, but they just glossed over the point entirely or it’d just kind of miss the mark.

I'm really glad I did it, but I'm really glad I don't have to do it anymore because it was 4-hour sessions 3 times a week. It wasn't even that much money. Honestly, people were very generous with tipping but Twitch takes 50% unless you’re a huge streamer, and I just feel like that’s a lot.

But I enjoyed it. I think it was a way for me to interface with my fans and my audience in a more meaningful way and I still really appreciate it for that. Sometimes people will talk about that time very fondly, but I think everybody understands why I don't do it anymore.

KC: You’re an artist that’s always been at the forefront of technology. Part of your early legacy is transforming remixing. In the past few years, remixing has exploded even more because of things like TikTok and AI. As an artist, what’s your relationship to technology and these types of tools that make music creation easier?

AA: Obviously, I love technology. It'd be pretty hypocritical of me to criticize it because I feel like I owe a lot of my career to it. Recording technology, the rise of the internet—these things have enabled me to reach a wider audience than I ever could have as a local artist playing local gigs.

Back when I started, the remixing technology that existed was somewhat limited. Now, you can make a fully finished song in a day—it's crazy how things have changed. Especially with AI these days, I'm almost glad it wasn't around when I started. Picking a path like remixing and going deeper and doing different things with it than most people at the time, it gave me an edge in my niche.

Now, my moat around that was simply the fact that most people didn't have access to these files. Because, and I don't know if everybody knows this, every remix I've done has always been officially through the artist. I've never put out anything bootleg, even though that's a big part of remix culture. And I'm not saying that's a bad thing. I just mean that it was a source of income for me. I was paid by the artist to do it, which was, again, kind of a weird path to take. But now with some of these AI tools where you can extract a vocal from any track on YouTube, it’s hard to make a living from that type of work.

Technology, especially crypto, has served me well. It's become an interest of mine, a tool for expression. It doesn't feel forced—it's just something I'm genuinely into.

KC: Many of the original attempts to build at the intersection of music x crypto started with music NFTs. What are your thoughts on how music NFTs evolved and what are your thoughts on them today?

AA: Music NFTs were, I think, piggybacking off the narrative of scarcity that was happening in the fine art world at the time, and it didn't quite connect in the same way because people think of music differently.

People don't think of music as being a scarce asset. So why do I have to buy a music NFT? I don’t think anybody would deny that music is art, but people just think about it differently. They think about music as being everywhere, pervasive in the background. It's just not something that people think of as a scarce thing. So I think that narrative just didn't quite click for people, and there wasn't an analog to compare it to. With art NFTs, it's like a digital painting, there's only 10 prints and editions. But with music, there was never that same analog that resonated for people.

Aside from that, I really like the emergent behavior of uploading a song on the internet that is free of a platform and is universally compatible with everything else. That is exciting. That, I think, is a novel behavior that is so much more interesting than what we have currently. There are so many problems there to solve with that structure, but I think that is really cool. It’s just not yet compatible with the ways music works today with copyright, government-mandated royalties, performance royalties and all these other structures already in place. I still think there's merit to these ideas. We haven't quite figured them out yet, but I'm not ready to throw them away.

KC: What are your learnings and reflections from your own crypto projects like $TAPE and $RAC?

AA: They've all been unique in their own way.

$TAPE started as a concept I worked on with Jacob from Zora and Jack, my Co-founder. We basically came up with this question: If you let a market decide the price of music instead of a platform like Spotify, what does it become worth?

RAC

@RAC

$TAPE is apparently the most expensive cassette tape of all time. I'm speechless. Thank you.

145

3:32 PM • Sep 30, 2020

$TAPE is a physical cassette tape and the supply was completely arbitrary, but there's a hundred cassettes. Let's put them on a bonding curve and see how much they’re worth. People can trade $TAPE back and forth, and the trading created crazy price action. It made us realize that music is worth way more than a fraction of a cent. I don’t know the exact number, but it’s definitely higher. We learned that if you introduce some kind of artificial scarcity, you can use markets to kind of find a price. We also created a function where in order to get the actual physical cassette, you had to burn the token. Out of 100, only 31 were redeemed which is also interesting. Maybe the token is even more valuable than the actual thing.

With the $RAC token, it was basically like an airdrop to fans. It was a retroactive thing. Everybody that bought merch at a show or heard about tickets, bought an MP3 on Bandcamp in 2008, etc., got an email with some tokens. We didn't want to sell the token, it was meant to just be a community token. I thought it was a really interesting experiment. We ran into other issues such as people constantly in my DMs or in my replies asking why the price was moving.

ERC20s, for better or for worse, come with some baggage. No matter how often I try to tell people that the $RAC token is not meant to be financial, people still want to attribute a financial value to it. It was around this time that social and creator tokens were a big conversation, and the focus soon shifted to NFTs. Unlike ERC20s, NFTs don't carry the same liquidity provision baggage. I realize now that I may have been a little early with ERC20s. An NFT membership model may have been more sustainable for a smaller community. It's interesting now to observe the resurgence of memecoins and watch how narratives cycle.

KC: How are you taking those learnings to build Oscillator?

To back up a bit, when I first discovered Ethereum, I was excited about the programmatic nature, the ability to replace institutions with code, or middlemen with code. I think this idea is often overlooked in crypto discussions. While some focus on tokens and financialization, I think there's actually so much more interesting stuff you can do with blockchains and just programming business logic, if you want to think about it that way. That's the part that's more exciting to me than anything else. That’s where I'm coming from initially, and that's the future I want to see. I think this will be a better ecosystem to play in because open data keeps people honest. That's the vision I want to work towards, and it's what we're trying to do with Oscillator. In general, I think we just want to attract people that believe in that vision and that want to help. We want to recreate what a modern music industry could be onchain and that is built with the right foundation and built for artists. We want to cut out and trim a lot of fat and baggage that comes with the traditional music industry.

So if any of that resonates, reach out. We even launched our company with just a manifesto that said if you believe in these principles, come talk to us, like we want to work with you. We want to find the right people because there's no way we're doing this on our own. We want to put this vision out in the world and meme it into reality.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

The best part of ETHDenver 2024 was watching two opposing groups collide in a classic left-curve / right-curve debate.

One arrived with a risk-on, full send mentality in anticipation of a coming mega cycle.

The other came in struggling to find the fundamentals to make sense of today’s market euphoria.

In the short term, I think both are probably right. The setup for a massive, prolonged run up in crypto looks pretty good, even if the current momentum in public and private markets has almost certainly lost sight of fundamentals.

But a closer look suggests something different might be going on this time.

Where previous bull markets drove users, those cycles tended to look less like true adoption and more like a levered macro beta.

Today, those fluctuating waves of liquidity are set to give way to a permanent shift in the tides thanks to three major tailwinds:

Macro x Crypto’s Mainnet

The explosion in public and private markets—thanks to a mix of AI hype and economic optimism—masks several secular changes playing out.

In fact, rather than reverting to pre-pandemic norms, a very different geopolitical paradigm is emerging: one centered on fragmentation and cross-border competition.

While these new dynamics will materially challenge incumbents, they present a generational opportunity for crypto to make the jump to global adoption.

Crypto’s first decade was an extraordinary testnet marked by grassroots development, exuberant highs and difficult lows. Now in its second decade, crypto is ready for mainnet as the world’s interoperability layer: a neutral home for economic exchange and technological innovation in an age that desperately needs it.

Why?

For crypto, secular changes are positive tailwinds arriving at the perfect time.

Real structural shifts in the pipes that have underpinned global trade for the last two decades are being accelerated by very serious fractures in international relations. A restructuring of global supply chains and trade rails accelerated in 2023 and has become a core focus for companies and governments alike.

Growing Focus on Reshoring by Corporates:

Meanwhile, over 54% of the world’s population and close to 60% of global GDP are undergoing an election cycle where protectionism is already a major focus as several major military conflicts play out in real time.

Equally relevant to crypto is an evolution in the art of international competition. Countries no longer rely solely on rockets and bullets.

America weaponized international finance in response to Russia’s invasion of Ukraine, while OPEC and Russia have shown they are happy to respond in kind by manipulating energy supplies.

In parallel, techno-nationalism around semi-conductors and other critical inputs has seen rhetoric replaced with action in the form of sanctions and subsidies.

This fragmentation of global commerce is damaging to margins in the developed world, while for the 40% of the world living in Lower-Middle Income (LMI) countries, the impacts are even more existential.

For individuals in LMI countries in particular, crypto provides vital solutions to everyday problems, and its importance only grows alongside systemic challenges. The data confirms this story: grassroots crypto adoption is not only reinforced but accelerated when geopolitical strains rise.

Likewise, for private enterprises this confluence of factors introduces significantly greater costs and restricts access to new consumer markets, all of this in a world where the cost of capital is no longer zero.

The trade wars & tariffs of the last few years have already proven damaging to businesses and domestic economies, and those negative impacts risk compounding as companies adapt to a new reality.

As a result, enterprises and individuals increasingly need to make a choice:

This crossroads reminds me of an intriguing historical parallel:

When Constantinople fell to the Ottomans in the 1400s, its new conquerors inherited control over a geographic nexus connecting global commerce across the ‘Silk Road’.

Famously, the Ottomans would soon move to restrict the overland trade routes that had thrived for centuries. The result pushed European powers to take to the seas in search of new trade routes, thereby igniting the ‘Age of Exploration’ that shaped the modern world.

This time around, it will be blockchains where the riches and perils of the New World will be found by those brave enough to set sail.

Reaching Enterprise Scale

There’s another important point worth hitting on: crypto has historically been a staging ground for large companies looking to experiment with a new technology.

Because of the factors covered here, the traditional enterprise’s exploration of crypto is now transitioning from R&D to production grade.

• Companies will accelerate efforts to explore digital assets and onchain ecosystems as a key source of greenfield markets. What was once a vanity project will increasingly become an existential mission.

• Capital allocators will scale up crypto-native deployment and participation to insulate against ‘beta’ exposure to global risk and uncertainty. There won’t be many places to hide in the old world.

• Systemic challenges in different regions (inflation, capital controls, cold / hot conflicts) will drive even greater relevance and need for digital assets. Permissionless blockchain infrastructure will start grassroots before going global.

To be sure, economic and geopolitical challenges have always driven users to crypto, particularly in developing markets.

But the scale and scope of challenges the world must navigate in the coming years presents a unique window for crypto to become the de facto system for free commerce and culture.

Institutional Flows

Of course, all of this is contingent on the final piece of the puzzle: bringing institutional capital onchain.

The approval of spot BTC ETFs marked a major turning point on this front, and it would seem a similar outcome is on the horizon for ETH.

Already, BTC ETFs have seen more than $7.5B in net inflows, and those launched by BlackRock and Fidelity marked the largest debut months of any ETF in the last 30 years.

This incredible momentum is what will finally allow the biggest institutions to join over 52M Americans and another 500M around the world in the onchain economy.

If the tectonic shifts in macro are providing the spark, the flow of institutional capital represents the gas that will light the fire underpinning crypto’s prime-time moment.

Middleware & Infrastructure Upgrades Drive Growth

The exogenous setup is outstanding. Are we ready to seize the moment?

I believe the answer is yes.

Following the crashes of 2022 that sent tourists scattering, crypto natives worked through self-reflection of the excesses and shortcomings that enabled the bubble in the first place.

As capital and talent consolidated around what felt like a full systems upgrade, colossal progress was made across all layers of the stack, setting the stage for a major coming cycle of adoption at scale.

The flow of private funding throughout the year reflected that story. The year started with financial infrastructure commanding the largest share of funding, followed by wallets, and ended with the former dominant once again and L2/interoperability projects in second place.

Amidst all of this, what is particularly fascinating is the infrastructure-application flywheel is starting to catch fire with more concentrated purpose than ever before.

The needs of a growing class of crypto-native consumers are driving directional and focused improvements in the tech stack, which in turn is yielding new use cases and applications.

Superior UI/UX is Fueling Adoption

Q1 2023 saw the release of the ERC-4337 standard, designed to transform externally owned accounts (EOAs) into smart contract wallets to enable customizability, better private key recovery mechanisms, and a materially more streamlined user experience.

Even more impactfully, teams like Privy* made massive progress by simplifying onboarding with embedded wallets to minimize user friction while letting developers design more contextual experiences.

Privy’s efforts helped streamline Friend.Tech’s explosive early capture of 100K addresses in a matter of weeks, and they’ve gone on to power onboarding for OpenSea, Zora, Blackbird and others to the tune of 2M+ users across 150+ countries in the last 13 months.

Meanwhile, Farcaster’s* launch of Frames—a new primitive allowing people to embed interactive experiences directly within Casts—is a transformative move that is already supercharging activity on the platform.

Farcaster has over 8M+ reactions across 4M+ casts and may well represent early signs of a crypto-native consumer app hitting escape velocity.

New Design Spaces Are Getting Bigger and Better

Just as Ethereum sought to go beyond Bitcoin’s functional limitations, a new generation of projects are now targeting Ethereum’s own structural shortcomings in a wave of modularity.

Alternative L1s & side chains have been a feature of previous cycles, but none have succeeded in disrupting Ethereum’s (mainnet) dominance of users, TVL, developers and activity.

This changed with the launch of rollups like Arbitrum and Optimism—projects designed to enable better throughput and lower fees by offloading computation from Ethereum.

While these new layers have reached scale that rivals, or in some cases surpasses, Ethereum itself, builders are dreaming bigger as they seek to further optimize the L1 stack.

That’s because even though the number of daily active users on L2s has risen 8x in the last year, much of what users are actually doing looks largely similar to historical L1 activity. As a result, the emerging consensus is that offloading transactions to cheaper execution environments isn’t enough to enable truly novel onchain experiences.

We need to actually rearchitect the components underpinning blockchains, from Data Availability (DA) to state access bottlenecks and parallel execution.

Standalone data availability (DA) layers built to scale to web2 performance parity (i.e., EigenDA, Celestia, Avail) are coming to market alongside upgraded virtual machines, some based on the EVM and others using alternate engines like Move’s (Movement Labs*) or the Solana VM (Eclipse). Some of these are building L2s to optimize execution only (MegaETH), while others are launching net new L1s from the ground up (Monad).

Meanwhile, EigenLayer’s mission of providing a shared security layer via restaking is making it possible for a new generation of projects to launch in a way that minimizes the need to bootstrap native liquidity and therefore deviate from the core security model of Ethereum itself.

All of this means the underlying infrastructure, tooling, and design optionality on offer to web3 builders is approaching unprecedented levels of maturity and performance.

As crypto’s ethos increasingly resonates alongside upgrades in infrastructure, tooling, and middleware, we’re seeing a promising story in the data underpinning crypto’s paramount leading indicator: developer flows.

Building on blockchains should not just be a more meaningful exercise, but a more technically performant one that effectively empowers developers to design the future of the open internet.

Significantly better retention of existing developers and onboarding of new ones amidst difficult market conditions speaks volumes to the work that’s been done on this front.

Open-Source AI & Crypto Rails

And finally, our belief is that the intersection of crypto and AI—two paradigm shifts in their own right—represents one of the most transformative moments in modern history.

Blockchain rails have successfully undergone a multi-year pressure test in the design of a permissionless system fit for a digital age, and especially an age shaped by generative AI.

Crypto’s toolkit of solutions tackles pertinent problems ranging from resource and liquidity coordination, asset ownership, data provenance, attestations, and much more. Crucially, the maturity of the ecosystem and technology stack are arriving just in time to meet the demands of the AI revolution.

While there is ample room for blockchains to streamline existing machine learning (ML) flows, the most exciting opportunities will emerge where crypto and AI converge to enable completely novel outcomes.

The most exciting new design spaces will span areas like:

• Decentralized storage underpinning shared, permissionless repositories for data to enable better training or more performant models via Retrieval-Augmented Generation (RAG)

• Zero Knowledge Proofs for model or content verification, training or user data privacy, or enabling edge and local (client side) inference

• Novel information markets and better mechanisms for collecting higher quality data as foundation models require more specialized data inputs to continue evolving

• Autonomous agents transacting on smart contracts that uniquely enable them to accumulate resources, knowledge and assets using machine-operated private keys

As crypto rails impact everything from the supply of compute to markets for data to the collective creation and monetization of powerful foundation models, open-source AI/ML will be supercharged by crypto, fueling a renaissance in human productivity and open collaboration in the coming years.

Crypto will be the best way to get exposure to the rise of AI, as a proxy via blue chip assets like ETH or directly through ownership or speculation across agents, models, networks, and datasets.

What’s Next?

We’re at an inflection point for this industry. After years of grinding in spite of market, incumbent, and regulatory resistance, the tides are finally shifting.

Crypto’s moment has arrived thanks to the convergence of a few significant tailwinds that are finally ushering in the decentralized future. Amidst this renaissance of infrastructure and middleware, which is already enabling a step function evolution in the onchain experience, there’s an important point we should keep in mind.

In a perfect world, modularity enables not just specialization but also a dissemination of control and points of failure across multiple contributors. However, each of these new puzzle pieces involve different technical and security assumptions, incentive mechanisms, token dissemination roadmaps, VCs, foundation setups, and internal politics.

I don’t want to detract from the inspiring efforts of builders across the space last year, especially in the face of a brutal economic downturn. But as activity and excitement pick back up, it’s imperative we escape the echo chambers and false signals of token grants masquerading as logical integration partnerships, effective PR as community approval, or incentivized behavior as a proxy for organic adoption.

In the coming years, we have a shared responsibility to keep this growing number of projects accountable across technical design choices, token concentration, value dissemination, ideology, and governance.

That’s how crypto wins the endgame.

*denotes an Archetype portfolio company

Thank you to my Archetype colleagues Katie Chiou, Benjamin Funk, Aadharsh Pannirselvam, Ash Egan, Tyler Gehringer, and Dmitriy Berenzon for thoughtful review and feedback.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Our brains, books, and databases serve as both the recipients and creators of humans’ ever-increasing propensity to produce data. The latest in this long lineation—the internet—generates and stores about 2.5 quintillion bytes of data per day. While it’s easy to look at this number in awe, data points in and of themselves offer little value. They are akin to scattered pieces of vast puzzles that require careful gathering, processing, and contextual integration for them to become valuable information.

Many of today’s internet giants have centered their entire business models around doing exactly that, and few companies have done so more successfully than Google. Their process is as follows: extract massive deposits of invaluable raw materials—billions of peoples’ “digital exhaust” in the form of private data—and feed them through pipelines of proprietary algorithms to predict the choices individuals are likely to make. The more data Google extracts and processes into information about us, the better the insights they can give to advertisers, and the more those advertisers bid in Google’s ad auctions in an attempt to convert us into customers.

As a result of these processes, Google generates $240B of advertising revenue per year.

While Google intentionally removes human beings from this process, there is another way to produce and monetize valuable information that may be even more powerful—by engaging humans as players in games centered around our inherent desire to create, search for, and speculate on information. From sports betting to MEV to social deduction games like Among Us, we are already naturally drawn to "information games" that center on competition and coordination, requiring us to skillfully hide and uncover information.

Some information games are just that—games. But as we’ll see, others can be used to generate and monetize new, valuable information and serve as the backbone of a new generation of products and business models.

However, information games have always had an Achilles heel: trust. Specifically, players need to trust that other players can’t share or act on information in such a way that violates the rules of the game. If a player of Among Us can switch from a crewmate to an imposter mid-game, or a block builder can calculate bad state roots and still be accepted by validators, no one will want to play the game anymore. To solve this trust problem, we turn to trusted third parties to create and mediate information games for us.

That’s fine for games with low stakes like Among Us, but restricting game creation and mediation to a centralized party limits the trust in and experimentation around the types of information games we play, and consequently the types of information we can collect, leverage, and monetize.

Simply put, there are many information games that haven’t even been attempted because we haven't found a way to keep them fair and trustworthy in a decentralized context.

Programmable blockchains and new cryptographic primitives are fixing this by allowing us permissionlessly create and coordinate information games at scale, without having to trust third parties or each other.

In turn, crypto-powered information games can rapidly accelerate the quantity and quality of information available to the world, increasing our collective decision making-abilities and unlocking efficiency gains on the scale of global GDP. Imagine prediction markets accessible across the globe, serving as a tool to allocate capital for internet-native megafunds. Or a game that allows individuals to pool their private health data and be rewarded for any new discoveries that result from its use, all while preserving their privacy.

As this piece will show, though, crypto-centric information games may not be ready for these high-stakes use cases just yet. But by experimenting with smaller, entertaining information games today, teams can focus on attracting players and building trust before potentially scaling to create and monetize more lucrative information markets tomorrow.

From prediction markets to game-theoretic oracles and TEE networks, this piece will cover the design space for creating these crypto-powered information games, and the infrastructure critical in bringing them to their full potential.

Permissionless Markets: A Prerequisite for Information Games

From futarchy to information marketplaces, blockchains allow developers to create customizable, automated financial devices that underpin permissionless, unstoppable markets. As a result, anyone can now create mechanisms for incentivizing, coordinating, and settling the exchange of value and information. This underscores blockchains' critical role in enabling us to rapidly experiment on how we can best configure games that maximize value for everyone involved.

It's very difficult to convince centralized intermediaries to adapt at this pace or allow their users to participate in these experiments. For that reason, permissionless markets will serve as the medium through which fringe theories and cutting-edge research papers will come to life. We’ve already seen this happen in the context of prediction markets, where theoretical automated market-making strategies conceptualized to deal with prediction markets’ low-liquidity have been implemented as CPMMs on crypto-rails, and tested with real money.

Permissionless markets serve as a vital enabler of tools to better produce new information and monetize its value.

Information Games for Information Production

Many information games produce new information to be used by players to make better decisions.

These information games create incentive mechanisms to extract raw materials (public and private data) from people, databases, and other sources, and then aggregate that data through the best information-producing machines (markets and algorithms). Ideally, in aggregating this information, new information is produced and monetized by helping another player make a good decision. For example, an investment DAO using the outcome of a prediction market to determine whether or not to invest in a new startup.

The games and tools leveraged by designers of information games vary depending on the type of information they may produce, and we have a vast design space of different challenges and opportunities to explore.

But let’s start with the most actively developed and discussed information game today–prediction markets.

Game #1: Prediction Markets as Tools for Generating Information

One of the most popular information games we’ve seen in crypto (and beyond) is the prediction market. Polymarket is the world’s leading prediction market, and is leveraging crypto rails to facilitate over $400M in cumulative trading volume (and growing fast).

Prediction markets operate by incentivizing players to wager on the outcomes of various events using their own capital (or play money). This requirement of a personal financial stake, or "skin in the game," helps to guarantee that participants are genuinely committed to their predictions. As traders act on their insights, by buying shares in undervalued outcomes and selling shares in overvalued ones, the market dynamically adjusts. These adjustments in market prices reflect a more accurate collective estimation of event probabilities, effectively correcting any initial mispricings.

The more people with disparate but relevant pieces of public and private knowledge who place bets in the market, the more closely the price will reflect the truth. Ultimately, prediction markets harness the “wisdom of the crowds” by leveraging financial stakes to drive the accurate aggregation of information.

Unfortunately, prediction markets suffer from a few critical challenges, many of which come down to various scalability problems.

Truth Bottlenecks

Keynesian beauty contests—contests where judges aim to select the option they think other judges will also select—are not unique to prediction markets. However, their negative impacts here are more pronounced than in traditional markets, as the very goal of prediction markets is to create accurate information. Moreover, unlike traditional financial markets where profit maximization predominantly drives participant behavior, bettors in a prediction market are more likely to be influenced by personal convictions, political leanings, or vested interests in certain outcomes. As a result, they are more willing to incur financial losses in the markets themselves if their bets resonate with their personal values or expectations of profit derived from actions outside those markets.

In addition, the more people look at any market or algorithm as a source of truth, the higher the incentive becomes to manipulate that market. This isn’t too dissimilar to the problems social media experiences. The more people trust the information goods produced by our social media platforms, the higher the incentive becomes to manipulate them for profit or sociopolitical gain.

Some players might even leverage the signals and incentives created by prediction markets to reprice collective beliefs and encourage collective action. For instance, imagine a government using a form of “quantitative easing” to influence prediction markets on critical issues like climate change or war. By purchasing large volumes of shares in a relevant prediction market, they could shift financial incentives towards desired outcomes. Perhaps they have determined that the systemic risk of climate change is undervalued, so they buy a significant number of "No" shares in a market predicting climate improvement by 2028. This action could encourage more climate startups to develop technology that gives them an information edge in betting on "Yes" shares, thereby accelerating efforts to find solutions.

While the factors above have been shown to negatively affect the quality of information produced, it has also been shown that instances of manipulation actually increase the accuracy of the market, because market manipulators are noise traders that informed market participants can make money by trading against.

As a result, we can deduce that the problems above are a result of an insufficient amount of well-capitalized, informed traders to help correct markets. Allowing these informed traders to borrow and short could be a critical means to making these markets more efficient.

Moreover, in markets with longer timelines, it's more difficult for informed traders to counteract manipulation, as manipulators have more time to reflexively influence both market sentiment and the actual outcomes through their trades. Implementing markets with shorter, renewing resolution dates could improve people’s trust in the game (and hence the quality of its information), but also make for more attention grabbing gameplay.

We’re also seeing early signs that, in some contexts, players enjoy information games where the resolution of the markets is manipulable. Perl, the #1 account on Farcaster at the time of writing, has leaned into this model and created an in-app platform to speculate on user engagement. Prediction markets like “Will @ace or @dwr.eth (co-founders of Perl and Farcaster respectively) get MORE likes tomorrow?” are initiated, and the trolling one can expect from football teams and their fans begins. Only here, the game happens asynchronously and is measured in likes instead of touchdowns. While Perl’s game intentionally subverts the information-producing quality of prediction markets, a fun meta-game emerges from coordinating to resolve the oracle in one’s favor.

Prediction-based games can reduce manipulation and boredom by using shorter, potentially renewing rounds. However, in low-stakes games, allowing player manipulation can add to the fun and become an integral part of the gameplay.

Finding the Right Judges & Oracles

Another challenge of prediction markets can be found in adjudication—how do you resolve the market correctly? In many cases, we can rely on oracles secured by reputation and collateral that can plug into offchain data feeds. To solve this, prediction market designers can lean on game-theoretic and cryptographic oracles to plug into a wider set of topics, including players' private information.

**Game-theoretic oracles, **or schelling-point oracles, assume that in the absence of direct communication, participants (or nodes) in the network will independently converge on a single answer or outcome that they believe others will also choose. These oracles,  pioneered by the likes of Augur and later on by UMA, encourage honest reporting and deter collusion by rewarding participants according to their degree of proximity from the “consensus” answer.

Still, there are many challenges in making these oracles reliable in adjudicating bets across a small number of players, where identifying and communicating with each other to collude becomes a potential threat. While encryption is touted as a critical tool to avoid collusion between voters, it can also be wielded as a tool to enable collusion and prevent prediction markets from resolving correctly. We can see this through the potential for DarkDAOs leveraging trusted execution environments (TEEs) to engage in programmatic bribery and coordinated price manipulation. One of the teams working on balancing these incentives is Blocksense, which uses secret committee selection and encrypted votes to prevent collusion and bribery.

It’s also possible to tackle the oracle challenge by leveraging onchain data. In MetaDAO, players are rewarded if they correctly predict how a specific proposal would impact the price of its native token. This price is served by the UniswapV3 position, serving as an oracle for the token’s value.

Even then, these oracles are limited in resolving markets based on publicly available data. If we can resolve markets based on private data, we can unlock entirely new types of prediction markets.

One of the ways we can resolve markets based on private information is by using the outcomes of information games themselves as oracles. One such example is the Bayesian Market, which leans on the principles of Bayesian reasoning to derive bettors' own beliefs about their private information by getting them to bet on others’ beliefs. For example, setting up a market where people are betting on “how many people are satisfied with their lives” reveals the bettor’s own beliefs about others’ life satisfaction. As a result, we can come to accurate conclusions about a player's private information, which would otherwise be an unverifiable truth.

Another solution we can lean into is to leverage oracles that leverage clever cryptography to “import” data from private web2 APIs. Some of these existing oracles are showcased in the “Oracles for Public & Private Information” section of the market map. Using these oracles, it’s possible to create prediction markets around some players' private information, incentivizing the holder of private information to verifiably resolve specific prediction markets in return for claiming trading fees from people betting on it. More generally, the ability to securely access a richer set of people’s offchain data onchain can serve as an identity primitive that helps us better identify, incentivize, and match players across information games much more efficiently, helping us bootstrap the necessary information to make information games relevant to players.

Innovations in oracle design will increase the scope of data we can use to resolve prediction-markets, expanding the design space for information games around private information.

Liquidity Bottlenecks

Attracting liquidity to prediction markets is hard. First, these markets are binomial markets, where players bet  “Yes” or “No” on a particular topic and either receive a fixed monetary amount or nothing at all. As a result, the value of these shares can shift drastically with small changes in the underlying asset's price, especially close to their expiration. This makes predicting their short-term price movements very important, but challenging. To handle the significant risk of these sudden changes, traders must use advanced and constantly adjusting strategies to protect against unexpected market movements.

More importantly, it becomes even more difficult for prediction markets to attract liquidity as they expand the scope of their markets to more topics and increase their time frames. The higher the variety of markets beyond politics and sports, and the longer their duration, the less people feel they have a perceived edge in betting on them. As a result, less people bet, and the quality of the produced information degrades.

Prediction markets inherently face these liquidity issues because forming prices requires uncovering private information and making bets based on that information, both of which are costly activities. Participants need compensation for their efforts and the risks they take, including the cost of gathering information and locking up capital. This compensation typically comes from others willing to accept worse odds for reasons like entertainment (i.e., sports betting) or hedging risk (i.e., oil futures) which help drive significant liquidity and volume. However, prediction market topics with narrower interests have less commercial appeal to players, leading to less liquidity and volume.

Economic Improvements: Overlays & Diversification

We can work towards solving these problems by recycling ideas from traditional finance and other existing information games.

Notably, we could make use of the overlay which Hasu covers in “The problem with prediction markets.” In gambling tournaments, the concept of an overlay—additional value added to the pot by the house to encourage participation—serves a similar purpose to the subsidy proposed for prediction markets. The overlay effectively reduces the cost of entry for players, making the tournament more attractive and thereby increasing participation from both novices and seasoned players.

Just as an overlay in a gambling tournament acts as a catalyst for player engagement by enhancing the potential return on investment, a subsidy in prediction markets incentivizes participants by lowering the barriers to entry and making participation more financially appealing. The subsidy also serves as a beacon, drawing in a multitude of perspectives and insights from both uninformed and informed traders who stand to profit from correcting them. Teams operationalizing this strategy will have to systematically identify and engage with potential subsidy providers and create markets around their needs, as they are the ones willing to provide the necessary liquidity.

In a similar vein, it’s possible that a fund-like structure could be implemented to achieve time and sector diversification, and increase the liquidity in prediction markets across a broader set of questions and time horizons. For example, many companies might find value in markets centered around how particular lawsuits might resolve. These companies could lower costs for legal experts to participate by lending them capital, allowing them to diversify across a wide set of markets, and then rewarding them according to their performance over time.

In this setup, traders would be able to borrow money to make markets, the amount which could be parameterized according to demand for the information that would be produced and the trader's reputation on the subject. This could be combined with management fees that serve as an additional overlay across each of the markets.

On the side of liquidity providers, they would receive exposure to traders incentivized to bet on these markets correctly, diversified across a large basket of uncorrelated assets with different durations. While the principal-agent problem would have to be considered, this system could increase the magnitude of liquidity provided in these markets and the variety of pools across which they are allocated. As a bonus, the quality and variety of information goods could be increased while creating new information about trader’s skills and knowledge across different markets, accelerating returns for liquidity providers through reputational byproducts.

When the value of the information that players could produce is large, integrating composable financial markets like lending and liquidity mining into gameplay can serve as critical tools to lower barriers to entry.

UX Improvements: Simpler Interfaces & Flexible Incentives

The default, exchange-centric UX and limited reward types across today’s prediction markets can push out those who are motivated by other types of interfaces and incentives, further limiting liquidity. On the side of bettors, there are many interesting ways to improve the quality of prediction markets, all of which center around increasing reach and accessibility to different types of players.

First, we can improve prediction market UX by integrating them within larger social platforms. Perl and Swaye have demonstrated how, by plugging into Farcaster’s data, users are spared the cognitive load of opening up a separate app, and information game designers can identify and direct players to markets they are uniquely situated to play (i.e., top participants in the channel for /nyc-politics).

There is also opportunity for experimentation around increasing the scope of the rewards distributed to bettors and creating looser requirements for the capital they put at stake. This could look like rewarding individuals with attestations, or increasing the scope of financial rewards to “in-app utility” or equity represented through points or tokens.

While monetary incentives are important to make prediction markets work, some literature indicates that play money can create prediction markets with equivalent quality. Practically speaking, this tells us that we can be flexible in our assumptions about the types of “skin in the game” bettors would put at risk and be compelled to gain.

Moreover, there are different types of market mechanisms that can be used to make the UX more poll-based, which would further minimize friction and lower barriers to entry. A study by Cambridge evaluated this hypothesis and found that polling mechanisms led to more accurate outcomes compared to prediction markets during periods of low trading activity, wide bid-ask spreads, and in markets that resolve quickly. The study also found that combining poll-based prediction games with the monetary incentives of prediction markets yielded significantly more accuracy than prediction market prices alone. Additionally, to solve the potential challenge of stagnant information, polls could “renew” periodically according to some push or pull-based system, incentivizing the dynamic reproduction of information based on new information.

Crypto information games used to deter all but the most dedicated power users. Now, with lower costs, improved usability, and richer data, there's an opportunity to develop more varied and accessible games that target specific audiences.

Game #2: Privacy-Preserving Computation to Produce Information

Imagine a game played by solidity devs, where players leverage multi-party computation (MPC) to reveal their salaries and compute the average, all while preserving confidentiality of their individual salaries. This would be a valuable way for crypto professionals to negotiate with their respective employers, while also serving as a source of entertainment.

More broadly, information games can leverage privacy-preserving technology to broaden the range of raw materials—specifically private data and information—that can be analyzed to generate new insights. By ensuring privacy, these tools can increase the variety and propensity for people to share data and information, as well as compensate those data providers for the value derived as a result.

While this isn’t all encompassing, a few of the tools information producers use to do this are zero knowledge (ZK), multi-party computation (MPC), fully homomorphic encryption (FHE), and trusted execution environments (TEES). These technologies differ in their core mechanics, but they all arrive at a similar place—enabling individuals to provide sensitive information in a privacy preserving way.

Still, there are many serious challenges to using both software and hardware-based cryptographic primitives for use cases that require strong confidentiality guarantees, which we’ll discuss later.

Privacy-preserving cryptography significantly widens the design space for new information games that couldn’t have existed before.

Game #3: Competition Between Models to Improve Information Production

Imagine a game where data scientists compete against each other by developing and betting on trading models for a decentralized hedge fund. Blockchains then come to consensus on the scores of particular models, and reward or slash participants depending on the correctness of their model’s predictions and their impact on the fund’s returns. This is the approach taken by one of the earliest information games on Ethereum, Numerai. In this game, Ethereum’s consensus is leveraged by global competitions between different models and their creators, effectively incentivizing AI to play information games that result in the production of valuable returns.

Taking this a step further, we could also incentivize AI to play information games for us much more directly, leveraging their encyclopedic knowledge to compete with each other in making predictions. While they might not necessarily be having fun playing these games, using intelligent machines instead of humans would significantly decrease the cost of labor needed to produce information. As a result, these AI models could increase the amount of liquidity in much more niche prediction markets where humans would otherwise be unwilling to play. As Vitalik put it:

“If you make a market, and put up a liquidity subsidy of $50, humans will not care enough to bid, but thousands of AIs will easily swarm all over the question and make the best guess they can. The incentive to do a good job on any one question may be tiny, but the incentive to make an AI that makes good predictions in general may be in the millions.”

Alternatively, we can leverage consensus between ML models to create competition between them around the value of the information they create. Teams like Allora and Bittensor TAO are working on coordinating models and agents to broadcast their predictions to others in the network, who, in turn, are responsible for evaluating, scoring, and broadcasting their performance back to the network. At each epoch, the collective assessments between models are used to distribute rewards and/or power to the different models according to the quality of their predictions. As a result, entrepreneurs can leverage self-improving networks of models to improve the quality of the information flowing through their marketplace.

It’s entirely possible that there are information markets for which the use of models leads to a quality of information goods that information games amongst humans simply cannot match.

Monetizing Information Games

Some information games can sustain themselves purely from the fun that users derive from it. But for those who want to monetize the value of the information they produce, things require a little more thought. Unfortunately, the qualities of information as a good lead to critical market failures that prevent their seamless monetization:

• Information can be valued only after its consumption, making it hard for buyers to assess whether a seller’s price accurately reflects the value of their information.

• Information is non-rivalrous—its consumption doesn't reduce its availability, meaning it doesn’t have the scarce properties that make it interesting to buyers.

• Information’s non-excludable nature, coupled with low reproduction costs, makes it hard for sellers to prevent unauthorized access, despite high initial production costs.

These economic characteristics create challenges for both buyers and sellers in profiting from information, potentially leading to its underproduction. If information is quickly known by everyone who can exploit it at the same time, then the opportunity for an information buyer to exploit an information asymmetry shrinks due to increased competition or a collapse of the scheme they were going to use. Thankfully, there are a couple of crypto-tools that can be used to solve these problems, and already are.

Game #4: Exchanges—Monetization Through Speculation on Information

One way to monetize information production without keeping said information confidential or limiting the set of actions that can be taken on it is to simply keep that information public, but create a vehicle for people to bet on how it will change – also known as derivatives.

One company actively doing this is Parcl*, whose exchange enables users to speculate on rising and falling real estate markets. Parcl’s markets are powered by real-time price information that Parcl Labs sources from vast real estate data reservoirs and feeds through proprietary algorithms to produce fine-grained, accurate information that surpasses the quality of traditional indexes of real estate prices.

While Parcl does monetize this information more directly through an API, they’ve created an additional monetization layer by allowing traders to bet on how that information will change over time. Other projects, such as those mentioned in the “alternative information markets'' section of the market map like IKB and Fantasy, focus on monetizing through speculation or hedging on how existing public information will change, from an athlete’s performance to a creator’s social engagement.

If you can sell the right to speculate on the information you produce, you can monetize it without keeping it confidential or restricting what buyers can use the information for.

Game #5: Marketplaces for Discovering Confidential Information

Picture a game that lets you discover curated alpha on the latest onchain activity and brand new crypto startups before they become known by the whole world. For this to work, information would need to remain confidential in order to solve for the issues of non-rivalrousness and excludability that come with public information. For this reason, the next-generation of information markets are facilitating the exchange of confidential information, while leveraging blockchains to discover and regulate access to all the players that could pay to access it.

Freatic’s* decentralized marketplace for confidential information, Murmur, exemplifies this approach by gating exclusive access to information through NFTs and a queue system. Information buyers first subscribe to a particular topic by buying an NFT represented as a coupon. This then grants them a slot in the queue to redeem confidential information from publishers and, for an additional price, allows them to pay to slow down its rate of dissemination. Buyers can also vote on the quality of that information afterwards. Through this process, Murmur ensures information remains confidential and valuable without having to limit its sale to one entity.

In contrast, Friend.tech uses keys and bonding curves to manage access to confidential information in group chats, making entry more expensive as demand increases. As a result, one can think of a Friend.tech key as a proxy for the average value of information from a person (assuming the market for keys is efficient). However, players have always “priced in” some notion of the person’s “value” when trading keys, making it difficult for buyers to price the information's worth. Maybe this serves as another datapoint to support the claim that the most valuable “information markets” to date have actually been the markets for memecoins, serving as prediction markets around the symbolic value of particular trends or people if you squint hard enough.

Memecoins aside, one direction that teams gating information access could pursue is to allow information sellers to design bonding curves that better correlate the access price with the information's value. For example, pricing for information that quickly loses value as it becomes known could be determined by a bonding curve that reflects the rapid depreciation of the information's value over time.

Decentralized money exchange is challenging due to trust issues and finding double coincidences of wants. Blockchains have resolved this for money (Bitcoin) and are set to do the same for information, catalyzed by fun games centered around seeking hidden information.

Game #6: Futarchy—Monetizing Prediction Markets

One major way of monetizing information without keeping it explicitly confidential is to produce and sell the information that only one organization can and will make use of. This playbook isn’t new, as many companies already monetize information by limiting access to particular buyers through auctions or confidentiality agreements. However, we’re seeing a new business model for selling information goods—producing public information that’s only relevant and valuable to organizations making specific decisions.

In fact, we’re just now seeing prediction markets being built on crypto rails in order to experiment with Futarchy as an alternative mechanism to monetize the information they produce.

Futarchy offers a novel approach to improve decision-making, centered around harnessing the information created by prediction markets. The information produced by the prediction market is used to make decisions, and when prediction markets are resolved, the players with the best predictions get rewarded.

On their own, prediction markets are zero-sum games for players, limiting incentives for informed traders to participate in them and worsening their existing liquidity bottlenecks. Futarchy can solve this, as the wealth created by better decisions can be redistributed back to traders.

Crypto-native entities like MetaDAO are already experimenting with Futarchy. When a proposal is made, such as Pantera's proposal to purchase MetaDAO governance tokens, two prediction markets are created: "pass" for support and "fail" for opposition. Participants trade conditional tokens within these markets, speculating on the proposal's influence on the DAO's value. The resolution hinges on the Time-Weighted Average Price (TWAP) comparison of the "pass" and "fail" tokens after a designated period. Should the "pass" market's TWAP surpass the "fail" market's by a set margin, the proposal is approved, leading to the execution of the proposal's terms and the annulment of transactions in the losing market. This system employs market dynamics to drive governance decisions, aligning them with the collective projection of the proposal's effect on enhancing or reducing the DAO's value.

There are still some cases where Futarchy must be designed around confidentiality. For example, if prediction markets are used to determine hiring decisions around a specific person, that information would become publicly available and turn into an information hazard—a competitor might be interested in poaching the hire based on the market’s prediction.

Another reason to keep information confidential is its impact on motivation and organizational culture. As Robin Hanson notes in his Future of Prediction Markets speech, Google's own internal experiments met resistance due to executives’ fears that public performance indicators could demotivate employees. Naturally, managers aren’t inclined to implement something that might reveal the emperor has no clothes, and we’re seeing this in practice today. According to MetaDAO’s founder, @metaproph3t, some people decide not to submit proposals because they don’t want to be evaluated by a market.

Both of these issues could be solved by limiting the availability of prediction-market information to specific decision-makers. However, by empowering these decision-makers with autonomy over their actions based on this information, bettors will incorporate these biases into their bets, reducing the quality of the information generated.

In other cases, Futarchy may just be better applied in specific industries where its advantages outweigh cultural impacts, like Bridgewater's hedge fund. Integrating blockchain could further enhance Futarchy's integrity to prevent manipulation (looking at you, Ray Dalio 🙂).

So far, prediction markets have been limited to monetizing by allowing for speculation or hedging. In being used to help organizations make better decisions, prediction markets can unlock an entirely new market, though open questions remain around the role of confidential information.

Game #7: Credible Commitments for Programmable Information Games

As mentioned at the beginning of this piece, Google monetizes information by leasing its use to advertisers while limiting their use of this information to Google’s ad auctions. Similarly, credible commitments help information sellers monetize by restricting the actions that buyers can take based on said information.

Cryptographic methods like MPC, TEEs, and FHE can be used by information sellers to secure credible commitments about the computation buyers will take on top of private data. As a result, sellers can delegate their information to buyers, giving them specific control over future actions around their private information without revealing the information itself.

This primitive unlocks all kinds of information games. Imagine enabling traders (information sellers) to sell the right to order their transaction to information buyers (searchers) only if the buyers commit to simulating the order of their transactions a capped number of times. Taking things a step further, imagine allowing Netflix users to delegate the right for others to watch Netflix movies from their account, allowing them to “yield farm” rewards from their account without leaking its login details. In turn, buyers can unlock value from sellers’ private information, without sellers having to deal with the challenges of selling the information itself (information is a non-rivalrous, non-excludable, experience good).

Unlocking Google-Scale Monetization for Information Game Designers Today

TEEs present a practical choice for implementing such controls today, albeit with limited confidentiality guarantees. While not fit for securing large assets or sensitive data, TEEs are suitable for use cases that require more time-limited access to confidential information, such as front-running protection. SUAVE, a project created by the Flashbots team, is building a network of TEEs that developers can already use today, with the long-term vision of enabling app developers to find new ways to better monetize the value of their and their customer’s information.

In SUAVE's design, integrating blockchains with TEEs addresses three critical TEE limitations essential for advancing information games. First, blockchains eliminate the need for trust in communication between hosts and players, who could censor or behave maliciously. Second, blockchains provide a secure mechanism for state maintenance, protecting against the rollback attacks that TEEs are prone to. Lastly, blockchains are critical to ensuring the permissionless, censorship-resistant creation of TEE-based information games (SUAPPs), whose smart contracts, inputs, outputs can be trusted by all players.

While many early information games using SUAVE will clearly center around MEV, they have the opportunity to be used in information games that extend far beyond trading.

Game #8: Reputation & Zero-Knowledge to Facilitate In-Game Marketplaces

A key challenge to monetizing information is the inherent nature of information as an "experience good." The value of an experience good is only recognized upon use, complicating the seller’s ability to set a price for it beforehand. In creating mechanisms to solve for this, we can also create fun gameplay for users. Some games center heavily around enabling players to build a reputation that distinguishes them from other players, like WoW, which can be a source of fun but also a critical way for players to decide who to coordinate with. Other games might want a seller to commit to a price for some intelligence (i.e., enemy locations, secret plans) without requiring them to reveal the information beforehand.

To overcome this, designers of information games can leverage cryptographic solutions like Zero-Knowledge Proofs (ZKPs) to verify the characteristics of computational information goods—such as the efficacy of a trading algorithm—without disclosing the actual data or code. This can be achieved by creating a cryptographic commitment, timestamping it on a blockchain, and providing a ZKP of the algorithm's performance. However, this method is only effective for information goods whose value derives from the properties of its computation and can be tested on verifiable inputs.

For other types of information goods, reputation and identity become crucial. It’s possible to leverage consensus mechanisms between information buyers to create reputation around the value of the information that sellers are trying to sell.

Systems like Murmur's leverage subscriber voting within exclusive windows to establish a publisher's reputation, elevating them from unverified to verified status based on community feedback. This process creates a transparent and immutable record of interactions, building a trusted reputation for sellers that gets created with a tight feedback loop.

Alternatively, the** Erasure Bay** protocol requires sellers to stake money as well as their reputation as a signal for their information's reliability. The protocol determines a "griefing factor" that allows buyers to destroy a certain portion of the seller’s stake if the information proves to be of low quality, thereby ensuring sellers are incentivized to offer high-quality information.

To avoid market failures and maximize volume, game designers need to give sellers cryptographic-tools to prove their information’s worth, or credible, quick mechanisms for building reputation around what they’ve sold before.

Conclusion

Information games aren’t new. However, until programmable blockchains, game designers were limited to asking centralized intermediaries for permission, and players were limited to games that could be mediated by trusted third parties.

Now more than ever, the dramatic reduction in the cost for blockspace means that anyone can create a futarchy-inspired DAO or a protocol for confidential information, and plug into an endless amount of tools for verification, adjudication, monetization, and more. The games we’ll see unlocked by low barriers to participation and open innovation on permissionless financial rails are unimaginable.

This piece showcases the early signs and challenges in implementing this new wave of information games, and the potential of using crypto-tools to solve these problems. With these tools in hand, some game designers will improve information games we already play, like trading and MEV, while others create games that simply couldn’t have existed before.

Still, each of these crypto-powered information games represent mini-games that need to be composed with each other to form a complete game. The joy and thrill players gain from building reputation, collaborating with a team, and vying for influence within an organization all act as components of a larger whole.

If you’re creating a fun, crypto-powered information game, please reach out to chat. I’d love to try it, learn more, add your project to the market map, and brainstorm ideas!

*denotes an Archetype portfolio company

Special thanks to all the people in the arena for the thoughtful conversations & feedback that all went into putting this piece together: Arthur Roing Baer, Soumya Basu, sxysun, Tyler Gehringer, Katie Chiou, Dmitriy Berenzon, Quintus, Harry Jones, Alex Kwon, Spenser Huang, E.S, Proph3t, Trace, Hactar, and Connor McCormick.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Communities and cultures are nourished by rich context. Historically, context was a natural product of communities that were local and physically constrained.

The internet ushered in a Cambrian explosion of connection, allowing us to find each other while thousands of miles apart. People began forming communities around factors beyond location, such as personal interests and ideologies. In some ways, the internet felt like the only venue where one could fully and freely express themselves and find other like-minded people without borders.

Developing rich context in mercurial cyberspace, where both people and information move at the speed of light, is a challenge. Digital communities are tight knit, but are often more narrow—focused on one or two key traits rather than any holistic, shared identity.

Platforms and their Algorithms are the all-powerful gods in this climate, the providers of context. While feeding you niche, personalized recommendations, platforms control your rapidly individualized view on the world and alienate you from everyone else. I’ve written about this before in the context of music discovery:

“Constructing identity—both as an artist and as an average listener—is complicated in a Big Algorithm society. I wrote earlier that broadcast and algorithmic feeds are isolating, but the method by which these platforms isolate you is not just by literally cutting you off from local content or discourse. Endless feeds and global search offer access to an infinite scroll of new contextual layers through which to view yourself. While then identity becomes more dynamic and fluid, finding platforms and communities that support these nuances becomes another challenge.”

When you’re alienated from other people, you become even more reliant on the platform to feed you information—namely the creators specifically anointed by the algorithm. There are a few problems that result from this dynamic:

Creators must play by the platform’s rules in order to participate. Pay the platform for amplification, be subjected to their business models and take rates, adhere to the platform’s arbitrary policies and optimizations—or fall through the cracks. As platforms gain more power over who gets elevated, they eventually gain more influence over general cultural production. Kyle Chayka shares an anecdote about an artist, Hallie, in his book Filterworld that illustrates this point:

“Hallie also realized that the Instagram feed rewarded specific qualities. She had always combined visual art and writing, but posts with clear written messages got the most engagement. [...] It was a meme-like assembly-line process perfectly suited for Instagram: the bright colors and simple text added a little spice to her followers’ feeds along with simple moral messages. Followers came to rely on her account for those pieces alone. [...] The pressure that Hallie felt to make the rest of her artwork similarly bright, clear and simple is much like the pressure that a musician feels to frontload the hook of a song so it succeeds on TikTok or a writer feels to have a take so hot it lights up the Twitter feed.”

Culture submits to power-law like distribution. The only semblance of any “shared” culture in an Algorithmic society values mass scale and commercialization above all else, coalescing around major IP from a wildly different past or breakthrough moments that dissipate as quickly as they appear. Attention consolidates around a few mega stars/trends like Beyoncé or Taylor Swift (whose careers took off pre-Youtube) and bottoms out drastically. You see the ramifications of this dynamic perhaps most clearly in the resurgence of reboots, spinoffs, and sequels. According to Adam Mastrioanni:

“Until the year 2000, about 25% of top-grossing movies were prequels, sequels, spinoffs, remakes, reboots, or cinematic universe expansions. Since 2010, it’s been over 50% every year.”

In 2022, the top ten highest-grossing movies were all reboots or sequels.

Consumption-focused design patterns make social platforms less sticky. There has been growing concern about why new consumer social applications have struggled to break out, with theories for cause ranging from domination by existing giants to the influx of tools has made it almost “too easy” to build an app today. One of my personal theories is that it’s a result of the fact that “social platforms” today have become more broadcast-based, rather than actually social. Eugene Wei refers to this development as the transition from “social networking” to “social media.” Platforms are increasingly designed to be consumed, rather than to foster interpersonal connection.

The result then, ironically, is that the platform's network effects diminish, as you’re less bound by the social graph of the platform. Instead, people flee from app to app en-masse without any actual loss or consequence. Social graphs retreat to what Venkatesh Rao refers to as the “cozy web,” insular group chats that retain intimacy, but lose the benefits of discovery or legible digital presence.

In a climate where users are more wary of platform influence than ever before, many are also retreating to physical spaces: underground local clubs, independent studio/community spaces, physical pop-up stores and experiences, etc.

The likelihood of a true retreat to the physical world is unlikely, given how intertwined the digital and physical worlds have become. Digital spaces have undoubtedly become first-class citizens alongside physical spaces, but an interesting tension emerges here:

How do we leverage the expansiveness and connectivity of the digital world, while avoiding manipulation by platform monopolies and maintaining the strong context and intimacy of the physical world?

To answer this question, I believe the next generation of consumer platforms and networks will have to account for the following considerations into their designs:

Online/Offline Fluidity

The user journey across online and offline is incredibly fragmented. As the boundary between the physical and digital planes continues to blur, the ability to form a truly legible, representative identity disintegrates. There is a massive opportunity to build experiences that seamlessly blend the physical and digital, particularly where social graphs can move freely between online and offline.

For individuals/communities: How can my favorite channels online better inform and tie to the people, groups, and places I spend time with offline and vice versa?

For brands: How can a brand better understand how a consumer who bought a product in-store spends their time online and engage them on those platforms?

As a firsthand example, I have a friend who is an interdisciplinary artist whose work generally focuses on the intersection of art and technology (and therefore, is often digital). He works primarily with global, blue-chip art galleries and institutions to exhibit his work, and he also distributes his films on digital platforms such as Metrograph’s streaming service. Because his direct client base is institutions and galleries, he has very few ties to the individuals who actually engage with his work through these channels.

This lack of connectivity creates a few major issues. His work is siloed into infrequent, large-scale exhibitions without clear opportunity for smaller, more intimate releases that directly target and engage his audience; lack of ability to connect with his audience also diminishes the opportunity to take advantage of organic, word-of-mouth marketing that is native to digital, social channels. For example, if I, as an artist, am able to understand what other artists my fans engage with through social channels, I can find ways to collaborate with those artists and increase my own reach.

Today, digital social channels are great for organic distribution that underpins more material monetization opportunities such as partnerships with institutions. A future where creators, influencers, and artists can directly monetize within these channels in a way that improves the relationships between artist and audience and funnel this distribution to other monetization opportunities IRL without relying as heavily on intermediaries would be game-changing unlock.

The concept of “digiphysical” isn’t new, and there are plenty of primitives that were built to bridge users from online to offline: attestations, NFT tickets, POAPs, NFC-chipped objects. General, extensible primitives such as these are useful from a composability and open experimentation perspective, but primitives hold the most power and utility when imbued with context and meaning, such as clear vertical focus or closed product loops. For example, collecting points in a game is only compelling when I know that those points will bring me closer to some certain goal or reward.

An example model to look towards is Blackbird, which has designed a closed product loop for user loyalty and rewards. Power diners/foodies attest to being at physical restaurant locations in order to claim $FLY, which they can redeem for perks at other restaurants.

Decentralized Reputation & Curation

In regards to user-generated content and media, there are two high-level trends currently occurring in parallel: Individual users are gaining more cultural power, thanks to social media, and users are more wary of monolithic media platforms.

These trends are uncomfortably contradictory, but we see them come to a head clearly in the decline of cultural criticism and the toppling of existing media institutions. Social media platforms have given the average user enough distribution to become their own cultural journalist and critic. Simultaneously, longstanding media institutions and platforms lose prestige and cultural credibility. Yancey Strickler writes about this in, “The Prestige Recession”:

“The death of Pitchfork and cultural criticism is evidence that the mainstream is going through a prestige recession. [...] Rather than prestige, this cultural moment is dominated by metrics, [...] It was once critics who helped shape cultural values – spotting a trend here, putting a scene on the map there – but now the process is driven by metrics. Context, the land of the artist and the critic, has been determined valueless (unless algorithmic) by the mainstream, which honestly never much cared for it to begin with. Instead, art and culture have been safely neutralized as interchangeable commercial objects just like everything else.”

To tie these threads together a bit more tightly, algorithmic feeds have both given users more cultural power and also stripped away power from traditional media institutions that have historically curated and distributed media. However, users have quickly caught onto the fact that algorithmic feeds have become a new form of monolithic platform, coalescing curation around scalable metrics and flattened cultural sensibilities. Essentially, the skepticism towards Big Media has shifted power towards Big Algorithm.

The next generation of consumer social platforms will likely follow a structure where individuals are encouraged to co-create and self-curate their own digital community spaces, abstracting away more centralizing platform mechanics that veer towards explicit scale and monoculture. Every user is able to actively build a community and elevate their own “digitally local” curators, rather than passively consume a single mega-feed that is designed to scale to the masses—think Facebook group pages, Subreddits, NTS channels, Discord servers, etc. A newer platform aligned with this co-creation model is Metalabel, which enables groups to coordinate and publish creative work together.

In order for these digital spaces to grow, the platform will still have to provide sufficient mechanisms for identity and reputation so that users can discover relevant collaborators, communities, and audiences. This process emerges rather naturally in physical, local environments, but bringing these mechanics online is tricky given the natural scale.

To date, digital social discovery and reputation mechanisms have been markers such as “likes” and “followers,” but I imagine we’re bound to soon see more creative mechanisms for users to signal taste and cultural capital to each other, as platforms and communities trend smaller and more intimate.

Speed & Transience

Consumer social applications have become fleeting almost by default—see Poparazzi, BeReal, Dispo, etc. This is due to a variety of factors including patterns trending more consumption-based versus social discovery-based and therefore less sticky (as I mentioned in the beginning of this post), the arena for attention has become infinitely more competitive in the last decade, etc.

The next generation of consumer social platforms will have to account for and remain resilient to [sped up] trend cycles. Such resilience will become even more crucial as AI-powered, generative content becomes more widespread and accessible.

Umbrella-like or channel-like product structures with sandboxed subspaces are typically more durable because spaces can develop their own unique color, but remain isolated so that they may “burn brightly, explode, and fade away,” to use Packy’s analogy. The caveat to this structure is that it adds complexity for the user in terms of navigation and discovery, and also is a less plausible structure for platforms that are already vertical-specific or particularly narrow in scope.

There are also different schools of thought related to whether a more broad, extensible approach or a more focused approach with a single, unified product experience is more effective. For example, should you build a platform that communities of all shapes and forms can use or should you specifically target and build for one community? You also see this type of question arise around gaming-related platforms: Should you build an app store-like platform or focus all your energy on a single game? My general belief is that while developing a generalized platform is tempting, opinionated, narrow products with closed feedback loops can better retain initial users. Once user trust has been established around a single, quality experience, it becomes easier to expand scope if later desired.

The “drop” model is also another option to consider to keep users consistently engaged, serially launching small-scale products that are generally differentiated, but contained by the overarching brand/ecosystem (ie: MSCHF). The challenge with a drop model is managing retention across drops and making sure the social value of each drop still directly ties to the overall brand equity.

Other writers I admire have offered their own commentary and solutions to this phenomenon.

In “Small Applications, Growing Protocols,” Packy McCormick cites protocols as the layer that may ultimately yield the most long-term value from ephemeral apps that contribute users and data to the base protocol/network. Protocols can then incentivize app developers to build strong products on their protocols through tokens.

Li Jin extends this point further in her post, “Multi-Hit Wonders: Embracing Apps With Short Shelf Life” where she posits that teams can use a token model to capture value from a series of more transient, experimental apps. Moreover, tokens can be used to bootstrap product usage, seize attention and conduct self-marketing, and to create composable ties between different products.

A challenge to designing token models is that financial primitives can often create emergent incentives for other stakeholders that may be difficult to manage. For example, evaluating early product-market fit is extremely difficult when a material percentage of users may be driven by token incentives, rather than product affinity. Effectively leveraging tokens without compromising sustainability of the product will be a key design consideration for teams looking to use tokens as mechanisms for threading value across multiple products.

Culture is Moving at the Speed of Light

Throughout this post, I’ve touched on a few different topics that would each be deserving of its own dedicated exploration: identity, social media, traditional media, subcultures, communities, brand loyalty, etc. However, the overarching narrative remains abundantly clear.

Users, culture, and value are flowing more rapidly, more abundantly, and more expansively across physical and digital spaces than ever before. We are still in the very early innings of developing new tools for individual identity, community-building, and culture creation for this new paradigm. The products of tomorrow will look wildly different than the products of yesterday, and they’re already being built today.

Thank you to Patrick Rivera, Eugene Wei, Brian Kim, Jose Mejia, Jad Esber, and Archetype colleagues Ash Egan, Benjamin Funk, Danny Sursock, Tyler Gehringer, and Dmitriy Berenzon for thoughtful review and feedback on drafts of this post.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

The word that best sums up consumer technology in the 2010s is gamification. In retrospect, this makes sense, given where technology was at the time. We were entering the mobile and the social eras simultaneously, where everyone essentially now had a connected, networked game device in their pocket at all times.

The early gamification trend ushered in a wave of companies that sought to make games out of normally mundane activities and transform them into booming businesses. It turned visiting places into a game (Foursquare, 2009), monitoring traffic into a game (Waze, 2008), language learning into a game (Duolingo, 2011), the list goes on. What these companies realized was that gamification was an effective strategy for generating promotion, marketing, engagement, and loyalty with their users.

One of the common elements of gamification is a point system, where you can translate qualitative measures of progress into quantifiable metrics. Point systems fundamentally accomplish two ends: binary, legible outcomes (number go up, number go down) and channels to easily direct intrinsic motivation toward extrinsic motivations (perks, streaks, and rewards).

Blockchains are natural infrastructure for point systems because they’re designed as a universal ledger of entities with rails that can programmatically distribute value to these entities based on certain actions.

Historically, this value has largely been distributed through tokens on Ethereum (ERC20s)—financial assets whose value adjusts in real-time in open markets. Tokens are powerful tools for identifying, coordinating, and compensating users who contribute productively to a network with financial rewards and/or shares of ownership.

Token incentives have been critical for blockchain usage. The promise of tokens as financial rewards act as a counterbalance to the relatively high costs and often high risks of transacting on L1s like Ethereum. However, this dynamic can create a vicious cycle. The high cost of onchain transactions means that rewards often go to the users who are willing to pay high fees (often mercenary capital) and are generally not favorable to participants who are less willing to pay high fees or are more risk averse (often new users).

As blockchains transactions rapidly become cheaper—through the proliferation of L2s and L3s—broader classes of non-financial actions become feasible to bring onchain without the urgency and expectation to compensate users with requisite financial rewards. This new paradigm signals the emergence of new onchain primitives such as attestations to identify, coordinate, and engage a complex, decentralized network of users.

Onchain attestations are a method for identifying and classifying users, allowing users to self-attest to their own attributes and to attest to those of others. However, attestations have their own limitations. Attestations are often qualitative, which makes them difficult to use in a low-context, computational environment such as a blockchains. For example, it’s generally much easier to compare a player with 20 kills in a game vs. a player with 12 kills in the same game than it is to compare a player who killed the Green Boss vs. a player who killed the Blue Boss in the same game. This can be improved by increasing the context of the environment, and further scaling combined with developments in AI and LLMs will also make this type of analysis easier. However, given these limitations, it’s likely that more quantitative forms of attestations are most appropriate for where blockchain scalability is today.

We’ve seen experimentation with point systems in crypto begin to take off such as Blur points, which utilizes forms like “Listing Points” and “Lending Points” to incentivize specific actions and to distribute rewards that may include $BLUR tokens. More recently, Rainbow began issuing Rainbow Points to reward users for making transactions in the Rainbow wallet. To date, these point experimentations have largely been offchain, which makes them relatively similar to web2 points programs, at least implementation-wise.

Beyond traditional point systems, onchain points present an interesting opportunity to use points trustlessly within blockchain environments for purposes such as token redemptions for ownership distribution, access gating for sybil resistance, or improving market functionality in DeFi.

The rest of this post serves to illustrate the differences and tradeoffs between tokens, offchain points, and onchain points, and to explore the extent to which onchain points can serve as an additional primitive for builders and users with its own unique benefits and challenges.

Why Points

In the case of tokens, there are many characteristics that are carefully scrutinized pre-launch and can materially impact the resulting traction of the project and price of its token. Some of these factors include, but are not limited to:

• Supply and Issuance: Will the token be inflationary or deflationary?

• Usage: Will the token be used for governance, and if so, will holding governance tokens represent a claim on any fees generated by the project and control over the allocation of the project’s treasury? Or will the native token be used for utility? Will it be the unit of account/integral to using the project?

• Value accrual: Are there staking mechanisms or lockups? Are tokens spent and/or burned as a means of scarcity and value accrual?

• Distribution: Will tokens be distributed through airdrops or emissions? Will there be vesting schedules?

In the case of points, they are usually non-financial, mutable, and controlled by the issuer, meaning that point systems can be easily adjusted without immediately impacting any market dynamics. Point supply can be unlimited, and the method by which points are used/redeemed can be modified. Moreover, the tradeability of points is also determined by the issuer, whereas tokens are tradeable by design.

Being able to adjust point systems and receive community feedback in real-time without fundamentally changing market dynamics, product mechanics, or user behavior gives teams much more time and awareness to understand and better retain users. In the case that points are used as precursors to tokens, points help remove the urgency for a project to define its token model and distribution too early since it can determine later what proportion of the token supply will be allocated to the aggregate point pool.

Notably, of course, because point systems have established precedence in web2, evaluating them from a regulatory lens is arguably less questionable.

Not only are points more simple to design and execute for builders, they’re much simpler for users. Given the dynamicity of token price, users can find it difficult to know how to conceptualize a certain token: Should I treat it as an investment or as a utility/access tool? For example, imagine an arcade game where you have to pay a quarter to play the game. If you knew that tomorrow that quarter could be worth $10, you might be more hesitant to feed that quarter to the machine.

Points, alternatively, can be thought of as “meta-currencies,” where points can convert into financial value and influence usage, but this conversion can be designed to be less or more direct, depending on the situation. In this model, the redeemability of points becomes much more flexible.

In terms of point utility, points can be redeemable for a variety of options including direct product perks, ownership/equity of the project, governance rights, and/or directly swapped for income. These configurations can also be based on an opt-in basis for users.

Why Onchain Points

The more flexible nature of points raises an obvious question of what differentiates onchain points from offchain points. A key tension that emerges when thinking about tokens vs. points is that ERC20 tokens maximize composability and minimize issuer flexibility, while offchain points minimize composability and maximize issuer flexibility.

Implementing points onchain, rather than offchain, will likely sit somewhere between these two ends, allowing for flexibility while maintaining the benefits of blockchain auditability and composability.

But in practice, what does this really mean and why does it matter?

Composability

In a way, we can regard onchain points as quantitative attestations that people can view and leverage globally. Anyone can issue points to anyone else onchain, as well as build point systems based on other parties’ product usage or native point systems. Onchain points can add a new dimension to a user’s onchain identity, similar to accruing other onchain credentials, that can integrate into various modular protocols. With this framework, onchain points become a powerful tool that projects and brands can use to identify power users across products, and even attract prospective customers with discounts and airdrops.

Provenance

Onchain points also guarantee provenance and auditability, enabling transparency into the total allocation of points in the system, as well as a historical account of the methods of allocation. This transparency is vital to the extent that the point system becomes valuable to the project’s community and the demands for fairness in the allocation process.

For example, brands and agencies often work with influencers based on engagement metrics across platforms like YouTube, TikTok, Instagram, etc. However, these platforms configure and manipulate their algorithms for amplification and distribution in black box environments, making the logic behind metrics indiscernible.

Trust Guarantees

Blockchains allow for explicit guarantees on a user’s current point allocation and redemption options. These guarantees enable safe redemptions of points for other onchain assets with minimal trust assumptions, imbuing onchain points with a potential for value unprecedented in web2 point systems. Without blockchains, point systems that attempt to bridge value will suffer the same criticisms in the crypto community that we levy against web2 platforms—i.e., that they fail to satisfy a level of trust commensurate with their value—and that any stated redemption mechanism can be “rugged” without notice or historical trace.

Sybil Resistance

Point systems are also likely to impact “farming” activity that often accompanies web3 product launches. Bots can farm points just as they can tokens, but point systems can serve as a helpful communication mechanism between project teams and early adopters by explicitly signaling types of rewards that aren’t associated with a token and be used to encourage certain contributions to the product or network—for example, providing liquidity to a protocol or stress-testing certain features.

Community Accountability

Point allocations can also be held to community scrutiny before any redemption mechanisms are revealed in a more explicit manner than traditional airdrops, reducing risk of post-airdrop controversy. Onchain point allocations can even be audited, with a timestamped verification from a third-party.

Implementation

As we mentioned earlier, points can be designed for a variety of types of rewards ranging from discounts to product perks to ownership/equity of the project to governance rights to direct income. Similarly, points will likely differ widely in implementation across projects, ranging from some form of attestations to modified ERC20 tokens to soulbound tokens. While each method will have its own benefits and tradeoffs, we’ll walk through a likely common flow: redemption for ERC20 tokens.

While ERC20 tokens are the most composable method for distributing rewards, they generally minimize issuer flexibility and maximize speculative behavior. You can make modifications to effectively make them non-transferable or in unlimited supply; however, you still run into the common conflation of the token with a form of currency.

There’s also a cost consideration for implementing points as ERC20 tokens. The transaction costs of transferring ERC20 tokens onchain every time a user joins and/or a point balance is updated can get prohibitively expensive for the issuer. Alternatively, you could accumulate points in an offchain database into a Merkle tree and periodically publish the Merkle root onchain in a smart contract. When a user wants to claim the tokens, they submit a transaction to the smart contract that includes a Merkle proof that, when combined with the user's address and claim amount, can be verified against the published Merkle root (this is essentially how Merkle airdrops work).  This is a common method of distributing tokens because it pushes transaction costs to the end user instead of the project—thereby distributing the total cost (which might be in the millions of dollars) across all the token holders.

Stack* has built a solution for redeeming points for ERC20 tokens trustlessly on any EVM chain, with a distribution method that is cheaper than traditional Merkle airdrops.

While the exact specifications of a point or token system can and will vary on a case-by-case basis, we’ve included a general delineation of characteristics of offchain points, onchain points, and tokens below for guiding reference.

Apart from any technical or crypto-specific implementation considerations, there are still plenty of other crucial design decisions for creating a point system. A few thoughts:

The primary goal of a project’s point system should be to encourage product usage, not to encourage point accrual. Ensuring that point schemes eventually drive users back to your own product ecosystem is key to successfully kicking off a flywheel driven by points, rather than encouraging farm and churn behavior. This is particularly important for value sustainability. Any value lost by offering rewards must be compensated by value elsewhere—more users, higher value transactions, upsells, subsidies through ads, etc. Channeling points directly into product benefits is particularly helpful for maintaining a closed feedback loop and testing the success of specific features/products. An example of this is Farcaster Warps, whereby points earned in the app can be used as gifts to other users, or used to discount in-app NFT purchases. This explicit use-case for points within the product reduces the risk that points are viewed primarily through a speculator’s lens; i.e. only as the basis for some future financial incentive.

An effective points system also requires an intuition of what will move the needle for both your users and your product. For example, if your users are relatively price insensitive, discounts may not be as interesting; other levers like personalization or social access/rewards may be more compelling for products that benefit from strong network effects. If your product is driven by time-in-session, dripping smaller rewards often and consistently may be more productive vs. products driven by large volumes may benefit from issuing higher-value rewards less often.

The Future of Points

The story of gamification is not a new one, and there are many case studies that demonstrate that gamification can lead to positive habit-forming, incentive alignment, and increased loyalty between brands and users.

When we look to the future, it becomes clear that decentralized, user-owned networks will define the new internet. In an onchain world, gamified points can serve as a unique way to identify and reward users for their actions and contributions in an even more powerful and holistic way than in web2. Therefore, it’s important to understand the goals and roles of decentralization and ownership in your product and design point systems with those goals in mind. While tokens are incredibly powerful tools to coordinate and govern these networks, they’ve proven to also be more rigid than originally conceived. Onchain points serve as a potential new primitive for teams to use alongside tokens to explore paths to better user identity, user ownership, and incentive alignment. However, points will only be conducive towards these goals to the extent that they’re carefully leveraged with these ends in mind. We’re excited to explore the possibilities of this new primitive with you.

*denotes an Archetype portfolio company

Thank you to Sina Habibian, Andrew Hong, Shreyas Hariharan, Linda Xie, Patrick Rivera, and Archetype colleagues Ash Egan, Benjamin Funk, Danny Sursock, Nick Pai, Tyler Gehringer, and Dmitriy Berenzon for thoughtful review and feedback on drafts of this post.

Reference list of sources instrumental to this post:

• “Beyond Tokens: The Era of Onchain Points” by Graeme Boy

• “Lessons on Points Programs for Crypto Apps” by Li Jin

• Addiction by Design by Natasha Dow Schüll

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Gaming has always operated at the absolute cutting edge of technology. The earliest video games (think Pong, Space Invaders, Doom, Quake) were deceptively simple, but pushed the limits of early computer graphics and network capabilities—eventually birthing companies like NVIDIA, which would go on to revolutionize not only the gaming industry, but technology entirely. Gaming continues to be a meaningful playground for frontier technologies today, from AI/ML to blockchains to VR/AR. However, to understand exactly where and how these technologies can fundamentally improve games requires a first principles understanding of games themselves. In this post, I’d like to deconstruct the elements of a game and explore opportunities to actually make them more immersive and ultimately, fun.

“‘What is a game?’ Marx said. ‘It's tomorrow, and tomorrow, and tomorrow. It's the possibility of infinite rebirth, infinite redemption. The idea that if you keep playing, you could win. No loss is permanent, because nothing is permanent, ever.’” - Tomorrow, and Tomorrow, and Tomorrow, Gabrielle Levin

Games are a unique art form because they engage one of the most dynamic aspects of humanity—agency. Consider other modes of media such as image or film. While the resulting form of those mediums is the final object, the object of a game is the action itself.

In his classic study of the art of play, Homo Ludens, Johan Huizinga coined the term “the magic circle” which refers to the theory that a game should draw clear distinction from reality. Games create spaces where players willingly subject themselves to new rules and constraints. In this way, games are fundamentally irrational, a suspension of reality. However, these “arbitrary” limitations enable players to experience new forms of agency, allowing them to pursue world conditions that suit their own goals and desires.

While the magic of a game is its abstraction, to understand what truly makes a game tick, you must inspect its wires. Successfully plucking a player out of reality and wholly submerging them in a new world is not a simple or easy feat. Let’s turn our attention to video games which we’ll focus on in this post. A video game is a complex, cybernetic software system in which the player interacts with a machine (console, PC, mobile device), which typically consists of some input device (keyboard, controller, screen) and some sort of UI (screen) which is loaded with the game’s software. Designing a video game then requires careful equilibrium between three principal actors:

1. Player(s)

2. Environment (the game itself)

3. Machine

The Player - Environment - Machine triad is a powerful schema we can use to investigate how to craft lively, conscious games (Part I), before exploring how onchain elements can help further imbue games with new life (Part II).

Part I: A Framework for Understanding Games

The Player(s): Autonomy and Emergence

The relationship between game developer and game player is a delicate one. The game, a canvas of rules and boundaries, unfolds as a collaborative act, scripted by one and performed by the other.

We can categorize the player experience in two categories: player-to-game and player-to-player.

The player-to-game relationship chiefly concerns itself with how much autonomy is given to a player. After all, we stated earlier that the best part about games is their ability to explore agency. In exchange for willingly submitting themselves to a game’s landscape, the player must be given assurances that in this fictional world, their journey is unique and that their choices matter.

However, there is an interesting tension here. Granting unlimited agency to players actually diminishes the quality of gameplay. Fun emerges from the challenge to creatively problem-solve and push against a game’s boundaries, rules, and limitations. Therein lies the difficulty of giving players autonomy. The relationship between developer and player is a pact forged in code and design. The player seeks the sensation of agency, the belief that their actions have consequences. The developer, meanwhile, must shape these actions, guiding the player along the chosen path, yet artfully concealing the strings that bind.

Emergence is a game design principle closely aligned with autonomy, a balance between structure and agency. Emergence essentially refers to a structure where there are a few, simple rules, but the overall gameplay is designed such that unexpected patterns can still organically surface. Emergent gameplay gives players a sense of ownership and power over a game, allowing them to develop their own unique tools, strategies, and experiences.

The Legend of Zelda: Breath of the Wild is a popular open world game where players are given goals, but can explore and interact freely with the entire map and its objects in whatever combination, sequence, and pace they choose. There is never only a single way to achieve goals within the game. If you look up walkthroughs (guilty) for any quest, there will be a wide variety of methods and items players use to accomplish a task.

Alternatively, emergence is often designated to player-to-player interactions in multiplayer games. Give players as much ability to self-organize and coordinate as possible, while arming them with the structure and tools to do so. Different permutations of people then use different tools to make different decisions resulting in different outcomes. EVE Online is a game famous for its emergent properties, designed around how players collaborate.

Aside from collaborative gameplay, emergent player-to-player interaction can be facilitated through virtual economies. A basic virtual economy can be rather simple, but as games get more multiplayer, interactive and complex, such as massively multiplayer online games (MMOs), economies must become more robust in order to function properly.

For example, the virtual economy must be able to adapt to the fluctuating number of players and the resulting inflation from constantly inflowing currency. Games then must create needs to spend currency (completing tasks, buying/trading assets) and the requisite opportunity costs (time/risk to craft/obtain assets or complete tasks). The game also then needs to create mechanisms to manage the supply and demand of resources and crafted assets (rarity, item upgrades/destruction) and their pricing (centralized (NPC) vs. dynamic (player-to-player)). These are just a few examples of potential considerations for virtual economies, but evidently, these economies can be just as, if not more lively, than real-world economies.

Players can also generate new user-generated content (UGC) as a means of emergence. This could include new maps, new assets/items, as well as new markets/economies for newly-introduced features, objects, and mods. A classic example of a game that thrives from UGC is Minecraft where players can spin up new mods/servers and design them with entirely distinct aesthetics, assets, and goals. At some point, the modding may even become the gameplay itself, rather than the original designated game objective.

UGC even includes creating entirely new games, a major driver of innovation in the video game industry. Take Counter-Strike, which is considered one of the best games of all time on Steam (according to user reviews) and one of the most played on the platform (Counter-Strike 2). Counter-Strike began in 1999 as a mod of Half-Life. Valve (creator of Half-Life) would then acquire the rights to the mod and release it as a standalone game in 2000, going on to include several games in the series since then. Dota 2, also currently one of the most played games on Steam, began as a mod for Warcraft III called Defense of the Ancients (DotA) in 2003. DotA would then go on to lay the foundation for the entire multiplayer online battle arena (MOBA) category. Valve acquired the rights to DotA (even though Warcraft was originally created by Blizzard) and released Dota 2 in 2013. Auto Chess then started as a community-built mod for Dota 2 called Dota Auto Chess, later becoming a standalone game and the original inspiration for many autobattler games that would soon follow—i.e., Dota Underlords from Valve.

Types of Players

Before putting principles like autonomy and emergence into practice, we must consider that there are different categories of players, each with unique goals, motivations, and preferences for gameplay. Jon Radoff designed a general framework for classifying players, drawing upon the classic work of Richard Bartle.

In this simplified framework, Radoff uses two dominant axes for types of games: number of players and type of rewards. He then identifies the dominant mode of player motivation for each style of game.

Based on these player archetypes, game designers can better assess how different types of players may engage with different types of games and design experiences accordingly. For example, a player who values a more immersive, multiplayer experience may value autonomy and coordination tools more than a player in an achievement-oriented single-player game who may want explicit rewards and point systems. A game might appeal to only one of these player quadrants, while another might try to appeal to a variety of players.

The Environment: Narrative, Gameplay, Physics

The true essence of a game derives from the interplay of narrative and gameplay in its overall environment. Different genres of games may focus on these components with varying intensity, but they go hand-in-hand.

Adventure games, role-playing games (RPGs), and visual novels usually focus on narrative, the winding story that unfolds as a game narrates lore to the player. Meanwhile other games, like platformers and casual mobile games, tend to be more concerned with gameplay, giving players clear tasks and rules and the means to accomplish them. Different players with different preferences may value narrative and gameplay differently.

In both cases however, the ultimate goal of a game is immersion—successfully drawing the player into the experience:

Narrative immersion: The goal of narrative immersion is to make the player feel like a living entity in the game world; the transition between character and player becomes practically seamless. The keys to narrative immersion are rich storylines and characters with strong personalities or backgrounds. An example of a game with strong narrative immersion but perhaps limited gameplay is Persona 5 Royal, which is structured as a visual novel where the player is taken through a rich, manga-style storyline, but active participation is basically restricted to turn-based combat or limited choices.

Gameplay immersion: Gameplay immersion relies more on engaging tasks, clear rules, and smooth game mechanics. Bad UI and clunky tech are absolute killers of gameplay immersion, where the immersion goal is often to encourage “button-smashing.” Examples of game genres that usually prioritize gameplay include casual mobile games, certain types of combat games, or racing games.

We make the distinction between narrative and gameplay immersion not to say that one is objectively better than the other, but to understand that each form may have different design priorities and tradeoffs. Whether through narrative or gameplay, achieving full game immersion is crucial because this is where players choose to buy into the rules of the game. The player enters willingly, suspending disbelief, embracing the new reality of the virtual world. But the developer must set the stage and provide the tools in order to compel the player.

Digital Physics and Map Design

The way a player fundamentally interacts with a virtual world is determined by its digital physics (a crypto-native term, but we’ll use it more broadly here) and map design.

Digital physics is similar to real-world physics insofar as it creates the baseline constraints for how objects appear, interact with each other (collide), and move through space and time. Sound digital physics are instrumental to keeping players engaged with the narrative, coordinating players, and maintaining a game’s mechanics and performance.

Timekeeping is an important aspect of designing a virtual world not only for sake of gameplay immersion, but also as a core mechanic for how games update. For sake of simplicity, you can generally think of games as turn-based or real-time games. In turn-based games that are basically “I go, then you go,” configuring time is straightforward; time progresses only when certain events and actions occur. However games that are based in realism usually simulate real-time. Effectively simulating continuous time and motion requires the game to update its internal logic/state and the player's console and display to render frames to the player at smooth speeds—known as ticking (tick rate) and framerate (frames-per-second), respectively. These mechanics, usually determined in a game engine, are essential game primitives that can be technically challenging to design across players with different hardware, software configurations, internet speeds, and geographical locations.

In regards to map design, an open world’s topography crucially informs its gameplay. For The Legend of Zelda: Breath of the Wild, the open world design was actually incredibly difficult to pull off. During development, the team had to completely redesign the terrain of the game world in order to nudge players towards different parts of the map without giving them marked paths.

Climate is also closely related to topography and can be used as a core mechanic of a game. A game’s climate can trigger certain characters, events, and challenges, or it can even dictate a player’s pace. In Zelda, weather is a very contentious topic that drives discussion and community among players. Many players appreciate how immersive and realistic the effect is, while others find it extremely inconvenient and unnecessary.

The Machine: Format and Crossplay

The machine is the primary conduit through which players can express physicality in a virtual world. Remember, we began this exploration by describing a video game as a cybernetic software system in which the player and machine play each step of the game together. Innovation at the machine-level inevitably bubbles up to its environment and players.

Machines, historically, informed much of game design. Consider the arcade game—often played in a public place where play is restricted and timebound. These games often are designed around the concept of finite lives whereas computer/console games require a single purchase and then can be played freely on one’s own device. These games are often measured by health score, rather than finite lives, allowing longer, more flexible gameplay including incremental power-ups, pauses, and saves. Similarly consider the advent of mobile games, where games are often designed around scrolls or fluid finger motions rather than clicks or button-smashing combinations.

As a counter-example to the importance of the machine-game relationship, mobile gaming and multi-platform/cross-platform support have emerged over the past few years as paradigm-shifting developments.

Crossplay will likely prove instrumental to the next decade of video game development at-large due to the pressure it places on gaming studios and publishers both from technical and distribution standpoints. Technically speaking, each platform from PlayStation to Nintendo consoles to mobile has different technical and graphics requirements that make adapting games to be multi-platform a nontrivial lift. From a distribution standpoint, platforms historically develop moats from the games/IP they offer. For example, most Nintendo games can only be played on Nintendo consoles, meaning Nintendo gains massive IP moats from series such as Mario, Zelda, etc.

While still nascent in development, the convergence of meta-trends like crossplay and mobile gaming drastically limit moats that gaming companies traditionally rely on such as tying IP to consoles, making the probability of a more open, player-driven future for gaming much higher.

Teyvat Tabloid ⚡ LORE & MEMES ⚡

@teyvattabloid

Genshin Mobile vs PC

What's your device and can you spot the difference ?

Mobile (Max Settings, And Scaled To 1080p)

PC (Max Settings, 1080p With Render Scale Set To 1.5)

Ghostyriah

#Genshinlmpact #Genshin #GenshinMemes

2,235

6:44 AM • Jun 7, 2023

Part II: Opportunities for Onchain Gaming

When discussing the intersection of crypto and gaming, the framework often looks something like this:

• Web 2.5: games bring assets (i.e., skins) onchain with NFTs; use ERC20 tokens for in-game currency, and build onchain market infrastructure for assets and tokens; most, if not all, game state/logic is offchain

• Web3: all game state and logic is fully onchain

The reality is that different types of games with different structures and goals are probably better suited for different onchain elements (if any), and it may be better to structure the conversation around game-technology fit rather than rely on a purity spectrum. With the high-level framework for game elements we developed in Part I, we can explore ways that onchain components can improve and revolutionize games.

Physics/Engine + Gameplay

The absolute primitive components of a game are its engine and physics—the substrate for designing a virtual world. The more open this level is, the more malleable the levels above it become.

Games built with blockchains at the base layer enable truly open modding, rather than relying on companies to issue composable tooling that they can close at any time or underground hacking (jailbreaking, reverse-engineering) that cannot easily scale. With natively open modding, players can permissionlessly and directly contribute to virtual worlds they already love, rather than having to create separate, derivative environments from scratch—and they can continue to do so until the end of time, regardless of any permissions or whether or not the original client continues to exist. Similarly, bringing all game state and logic onchain opens the design space for permissionless new clients which can even be built cross-platform, a usually centralized and demanding task as discussed earlier.

We’ve already seen onchain modding manifest in games such as Dark Forest. Without core developer intervention, players have created new marketplaces, guilds, game modes, and an entire library of plugins, all of which are onchain.

Compare this to earlier examples of modding we considered, where a general pattern appears: company provides game tooling (Warcraft World Editor, Half-Life SDK, Dota 2 Workshop, etc.) → community builds mod (Counter-Strike, DotA, Dota Auto Chess) → IP gets acquired by a big company and is released as a stand-alone, closed game. The overall effect of this dynamic is fragmentation and cold-start problem for newer games, versus having the ability to seamlessly extend an existing world.

There is a more existential question to be asked, however: Do players genuinely want more control and autonomy over their games? The nuanced answer here is that there will likely be players that prefer games that look more like curated theme parks and players that prefer games that look more like buildable cities. Neilson has an excellent meditation over the complex role of player autonomy in games here:

“Players are becoming designers, creating new rules for the game, automating new players, creating new autonomies—a process of positive feedback. [...] There are games in which players play at design while still remaining firmly inside the game, RollerCoaster Tycoon and Fortnite are somewhat like this. That is ok. However, when a player is fully able to change the rules of a game, they are no longer inside it. That is also ok. To truly make players into designers is to upend the game entirely.”

Perhaps the most material implication of open modding for core game developers is that they must pay careful, close attention to determining which constraints and details of a world must be formally and strictly defined (i.e., laws of time and physics, alchemy of materials, etc.) and which details may have softer boundaries and be open to change. The immutability of blockchains means that the original developer can still instantiate permanent rules that ensure the integrity of the virtual world; after the introduction of these rules, no party can alter them.

In some ways, these capabilities bring a game even closer to reality than further from it—allowing players and developers to build, interact, and coexist in a lively world that still remains stable and unyielding in its core fabric of reality. These ”Autonomous Worlds” may very well defy our perceptions of not only games, but of digital spaces entirely.

There are many complex challenges to bringing game physics—namely time, state, and collision—onchain. We covered some of the basics of timekeeping and the challenges it presents, but bytes and yonada detail in-depth considerations for onchain timekeeping in their post “Timekeeping For Digital Gods.” To summarize: implementing ticking in a blockchain context based on discrete blocks and transactions is not a trivial task.

There are currently workaround solutions to these challenges such as custom rollups to implement ticking, designing games around unsynced time, or more broadly “lazy updates,” but each approach comes with its own complexities and is still in early experimentation.

Emerson Hsieh

@intemerson

How we built lazy updates in our fully onchain game @primodiumgame:

A fully onchain game is gas-intensive, so we need ways to not have players sign as many transactions. Enter lazy updates where we calculate how much game resources are needed only when players perform actions.

59

11:30 AM • Dec 29, 2023

Another interesting consideration for building onchain games is the challenge of incorporating hidden information into a global, shared environment (blockchains). Hidden information is a major cornerstone of gameplay. The thrill of a game is making strategic decisions despite not knowing the statuses of your opponent(s) or what treasures/traps the virtual world may hold in store. This becomes especially crucial when you consider that developers have open access to onchain game state and data and can build other features atop them.

Dark Forest used ZK to tackle hidden information, enabling the game to verify critical information, like player and asset location, while keeping the information hidden from other players. This method of verifying and storing game state without publicly sharing that information is critical to making games like MMOs and real-time strategy games (RTSs) work onchain.

While there are certainly challenges to hiding information onchain, there are also net-new unlocks that are enabled by a new form of data privacy. Intuitively speaking, the more data you have to work with, the better you can program games dynamically and expressively. And the more that that data is also able to be kept secret, the more that the data can be utilized without disrupting gameplay or compromising user privacy (i.e., more robust third-party clients because game states are provable).

The set of privacy tools for games is quickly expanding, incorporating other cryptographic mechanisms like MPC, TEEs, and interactive shielding. However these fields remain nascent, expensive, slow, and overall difficult to implement.

Narrative/Lore + UGC/P2P Coordination

An immersive world fosters an engaged community that desires to contribute and proliferate its narrative. The popularity of fanfiction sites like Wattpad (acquired by Naver in 2021 for $600M), Fanfiction.net (12M+ registered users), or Archive of Our Own (6M+ published stories) make this clear.

With the ability to co-author a virtual world comes the power to build its lore in much more meaningful ways. Related to our earlier discussion around emergence, players can contribute to a game’s lore in two principal and often overlapping ways:

1. User-generated content (UGC)

2. Player-to-player (P2P) coordination

We spoke previously about UGC mostly in terms of building entirely new games and assets, but UGC can also materialize in much simpler ways—creating new stories, theories, memes.

Another Zelda example: Koroks, although a relatively small character in the game, became a major theme for many players. In Tears of the Kingdom, players were able to interact with Koroks more “creatively” than in Breath of the Wild, unleashing a swath of player engagement around a feature mostly inconsequential to the main game.

Traditionally, these forms of UGC exist in siloed, side forums like Reddit but more decentralized, open games offer opportunities for players to integrate UGC more directly and expressively into games without permissioned access. However, it’s important to note that open access does not render incremental features such as creator tooling, content-exchange, and content moderation irrelevant. If anything, making these tools accessible to a wider audience becomes more essential.

Player-to-player coordination is another primary channel of lore-building. Player-to-player coordination can vary from chatting on side forums in Reddit to game-native chat rooms or collaborative core gameplay. Much of the lore in EVE Online was developed by stories about its players, rather than the game itself. There’s even entire books detailing the various wars players have independently waged against each other in EVE Online. It’s also worth mentioning that EVE Online announced the development of a game in 2023 called Project Awakening, which will leverage blockchain technology.

Making community-developed and community-owned features like UGC and P2P coordination first-class game citizens creates a core dynamic shift between game stakeholders. Usually, major companies and game developers are trusted to design a game with adequate guardrails and considerations for fun and safe gameplay. However, the more a game is community-driven, the more that power shifts to players, creating the need for strong mechanisms around player governance and identity.

Governance tools are paramount for games that are team-based or inherently collaborative in nature, enabling players to effectively self-organize into teams, organizations, guilds, etc. Informal governance tools largely already exist in games today but making these tools more powerful and programmatic becomes more crucial as games become increasingly open and community-driven.

Governance tokens are an existing model for more open and organized governance that have yet to be fully utilized in games, specifically. A more exploratory vision for game governance tools are mechanisms that allow players to directly govern a game’s code. For example, players that obtain a level of achievement in a game (through the accrual of tokens or otherwise), could be programmatically granted more access to the core mechanics of a game.

A precursor to effective player governance is robust tooling and mechanisms for identity and reputation. Most gaming platforms have methods for players to build profiles, but in a world in which 1) games persist beyond single platforms (cross-platform play), 2) more independent and community-developed games/forums gain industry power (driven by more open tooling and open play), and 3) games generally become more player-driven, identity must become extensible and interoperable.

Onchain identity tools—ranging from cross-platform wallets to soulbound tokens (SBTs) to attestations and flexible signatures—enable players to build holistic reputation across games and platforms, to more effectively self-organize and socialize with other players, to discover new relevant games and be identified by new game developers, and to build stronger trust with each other, rather than relying on a central authority to create restrictions or guardrails within single games (increasingly difficult in an open-game paradigm).

Open Economies

The most experimentation in onchain gaming has undoubtedly been in onchain economies—colloquially known as GameFi. A very direct method to drive player engagement is to create monetary incentives where upside and downside are demonstrably high. For sake of scope, we’ll focus on games outside of casino game genres.

Game assets such as skins have proven over the past few decades to be a robust market, with certain assets selling for thousands of dollars, and some sources estimating the size of the gaming cosmetics market to be $50B. Platforms like Steam that run asset marketplaces can take fees on these sales and can also direct portions of sales revenue back to game developers. Trading points or in-game currency within games or gaming marketplaces is a much less common and usually banned practice unless trade is native to the game’s economy; doing so is also usually technically impossible without trading account credentials.

Recall that earlier we discussed the complexities of designing a sustainable virtual economy. Currency inflation/deflation, asset supply/demand, and pricing all must be carefully manufactured in order to keep a player engaged and active in the game’s economy. While these parameters are usually strictly and manually controlled by the developer (often for good reason), the appeal of tokenizing in-game assets is that it makes these assets more dynamic. Onchain assets and in-game entities (NPCs) can interact more programmatically and autonomously, making the economy more robust and lively.

Onchain game assets are also more easily updatable, composable and interoperable across games. This is ideal in an open, Kingdom Hearts-type multiverse where players can combine game lore—technically possible in a fully onchain world, but unclear today how narratively feasible that is. Perhaps more realistically, bringing game assets onchain makes it easier to combine, upgrade, and trade assets and characters between players. Onchain game assets also present possibilities for game access controls based on asset ownership.

As a more meta note on tokenizing in-game assets: in a world in which “canon” is fluid and player-driven, it’s perhaps easiest to canonize the introduction of a new asset by attaching “value” to it in the form of a market price (whether purely in-game or real-world).

Early play-to-earn games like Axie Infinity explored at-scale the prospects of bringing game assets onchain (NFTs) in order to earn in-game currency (ERC20s) that could be swapped and withdrawn for fiat currency. However, designing open, sustainable in-game economies is a difficult endeavor when gameplay is bifurcated into players and farmers and retention is highly unpredictable. At its peak, Axie Infinity had almost 2.8M monthly active players, whereas today that number is closer to 300K.

As a traditional gaming example, Diablo III (2012) introduced an auction house for official in-game trading, but was quickly lambasted by players as they felt that the core game loop was completely broken to focus on the auction house. The auction house was later removed.

Other areas for experimentation in onchain gaming economies include prediction markets and more nascent mechanisms like staking and AI-powered components like sentient NPCs and AI-generated assets. For example, ZK and ML could be combined to leverage ML models verifiably in games, creating “sophisticated agents/NPCs and concurrent storylines” as outlined in EZKL’s “ZKML and Autonomous Worlds.”

Onchain game economies also enable powerful flywheels for mod incentives. While most mods in traditional gaming are separate instances, onchain mods allow value to flow programmatically through the core game and back to its contributors—both core game developers and mod developers—creating powerful social and economic network effects. Designing aligned incentives between large gaming companies and players may seem like a secondary concern, but it becomes more important as you zoom out to the meta trends we’re seeing in gaming at-large. Gaming is becoming more player-driven. The moats companies traditionally rely on are being chipped away by developments like mobile gaming, crossplay, and more indie game development and subsequently more digitally-native game discovery. Creating strategies for value alignment between core game developers and their communities will become a crucial requirement for sustainable game development.

Looking Ahead

Games are a fantastic playground for exploring the applications of frontier technologies, but games first and foremost have one mission: to create fun, immersive experiences for their players.

We built a general framework for understanding the elements of a game, examining the intricate relationships between players, game environments, and machines.

Understanding player autonomy and emergence as fundamental principles highlights the importance of giving players agency while still maintaining the integrity of gameplay. Within a game itself, the interplay between narrative and gameplay must be carefully crafted to create game immersion, with different genres of games prioritizing these components differently. The machine, as the conduit for player interaction, has historically influenced game design, but the rise of cross-platform play is reshaping the gaming landscape. This development reduces the influence of platform-specific IP and challenges game studios to adapt to a more player-driven future.

Looking ahead with this framework in mind, opportunities for blockchain rails to revolutionize gaming emerge:

For games that are open and built around modding

Robust onchain game engines allow mods to become more composable, enabling players and mod developers to combine and expand upon existing game mechanics and assets seamlessly. This composability creates a flywheel effect for value creation, as modding becomes more accessible and the community grows. Core game developers benefit from the influx of creative content and can even monetize successful mods. Simultaneously, mod developers are still able to benefit from the network effects of existing games, while earning social capital and programmatic rewards for their work.

For games driven by player-driven lore and collaborative gameplay

Onchain games empower players to become co-authors of their favorite games' lore. They can actively shape and expand the game's narrative, which was previously confined to external venues like fanfiction sites or community forums. This newfound power to contribute directly to a game environment adds depth and new life to the storytelling, enhancing player engagement and immersion. With more player power comes the need for extensible identity and governance tools, which can also be facilitated by onchain mechanisms such as attestations, wallets, and DAO tooling, just to name a few. These tools improve players’ abilities to self-organize and collaborate, whether it's forming teams, guilds, or engaging in cross-platform play. Trust and recognition become more transferable, fostering stronger social bonds among players.

For games driven by upgradeable, dynamic gameplay

Games that rely on dynamic, ever-evolving gameplay can benefit immensely from composable onchain assets. These assets enable faster-paced generation and upgradeability (perhaps aided by generative AI), allowing game developers to introduce new content and mechanics rapidly. Players can adapt and customize their in-game experiences in real-time, keeping gameplay fresh and exciting.

For games driven by markets and economies

In games where markets and economies play a pivotal role, onchain economies enable markets to dynamically evolve and thrive. Players can trade assets, purchase items, and engage in sophisticated economic activities permissionlessly, with clear real-world value. The ability to convert in-game assets into real-world value adds an exciting dimension to gameplay, attracting a broader audience while providing tangible rewards for players' efforts.

These opportunities were identified within the scope of how we understand games today, which I felt was particularly important for setting the stage and context for what current users and players expect from their games.

However, it’s important to acknowledge that new developments in crypto and other rapidly emerging technologies may create opportunities that we can only dream of—truly testing the limits of not only the question “What is a game?” but also “what is a world?” While the exact details of the future are unclear, it is glaringly obvious that we are hurtling towards a more user-driven paradigm of play. As players and technologists ourselves, we’re excited to explore that future with you.

Thank you to GVN, baz, Tarrence, Lia Zhang, Neilson Koerner-Safrata, Dante Camuto, guiltygyoza, Emerson Hsieh, Brian Cho, Vivek Bhupatiraju and my Archetype colleagues Katherine Wu, Benjamin Funk, Danny Sursock, Nick Pai, Aadharsh Pannirselvam, and Dmitriy Berenzon for thoughtful review and feedback on drafts of this post.

Reference list of sources instrumental to this post:

• Games: Agency As Art by C. Thi Nguyen

• Gaming: Essays on Algorithmic Culture by Alexander R. Galloway

• Gamecraft by Mitch Lasky and Blake Robbins

• “The Open Problems of Onchain Games” by Charlie Noyes and Doug Feagin

• “Autonomy & Automation” by Neilson

• “Infinite Modding” by Rafael Morado

• “Timekeeping for Digital Gods” by bytes and yonada

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

This article builds off of the previous MEV & The Evolution of Crypto Exchanges: Part I. While I recommend first reading through Part I if you haven’t yet, here’s a quick recap in case you’re short on time (or just want to ape into Part II).

Part I Recap & Reflections

Part I laid out the critical challenges and tradeoffs between between order matching systems in crypto today, including:

1. Limitations of futuristic AMMs

2. RFQs vs order books from a market microstructure perspective

3. The fundamental MEV-related tradeoffs and limitations of onchain order books

4. Design challenges for rollup-based exchanges

5. Batch auction analysis from a market microstructure perspective

Coming out of the other side, a few things become quite clear.

Many order flow aggregators using onchain order matching systems today generate, and will seek to internalize, MEV, as competition compels them to do so.

However, due to the technical challenges highlighted in Part 1, and the intensifying competitive landscape around exchanges, order flow aggregators new and old will accelerate the development and integration with platforms that leverage offchain order matching to enhance counterparty discovery across users. A few years down the line, the markets and assets that receive a stamp of approval from the global financial-regulatory-complex will leverage offchain order matching engines because they are a minimum viable architecture required for market makers and traders to engage with reliably.

As an industry often driven by short-term incentives, but also a pragmatic approach to solving today’s user needs, it’s easy for us to make assumptions about the evolution of the product landscape based on the current profiles of crypto users. However, we’re not even on the market's ground floor for crypto adoption or retail trading, let alone institutional participation. In five to ten years, the relatively price-insensitive MetaMask swappers are likely to be just a sliver of the millions of people that want to self-custody tokenized assets on blockchains and trade them efficiently.

As a result, we should be wary of the emergence of incumbents entering the crypto exchange market, but open to the opportunity to build and invest in companies that utilize offchain order matching engines while leveraging blockchains for settlement and self-custody.

While the order matching systems covered in Part I have their critical limitations, they also have many advantages for different markets. There will be markets for many valuable assets that the global financial-regulatory-complex won’t be able to compete in even if it wanted to, and onchain order matching venues will be the platforms to serve them. Moreover, the pace of innovation and diversity of thought taken by teams in building onchain financial products will create all sorts of interesting and also practical new paradigms that are worth exploring — as technologies to facilitate exchange and their impact on market structure and MEV.

After releasing Part I, I received a lot of feedback along the lines of “Great article; the challenges and tradeoffs in exchange designs are clear; how do we solve MEV though?”

While it’s difficult to answer that question, it’s worth considering that the more important questions might be how do we create the best exchanges and for what markets can the onchain exchange primitives we’re building be competitive? While the structure of MEV does shift according to blockchain design — such as mempool configuration, encryption schemes, and fee markets — the biggest contributor to the shape of MEV will be the exchanges where users trade crypto assets. With this in mind, we need to examine the impacts of new primitives around exchange design and their impacts on market structure.

Using our framework from Part I, this article will attempt to break down these questions by evaluating how new onchain primitives might impact market structure and MEV. Assuming much of MEV is downstream of adopted exchange designs, this will help give us a clearer picture of how the future of MEV might play out.

Intents and DEX Design

Beyond the LVR-minimizing AMMs and privacy-preserving batch auctions described in Part 1, there are new tools that plug into and build on top of onchain order matching systems. These tools don’t directly change how order matching *works*, but aim to help users and the dapps that aggregate their order flow better express and redistribute the value of the MEV they generate. In this section, I break down their unlocks and potential merits as means to improve the state of onchain exchange.

Intents

There are three contexts in which intents are discussed today:

1. UX-improving architecture separating offchain computation with onchain verification

2. New order matching system to compete with exchanges of today

3. Technological substrate to improve our ability to coordinate

For the scope of this piece, I’ll focus on just 1 and 2. However, intents as a technological substrate to improve our ability to coordinate, taken to their hypothetical limits, are one of the most exciting and wild frontiers in crypto today. In this context, we need to take a step back and see how intents can compose with the variety of cryptographic and financial primitives we’ve already created onchain.

The article that intent-pilled me on this is Christopher Goes’s Towards heterotopia - the prerequisite cultural and technological substrate for a return to a world of scale-free credit money. I’m pretty convinced we’d be better off redirecting techno-futuristic e/acc energy into investing time, energy, and capital to render the vision in this article alive than whatever we’re doing now. Alas, best to give this subject the scope it deserves.

Enhancing UX & Prices: Decoupling Offchain Computation and Onchain Verification

With this framework, intents are a way for developers to design decentralized applications and deliver better UX for users. Intent-centric smart contract design leverages offchain computation to perform onchain actions that are constrained in the set of potential state transitions they can trigger on blockchain through some verification mechanism. As a result, dapps and DEXs can help users achieve their “intents” faster and at a lower cost in terms of price and gas.

My colleague Nick described what this looks like on blockchains like Ethereum today better than I could. Many dapps on Ethereum, such as Across, UniswapX, and CoW Swap, use an intent-centric smart contract design in that they leverage offchain computation and optimistic verification to minimize latency and costs while preserving safety.

This design pattern is especially helpful in the context of onchain exchange. For example, the resulting price improvements from leveraging intents on UniswapX on a single chain are the product of enabling fillers to leverage offchain logic, access offchain liquidity, and incorporate offchain prices into their quotes. This also allows fillers to quickly assess and provisionally fulfill users' intents for cross-chain swaps by lending funds, all while some oracle (like the UMA Oracle in Across’s case) ensures the security and finality of these transactions. Intents systems also lean on the gas savings that come with “coincidences of slots'' from submitting a collective batch of intents onchain, as opposed to individual orders. It’s cheaper to verify 1 batched transaction than verify that 100 different transactions followed a correct state transition.

New, intent-centric blockchain architectures and virtual machines also benefit from separating offchain computation and onchain verification. In this regard, their differentiation stems from how these new designs enhance multi-party coordination, and whether state transitions are verified through the use of invariants or oracles.

A New Order Matching System to Compete with Exchanges of Today

Taking things a step further, some believe that by operating outside of the constraints of the Ethereum transaction format, users will be able to more granularly specify conditions of exchange and receive better outcomes in terms of trade execution.

There’s good reason to look at the state of order books today, then look at an intent-centric system for exchange, and conclude that intents could represent a needed evolution from order books. Specifically, the fact that order books are characterized by an order to trade ratio of over 100-to-1 could be an indication that limit orders aren’t cutting it as a means of expressing traders’ preferences. Moreover, it’s possible that in leveraging a common language for expressing trading preferences across disparate financial domains, intent-centric exchanges could clear trades across broader available liquidity than a traditional exchange, ultimately improving the efficiency of counterparty discovery.

In the context of exchange, intent-centric architectures can be implemented to tackle this by allowing users to add arbitrary database transactions to their order, enabling them express their preference functions for how state (probably the state of their balance of an asset) can be updated in the future. These database transactions specify additional conditions of exchange beyond the quantity and price of the desired assets. Solvers then calculate the pareto-efficient frontier of potential new states expressed across multiple parties, but are constrained in how they update the distributed ledger serving as a settlement layer. This ensures that the solvers abide by the validity predicates that represent the constraints on state transitions specified by all users' intents. As a result, nodes are enabled to verifiably update their copy of the distributed ledger without having to engage in the computationally heavy task of solving intents in a pareto-optimal way.

There are a few reasons to be skeptical that intents represent the solution to the challenges mentioned above. The core reason behind the symptom, which is a 100-to-1+ order-to-trade ratio, is that public order books reveal traders' preference functions in the first place, forcing them to break up trades into smaller orders to conceal the implied information they may reveal. The biggest reason traders are forced to submit many orders to express their preference function isn’t generally because their demand function is changing, it’s because they don’t want to reveal their demand function to the market. By flashing multiple orders, traders can effectively keep their demand hidden from the market.

Moreover, intents aren’t a new way of order matching, and could even deepen the existing challenges in creating competitive onchain exchanges today. Intents can be applied on top of onchain or offchain order matching systems, including order books and RFQs, but don’t represent a new way of operating an exchange that can solve the core challenges of onchain order matching. Although they follow very different designs, both Anoma and SUAVE represent architecture conducive to building onchain order books. Both allow users to specify arbitrary database transactions on top of their limit orders, and enable developers to create applications that guide users to do so. However, they don’t overcome the critical challenges of onchain order books already highlighted in Part 1.

Long-story short, there is a fundamental and inescapable tradeoff in the type of MEV onchain market makers have to deal with between order book configurations where multiple leaders provide input into the sequence of orders or a single leader decides the sequence of orders.

In all intent systems, the computationally complex work of clearing intents is delegated to the solvers, who run heavy algorithms to clear these intent-based orders optimally. The higher the breadth and variety of database transactions takers can define, the more an intent system looks like a complex combinatorial auction — participants can place bids on combinations of discrete heterogeneous items, or “packages'', rather than individual items or continuous quantities. Ultimately, intent-based architectures create more computational complexity for solvers than a traditional order book, akin to the time taken to solve a combinatorial auction. This results in the issues core to the MafiaEV problem described in Part 1, whereby the liquidity and user uptake on the exchange would likely suffer from the slower updates to prices.

It’s worth pointing out that some intent systems, such as Khalani, allow developers to create composable modules for their users to specify intents, enabling them to be compiled into more fungible atomic units. As a result, these modules help direct the runtime search for solvers, decreasing the heaviness of the computation solvers must undergo. However, the more fungible intents are for solvers, the less expressive they are.

While there’s no reason to think order books represent some pinnacle of evolution for exchange that can’t be surpassed and enhanced by new technologies, there’s reason to be skeptical that specifying one’s preference function more granularly will change the fundamental problems that challenge the scalability of onchain order matching systems today.

Along with this, it’s unclear to what extent traders want to, or should, go beyond defining slippage tolerances and limit orders. There’s not much evidence to support the claim that retail traders want to express very complex preference functions in the context of financial transactions. Either Joe wants to buy an asset because he wants to gamble on something, or Joe doesn’t want to gamble and outsources thinking about his preference function to someone who does things for him. Within an institutional context, sophisticated trading firms likely have internal systems that add some version of database transactions on top of their orders today, and have developed their own internal infrastructure to do so.

While intents represent an exciting frontier for designing distributed systems, enabling net new applications that transform how people coordinate, it’s unclear what their architectural edge is in creating the onchain exchanges of the future to compete with their offchain counterparts. It is possible that intent-centric architectures may help reduce MEV for assets whose order matching needs to be onchain by empowering users to better define constraints around trade execution. Expressing those constraints as validity predicates may be a more effective way of preventing MEV than retroactively enforcing a punishment on a solver that has violated a user’s intent.

The opportunities and risks associated with different configurations of the mempool by which users submit intents is also another major factor to take into account when evaluating how they might improve or harm the quality by which intents get executed. This topic is covered and debated in great depth by @0xQuintus, @gakonst, and @cwgoes in the following pieces: Intent-Based Architectures and Their Risks and Towards an intent-centric topology.

OFAs & DEX Design

Unlike batch auctions covered in Part 1, where a user submits a trade without specifying a price, order flow auctions (OFAs) offer the right to execute a transaction at a price the user has already specified. In an OFA, the searchers/solvers/fillers who won the auction can execute a trade if it's within the user-specified restrictions.

Reintroducing the key stakeholders in matching a trade:

1. The Buyer (Party A)

2. The Seller (Party B)

3. The Market Maker – the intermediary facilitating the interaction

Bidders and the OFA platforms that host them (onchain or offchain) make money from pushing Party A and B on the opposite sides of a trade closer to their breaking price.

Just as intents need to be delineated from the systems that match orders, OFAs compose with and do not represent a new, standalone way of handling onchain order matching. In practice, the bidders in many of today’s OFAs enable bidders to take on counterparty risk by serving as market makers. In this case, those OFAs are implemented on top of order matching systems like RFQs or order books, where the searcher/solver takes on some liquidity risk in filling the user's trade as well.

In this paradigm, the value OFAs extend to order flow aggregators and their users goes beyond what the dapp and user could ever do alone, making the business model of OFA providers more sustainable long-term. However, it’s worth assessing whether the solvers are market makers taking on liquidity risk, or arbitrageurs taking none, to assess their long-term competitive advantage.

Smart Contracts as Order Flow Segmentation Devices

In tradfi, OFAs have emerged as a function of retail-friendly brokers leaning into order flow segmentation as their core business model. Naturally, retail-friendly brokers like Robinhood have a high concentration of retail traders who are much more likely to submit trades uncorrelated with broader market movements.

As a result, market makers (such as Citadel) pay those retail-friendly brokers to access that flow (meaning Robinhood auctions it off), because it helps them delineate between uninformed or informed traders, and gives them better guarantees around the non-toxicity of the flow sent from those brokers. As a result, unsophisticated traders coming from these retail-friendly exchanges could be profitably served by market makers at narrower spreads than on a regular exchange.

If the retail-friendly exchange retrospectively observes that a market maker doesn’t provide prices that compete with the NBBO, the retail-friendly exchange will auction the order flow off to another market maker (this is a simplified explanation). On Robinhood, this OFA is paired with an RFQ as the underlying order matching system.

Decentralized exchanges could leverage, systematize, and accelerate order flow segmentation by opening up the access and development of smart contracts that delineate the extent to which a trader is considered informed. Onchain market makers (passive or professional) could then offer different spreads to traders depending on the smart contracts or wallets they initiate trades from.

To visualize how this works, let’s look at an oversimplified runthrough of a switch from pre to post-order flow segmentation. In this example, let’s assume that Exchange 1 is implementing order flow segmentation, while Exchange 2 is not implementing order flow segmentation.

1. Pre-order flow segmentation, let’s imagine the breakeven point for market makers to service aggregated informed and uninformed traders is 5bps.

2. The competition between market makers is so high that all flow is serviced at this breakeven price on both Exchange 1 and Exchange 2.

3. New smart contracts are created on Exchange 1 that enable retail users to identify themselves as uninformed. Now, those retail traders are served at 4bps, and any trader seeking onchain liquidity through other means is served at 6bps.

4. Exchange 1 will likely profit off of increased retail volume.

However, we also need to think about the stable equilibrium of order flow segmentation. In practice, it’s likely that order flow segmentation leads to a series of adjustments by market makers that ultimately increase spreads for all market participants. Coming back to our example:

5. Because Exchange 1 has implemented order flow segmentation, sophisticated traders migrate from Exchange 1 to Exchange 2 — 5bps on exchange 2 is better than 6bps on Exchange 1.

6. Now, market makers on Exchange 2 deal with an increased concentration of sophisticated traders, and widen their spreads as a result.

7. Observing this shift, market makers on Exchange 1, initially operating at narrower spreads for retail traders, begin to increase their spreads, not due to a change in the risk profile of their traders, but just because they can from a competitive standpoint.

The end result is a market where spreads are generally higher for all participants, contradicting the initial expectation of efficiency gains through segmentation.

It’s interesting to consider whether AMMs could make credible commitments to servicing flow from smart contracts at unchanging spreads over time, potentially overcoming the competitive dynamic above. It’s also possible that identity projects or third party data providers issuing credentials to wallets could be an alternative to relying on different smart contracts that delineate different order types or order size.

However, much more work needs to be done to evaluate whether any implementation of a model for order flow segmentation could ever be resistant to sophisticated traders “tricking the system” by masking their use of smart contracts, trade size, and wallet credentials.

Competitive Dynamics Along the MEV Supply Chain

If bidders in the OFA do not take on liquidity risk, it’s fair to characterize them as taking a fraction of every $1 of MEV created by the user <> dapp relationship. For this reason, these businesses could be characterized by potential races to the bottom and the risk of vertical integration by the order flow originators they partner with. Products having already aggregated demand (order flow) and supply (market makers), like UniswapX, have gotten over the cold-start problem and may end up being the service providers to external dapps long-term beyond hooks and V4.

However, even if the bidders in an OFA take on no liquidity risk, OFAs could be characterized by potential network effects — each additional dapp participating in an OFA should create additional revenue per dapp through higher bids from searchers wanting to capture MEV from flows across dapps. Moreover, there are ways in which OFAs can develop moats by overcoming the cold start problem that creates a barrier to entry for new OFAs in the market. We can see this by going through an example:

Consider a new OFA attempting to enter the market. Because they’re new, their customer base and order flow will likely be insufficient to sustain searcher interest and activity. As a result, competition between searchers in the auction suffers, as do revenues distributed to dapps.

Alternatively, suppose a new OFA integrates with a dapp, capturing 20% of auctionable MEV in a batch through an auction because there are too few searchers for the auction to be competitive. To attract traders and solve this problem, the OFA might offer 0% fees to undercut existing OFAs. Yet, an established OFA, having already built demand, has attracted a sufficient number of solvers to ensure competitive auctions. If this OFA captures 90% of MEV in such auctions for order flow originators, even a 5% fee on bids would yield 5 times more revenue for originators than the feeless newcomer.

The competitive dynamics mentioned above are already driving OFAs to become hybrids between OFAs and exchanges. They are intentionally seeking to attract solvers who are willing to take on liquidity risk, not just extract arbitrage.

Additionally, in the search for higher profits and sustainable moats, OFAs are shifting their focus to attract order flow aggregators (i.e., dapps) over individual users. This is driven by two main factors. First, order flow from individual users sophisticated enough to use a specialized RPC isn’t that attractive to monetize. And second, targeting order flow aggregators is a more effective way of achieving the network effects and moat around the cold-start problem mentioned above.

While RPC solutions are technically compatible with dapp-centric OFAs, the fact that they add additional steps for users in requiring them to change their RPCs to a new network has limited their adoption, given that dapps tend to prioritize smooth UX over optimizing execution quality. To overcome this, emerging SDKs and APIs are leveraging both off and onchain infrastructure, such as those developed by the likes of Flood and Atlas, as well as enabling order flow aggregators to monetize their flow without requiring users to switch RPCs. Along with this, these SDKs and APIs are maintaining composability with a variety of wallet architectures that users may utilize in the process.

Order flow aggregators, including exchanges, are compelled through competition to more effectively match Parties A and B. The better they are at matching the orders from these parties, the smaller the market for capturable MEV for OFAs to compete in. The notion that solvers and the marketplaces for them provide a sustainable, value-added service presupposes that the order matching systems used by their customers need to generate MEV in the first place. That will be the case for some segment of crypto assets, but is not an assumption that can be uniformly applied across all.

MEV Redistribution — From Proposers to Order Flow Aggregators

In current and potentially future enshrined implementations of PBS, arbitrageurs (and the proposers that benefit from including their transactions in blocks) capture most of the LVR generated on Ethereum. As a result of OFAs’ technical progress mentioned above, the MEV landscape is likely to see a shift whereby these profits are captured upstream of proposers and programmatically distributed to dapps and wallets stakeholders as they see fit. The fact that orderflow auctions are currently happening through MEV-Boost means that the proceeds are currently going to the proposers. However, the new OFA architectures described above can instead redirect the proceeds of those auctions to the stakeholders of the underlying order flow aggregator, such as LPs.

Creating an efficient way for arbitrageurs to compete on paying LPs to arbitrage them could be a more effective way to improve the returns for LPs than to enable them to anticipate toxic flow from those arbitrageurs, and position themselves ahead of time. This is because, in attempting to rebalance their liquidity to avoid adverse selection, LPs need to outbid the arbitrageurs who are also competing for the inclusion in the block. Arbitrageurs are funding their bids out of LPs’ profits, meaning that the system managing LPs’ capital needs to bid as much as the losses they are trying to avoid.

While it’s possible to implement tools that redirect this value to order flow aggregators, it’s also possible to design exchanges to minimize LVR in the first place. By leveraging systems that are cleared at the prices bid by arbitrageurs, such as batch auctions, new onchain exchange designs could play a role in minimizing the amount of money paid to arbitrageurs and proposers to achieve price discovery, improving returns for passive liquidity providers.

Nevertheless, given the existence of offchain order books, the theoretical and empirical research points towards a reality where batch auctions, no matter how fast, will not be the means by which price discovery occurs onchain. This ultimately results in lower levels of liquidity, lower volume, and higher relative payouts to arbitrageurs relative to offchain alternatives (see more in Part 1 on batch auctions).

For this reason, it’s unclear whether or not these exchanges (despite being more trust-minimized) will compete in attracting passive LPs who would have a choice to lend their assets to market makers operating on offchain venues. If this is true, we should see the average losses to LVR decrease significantly over the next few years, which should in turn decrease profits to block builders and proposers downstream of the MEV supply chain.

The Future of Onchain Order Matching Venues

What does the future of onchain order matching look like?

At a high level, most AMMs and order books today are leveraging new scaling solutions to reduce the cost and speed of submitting and canceling trades for market makers and traders. Their main differentiators are the low-level details around the mechanisms by which asset holders can contribute to a market-making strategy — either directly into a pool or by lending them to a sophisticated market maker — and the mechanisms by which they bound the losses of the respective market maker.

As covered in Part I, in attempting to make AMMs sophisticated there is always a tradeoff between keeping them transparent enough for liquidity providers and trying to make them more competitive with their offchain counterparts. While onchain order books allow for more flexibility, they suffer from an inescapable tradeoff between MafiaEV and MonarchEV that challenges their trustworthiness and competitiveness relative to offchain exchanges. Following this, it’s worth considering how onchain order books and AMMs may differentiate themselves in the future.

While onchain order books face many challenges in competing with offchain alternatives across makers and traders, the risk/reward ratio for participating in them will persist long-term, given that they will host large markets that won’t be available to engage with in offchain environments.

Finding AMM Customers

As we start thinking about AMMs and onchain order books through the lens of the customers they serve, as opposed to the assets they host, we reveal the drawbacks of framing AMMs simply as hosts for long-tail assets.

Portfolio Management for Asset Issuers

AMMs can offer a differentiated value proposition to order books in the long term by serving asset issuers looking to manage their portfolios in 1) an optimized way against some critical part of their product or onchain organization and 2) an automated, low-overhead way.

We could see AMMs evolve to tailor their needs to asset issuers or large holders who want to optimize their portfolios for a different preference function than the classic market maker’s desire to take on liquidity risk efficiently. Our conclusions in Part I make it clear that AMMs are not necessarily the best exchange venues. However, they can be used as non-custodial asset management platforms, enabling asset issuers to design and engage in automated portfolio management strategies that could create a plethora of permissionless, sophisticated ETFs.

Here, LPs would commit to a predefined strategy defined by a custom bonding curve that deliberately incentivizes arbitrageurs to optimize the value of a specific portfolio that, for example, optimizes asset management by targeting a specific value for volatility. Taking this approach involves deliberately modeling and framing the expectations around LVR as a cost to achieving specific outcomes for LPs, who will likely be the asset issuers themselves wanting to choose their strategy flexibly.

In the future, onchain organizations will likely want to make their strategies composable with the smart contracts related to their core products in such a way that that optimizes their portfolio according to critical onchain metrics. In turn, onchain organizations could create systems that automatically adjust their financial position according to some element of their core business.

As a byproduct, they will create liquidity for arbitrageurs to pick up, much like how ETFs get arbitraged today. Moreover, these systems can coexist and be reinforced by the existence of onchain order books, which can be tailored to market makers’ needs in taking on liquidity risk effectively, while ensuring competition among them. This is the approach to AMM development taken by the Primitive team, who have done some fantastic R&D on this subject.

Comparative Advantages & Lazy Markets

Market-making is hard, and digital goods issuers need help to optimize it to the extent of gaining an edge in making direct profits from managing liquidity. In this context, AMMs represent infrastructure that expands the scope of what can be tradable on an exchange that would have otherwise traded P2P, and are poised to achieve another differentiated value proposition to onchain order books by reducing the time, cost, and complexity of market making.

As a result, a new generation of asset issuers could (and have already become) market creators. Companies issuing tokens with utility inherent to their business, such as brand loyalty points or skins for in-game characters, will increasingly leverage blockchain’s ability to frictionlessly and permissionlessly create markets around those goods. While some asset issuers will want to reduce their customer’s ability to trade their goods efficiently and actively, other digital goods issuers will want to integrate markets as a core part of the product or service offering.

For example, if an asset issuer wants to minimize their capital at risk while creating markets for their assets, an AMMs’ ability to constrain LPs to a public market-making strategy could be seen as a feature, not a bug. By attracting external LPs to commit to a specific, transparent, and tokenizable liquidity provision strategy, asset issuers could layer complimentary yield primitives to increase the per unit return for these external LPs while minimizing their capital at risk.

As highlighted in Part 1, the extent to which this is sustainable depends on the magnitude of these layered incentives and, more importantly, the market structure. Will faster, smarter market makers pick off the AMM’s stale prices? There are a couple of reasons why the market structure may not always play out such that competing trading venues exist.

First, it’s possible that in serving these particular asset issuers, other issuers and external LPs could engage with an AMM without worrying about being adversely selected from professional market makers on onchain order books to begin with. These professional market makers may not care to take on the risks of doing so for a less liquid or less volatile asset.

Another factor to consider is that asset issuers within this category could have natural monopolies on the market-making process. If the asset with liquidity in the AMM could be created for free by the asset producer, or if the asset's production came at a cost to the issuing entity or network, they may be able to monopolize the role of market-making said asset. Assuming the market maker wants to be lazy with their strategy, it’s possible to see a world where the issuing entity is much better positioned to take on the cost of market-making this asset on their balance sheet than other market makers.

The issuer would also have a lower hurdle rate because their market-making strategy would drive value to their core business. If the digital asset issuer wanted to, their comparative advantage at market making could effectively enshrine a passive AMM as the de facto exchange compared to an order book.

The asset issuers within this profile may even want to throw their assets into an offchain bonding curve that stitches together an offchain database with traditional banking rails. Whether or not these asset issuers will choose one or the other will depend on 1) the extent to which they want to access onchain liquidity and minimize the amount of money they add into the pool and 2) whether or not the driver of the asset’s utility feels the need to bootstrap trust from the public blockchain enforcing the rules of exchange.

However, the critical factor will come down to the user and developer experience. As capabilities for developers to abstract interactions with blockchains increase and onchain compliance tooling improves, onchain AMMs may likely be adopted in this market because they will provide a better experience for users and developers than stitching together internal systems with global banking.

If any of these paradigms materialize, and the pressures on applications to minimize LVR decrease, the impacts on MEV could be significant. If AMMs were adopted as vehicles for portfolio management, then LVR would be explicitly reframed and modeled as a cost desired by participants, who would pay for it in return for a specific, complex portfolio management strategy. If AMMs were adopted as lazy markets, LVR would become less likely to occur as liquidity consolidated on the AMM where the asset issuer already has a comparative advantage in making the market.

Conclusions on Exchange Design and MEV

Fundamentally, blockchains settle value for complicated ecosystems of intertwined financial markets. Each of these markets comprises different types of assets and, as a result, have different requirements. Traditional finance runs on T+2 settlement (soon to be T+1) because banks, hedge funds, market makers, etc., all need to reconcile trades across their databases at the end of the day and settle up. The only way to solve this is if everyone uses the same database, which can only be done with a distributed network. This is a perfect use case for public blockchains, which could also provide settlement-driven cost improvements as well as self-custody.

As blockchains become the settlement layer for an even wider world of assets, each market will likely evolve into a specialized ecosystem, catering to the unique needs of its underlying assets.

Exchanges require both the matching of orders and settlement. While we must recognize that there is no “one-size-fits-all” order matching system for all onchain assets, today’s onchain order matching venues face existential challenges that are likely to deter market makers and takers from adopting them over offchain alternatives.

Onchain, the battle will be between systems that enable market makers to reliably update their liquidity profile to frequently changing prices (dynamic AMMs and order books), systems that create competitive auctions for market makers to do it for them (OFA-exchange hybrids), and systems optimizing returns to passive LPs by clearing trades at uniform prices closest to the prices on offchain venues (batch auctions).

However, for dynamic AMMs and order books to compete, they will have to leverage a design that overcomes the critical challenge on Ethereum today — market makers need to outbid the arbitrageurs who are also competing for inclusion in the block. Moreover, as described in Part 1, the market design around onchain order books and some classes of AMMs are converging to allow for dynamic and programmatic inventory rebalancing, so the distinction between the two is likely to become less relevant over time. Still, the challenges across them highlighted in Part 1 will be large technical hurdles to overcome. In light of this, and given the fact that trust minimization as a feature in and of itself doesn’t represent a large market, it’s a strong possibility that trust-minimized solutions will lose market share to systems that enable lending those assets to market makers on offchain venues.

On the other hand, from our explorations into AMMs’ potential future customers, we can see another class of AMMs emerging tailored to asset issuers and holders with a different preference function than efficient market making. This includes AMMs characterized by custom bonding curves explicitly designed to pay arbitrageurs to manage their portfolio according to a specific goal or return profile. AMMs will also likely unlock value in creating new markets that wouldn’t have otherwise existed by pairing a lower overhead for creating more liquid digital goods markets with an experience for users and developers superior to web2.

Much of the MEV we are working towards mitigating as an industry stems from the existence of the same assets having liquidity both onchain and offchain, along with the arbitrage incentives this creates. While the development of complex new mechanisms built at the protocol layer must be created assuming that this persists, what Part 1 and Part 2 have attempted to show is that onchain exchanges and financial products will likely specialize in capturing markets very different than those that offchain exchanges do. As a result, we will likely see the liquidity for any given asset consolidate either onchain or offchain, which should naturally reduce much of the pressure MEV places on maintaining stable, credibly neutral blockchains.

As for the application layer, we should be wary of the potential impact that adoption of advanced onchain order matching systems might have on the incentives for specialization and centralization in the name of MEV extraction, potentially harming blockchain’s long-term viability as a successful settlement layer.

Special thanks to @soumyab8, @Autoparallel, @0xjepsen, @ThogardPvP, @fulminmaxi, @tylerinternet, @katiewav, @mountainwaterpi, @willkantaros, @AshAEgan, @DannySursock, and @dberenzon for their feedback and insights.

I also want to thank @cwgoes, @0xQuintus, and @gakonst whose insightful work has been referenced in this piece.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Revisiting the Exchange Landscape

As quickly as its discovery, MEV has cemented itself as an unshakeable part of public blockchains. With this, an acceptance of MEV as a part of the exchange process for crypto assets has followed and led to an explosion of new products and protocols focused on minimizing and mitigating its adverse effects. Whether these efforts are enshrined, aligned, offchain, or onchain, they’ve brought together teams of brilliant people and a massive capital infusion to improve the “MEV stack.” However, before mindlessly climbing the trees of Ethereum’s dark forest, I’d like to return to the forest floor and reexamine the ground we’re building on.

To get a glimpse of the future of exchange for crypto assets, we need to revisit the critical design challenges we face today. This piece represents the first of two articles exploring the path for exchange design in crypto, touching on the intersection between market microstructure and distributed systems to explore the current opportunities, challenges, and paths forward in creating the crypto asset exchanges of the future.

Where Are We Today?

Execution quality — how closely traders buy and sell an asset at a price reflective of the “true market price” — is generally a function of the liquidity offered by market makers. These market makers play a crucial role as exchange intermediaries and are compensated for taking on the risk of matching both sides over time. The exchange component responsible for interfacing with market makers and traders is an order-matching engine — effectively a digital system that pairs buy and sell orders according to specific rules.

Architecturally, an order-matching system’s effectiveness at matching users' trades is a function of the balance between the degrees of freedom and constraints they impose on the market makers submitting orders to the matching engine. Exchanges that design their order-matching systems in a way that enables market makers to update their quotes efficiently and reliably attract the most competing market makers and liquidity, leading to the highest execution quality for traders.

Despite having paved the way for permissionless trust-minimized exchange, the execution quality obtained through AMMs has yet to catch up with offchain alternatives. Even before incorporating front-running risks, user fees in DeFi today are substantially higher than in traditional finance (tradfi). The deepest Uniswap pool averages around 0.05%, or 5 basis points (bps), before gas, while the average markouts of retail orders across traditional exchanges average around 0.007%, or 0.7 bps. That’s almost 10x in terms of performance.

This poor quality of execution on AMMs stems from the fact that the environments they create don’t effectively attract sufficient, high-quality market makers to facilitate the exchange between buyers and sellers in the first place.

Over the past year especially, order flow aggregators ranging from exchanges like Uniswap to wallets like MetaMask have become more opinionated in designing order-matching systems that improve upon execution quality. Some of these aggregators are building out internal solutions to improve execution quality. At the same time, an array of traveling merchants are attempting to sell them software – from account abstraction SDKs to intent solutions to OFAs – that promise to solve their problems for them. The main force driving this trend is the inability of AMMs to provide users with high-quality trade execution.

A Brief Overview of the AMM Problem

These challenges of classic AMMs have already been discussed at length, so I’ll keep this short. On AMMs, the market makers (liquidity providers) who take on liquidity risk must openly declare their market-making strategies. These strategies, recorded on blockchains, specify how each trade influences the prices they offer for their assets. However, since blockchains update slowly, these market makers cannot adjust their prices quickly enough to avoid getting sniped by arbitrageurs. Consequently, market makers are disincentivized to participate in AMMs, and onchain order execution suffers.

One argument made in favor of AMMs is that passive market-making should be able to compete in terms of offering good prices to users compared to professional market-making because competition between professional tradfi market makers stems from latency wars to get to the top of the queue for execution priority. Consequently, the efforts made by professional market makers to win this competition don’t necessarily lead to better execution quality for traders.

If the only market maker in town for ETH were an AMM that takes money from liquidity providers, it wouldn’t be surprising to see more or less the same prices offered as if professional market makers were given that monopoly. Nevertheless, this correct observation around the realities of market making today can only be applied to good use in a vacuum, as the existence of faster, more expressive exchanges for professional market makers leads to passive liquidity providers on AMMs committing to a losing strategy anyway (through the means described earlier in A Brief Overview of the AMM Problem).

Hooks & TEEs – The Limits to Futuristic AMMs

Innovative designs, such as LVR (loss-versus-rebalancing)-reducing hooks through dynamic fees, are emerging to deal with adverse selection by helping market makers forecast future trades and reprice their inventory. Nevertheless, there remains a critical barrier to their adoption by market makers, which is that AMMs force market makers to publicly commit to a strategy, leading to frontrunning issues.

Generally, frontrunning occurs because adversaries know a market maker’s order before it gets executed. But when a market maker publicly commits to a market-making strategy, they tell the market the sequence of trades they will make given a set of inputs. This makes it possible for an adversary to front-run the market maker before they even place an order. The more sophisticated the market-making strategy, the greater the surface area for attack vectors exists.

For the same reason that it would suck to play poker while telling everyone what you’ll bid on, tradfi market makers like Citadel make their employees sign NDAs to keep their market-making strategies private. Keeping these strategies confidential is so essential to these firms that they don’t let their former employees work for a competitor for a year or two after they leave. They may even pay their full-time salary and benefits for them to just sit at home throughout this time.

Privacy solutions like TEEs (such as SGX) combined with tools such as Uniswap’s hooks could offer a way for AMMs to incorporate highly sophisticated strategies similar to those used by sophisticated high-frequency trading market makers while keeping these strategies hidden from the public. Despite these potential improvements, the challenge with this approach is that to stay competitive, sophisticated market makers would need to update their market-making algorithms constantly.

For example, a liquidity pool creator on UniswapV4 implementing their market-making strategy using SGX-based hooks would have to regularly change their algorithms running inside SGX to stay competitive as market conditions change. Moreover, the confidentiality of these algorithms doesn't guarantee protection against an adversary inferring and exploiting them, which would be another factor pushing liquidity managers to update their algorithms running inside SGX.

This creates a tradeoff: you can either have an exchange where liquidity providers can publicly verify the strategy before they put in money, or you can have customizability for adapting to changing market conditions. Consequently, the value proposition of AMMs as trust-minimized vehicles that define strict rules around managing assets would need to be reconsidered.

Does Your Order Matcher Know Too Much, Too Little, or Just Enough?

Because of these challenges, we’re seeing a partial move away from the AMM model and the re-emergence of order books and Request for Quote (RFQ) systems in crypto in an attempt to invite the virtuous cycle of market makers, liquidity, and high-quality execution that comes with it.

User-facing order flow aggregators, from wallets to dapps to exchanges, vary in their roles, incentives, and responsibilities. They make money by 1) providing the front-end that captures trader attention and trust or 2) creating the exchange that facilitates trades. Focusing on the latter, we need to understand the tradeoffs and challenges in pairing different architectures for order-matching with varying systems of markets and assets. The long-term competitiveness of exchanges depends on how effectively Party A and B can be matched, primarily downstream of the constraints (or lack thereof) placed upon the market-making intermediaries across these venues. One of the primary ways exchanges differ in this regard is in the choice of an RFQ or an order book.

Information Asymmetry in RFQs and Order Books

Stepping away from the blockchain context and looking at these systems in a vacuum, evidence across academia and industry overwhelmingly favors order books over RFQ systems for superior order execution. Order books enable efficient price discovery and reduced spreads for users by creating a much more dynamic equilibrium of supply and demand.

We can see this by breaking down the stakeholders in matching a trade:

1. The Buyer (Party A)

2. The Seller (Party B)

3. The Market Maker – the intermediary facilitating the interaction

In an order book system, price intentions are declared publicly by all parties involved. Users post their orders directly, and market makers compete to execute them. If Party A wants to buy 1 ETH for up to $10,000, and Party B wishes to sell 1 ETH for no less than $11,000, the spread between these two prices is visibly $1,000. With this information in the open, participants can make decisions based on real-time order depth and liquidity. If market makers or other participants place quotes that don't align with current market conditions, their orders will remain unfilled until they adjust.

Conversely, in an RFQ system, Parties A and B request a quote based on the amount of the asset they're eyeing without being able to specify their price limits. When the market maker receives this request, they are incentivized to widen their prices, anticipating that Parties A and B may tolerate some slippage.

In some situations, RFQs allow traders to better control the dissemination of information: what to show, who to show it to, and when to do so to limit adverse market reactions during large block trades in illiquid markets. In these situations, RFQs can be more effective than dark pools as they enable traders to outsource order execution to a professional market maker who takes a cut in return for ensuring the trader doesn’t screw up their order execution.

Generally, market makers in RFQs are positioned to make more than in order books because they don’t need to commit liquidity until Parties A and B specify how much of an asset they’d like. Without the pricing pressure from a transparent order book, Parties A and B will likely incur higher costs in an RFQ system, benefiting the market makers at the traders’ expense.

In light of these differences, we should be cautious to accept the narrative that the difference between order books and RFQs is insignificant enough to accept RFQs as the path forward for crypto exchange. Many participants in the crypto industry allude to the existence of zero fees in Robinhood’s RFQ system and the current dominance of RFQs in the bond market as evidence for their legitimacy. But we shouldn’t forget that these markets are characterized by uncompetitive behavior antithetical to crypto’s purpose.

Looking at Robinhood, for example, it is true that market makers like Citadel only get retail order flow if they are improving upon national best bid and offer (NBBO) across the numerous order book venues where equities are traded. However, if those Robinhood users collectively sent their trade on the NASDAQ, the spreads they would pay would decrease because Citadel would have to compete with everyone else.

We should not rely on evidence from oligopolistic industries to justify the existence of the order-matching systems they utilize. Opaque market structures, such as the bond market (controlled by JPM, Citi, and BofA), benefit people with more information. It should go without saying that when concentrated entities control a significant portion of a market, they have the information, leverage, and incentive to resist changes to that market’s structure that could affect their dominance.

Despite this, it’s clear that we’ve made strides towards improving how RFQs work as an industry. For example, RFQ systems in traditional markets are characterized by high-touch processes and inefficiency. A typical RFQ-based interaction between counterparties for derivatives on commodities will force Party A, Party B, and their market makers to set initial and variation margins. The financial contract is outlined through back and forths with legal contracts on legacy corporate ticket systems and manual, error-prone communications at expiry. This complex process, compounded by T+2 settlement periods, creates challenges across effective validation, reconciliation, and risk management, the negative results of which are all passed on to the end users. There’s a lot of room for improvement here that crypto can play a role in.

Within crypto RFQ development, we’ve seen some fast-paced improvements as well. In many crypto RFQ systems, market makers aren’t required to commit to liquidity to match against it beforehand and only have to improve against the AMM price from the previous block.

On the surface, it may seem that if the market maker decides not to improve that price, the worst possible price for the trader should be the same as if they had gone through an AMM directly. However, by looking at the example of a buy order, we can see that things aren’t so simple.

The market scenario in which an RFQ provider is likely to route the order to an AMM (instead of filling it themselves) is when the AMM price is lower than the offchain price. Why would an arbitrageur sell their assets to this RFQ trader at the price of the last block when they can sell it for more money on Binance? Consequently, the swapper gets routed to the AMM, where they must compete against specialized arbitrageurs to get to the top of a block. If they can get to the top of the block, the swapper can get the price they were originally quoted, but the swapper won’t win that battle.

In UniswapX’s RFQ implementation, the price at which user orders get filled is a function of competition between fillers which are not only able to, but forced to, compete based on the speed of on/offchain data ingestion, analysis, and order submission. Should a filler decide not to fill a trade they won after this offchain competition, the price they had previously committed to is used to parameterize an onchain Dutch auction. Getting back to the example of a buy order routed onchain because it’s unattractive to a filler (they can sell for more offchain), the swapper’s price will likely be better from a well-parameterized Dutch auction than from an RFQ where they have little chance of getting to the top of the block.

MafiaEV or MonarchEV? The Information Asymmetry Tradeoff in Onchain Order Books

So, if order books are better than RFQs, let’s put that onchain and call it a day! In both theory and practice, it’s not that simple.

An onchain order book is defined as a platform where:

1. Users post orders onchain

2. Order execution is prioritized according to the orders with the best prices and earliest submission times

3. Consensus or a leader selection algorithm is utilized for censorship resistance

There are some fantastic attempts at creating performant onchain order books designed to be competitive with their offchain counterparts. This is often achieved through operating in environments with cheap compute to reduce the costs of placing orders onchain and achieve faster block times, both of which reduce LVR. Even if these characteristics can be obtained such that onchain order books can compete with onchain AMMs, critical challenges still emerge from inherent blockchain constraints that challenge onchain order books vying for liquidity and volume from those available offchain.

Onchain order books don’t have a uniform architecture and will all look different depending on the chain they’re built on. But in all cases, the basic flow is similar — a retail user submits an order, the order goes through the consensus mechanism where the sequence of orders is decided, and then the order appears onchain.

"Goes through the consensus mechanism" is where all of the games can be played that put onchain order books at a structural disadvantage to competing order flow aggregators using offchain order books.

Onchain order books can choose one of two systems to determine the state of an order book:

1. Multiple leaders provide input into the sequence of orders

2. A single leader decides the sequence of orders

Either way, onchain order books will encounter one type of MEV – LVR, resulting from multiple leaders, or tx reordering, resulting from a single leader. These different types of MEV are well suited to @sxysun1’s framing in this talk, which classified MafiaEV and MonarchEV as two of three different types of MEV. MafiaEV denotes the extractable value achieved through coordinated strategies among network participants exploiting information asymmetries. In contrast, MonarchEV encapsulates the value extractable through centralized, authoritative control within blockchain protocols, particularly by entities with decisive power over transaction sequencing and state finalization, such as block builders.

MafiaEV: Designs with Multiple Leaders

In blockchain-based order book systems using multi-leader consensus, latency arises from three key technical aspects: conflict resolution, network delays, and transaction processing. Multiple leaders processing transactions simultaneously leads to conflicts, requiring time-consuming consensus rounds. Geographically dispersed nodes introduce significant network latency. Additionally, each node's independent validation and ledger state replications add processing time.

Regardless of the precise details of the consensus mechanism, onchain order books that multiple leaders update must deal with this MafiaEV stemming from adversaries taking advantage of market makers' inability to update how they distribute liquidity across the order book quickly. While the absolute latency a market maker experiences when interacting with an order book is important, it’s critical to emphasize that an exchange’s survival depends more on its latency relative to other exchanges.

Suppose the fastest onchain order-matching engine Y takes 10 seconds to trade, but offchain order book X takes half a second. In that case, price discovery will happen offchain, and all arbitrage will be from the offchain order book to the other exchanges. Suppose the onchain order book Y lowers latency to half a second, but Coinbase takes 10 milliseconds. In that case, onchain order book Y’s prices will be stale, as will its liquidity and user uptake.

Block times, costs, and the latency to submit and cancel quotes can certainly be reduced, and consensus and network layers can be innovated and push the boundaries such that the relative latency between onchain order books gets close to their offchain counterparts.

Still, we must also consider latency guarantees across time and order type. On any order book, if order cancellations are slower than order submissions (or vice versa), market makers lack guarantees around how they will be able to handle various market conditions. While the latencies they bear in submitting orders might be suitable, market makers can’t rely on that information to infer how quickly they could cancel those quotes in the future if they become stale. In an onchain order book, consensus mechanisms’ unpredictable latency magnifies this problem.

On top of this, participants depend on the block builder to not order trade requests in such a way that is highly beneficial to them at some point. In fact, should an onchain order book attract significant volume, block producers specialize in capturing the generated MEV. This could have a centralizing force on the underlying blockchain, potentially harming its value proposition as a credibly neutral settlement layer.

MonarchEV: Designs with a Single Leader

Onchain exchanges will likely remove consensus from as many parts of the order-matching process to combat this latency issue. One of the simplest solutions to this issue is to grant a single leader the ability to decide the order sequence.

These single-leader venues experience MonarchEV. In this context, MonarchEV originates from temporary monopolies granted to single market makers in permissionless environments, allowing them to reorder transactions.

Teams like dYdX aim to counter this by requiring market makers to put up collateral before granting them these monopolies, keeping them in check. However, this requirement increases market makers’ necessary capital costs and, more importantly, increases the venue’s risks of mispricing the collateral required to keep block producers in check. This ultimately creates scalability issues for the venue as asset variety, volume, and volatility grow.

How an exchange sets the slashing costs also becomes a significant challenge. Slash too little, and manipulation becomes profitable, even considering the slashed stake. Slash too much, and more capital is put at risk, making benign failures (i.e., misconfiguration) more costly. If an exchange wants to figure out the "right" amount to slash, they would have to do something akin to an auction, and then they’re back to latency problems.

Implementing SGX or threshold encryption can also constrain the power the monopolist(s) holds over transaction ordering. Still, these implementations can only guarantee that, for a given set of transactions, the leader commits to not reordering them, inserting their own, etc. However, they cannot guarantee that every transaction is included fairly, so it still doesn't mitigate the problem – it just invalidates a subset of attacks.

Rollup-ing the Exchange

One way to work around the difficulties that front-running poses is to design a system that holds a check on the single operator of the order-matching system. This could be achieved by forcing the operator to commit to rules around issuing order receipts to users upon trade submission and posting trade history to a data availability (DA) layer.

An exciting approach worth highlighting here is to turn the exchange into a rollup like LayerN has done. By rollup-ing the exchange, the order-matching system could execute offchain while keeping its operator in check through a verifiable proof on a DA layer. At a high level, this system could guarantee market participants that if the sequencer orders trade in such a way that violates the rules of the matching engine, traders could submit a fraud-proof and rely on a DA layer filled with trade history to do so. This also means that the exchange’s throughput would be limited to the performance of the underlying DA layer.

Combined with a leader selection algorithm (automated or governance-based) that can replace a censoring sequencer, this exchange model could maintain the censorship resistance required for permissionless market creation while freeing itself from the constraints of a consensus-based orderbook. Moreover, exchange rollups could improve the security model for mitigating censorship and front-running from an honest majority assumption to an honest minority assumption through fraud or validity proofs.

However, minor latency manipulations by the exchange operator would be undetectable by fraud proofs. Consequently, affected market makers wouldn’t be able to discern if the latency issues they are experiencing result from uniformly distributed network issues or targeted actions on the part of a misaligned sequencer. While all participants experience some variance in latency, a consistent disadvantage of a few milliseconds can critically impact a market maker's survival. For this reason, exchanges underpinned by these single-sequencer order-matching systems might struggle to gain adoption from market makers who expect regulations and reputation to give them guarantees around these risks.

It’s important to note that, unfortunately, SGX doesn’t solve this. Yes, if the bits of information containing orders could be sent directly to the order-matching engine running inside SGX, market participants could get guarantees that latencies were applied impartially. However, these packets of information that contain trade orders don’t go directly from users to the enclave. They rely on some untrusted computer, like a router, to deal with communication between them. For this reason, it’s always possible for the sequencer to manipulate the timing around when orders are seen by the order-matching engine running in an SGX enclave.

What About Auctions?

One of the exciting solutions to the MafiaEV vs. MonarchEV tradeoff that occurs in onchain order books is to combine a batch auction with sufficiently low latency, solving for MafiaEV, which uses encryption, solving MonarchEV.

In contrast to continuous trading systems like orderbooks, where transactions are processed sequentially and immediately as they occur, batch auctions operate by accumulating a series of buy and sell orders within a predetermined time frame. At the conclusion of this interval, the collected orders are executed simultaneously at the same clearing price.

Noteworthy developments have been made in enhancing the efficiency of batch auctions through privacy. For instance, In Penumbra’s sealed-bid batch auction implementation, orders are first encrypted, and block builders commit to including these encrypted orders within a block. Only then are these orders decrypted and executed through a batch auction.

However, batch auctions struggle with real-time price discovery, largely due to the time needed to integrate new market information. This delay, inherent in their interval-based execution, contrasts with the continuous, immediate processing of onchain order books, which better suit high-frequency traders' need for quick liquidity injection.

When market consensus on an asset's value changes rapidly, batch auctions can't keep up, leading to a mismatch between the real-time market valuation and the batch auction price before the next interval begins. High-frequency traders (HFTs) who, in part, capitalize on short-term price differences, find this delay unappealing. As a result, they shy away from these platforms, which potentially leads to reduced liquidity and slows integration of new price information into the market. While the positive outcome is that latency arbitrage becomes less profitable, traders looking for more immediate prices also shy away from placing orders in these batch auctions.

This phenomenon is backed up by empirical research on transitioning from batch to continuous trading conducted on the Taiwan Stock Exchange. The study found that continuous trading significantly enhanced price efficiency for medium and small cap stocks, indicating the importance of the market's ability to quickly integrate new information. Notably, this increase in trading activity was not attributable to latency arbitrage, suggesting that the improvements in price efficiency resulted from the incorporation of continuous trading.

Despite this, the questions around batch auctions' relative merits and drawbacks seem far from reaching a close, at least in academia. While batch auctions might only take over tradfi markets if they receive support in the form of regulatory pressures against HFTs, they could become an integral part of the solution for onchain exchange due to their attractive properties in eliminating sandwich attacks and reducing gas costs.

Wen Part II?

This article aimed to lay out the challenges and opportunities around crypto exchange and MEV that we are currently presented with, including the shortcomings of AMMs, the arrival of order books and RFQs in crypto, and the design space for their implementation off and onchain. When looking at onchain order books, these tradeoffs can be framed through the lens of MafiaEV and MonarchEV. At a higher level, it seems that any attempts to make systems for onchain exchange more sophisticated lead us to a battle between efficiency and integrity.

In Part II, we further explore the opportunities, challenges, and implications of rapidly emerging primitives across cryptography and systems design, from intents to OFAs and net-new financial products. From this point, we’ll hopefully be able to paint a clearer picture of how the future pipelines of onchain value might shape up.

We’re excited to see teams tackle these difficult design challenges. If you’re working at the cutting edge of these open questions, please reach out!

Read Part II.

Special thanks to @soumyab8, @Autoparallel, @0xjepsen, @tylerinternet, @katiewav, @mountainwaterpi, @willkantaros, @AshAEgan, and @DannySursock for their feedback and insights on Part I, and the many more who also helped with Part II (coming soon).

I also want to shout out the @thebellcurvepod and @MikeIppolito_, @danrobinson, and @hasufl. Grateful to be getting a front seat in learning from thoughtful conversations between all the fantastic hosts and guests on the podcast throughout the seasons.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

I recently wrote a post on my personal blog about the state of the music industry, spanning the economics of streaming to social discovery to the role of media publications in elevating artists. The general thesis was that the music ecosystem is extremely complex and esoteric, and we’re overdue for new models and new platforms.

The original post was not framed through the lens of solution-finding or crypto, but my research made me more excited and confident about the opportunity for crypto to enable next-generation social platforms and creator tooling.

So far, explorations in onchain social and creator platforms emerge from the following guiding question: “Should I build for web2 users or should I build for crypto-native users?” Consequently, a tension between two approaches arises–familiar and arguably skeuomorphic platforms like Lens or Farcaster vs. more experimental, “crypto-native” platforms like Friend.tech or Song.tech that are often explicitly financially speculative.

Rather than argue that one approach is better than the other, I’d like to explore the potential of onchain social through a different first principles question: “What types of platforms uniquely leverage the power of being onchain?”

To properly understand the “power of being onchain,” we must return to what I’ll call “The Principles of Onchainness,” drawing heavily from Jacob Horne’s seminal piece “Onchain,” and examine how onchain mechanisms can unlock net-new platforms and features for users, builders, and creators.

Permissionless + Composable

The permissionless nature of blockchains means anyone can participate onchain without requiring approval or permission from any central authority or platform. The public and open anatomy of the network also means that everyone has access to all existing onchain data and, consequently, onchain users.

Composability closely follows being permissionless. Permissionless composability is a strong value proposition for builders who can now build apps and interfaces on top of existing data and protocols–tapping natively into social graphs, media, assets, etc. without necessarily having to bootstrap data or users from scratch or acquire permissions from any closed ecosystem or API economy. An important aside is the ability to do this while maintaining user privacy and integrity through cryptographic mechanisms such as ZK.

From Chris Dixon:

“When Twitter changed their API in 2011 or so, there was a big wave of startups — including a lot of my friends — who built Twitter startups. That was a thing in 2009 and 2010, with Tweety, TweetDeck, and all sorts of API services. [...] at some point [Twitter] decided, ‘Hey, we need to control. We are going to have client software, have an ad-based model, and change the API,’ and that whole industry died. Same thing happened with the Facebook platform.”

Composability breeds developer innovation, but what does “permissionless” or “composable” actually mean to the average user?

Users don’t join platforms out of principle, they join for some sort of utility (whether social, economic, emotional, etc.). Web2 consumer social platforms, for obvious reasons, make it nearly frictionless for users to join, making the immediate utility of a “permissionless” platform perhaps non-obvious. Composability is only valuable for users insofar as if there’s another platform or app users want to use, they don’t necessarily have to start building their social graphs or assets or even log history from zero.

From Eugene Wei when initially reviewing this post:

“The example I like to use is that every app that wants social features forces people to go through a friend discovery process and to build their network from scratch (Netflix and other social viewing experiences, for example). But scaled services like that probably already have enough nodes for a scaled graph, it's just the friction to create the graphs again that prevents us from experiencing what a social version of that service would be.”

At its weakest, onchain composability can eliminate the possibility for user data to be locked into a single platform. Threads is an example of unsuccessful composability, where the platform was able to attract signups at mind-blowing speed due to its native integration with Instagram, but quickly lost the majority of its daily active users because the Threads experience was not considered fun or differentiated.

At its strongest, symbiotic platform relationships can actually incentivize users to traverse platforms. When I think about a rich digital social ecosystem, the first example that always comes to mind is Neopets. Imagine a world in which every spot on the map–whether a game, store, quest, battle, etc.–was built permissionlessly by a different team, each not knowing the other, leveraging the same universal player data and identity. These types of open, interoperable world-building experiences are not possible in web2 where data access is constrained to a single team or permissioned access.

Ownership + Autonomy

Ownership is probably the most straightforward value proposition of onchainness. In web3, whatever you own onchain, you own everywhere, forever.

However, knowing when ownership is a must-have versus a nice-to-have is crucial when designing a consumer experience–and the answer is more complicated than one might think. While the value of onchain ownership is usually discussed in the context of owning assets themselves, this neglects a far bigger picture.

For the sake of this post, we can think about ownership in two different, often interlocking, forms: owning assets and owning distribution.

Owning an asset is important when it has persistent, long-term value–money, art, “real-world” assets or representations of such. These assets have clear utility, aesthetic value, and/or durable markets. Owning assets onchain is usually most important in the context of security, ensuring that these high-value assets are secure, self-custodied, and can’t be siphoned away from the user.

Owning distribution is important when the assets themselves either aren’t objectively valuable or are only valuable within a specific context. To extend the earlier Instagram/Threads example, if Instagram suddenly shut down tomorrow, all my posts would disappear. How much does this matter? I may care about my posts emotionally, but I could very easily save the pictures in other places (like a hard drive). However, losing Instagram as a distribution channel for my pictures means Iosing 1) the social context in which my pictures became valuable, 2) the social graph/audience I crafted from the platform, and 3) history/proof that I was ever on the platform at all. This is a similar case for Twitter. If I were really that attached to my tweets, I could just screenshot them or write them down. I can even still download all my Twitter data (though this feature is permissioned/could be killed at any time). However, losing access to Twitter as a platform would mean losing access to my audience, my distribution, and the years of history and social capital I’ve accumulated from using the platform.

Owning distribution can be just as, if not more, valuable than owning an inherently valuable asset. Distribution → social capital → economic capital is often a powerful revenue flywheel for creators, with the monetization features either baked into the platform directly (YouTube) or indirectly (Instagram + brand partnerships).

As a holistic, real-life example, take this Reddit post from a DistroKid user. For level-setting, DistroKid is a creator platform that allows artists to upload their music to streaming platforms like Spotify and Apple Music and manage their streams earnings across those platforms.

In this example, it’s important to recognize that the music files themselves aren’t the issue, it’s the distribution and revenue tied to the uploads, facilitated by DistroKid. Without DistroKid, the artist has little clarity or control over their uploads, their metrics on streaming platforms, and the revenue they’ve accrued and are due. There are other platforms similar to DistroKid like TuneCore and CD Baby that artists can switch to, but interoperating between these platforms is a whole new challenge in itself. This isn’t to say that platforms like DistroKid are inherently bad, power structures simply exist in a web2 ecosystem that give platforms indiscriminate control.

In an onchain ecosystem, while the artist may still interact with a DistroKid-like interface, they’d have much more autonomy. They would be able to directly check and manage their assets onchain, the value accrued to those assets could flow programmatically through the protocol, and the artist can directly view and interact with their onchain audience. DistroKid wouldn’t be able to arbitrarily change the mechanics of the underlying protocol without being held publicly accountable–platforms are more incentivized to maintain credible neutrality.

The amount of autonomy an “onchain DistroKid” exactly enables still depends on many core design decisions, but at the very least the mechanics of the platform would be much more transparent to its users. If the interface becomes untrustworthy or a “bad actor,” the public is able to keep entities accountable and/or another interface could directly compete while leveraging the same data.

A few onchain-specific features to highlight that also strengthen ownership:

Protocols: Protocols enable distribution channels, social graphs, and transaction history to be permanent and programmable. Even if a platform shuts down, the underlying protocol keeps all the data, assets, and rules intact allowing other interfaces to easily emerge.

Markets: Onchainness makes it simple to permissionlessly create markets around assets, thereby making those assets more attractive to own. We’ve seen examples of consumer social platforms creating markets for their native assets in attempts to make them more durable and objectively valuable. For example, Friend.tech keys only have utility in the context of Friend.tech as a platform, but the markets around keys make them more valuable as singular assets.

Provenance

To date, provenance has been one of the more legible and compelling value propositions of being onchain to artists and creators.

In a world of 1) digital nativity and transmission, 2) permissionless remixing, and 3) synthetically-generated content, tracking authorship and attribution becomes a significant challenge.

Josh Benaron, the founder of Bundlr, describes this problem space well:

“In the Web2 era, internet users gained the ability to easily ‘write’ to the internet. While this empowered users in new ways, the explosion of content came with downsides: murky attribution, lack of metadata to describe the content, unverifiable authorship, and weak assurances around who created content and when. AI has only magnified these downsides with its ability to facilitate counterfeit and derived content at a scale never before imagined. This trajectory sets us on a collision course where inaction will have grave consequences.”

A few real-world examples:

1. In an interview with the CEO of YouTube, creators Colin and Samir (1.4M subscribers) share that the most popular Colin and Samir videos on YouTube are short clips of their original videos posted by third parties. Colin and Samir aren’t able to capture any direct view count metrics or AdSense revenue from these videos, even though they created the original content. YouTube has a system called Content ID and a newer tool called Remix that were created to better attribute value to original creators, but the systems are not highly sensitive or yet used, respectively. Colin and Samir have resigned to the fact that the shorts (hopefully) give them free distribution.

2. TikTok has created an entirely new way to discover, distribute, and trial songs. Artists often have to scramble to funnel the free distribution into ways that actually capture monetary value. For example, in 2020 Aly and AJ released a new, explicit version of their 2007 song “​​Potential Breakup Song” after the original song went viral on TikTok. According to an article from Variety, “the TikTok trend associated with ‘Potential Breakup Song’ led to the creation of about two million videos, with the two most popular posts collecting over 10 million likes each.” The re-released version has amassed over 50M streams on Spotify, as compared to the original which has about 150M Spotify streams.

3. Ghostwriter is a music artist known for creating songs with AI-generated vocals from artists like Drake and the Weeknd, gaining 15M+ views on TikTok for their song “Heart on My Sleeve.” Neither Drake nor the Weeknd were directly compensated for the usage of their vocals, and neither have publicly commented on Ghostwriter’s music.

Derivatives have been valuable art forms since the beginning of art itself (remixing, sampling, interpolation, etc.), and attempting to terminate those artforms is a losing battle. However, as content becomes digitally-native and the technical barrier to creating and distributing digital-native content plummets to zero, the context collapse around remixed content becomes an existential issue. Authorship and authenticity are crucial to creators and artists who want to capture the full value of their work, whether as social capital or economic capital.

Verifiable provenance makes the transfer of content between people and platforms seamless.

The Onchain Era - ZORA ZINE

Crypto is dead and few will mourn it. Now that the bubble has burst, the truly revolutionary work can begin

https://zine.zora.co

“By being onchain, information gets a provable provenance that establishes its origins, its collaborators and supporters, and its context in a way that’s permanently and publicly accessible without requiring any institutional stamp of approval or maintenance. This system of independent verification and dissemination of knowledge—whether it be ideas, artwork, or personal information—is truly revolutionary.”

- Yancey Strickler, Co-Founder of Kickstarter, Co-Founder of Metalabel

As technical barriers fall and tools/platforms launch unique ways for users to leverage AI, the social inclination to interact with these applications increases. However, today’s younger users are also both more socially conscious and emotionally attached to artists, meaning they are simultaneously less willing to engage with platforms and tools perceived as extractive.

For derivative, remixed, or synthetically-generated content to truly reach its mass potential, value from these platforms must tangibly flow back to artists. Value doesn’t have to be explicitly financial, but cultural capital often indirectly leads to financial capital, though the relationship is quite blurred. Being able to at least map these flows more distinctly becomes even more of a priority.

Coordination

Being able to track information between people is just as much a coordination issue as it is a source issue. Programmatic provenance increases the speed and granularity at which information and data can move.

In abstract terms, this means that 1) the upper limit of people that can be coordinated simultaneously disappears, 2) the number of layers added onto a piece of media or information (remixing) without losing context approaches infinity, and 3) the atomic unit of information or value that can be transferred between people or platforms approaches zero.

What does this actually unlock? I’ve shared this before when writing about the type of future I’m excited to watch crypto enable:

• Communities to self-organize and govern

• Communities to self-custody, coordinate, and deploy capital

• Users to own and selectively share/their identity/data

Strong coordination mechanisms create environments in which people are actually rewarded for collaboration, rather than zero-sum mechanisms. We see this clearly with the rise of freelance workers and experimentation around DAOs. From a 2021 post I originally wrote in collaboration with Station called “A New Genre of Work”:

“In his seminal 1937 essay, The Nature of the Firm, economist Ronald Coase explained why companies exist—to reduce the friction and transaction costs of contracting individual work on the free market. While perhaps a truth of the past, traditional corporations with bloated management and poor incentives for employees and users no longer effectively create value. Rather than focusing on practicing the craft at hand, tremendous energy is wasted optimizing for zero-sum games of equity vesting, salary negotiation, and organizational politics. [...] It’s clear that the most thorny problems facing humanity today—climate crisis, cybersecurity, income inequality to name a few—will not be solved by one corporation or one individual. These problems need to be addressed with the scale and efficiency of a corporation, without compromising on individual autonomy, creativity, and ownership. They require fluid and multidisciplinary collaboration that transcends the borders of institutions, from corporations to nation-states.”

With onchain mechanisms, networks can be orchestrated programmably, and value can stream atomically to contributors. Inherently decentralized fields like AI and media, that are rapidly approaching and inevitable, can be better approached from positions of encouragement, rather than fear.

The Power of Onchain: 1 + 1 = 3

Permissionless, Composable, Ownership, Autonomy, Provenance, Coordination

It’s worth examining how these lofty principles can actually be stitched together to unlock new, rich experiences.

The permissionless and composable principles of blockchains realistically benefit builders and developers more than they do end users. Think of this similarly to a perhaps more self-evident statement that users don’t care if the AI they’re interacting with was built with an open source or private model. However, the permissionless and composable nature of blockchain data encourages more rapid and open developer experimentation, which one must believe will ultimately result in richer end experiences for users.

As developer bases become more distributed and users traverse platforms more freely, value flows become much more complex and difficult to orchestrate. For developers, contributions become more granular and shared between more parties. For users, identity and inventory become increasingly fragmented. Distributed parties and information are one of the key trade offs of decentralization. In order to resolve these issues, data must adopt stronger provenance and coordination features–to both ensure verifiability and accountability of information and seamlessly orchestrate information and value flows between parties across these decentralized bases.

While the means by which permissionless, composability, provenance, and coordination affect users are more nuanced, the value of ownership and autonomy appear more clear.

In web3, whatever you own onchain, you own everywhere, forever.

However, as I outlined earlier, ownership and autonomy are contextually valuable.

Ownership of private and identifying information like PII and highly-secure assets are always important to users, but consumer social platforms are most powerful when they create new value substrates for users to own. I cannot stress enough that value does not have to be explicitly financial.

To steal a few excerpts from Eugene’s canon piece “Status as a Service (StaaS)”:

“Let's begin with two principles: People are status-seeking monkeys. People seek out the most efficient path to maximizing social capital”

“We have no such methods for measuring the values and movement of social capital, at least not with anywhere near the accuracy or precision. [...] Despite this, most of the social media networks we study generate much more social capital than actual financial capital [...] And, while we may not be able to quantify social capital, as highly attuned social creatures, we can feel it. Social capital is, in many ways, a leading indicator of financial capital, and so its nature bears greater scrutiny. Not only is it good investment or business practice, but analyzing social capital dynamics can help to explain all sorts of online behavior that would otherwise seem irrational. [...] What ties many of these explanations together is social capital theory, and how we analyze social networks should include a study of a social network's accumulation of social capital assets and the nature and structure of its status games. In other words, how do such companies capitalize, either consciously or not, on the fact that people are status-seeking monkeys, always trying to seek more of it in the most efficient way possible?”

My interpretation of these specific quotes is that humans are generally very good at finding ways to increase and leverage their own social capital. The opportunity for new consumer social platforms therein lies in either 1) creating assets that are newly valuable, 2) creating connections that are newly valuable, or 3) enabling new value chains for existing assets or connections.

Tying this more meta-point to crypto, crypto is uniquely good at 1) accelerating creation through permissionlessness and composability, 2) tracing connections and flows between people and platforms with strong provenance, and 3) enabling the creation of new markets.

The Future of Social

What has hopefully become clear through this exploration, is that creating new, sustainable value mechanisms is much more complex and nuanced than assigning a speculative dollar value to an asset. This is not all to say that financial capital is not an important lever for social experiences. Financial speculation and gambling can be very fun and very lucrative. Wealth will likely continue to be *the* status marker for a very long time. Creators should be able to directly channel social capital into financial capital. Creating new, sustainable value mechanisms also requires reaching beyond theoretical principles like “permissionless” and “composable” as means to reach users.

History tells us that the most paradigm-shifting markers for value originate not from dollar value or grandstanding, but from new experiences demanded by concentrated movements of people (often underserved) with new interests and new tastes.

I’m most excited about a future that uniquely enables new means of creative expression and value distribution. This future is already rapidly approaching. Creation is easier, faster, and more distributed than ever before, thanks to the globalizing force of the internet and more emergent technologies like AI, and people’s identities, connections, tastes, and interests are also increasingly more complex and distributed.

Onchain mechanisms are uniquely positioned to unlock collaboration, power experimentation, and ultimately track and create value for builders, creators, and users–all while maintaining user sovereignty.

At the beginning of this post, I mentioned a post on my personal blog about the state of the music industry. At the end of that original post, I proposed a few new models for tools for artists and social platforms more broadly that I would love to see come to life:

“Platforms that spotlight a track’s social provenance, highlighting its often unsung songwriters and producers, and create rich attribution graphs and experiences around those social connections.

Platforms that spotlight a track’s data provenance, gamifying remixing and experimental AI creation while preserving attribution and data provenance.

Decentralized media and curatorial platforms, where media is sustainably co-created and co-distributed by the local scenes and communities that create niche scenes and sounds.

Digiphysical platforms and experiences that sit at the interplay of live and digital.”

I’m more confident than I ever have been that the rails of these next-generation platforms and many more will live onchain.

Thank you to Eugene Wei, Reggie James, Andy Weissman, Seyi Taylor, and Ruby Justice Thelot for thoughtful and critical review and feedback on drafts of this post.

Thank you to Jacob Horne, Yancey Strickler, and Josh Benaron whose pieces I referenced in this post.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

Platform Shift, Meet Platform Shift…

The world is shaped by periods in which extraordinary upheavals in technology or infrastructure coincide, unleashing a generational step function in innovation. Think telegraphs and railroads, fiber-optic cables and the internet, or mobile phones and 3G.

Our belief is that the intersection of two groundbreaking frontiers – Artificial Intelligence (AI) and blockchains – represents a similarly transformative moment.

Three important pillars underpin this thesis:

Blockchains Can Offer a Superior Design Space

AI’s high-impact areas are numerous but can broadly be summarized into three main categories:

In particular, Generative AI introduces unique challenges and opportunities that we believe play to the strengths of blockchain technology.

To understand why, it’s important to consider the core inputs that drive the evolution of intelligent systems. Machine Learning (ML) is fundamentally powered by data (quantity but increasingly quality), feedback mechanisms, and compute power.

Dominant players in AI/ML like OpenAI (backed by Microsoft) and Anthropic (with Google and Amazon) are already consolidating resources and building walls around their models and data. But despite early advantages in compute, data, and distribution, this approach risks stifling momentum by fragmenting the collaborative development cycles that birthed the industry in the first place.

Offering a viable counter to this are blockchains like Ethereum, which have emerged as credibly neutral systems of data and compute fueling open-source innovation. Blockchains already underpin a range of digitally native primitives that are well positioned to serve critical roles in a world increasingly shaped by generative AI.

Our belief is that there is a major opportunity for blockchains to become the primary domain upon which open-source research & development in AI compounds.

The State of Today’s Market

A tremendous amount has already been invested in this year’s generative AI frenzy across core infrastructure, the model layer, and even user-facing applications like chatbots, customer support, and coding assistants. Despite that, where (and to whom) value accrues across the traditional stack in the long run isn’t obvious.

In the current paradigm, AI risks being a centralizing force that extends the dominance of web2 market leaders. At the infrastructure and model layers in particular, the name of the game is scale – in hardware and capital resources, access to data, distribution channels, and unique partnerships.

Many of these players – from cloud service providers like AWS to hardware manufacturers like Nvidia to longstanding heavyweights like Microsoft – are going full-stack, whether vertically via M&A or through proprietary partnerships.

The titans at the top are competing for scale and accuracy at the margin, but the market for ultra-expensive, high-accuracy enterprise API models may well be constrained by economics, emerging performance parity of open-source, or even a trend towards lower-latency workload needs.

Meanwhile, a large portion of the middle market is already seeing a commoditization in offerings resembling a collection of ‘OpenAI API wrappers’ with indistinguishable albeit sufficient functionality.

Building on Open-Source Momentum

Open-source datasets for pretraining, training and finetuning, as well as freely accessible foundational models and tools, are already encouraging enterprises of all sizes to get creative with open systems & tooling directly.

A leaked paper from Google outlined just how quickly the gap is closing between the closed and open- source worlds. Notably, 96% of today’s code bases already use open-source software, with the trend particularly evident across Big Data, AI, and machine learning.

Meanwhile, the cloud services oligopoly may be ripe for disruption anyway.

Historically, the big three of AWS, Google Cloud, and Azure have come to own the market by layering on tools and services to entrench themselves deep within the enterprise stack. This dominance has led to a number of challenges for companies, ranging from restrictive operational dependence to excessive costs associated with cloud infrastructure, especially given the premium charged by the major providers.

The pressure on incumbent companies to restructure operating expenses, coupled with a desire to experiment with and integrate the growing range of open-source AI, will create a window to reimagine the stack with decentralized alternatives.

The emerging intersection of open-source AI and blockchain technology therefore presents an extraordinary domain for experimentation and investment.

Crypto x AI: A Mutually Valuable Relationship

We’re profoundly excited by the potential symbiosis between AI and blockchains.

Crypto middleware can drastically improve inputs across the supply side of AI by establishing efficient markets for compute and data (provision, labeling, or finetuning), as well as tools for attestation or privacy.

In turn, decentralized applications and protocols will reach new heights by ingesting the fruits of that labor.

Undeniably, crypto has come a long way, but protocols and applications still suffer from tooling and user interfaces that remain unintuitive for mainstream users. Likewise, smart contracts themselves can be constricting, both in terms of manual workload demands for developers, but also in overall functional fluidity.

Web3 developers are a remarkably productive bunch. A peak of just ~7.5K full-time developers have built a multi-trillion-dollar industry. Coding assistants and DevOps augmented by ML promise to supercharge existing efforts, while no-code tooling is rapidly empowering a new class of builders.

As ML capabilities get integrated into smart contracts and brought onchain, developers will be able to design more seamless and expressive user experiences and, eventually, net-new killer apps. That step function improvement in the onchain experience will attract a new – and likely much larger – audience, catalyzing an important adoption-feedback flywheel.

Generative AI may prove to be crypto’s missing link, transforming UI/UX and catalyzing a major wave of renewed technical development. In turn, blockchain technology will harness, contextualize, and accelerate AI’s potential.

‍Using Blockchains to Build a Better Market for Data

Data is ML’s Foundational Input

Yes, huge improvements in compute infrastructure have been instrumental, but enormous repositories of data like Common Crawl and The Pile are what made the foundation models captivating the world today possible.

Moreover, it’ll be data with which companies refine the models underpinning their product offerings or build competitive moats going forward. And ultimately, data will be the bridge between users and personal models that run locally and continuously adapt to individual needs.

The competition for data is therefore an essential frontier, and one where blockchains can carve an edge – especially as quality becomes the prized attribute shaping the market for data.

Quality over Quantity

Early research suggests that up to 90% of online content may be synthetically generated in the coming years. While synthetic training data offers advantages, it also introduces material risks around deteriorating model quality as well as the reinforcement of biases.

There’s a real risk that Machine Learning models may deplete non-synthetic data sources in the next few years. Crypto’s coordination mechanisms and attestation primitives are inherently optimized to support decentralized marketplaces where users can share, own, or monetize their data for training or fine-tuning domain-specific models.

As a result, web3 may prove to be a better and more efficient source of human-generated training and fine-tuning data overall.

Compounding Progress

Decentralized training, finetuning, and inference processes enabled by blockchains can also better preserve and compound open-source intelligence.

Smaller open-source models refined using efficient fine-tuning processes are already rivaling their larger peers in output accuracy. The tide is therefore starting to shift from quantity to quality in terms of source & fine-tuning data.

The ability to track and verify the lifecycle of both original and derivative data enables reproducibility and transparency that will fuel higher quality models & inputs.

Source: Will Henshall / Epoch (TIME)

Blockchains can build a durable moat as the primary domain with diverse, verifiable, and tailored datasets. This can be particularly valuable as traditional solutions over-index on algorithmic progress to counter data shortfalls.

The Content Tsunami

The coming tidal wave of AI-generated content is another place where crypto’s early-mover advantage will excel.

This new technological paradigm will empower digital content creators at unprecedented scale, and Web3 offers plug and play foundations to make sense of it all. Crypto has homecourt advantage thanks to years of development around primitives that establish ownership and immutable provenance of digital assets AND content in the form of NFTs.

NFTs can capture the entire content creation lifecycle, but can also represent digitally-native identity, virtual assets, or even streams of cashflows.

As a result, NFTs make possible new user experiences like digital asset marketplaces (OpenSea, Blur), while also rethinking business models around written content (Mirror), social media (Farcaster, Lens), gaming (Dapper Labs, Immutable), and even financial infrastructure (Upshot, NFTFi).

The technology may even combat deep fakes and computational manipulation more reliably than the alternative - using algorithms to do the work. In one glaring example, OpenAI’s detection tool was shut down because of accuracy failures.

A final point: advancements in succinct and verifiable compute will also upgrade the dynamism of NFTs as they incorporate ML outputs to drive more intelligent, evolving metadata. Our belief is that AI-powered tooling and interfaces atop blockchain technology will unleash full-stack value and reshape the digital content landscape.

Harnessing ML’s Infinite Knowledge with Zero Knowledge

The blockchain industry’s search for technical solutions enabling resource-efficient compute while preserving trustless dynamics has led to substantial progress in zero-knowledge (ZK) cryptography.

Though initially designed to tackle resource bottlenecks inherent to systems like the Ethereum Virtual Machine (EVM), ZK proofs offer a range of valuable use cases related to AI.

An obvious one is simply an extension of an existing unlock: efficiently and succinctly verifying compute-intensive processes, like running an ML model offchain, so that the end product, like a model’s inference, can be ingested onchain by smart contracts in the form of a ZK proof.

Storage proofs paired with coprocessing can take this a step further, materially enhancing the capabilities of onchain applications by making them more reflective without introducing new trust assumptions.

The implications allow for net-new functions as well.

ZK cryptography can be used to verify that a specific model or pool of data was in fact used in generating inferences when called via an API. It can also conceal the specific weights or data consumed by a model in client-sensitive industries like healthcare or insurance.

Companies can even collaborate more effectively by exchanging data or IP, benefiting from shared learnings while still keeping their resources proprietary.

And finally, ZKPs have real applicability in the increasingly relevant (and challenging) realm of differentiating between human and synthetically generated data discussed earlier.

Some of these use cases are contingent on the need for further development around technical implementation and the search for sustainable economics at scale, but zkML has the potential to be uniquely impactful on the trajectory of AI.

Long Tail Assets & Latent Value

Crypto has already demonstrated its role as a superior architect of value flow across legacy markets like music and art. Over the last couple of years, onchain liquid markets representing offchain, tangible assets like wine and sneakers have also emerged.

The natural successor will involve advanced ML capabilities as AI is brought onchain and made accessible to smart contracts.

ML models, in combination with blockchain rails, will rework the underwriting process behind illiquid assets previously inaccessible due to a lack of data or buyer depth.

One method will see ML algorithms query a massive range of variables to assess hidden relationships and minimize the attack surface of manipulative actors. Web3 is already experimenting with creating markets around novel concepts like social media connections and wallet usernames.

Similar to the impact AMMs had on unlocking liquidity for long-tail tokens, ML will revolutionize price discovery by ingesting massive amounts of quantitative and qualitative data to derive nonobvious patterns. These new insights can then form the basis for smart-contract based markets.

AI’s analytical capabilities will plug into decentralized financial infrastructure to uncover dormant value in long tail assets.

Decentralizing the Infrastructure Layer

Crypto’s advantages around attracting and monetizing higher quality data address one side of the equation. The other side – the supporting infrastructure behind AI – holds similar promise.

Decentralized Physical Infrastructure Networks (DePINs) like Filecoin or Arweave have already built systems for storage that natively incorporate blockchain technology.

Others like Gensyn and Together are tackling the challenge of model training across a distributed network, while Akash has launched an impressive P2P marketplace connecting supply and demand around excess computing resources.

Beyond that, Ritual is building the foundation for open AI infrastructure in the form of an incentivized network and suite of models, connecting distributed computing devices for users to run inference and fine-tuning against.

Crucially, DePINs like Ritual, Filecoin or Akash can also create a much larger and more efficient market. They do this by opening up the supply side to a much broader domain that includes passive providers able to unlock latent economic value, or by consolidating less-performant hardware into pools that rival their sophisticated peers.

Each part of the stack involves different constraints and value preferences, and significant work remains to be done in battle-testing these layers at scale (in particular, the emerging fields of decentralized model training and compute).

However, the foundations exist for blockchain-based solutions for compute, storage, and even model training that can eventually compete with conventional markets.

What It All Means

Crypto x AI is quickly becoming one of the most inspiring design spaces. The respective fields are already impacting everything from content creation and cultural expression to enterprise workflows and financial infrastructure.

Together, we believe these technologies will reshape the world in the coming decades. The best teams are natively incorporating permissionless infrastructure and cryptoeconomics alongside AI to upgrade performance, enable net-new behaviors, or achieve competitive cost structures.

Crypto introduces unprecedented scale, depth, and granularity of standardized data into coordination networks, often without an obvious means for deriving utility from that data.

Meanwhile, AI converts pools of information into vectors of relevant context or relationships.

When paired together, these two frontiers can form a uniquely reciprocal relationship that sets the stage for builders of the decentralized future.

A huge thank you to Niraj Pant, Akilesh Potti, Jason Morton, Dante Camuto, David Wong, Ismael Hishon- Rezaizadeh, Illia Polosukhin, and others for their work at the forefront of this space, invaluable insights, and inspiration – all of which make possible not only this article but crypto’s bright future.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.

At Archetype, we believe that understanding where new interoperability projects are positioned within the infrastructure stack is crucial to making sound investment decisions. The technical stack for blockchain infrastructure has changed drastically over the years, demanding that frameworks for understanding interoperability adjust as well.

With this post, we want to reflect on the evolution of the blockchain interoperability stack, unpack each of its components, and share our mental model for thinking about what comes next.

Native Layer 1s (L1s): In 2017, the blockchain landscape consisted of isolated chains that could only be connected somewhat through centralized exchanges (CEX) as hubs.

“Alt” L1s: As Ethereum gained market share, it became table stakes for other blockchains to have native bridges to Ethereum, rather than having to always go through a CEX. These connections helped bolster Ethereum’s dominance over other chains even further, creating the “Alt L1” narrative.

Layer 2 (L2)-centric: The next evolutionary step saw new networks that not only connected to Ethereum but used it as a data availability layer. These networks, like Arbitrum and Optimism, offered an enhanced UX to end users without trading security, because they used Ethereum as their settlement layer. Naturally they were called “Ethereum L2s” or rollups.

This brings us to today, where L2s have gained a lot of momentum, the Ethereum roadmap is “rollup-centric,” and there are now connections between each of the L2s, a big step from the state of blockchain infrastructure five years ago where there was a handful of independently-secured blockchains that were only connected by CEXs.

Today, the design questions around blockchain interoperability are generally constructed around the following factors:

• L2s that offer very fast and cheap UX

• L2s and L3s (appchains) serving as the home to new user applications

• More chains leveraging shared security and infrastructure

• A focus towards communication between L2 chains rather than just L1 to L2 chains

The number of connections between blockchains has created a lot more obscurity for people who want to better understand how everything works.

How do my tokens transfer from L2 to L2?

Who is securing my data, and who am I implicitly trusting to keep my data safe?

To answer these questions, we need to understand each layer of the stack.

What we’re going to cover:

• L2s

  • Why are L2s fast and cheap and how are they secured

  • The relationship between L2 and Ethereum (L1)

  • The relationship between L2s and Data Availability (DA) layers

• Sequencers

  • What is a sequencer

  • Shared sequencers

• Fraud Proofs, Validity Proofs, and Proof of Authority (PoA)

• Bridges

  • Types of bridges

  • Bridges vs sequencers

• Intents

  • What are intents

  • Intents vs bridges

• CEXs and DEXs

Let’s dive in.

Layer 2s

Users interact with dApps, and users demand that those dApps are fast and cheap, which is why dApps today are often deployed on L2s like Polygon, Arbitrum, and Optimism, and not on Ethereum.

Why are L2s cheaper than L1s?

L2s process transactions offchain and periodically publish batches of finalized transactions to some onchain layer that ensures the data was in fact published and made available to verify (data availability). The reason why L2s can be so fast is that they record user interactions offchain on a private server, offering a “web2” experience. In the background, L2s publish batches of these transactions to a data availability layer, offering additional transparency and security.

How do L2s publish and store transaction data/history today?

Choosing a secure data availability layer is the most important security decision an L2 makes. If an L2 is an “Ethereum Rollup,” then it publishes its transaction data to Ethereum.

There are two reasons to choose Ethereum. First, Ethereum is the “most immutable” layer. Once a rollup posts transaction data to Ethereum, it is infeasibly expensive to reverse finality on Ethereum. So, L2s can publish their state to Ethereum and have a very high degree of confidence (following Ethereum’s finality period of ~13 minutes or 2 epochs of 32 slots taking 12 seconds each) that their transaction will not be removed. The second reason is Ethereum’s high availability of data. Ethereum has a large number of nodes replicating and verifying transaction data, making it highly unlikely that data would ever disappear or be entirely unavailable.

However, using Ethereum for data availability is famously expensive. While Ethereum is generally the most secure data availability layer, it wasn’t optimized to just store data—it’s a generalized computing machine.

Until EIP4844 is implemented, there is no way to request blockspace only; you have to pay the same gas as everyone else sending normal transactions on Ethereum. So if an L2 chooses to publish state on Ethereum, it’s choosing the blue chip option but it’s also paying a premium for its security.

What options are available to an L2 that wants to pay less to post state? Enter the DA layer.

Data Availability (DA) layers like Celestia are optimized to offer a place specifically for data availability, where demand for blockspace is lower and blockspace itself is cheaper.

The natural downside is that newer DA layers are less economically secure than Ethereum at launch, given the necessity and time needed to bootstrap a network. Another downside to a pure DA layer like Celestia is that you can’t naturally do computation on the data, which adds a layer of complexity when arbitrating or validating proofs. When sending a transaction on an L2, you should be asking: Which DA layer does this L2 publish its state to, do I trust it, and how can I later access that data?

Why do L2s pay to post data?

Posting state on an external DA layer removes a centralization vector for L2s insofar as it allows for permissionless, independent verification of state. In the case of optimistic rollups, data availability is needed to generate fraud proofs. In the case of ZK rollups, data availability is needed to ensure censorship resistance and liveness.

If an L2 never published its state onchain publicly, then it would just be in a private server. Users would be at the complete mercy of the L2 operator, which also introduces questions around liveness (what if the centralized operator goes offline?).

If the L2 publishes its state to a chain that is relatively insecure, then users would also be at the mercy of that chain. For example, if L2s posted their history snapshots to Tron, then Justin Sun would have the ability to change the L2’s history.

This is why most L2s choose to publish state to highly secure and battle-tested chains like Ethereum. Users want to know that they have control of their data and that their transactions won’t be modified or reversed.

A key point worth repeating: An L2 is only as secure as the DA layer/L1 that it publishes state to.

Rollups without DA:

Does that mean that DA layers are competitive with Ethereum?

They compete for blobs of data; post-EIP4844, these DA layers will compete directly with the Ethereum “Blob” market.

For other types of transactions like sending tokens or interacting with DeFi apps, Ethereum competes with other turing-complete blockchains like Solana, BSC, and Avalanche that don’t publish their state to Ethereum.

Sequencers

What is a Sequencer?

Sequencers are the parties that actually publish L2 state to the DA layer. They are in charge of determining the canonical order of transactions on an L2 and publishing snapshots of that history to a DA layer. Sequencers pay the DA layer fees to post the data, and sequencers earn revenue by collecting all fees paid by users who send transactions on the L2s. Sequencers can be single, whitelisted parties, or they can be competitors in a decentralized market vying to publish the next set of L2 state. Typically, sequencers needs to be aware of:

1. State on the DA layer so that it can publish data to it/mint bridged assets (if the DA layer supports smart contracts

2. State on the L2 layer so that it can sequence the L2 transactions

Most of these dynamics assume that the sequencer is also a block builder (which is true of sequencers in production today), but these roles could be decoupled depending on design/development around decentralized sequencers, shared sequencers, and PBS

What is a Shared Sequencer?

Generally speaking, proposals for shared sequencers today separate transaction ordering and execution; shared sequencers don’t execute transactions. This design decision makes shared sequencers scalable–they’re fast, sequencer node requirements are light because ordering is stateless, and it becomes relatively easier to decentralize the sequencer set.

However, because these sequencers don’t execute transactions, this places a centralizing force on block builders–the parties that actually execute the state transitions across the connected domains. The more rollups connect to shared sequencing layers, the higher requirements that are needed for builders to be competitive across domains. The higher the resource requirements, the risks of centralization at the builder level. The design space around decentralized block building and proposer-builder separation is vast and for sake of scope will not go down that rabbit hole here.

Another issue: shared sequencers alone don’t offer cross-rollup atomic conditional transaction execution.

For example, a common demand is for a user to bridge tokens from Optimism to Arbitrum and then swap tokens on Arbitrum. Ideally the user would want both of these actions to execute in order, or none at all. In the worst case, the user would end up with tokens on Arbitrum that are not swapped.

To enable this conditional execution for transactions between rollups A and B, a shared sequencer would need to sequence both A and B and publish a shared L2 state to a DA layer containing both conditional transactions.

This is likely where the shared sequencer evolution will go next: interdependent L2 state secured by either “shared fraud proofs” or ZKPs. To get there, shared sequencer behavior and system contracts will have to change to support publishing multiple L2 states in a single L1 transaction.

Prediction: If shared sequencing becomes the dominant paradigm over single sequencing, then state validity between L2s using a shared sequencer will be interdependent.

Here is an interesting proposal introducing this concept of “shared fraud proofs,” and here’s another recent article expounding on tradeoffs in running a sequencer.

Another potential concern around shared sequencers is that teams utilizing shared sequencer networks may lose control over certain operational parameters and value capture opportunities. For example, shared sequencers could extract intra-rollup MEV that would have been valuable to the rollup itself. Given that this would be an obvious race to the bottom in the shared sequencer market, the higher likelihood is that shared sequencers will focus on extracting inter-rollup MEV that wouldn’t have been as easily accessible to individual rollups anyway.

Fraud Proofs vs Validity Proofs vs Proof of Authority

We’ve established that an L2 should post its data somewhere transparent, but how do we verify the data?

If whatever data the L2 sequencer publishes to L1 is considered valid, then what prevents the L2 sequencer from posting an incorrect snapshot of the L2 history? Couldn’t the sequencer pay themselves a little bit extra ETH out of user wallets?

Fraud Proofs

In optimistic rollups, L2 state is published optimistically and is not considered final until it passes through a challenge window (e.g. a 7 day challenge window). During this challenge window, anyone can dispute an L2 state commitment if they believe that the commitment is missing the correct transaction history. To submit a dispute, one must publish a fraud proof that is used in an interactive process to resolve the dispute.

The main advantage of fraud proofs is that they only need to be generated when there is a dispute.

Validity Proofs

A validity proof claims that the L2 state is valid and proves it at the time of publishing. There is no dispute window; if the proof is verified by an L1 contract designed to verify such proofs, then the L2 state is valid.

The main advantage of proving L2 state via validity proofs is that the L2 state published to L1 can achieve immediate finality. This means that L1 contracts can instantly take the L2 state as “final” and act upon it. This is why L2 to L1 withdrawals for ZK Rollups are very fast (~24 hours), compared to the 7 day challenge window for optimistic rollups.

Proof of Authority (PoA)

The last way of validating L2 state is through a Proof of Authority (PoA) mechanism. This is when the sequencer publishing the L2 state basically is given the authority to claim: “This is valid because I am publishing it.” This is how many rollups work in practice today; sequencers run permissioned full rollup nodes that validate state via PoA.There are no challenge windows, there are no proofs. Users simply trust these sequencers to not modify the L2 state.

How do fraud and validity proofs actually work?

These proofs are very difficult to implement and expensive to verify. They essentially require simulating the L2 VM execution for a set of alleged L2 transactions and an initial state. If the resulting state that is produced following those alleged transactions on the initial state differs from what the sequencer published to L1, then the state is disputable.

These proofs must be verified on the same DA layer that the L2 state is published to, which makes their verification expensive to run.

The main tradeoffs between fraud and validity proofs:

• Cost to Validate Proof: Validity proofs are generally much more expensive than fraud proofs.

• Speed: Fraud proofs use a Dispute Time Delay (DTD) system–the challenge window–meaning transactions don’t reach finality on L1 until the window passes, whereas validity proofs are verified immediately in a single transaction.

• Implementation Complexity: Both types of proof verification contracts are difficult to build. Validity proofs rely more on the succinctness property of cryptographic tools so that they can simulate L2 state in a single transaction. Fraud proofs are interactive and therefore require fewer cryptographic tools but more offchain infrastructure is needed to support an interactive proving system.

• L2 VM Implementation complexity: L2 state that is validated via validity proofs usually requires a modification to the EVM in order to make the validity proof verification cheaper. L2 state validated by fraud proofs can more easily mirror the EVM exactly.

• Running Cost for Sequencer: Validity proofs require a payment for each submission to L1 while fraud proofs only impose costs when a challenge is submitted. In the case that a challenge is submitted, however, fraud proofs require payment for every interaction between the parties arbitrating the dispute (interactive), whereas each validity proof is generated in a single transaction (non-interactive).

• Upper Limit on Funds at Risk: If a fraudulent sequencer is not disputed, all L2 funds are at risk. If a sequencer does not submit a valid validity proof, then the L2 state is effectively frozen but no funds are lost.

• Operating Cost for Validators: In a fraud proof system, there should always be at least one honest validator watching the sequencer’s submissions. In a validity proof system, there is no need for external validators assuming the validity proof is submitted

It’s worth noting that there are designs for non-interactive fraud proofs in development, though more technically challenging to implement.

There is a lot more to dive into on this topic. We find these resources particularly helpful:

• Validity Proofs vs. Fraud Proofs Strike Back

• Optimistic time-travel

Bridges

Before diving into specific examples, it’s worth teasing out subtle differences between how bridges are discussed in different contexts.

An arbitrary message bridge (AMB) is a protocol that keeps track of arbitrary cross-chain state—arbitrary meaning anything from token transfers to data storage to anything else. Even more simply, message bridges essentially make state from one chain available on another chain. (This sounds a lot like a shared sequencer.)

Diving into the mechanisms by which bridges are validated/secured is out of scope for this post, important to understand. A few resources that may be helpful can be found here and here (DYOR!)

A token bridge is an application of a message bridge that uses the cross-chain state to transfer assets/capital between chains.

For example, if the token bridge sees that transaction A, a 10 ETH “burn” has happened on Optimism, then it instructs a contract on Arbitrum to release 10 ETH to the user. If the 10 ETH burn gets wiped from history, then the token bridge is likely held liable for this loss. A token bridge is built on top of a message bridge because it needs to be aware of state on both origin and destination chain and also needs to know when transactions get finalized on the origin chain before it acts on the destination chain.

The safest, albeit slowest, way to move assets between L2s is to withdraw via the origin L2’s canonical token bridge to the L1 and then deposit from L1 to the destination L2 via that L2’s own canonical token bridge. Using a canonical token bridge is often a slow process as it’s dependent on withdrawing from an L2 to L1.

A “canonical token bridge” is essentially a special service offered by L2 sequencers. Depositing to an L2 via the canonical token bridge means locking funds on a sequencer’s contract on L1 and requesting that the L2 sequencer mint an equivalent amount of funds on the L2 chain.

Withdrawing from an L2 requires sending funds to a special contract on the L2 to be “burned” and waiting for the sequencer to publish proof of this burn to L1.  Once that proof is confirmed, like any other L2 state published by the sequencer to the L1, the sequencer’s L1 contract can release tokens to the user.

Using the canonical token bridge is as slow as waiting for the rollup full nodes to finalize L2 state on L1, but it’s also as “safe” as it gets when interacting with the L2.

A faster way to transfer assets between L2s is to use a fast bridge. A fast bridge temporarily custodies your capital on a non-sequencer contract that then fronts you your capital on a destination chain. This means that the user is temporarily placing trust in the fast bridge to not modify their information or steal their funds.

How fast bridges work:

The diagram above shows the flow of funds that enable the user to receive their funds quickly, the fees that the bridge earns for providing the fast bridge service, and the finality risk that the bridge assumes for the user.

First of all, fast bridges credit users their bridged amount minus fees on their desired destination chain. In this case, this is 10 ETH minus a 0.1 ETH fee. The user walks away happy to have received their expected amount of funds so quickly.

Let’s pretend that the bridge prefers to custody assets on Ethereum. So, unbeknownst to the user, the bridge protocol will withdraw the user’s deposit via the slower canonical bridge to Ethereum. Once the canonical bridge withdrawal finalizes (e.g. seven days later in the diagram above), the bridge protocol has ended up with a fee surplus of 0.1 ETH on Ethereum (10 ETH user deposit minus 9.9 ETH credited to the user).

In exchange for getting paid for this service, the bridge assumes the finality risk of the user’s deposit: If the user deposit is reverted by the origin chain’s sequencer any time during the seven day finality period, then the bridge will lose funds. This is because the bridge has already credited user funds on the destination chain but will no longer receive the user’s original deposit via the canonical bridge.

Bridges vs Sequencers

How can users be assured that if they deposit 10 ETH on the origin chain that they’ll receive 10 ETH on the destination chain (minus fees)? This boils down to that respective bridge’s security mechanism, which we’ll see looks eerily similar to a sequencer’s security mechanism.

When the bridge credits funds to the user on the destination chain, the user needs a way to verify that the bridged funds were correctly received. Ideally, the bridge also offers a recourse option in the case of an error. One way the bridge can offer this assurance is to allow the user to challenge the bridge in a challenge window. Much like the fraud proof process, this allows users to take their security into their own hands and provide proof of fraud in the case that they never received their funds as expected. This is why it’s important that the bridge publish its transaction history to an immutable DA layer—to offer transparency to users and assure them that if they do ever challenge the bridge, that the bridge cannot modify its history and must correct the error.

So it should be apparent now that both fast bridges and sequencers require DA layers in order to give users confidence that they are processing transactions honestly.

A bridge is only as secure as the DA layer that it posts state to and its dispute mechanism.

These fast bridges serve as a way to provide economic atomicity to users without technical atomicity.  If users could express their preference to execute a transaction on one rollup and conditional on another transaction on another, and is only willing to pay if both of these conditions are met, by providing these economic incentives fast bridges give users the properties of technical atomicity by outsourcing the execution of that to specialized parties in the absence of some actual communication or shared sequencing layer between different rollups.

ZK Bridges

Bridges can also offer security assurances to users by accompanying messages published to the destination chain with validity proofs. These types of bridges are called ZK bridges. The destination chain requires a contract to be deployed that can simulate the origin chain’s consensus and can verify that an event happened as claimed on the origin chain. The validity proof is used as input into this verification contract to prove to the contract that the message sent to the destination chain is an accurate representation of the request sent on the origin chain.

Of course, bridges can also use PoA if they don’t want to publish data anywhere transparently and immutably.

Intents

An intent is a set of preferences that the user wants to be fulfilled including the destination chain(s) where they want those preferences fulfilled.

A marketplace for intents offers to store a canonical history of intents across one or more chains. An intent marketplace allows users to signal their preferences, relayers to fulfill those preferences, and offers validation for the fulfillment of those preferences. If a user feels that a relayer did not correctly fulfill their intent, then the intent marketplace should offer a way for the user to dispute or challenge the relayer.

This is starting to sound a lot like a bridge.

Intents vs Bridges

Intents are innovative because they open design space for a different, more flexible mechanisms/language for expressing and executing transactions.

An intent marketplace does not have to be cross-chain but it’s well-positioned to be, given it’s already defining a new paradigm and language for expressing preferences and storing a canonical history of intents.

Cross-chain intent marketplaces can therefore be thought of as a more specific form of general message bridging.

The Archetype team outlines specific examples of intent systems (SUAVE, Anoma) in a previous post here.

CEXs/DEXs

We began this post by reflecting on how, in what feels like eons ago in crypto time, CEXs were the main hub for crypto assets. Given all the evolution since then that we’ve just outlined, where do CEXs/DEXs fit into today’s framework?

CEXs essentially function like both dApps and token bridges insofar as you can use them as a way to transfer tokens from one network to another. You can almost think of CEXes as Proof of Authority bridges. The exchange offers little recourse for users in the case that it steals your funds. The main recourse is to challenge the CEX legally, in person. GLHF. DEXs that exist on a single chain are dApps. Cross-chain DEXs are like token bridges plus dApps.

Conclusion

Let’s revisit the full diagram of the model that we’ve built:

A few general observations:

• Users interacting on L2s pay L2 sequencers

• Users looking to jump across chains pay Fast Bridges (and L2 sequencers)

• Bridges pay sequencers on the destination chain to carry out bridge activity

• L2 sequencers pay DA layers to store data

• Shared sequencers get first shot at extracting MEV between the L2s that they sequence Shared sequencers also get paid by multiple L2s and can reduce their costs by publishing both of the L2 states in a single, batched transaction to the DA layer

The blockchain interoperability stack has seen multiple iterations over the past several years, leaving trails of valuable takeaways for folks who have closely watched the space throughout history. Regardless of which cyclical buzz word you use (sequencers, bridges, and cross-chain intent marketplaces), interoperability infrastructure begins from the same first principle functions:

• Ordering transactions canonically (whether single chain or across multiple chains)

• Posting data to a transparent and immutable DA layer

• Offering data proof/verification mechanisms

Keeping these principles in mind, we can ask better questions and better assess where value will accrue in the interoperability stacks of the future.

Thank you to @0xFunk, @mrice32, and @pumatheuma for thoughtful review and feedback on drafts of this post.

Disclaimer:

This post is for general information purposes only. It does not constitute investment advice or a recommendation or solicitation to buy or sell any investment and should not be used in the evaluation of the merits of making any investment decision. It should not be relied upon for accounting, legal or tax advice or investment recommendations. You should consult your own advisers as to legal, business, tax, and other related matters concerning any investment or legal matters. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by Archetype. This post reflects the current opinions of the authors and is not made on behalf of Archetype or its affiliates and does not necessarily reflect the opinions of Archetype, its affiliates or individuals associated with Archetype. The opinions reflected herein are subject to change without being updated.