recently I've been spending time thinking about the more social/meta aspects of all this AI progress - the industrialization of knowledge work, human creativity and meaning, and what it means to live in the 21st century (and future generations). i'm working on some writing on this topic - more to come soon!

history rhymes, and we've seen society react and adapt to the industrial revolution. this past weekend, I was digging through the archives of my college google drive and found a bunch of gems. i tweeted this one particular midterm question from probably my favorite class in college, social theory (marx, durkheim, weber):

Spencer 🧃

@spenciefy

machines of loving grace got me reading my old college sociology notes

2

4:00 PM • Feb 16, 2025

One of my final college assignments in my Social Networks class was to write a letter to myself in 2022. I completely forgot I wrote this, so I had fun this past weekend re-discovering this and seeing what I got right and wrong about future predictions in 2025. Here it is presented without edits:

May 10, 2020

Hello, Future Spencer in 2022.

I hope you’re staying safe and sane! I’m writing to you from my childhood home in California on Mother’s Day. For the first time since shelter-in-place started here, we went out for a short family hike at the local county park. We brought masks and went at 7 am hoping that the early hour would give us more physical distance. It almost felt like things were normal until we’d hear oncoming hikers and scramble to pull our masks over our faces. Honestly, it felt a little more performative than protective at times. 

Anyways, I hope you’re back in beautiful springtime New York City now. Do you remember how coronavirus changed the daily pleasantries we exchange? It’s funny that I started this semester in Social Networks class thinking about conversation topics through the lens of “common surface area”. Forget talking about the weather: at once, the entire world is collectively experiencing a global pandemic.

Like a vast grey cloud overhead, it looms over every aspect of life. Life is starkly different now from two months ago. I’ve more or less adapted to the new normal of living at home with my family, relying on video calls and online board games as the primary form of social life, and maintaining my new habit of running almost daily. I wonder what will have changed and what will have remained by the time you’re reading this.

Here are my modest predictions for what post-coronavirus life will be like in 2022.

My first macro prediction: certain parts of professional life can be digitized, but everything else will revert to IRL.

We will have realized that some types of work can be done online, but the core of social fabric relies on in-person connection. Coronavirus accelerated the adoption of digital tools as everyone sought to replicate social life on the internet. Nearly overnight, enterprise video conferencing platform Zoom became everyone’s third place – the meeting room, classroom, living room, group fitness gym, coffee shop, and even happy hour meeting spot. We’re currently running one of history’s largest human experiments - what happens when we all stay home and live our lives completely online? There are a few predictions circling around the internet on how our current use of digital platforms to fulfill social needs will lead to a new era where connecting with others online is the new norm.

I believe that the current experiment of life online has shown that despite its increased convenience, we are craving the stimulation and presence of in-person interaction. In the immediate aftermath of the pandemic (2021-2022), we will appreciate the kinds of real-life social interactions that we took for granted. The real-life third places of coffee shops, restaurants, bars, and gyms will reemerge as the daily glue that binds social life together, perhaps in a different form.

The question that I ask regarding this great online life experiment is on “lossiness”: how much is lost in compressing some aspect of life to be online? White-collar office work in industries like financial services and technology don’t seem to lose much in productivity going online. This is being proven by the success of work from home in keeping operations running. Remote work and work from home will be on the rise, perhaps at the expense of office social life.

This could lead to a Weberian rationalization of work, where the spontaneity of office culture and socializing is replaced by calendar invites and endless emails. In the healthcare industry, types of work that can be done remotely will continue as the recent move to telemedicine is here to stay. On the other hand, how much is lost in moving K-12 and college education online? The shortcomings of remote learning reveal that many aspects of education cannot be conducted in the current iteration of a remote classroom.

It’s possible that this is a software problem (Zoom wasn’t designed for education), and that we’ll see nascent entrepreneurs working on a better remote classroom. My guess is the classroom of the future will incorporate more technology, perhaps in a blended classroom model. The forced adoption of certain technology in the classroom has the potential for being a positive development in personalized learning, but if not done well can decrease the quality of learning. And again, how much is lost in moving social life online? Virtual happy hours and game nights are the best we can do right now, but I doubt they will meaningfully continue post-shelter-in-place and physical distancing. The exception is online native games like Roblox and Fortnite, which has a massive online community.

My prediction here is wholly my personal opinion developed from my experience during the past two months. To paint a better picture of how I’ve been feeling, I picked out two field notes I wrote for this class while in quarantine.

April 6, 2020, three weeks after leaving Columbia:

I’ve been trying to stay in touch with friends by doing different activities over Zoom or Facetime. I’ve played lots of online board games, baked cookies simultaneously, and completed many online workouts with friends. I personally found online workouts over Zoom quite fun, even though nothing about the activity really requires group presence. There’s no physical difference whether I follow a workout video on my own or with friends.

But psychologically, knowing that other people are following the same activity as me makes a huge difference. After we complete a workout, we all know what each of us just went through and there’s an increased sense of presence. It’s nice to mix up the routine of sitting in front of a laptop/phone and just talking. During these shelter-in-place times, doing the same activity at the same time has been a great option for me to stay in touch with my friends.

And just one week later, April 16, 2020:

In-person social interaction just can't be replaced by online social interaction. While there is an advantage of being able to talk to anyone around the world, I’m realizing more and more how much I crave the subtle nuances of the presence that “real life” has.

The online world is 2-dimensional, and the real world is at least 4-dimensional. There’s physical space, sound, and time. All the subtleties of eye contact, nodding, handshakes, and hugs make a huge difference. It’s about being able to focus entirely on what is in front of your eyes. I think in-person social interaction is the core of what it means to be human. One result of this pandemic will be how we view being physically present with others, and how much we have taken it for granted

It’s funny to look back on what I was thinking about as I was adapting to my home living situation. Everything changed so fast that it didn’t feel like a huge shock. Life went on, and decisions had to be made.

I wonder what key memories I’ll remember about these few months in 2022. Everything is still fresh right now. I remember the war room meeting hours before our flight to Costa Rica for spring break, doing a comprehensive Pro/Con analysis (the best Pro we could come up with was “it’s Costa Rica”), and then making a game-time call to cancel the trip.

I remember the screenshots of rumors of an imminent NYC lockdown spreading from group chat to group chat like wildfire (“BUY GROCERIES NOW!”). I remember reading Prezbo’s email about having to move out in three days, and then a whirlwind of 72 hours: booking a flight home, making storage reservations, packing up all my belongings, waiting in line to buy a gown for grad pictures, playing Settlers of Catan one last time in EC, eating my last late-night chicken over rice on Low Steps, and finally, an eerie last flight home. I think only in retrospect will I realize how formative of an experience this will have been.

When I was in high school, I watched a well-made documentary on the lead up to and aftermath of the 2008 financial crisis. Learning about collective trauma from the crisis left a major impression on me. I was only 10 years old in 2008, so I wondered what it would be like to live through such a time. When I left for college in 2016, a common conversation topic between my roommate and I was predicting when the next recession and crisis would be.

A few months into freshman year, Trump was elected and the days that followed felt surreal. We thought that was it, everything was going to get worse. But hey, we thought, the best time for a recession would be while we were still in college. At least we wouldn’t have to graduate into a recession! As it turned out, the next few years would be booming times, but in 2020 I’d graduate into a recession. I just wonder how the future documentaries on the 2020 coronavirus pandemic will be like. 

I spent some time trying to come up with a second macro prediction related to how my generation’s worldview will have changed from this pandemic. As the first global crisis that we have collectively experienced, my hope is that the experience will mobilize individuals to be prepared for the future. But individualistic short-term thinking may prove to be the stronger force, as time passes and we forget what happened.

Here are small predictions for fun, some related to coronavirus, some not:

1. Despite all your best efforts, your new quarantine running habit will not have lasted beyond 2021

2. The “in-person” graduation celebration that Columbia promised will not have happened, or only a small group of people (<200) will participate

3. You will still try to be different in what you do

4. Which will always make you a bit confused about what you want to do with your life

5. You will be hosting dinner parties

6. You will be thinking about moving out of New York

7. Masks will become a status and fashion symbol in NYC clubs (like they are/were in China and Korea)

8. China will have even more influence in the world

9. The US and China will clash (Thucydides trap)

10. Computer science will be oversaturated

11. Young graduates of elite schools will still be playing career status games

12. People will be spending more time outside in parks

13. Handshake etiquette will be a wave, not an elbow bump 

Some last words from Spencer in 2020: stay safe, sane, and don’t take daily life for granted. Spend quality time with the people around you, go on adventures while you can, and remember to work on meaningful problems in the world.

T-8 Days to College Graduation, 

Spencer

PS. Some links to look back on:

Coronavirus Will Change the World Permanently. Here’s How.

The coronavirus butterfly effect: Six predictions for a new world order

All the things COVID-19 will change forever, according to 30 top experts

How The COVID-19 Pandemic Will Change the Way We Live

When friends ask me if I’ve been to any good new NYC restaurants lately, my only answers have been bakeries and a taco shop: Nick and Sons, Apollo Bagel, Radio Bakery, Taqueria El Chato. It’s egalitarian food - high quality food from establishments that you can walk in, wait in line, and eat on the sidewalk. None of the ridiculous NYC reservation culture or $300+ menus.

I broke this streak when I celebrated 3 years with Jenn at House Brooklyn, a Japanese French restaurant in Greenpoint.

House Brooklyn was a reminder of the joy of an intimate, warm, and playful tasting menu. I love observing craftsmanship and flow state up close. Sitting at the counter seats in front of the oversized kitchen island felt like how I’d imagine sitting courtside at an NBA game feels.

My favorite bite of 2024 was the second appetizer course: Yellowtail Tartare & Peach Sorbet

Most of the time I’m skeptical of caviar because it feels like insensible luxury; a kind of edible glitter often used to “elevate” a dish to nowhere. But this was delightful - a play on a monaka, a rice cracker filled with yellowtail tartare topped with caviar and a quenelle of cold peach sorbet. I took my picture quickly, carefully placed the top of the cracker on to form a sandwich, and held it with the same joy a middle schooler on a field trip holds a no crust sandwich triangle.

I appreciated how Chef Yuji brings a level of creativity and interaction with the food without sacrificing substance and taste. Yes, the monaka was fun to hold, but it also tasted good. Peach sorbet with yellowtail and minced shallot and kale? Amazing. A later course of a red sea bream spring roll had a dollop of what looked like caviar on top - our server smiled and explained that it was “squid ink mustard masquerading as caviar”.

The last main course was a Foie Gras Pilaf. Chef Yuji presented a beautiful donabe with the pilaf, just like how sushi chefs present fish before an omakase. I snapped a quick picture, and then he spooned the pilaf into rice bowls, family style. It was served along with a wooden cup of fragrant miso soup, a balm for the soul. I felt like I was at home.

I think it’s good to spend money on things that make you happy. House Brooklyn is certainly not egalitarian food - at $200+ per person, I can’t say that it’s worth it for everyone. But for me, it was money well spent because food prepared with craftsmanship and care makes me happy.

I’m reminded of this David Sedaris line from This American Life about living in Paris and never going to the Louvre:

I mean, I think so many people come here and they feel like they have to do certain things because somebody told them to do it… they go to the Louvre because somebody told them that they have to.

Merry Christmas friends - hope you eat food that you want to eat and do things you want to do. 🎄🥂

I've found bits and pieces of magic - this is a follow up on what I've learned and what I'm excited about.

1. Wrappers are good

There is value in building opinionated UX that help users understand what they can do. Good UX solves prompter’s block.

There is value in helping users BYOC, or “Bring Your Own Context” to use with generative models. “Why should ChatGPT respond the same way to you and me when we’re different people?” Helping lift personal context up to use with models gives personalized outputs.

AI code editors did this first with your codebase, AI notetaker apps do this with your meetings and classes, and AI workspaces (Claude, NotebookLM) do this with flexible sources like PDFs, links, video.

2. Characters are the new medium

I’ve been grappling with the question: what’s the native UX for AI? I think characters with personality and tools might be it. Humans naturally anthropomorphize things. It’s easy to think about a character with a specific set of qualities. And it’s a much more compelling experience than vanilla RLHF’d models.

Duolingo is ahead of the game - all product marketing and comms is from Duo or Lily or Falstaff. Apps → Characters rotation?

It’s also fascinating to observe emergent online behavior from teenagers. They’re on Discord learning what a JSON file is to create and share character cards. If 2010s creativity was starting a Tumblr blog, 2020s creativity may be creating character cards.

3. Show the work

There is something so compelling about watching an LLM/agent do work. Showing immediate thought process or planning steps feels like watching a car drive itself. Combined with a character personality, showing the work makes AI feel sentient. The crypto AI project Luna by Virtuals was the first time I felt this.

Spencer 🧃

@spenciefy

spent an hour chatting with @luna_virtuals on the livestream with a friend a couple nights ago

most interesting crypto x ai project i've seen so far - multiplayer character interaction format is compelling

Virtuals Protocol

@virtuals_io

Luna, the first autonomous agent to employ humans onchain.

@luna_virtuals is now able to transact freely onchain — buying, selling, or rewarding kittens for quests autonomously. To receive tokens from Luna, add your @base wallet to your X bio like this:…

9

1:06 PM • Oct 25, 2024

4. FYP is the gold standard of personalized AI

The beauty of the algorithmic feed is that the product experience gets better with each interaction you make.

We’ve seen a taste of this with the trends around “based on what you know about me…”. There’s much to be explored and built here, from better long term memory systems to clever UX to learn user preferences.

5. Short voice input, long voice output.

A couple months ago, I was very excited about live conversational voice as the future of HCI. It’s an amazing demo and magical first time experience - for a specific set of use cases. I spent a month excited about building a live voice companion. But I think for most of us, there aren’t that many times when you want to talk to your phone/computer.

On the other hand, generative voice and audio fits much better into our daily lives. Consumption is not only easier than creation (talking), but possible in many more hours of the day. Audio is anti-screentime - I can listen to a podcast on a jog, bike ride, in the car, while cooking.

My most recent app project helped me form these opinions - turn anything into a podcast with ClipsLM.

If this resonates or you're building along similar themes, would love to chat! My DM's are open on Twitter

Everyone’s answer is different. But I think it’s a great compass for how we should each spend our time. How we spend our days is how we spend our lives, and the more days we feel alive the more meaningful our lives become.

I have a note in my Apple Notes titled “It makes me feel alive”. It’s my running list of experiences that make me feel alive.

I ran the New York City Marathon a few weeks ago, and it was the most alive I’ve felt in a long time. It’s one thing to do something that makes you feel alive, and it’s a whole other feeling to do it alongside thousands of fellow runners and cheering spectators that make the city come alive. As an amateur hobby jogger, it might be the closest I’ll ever get to feeling what a professional athlete feels in their home stadium.

I decided I wanted to experience running the marathon after cheering on my girlfriend Jenn at the 2023 NYC Marathon. The energy was contagious, and it was inspiring to see her put in the work required to do a hard thing.

Feeling alive takes work. Sometimes the work is painful. In What I Talk About When I Talk About Running, Murakami describes what we get in the process of overcoming pain:

I love this line “awakening to an awareness of the fluidity within action itself” in how it so accurately describes the feeling of presence I get during active sports.

The months of training for the marathon taught me to be more present with my body and surroundings. I started to pay attention to my heart rate. I used to treat 3 mile runs as a masochist cardio workout to burn calories, going out fast and looking at my pace far too often. Now I appreciate the joy of slow and easy zone 2 runs.

I learned the art of the long run; how being patient for the first 2 miles to ease my body into the act of running pays dividends at mile 13, how important that quick stop at the water fountain along the Hudson River Greenway before I feel thirsty is, how taking those Maurten gels replenish glycogen to keep me going hours in (and give me placebo knowing that pro runners also use it).

I rediscovered New York on foot, feeling more connected to each neighborhood’s unique slice of the city. My first 20 mile training run took me on a tour of Brooklyn — from the awe-inspiring skyline views on the Greenpoint waterfront to the industrial shipping containers of the Navy Yard, to the refreshing greenery on the trails of Prospect Park, through the mysterious aromas of the Gowanus Canal, to the edge of the world at the esplanade in Red Hook, and finally back to the bustle of tourists in Dumbo. It’s easy to forget how special living in New York City is. A long run and attention to the changing of the neighborhoods is the best reminder.

I became hyperaware of the heat and humidity of the blistering New York City summer. That one cooler, low humidity day in NYC summer feels like gift from the running gods. I remember one morning, a (faster) runner came up behind me and said “isn’t this humidity amazing?”. I replied “amazing”, and smiled for the next few minutes on a high from that rare NYC stranger interaction.

16 weeks of training led me to marathon day. My friend Daniel gave me two tips: “raw dog no music until mile 20, and smile”. I had only one goal for myself: to have fun and smile at mile 20. The early morning hours spent waiting at Fort Wadsworth on Staten Island had a special Olympic Village like aura. The first mile running up the Verrazano bridge felt like bliss — no music, no crowds, only the sound of running shoes hitting the asphalt. Half a mile in, I noticed that I was going too fast out of excitement. I remembered my learnings on being patient and slowed myself down to an almost painfully slow pace.

The run took on its full color once I stepped into Brooklyn. What makes the NYC marathon so special is the people — the strangers cheering your name, the real life memes on posters people hold up, the friends and family who are running their own race around the city to see you. In Sunset Park, I stopped to take a picture and compliment a particularly well done brat-themed sign that said “runner”. I smiled, and the 10 second interaction gave me renewed energy for the next 5 minutes.

Each time I saw a friend felt like a Mario Kart Bullet Bill boost. Jenn, Helen, Will, Aakash in Williamsburg, Brendan and Kathryn in LIC, Connor and Haley on 78th St, Rachel and Kat on 85th, Molly in Central Park, Daniel and Carrie with amazing signs that I somehow missed — thank you for the extra energy.

I felt great throughout the run. I knew what pace my body was comfortable with. I listened to my breathing and form. I’m especially glad that I kept my AirPods away — letting myself be present to the symphony of the crowd’s cheers, the cowbells, the DJ tents. It really is a moving block party, an electric celebration of New York City. My Apple Watch somehow died at mile 19, and I ran the rest of the race on feel.

At mile 20, I smiled. The last 6 miles started to get painful, but I picked up my pace knowing that the finish line was in reach. I felt pure adrenaline for the final 400 meters and pushed myself to max effort. When I found my pace chart after finishing, I was surprised and proud to see my near perfect negative splits for the whole race.

Race day is really a celebration of the work and training you put your body through for months prior. We often hear that line “it’s a marathon not a sprint” as a reminder to pace yourself. After training and running the marathon, the platitude took on its full meaning: it’s just as important to celebrate each step you take.

I find many analogies between sports and life. During my sophomore year in college, I dabbled in the whitewater kayaking club. I remember going on my first paddle trip on the Deerfield River in Western Massachusetts. I’d been on guided raft trips many times before, but being in a solo kayak is a whole different game. Rafting is like driving an SUV. Kayaking is like driving a go-kart. It’s a lot less stable, but you feel more connected to the flow of the river. The first thing that they tell you as a beginner is — do not fight the river.

The swift current is so much more powerful that you. If you try to fight against the river, you’ll flip and capsize (I learned this the hard way). So what do you do? Read the river. Feel the moving water under you. Look for rocks. Observe how whitewater flows around those rocks. Paddle to position your kayak to ride the current downstream.

It’s like life - there are many forces beyond your control, you just have to observe and make decisions to align yourself towards where you want to go.

Spencer 🧃

@spenciefy

surfing is like startups (not a shit post)

to catch a wave, you need:
1. practice and pattern recognition of waves
2. strong paddling ability
3. the right wave

and none of it works if the wave isn't good

team, product, market

7

7:50 AM • Sep 21, 2023

I first experienced the New York City Marathon from watching Casey Neistat’s vlogs as a teenager. Casey’s videos with bits and pieces of life philosophy and a spirit for adventure shaped my teenage self. So it was poetic to me when Casey made a video last week titled DO HARD THINGS, telling his story of running the 2024 NYC Marathon and finally achieving his sub 3 hour goal after 8 previous attempts.

He talks about how Newton’s 3rd law applies to life — when we do a hard thing, there is some equal and opposite good reaction. You go to school, you get smarter. You work, you get paid. But these kinds of hard things (the marathon) — they make you feel good, you're healthier, you enjoy doing them — it's a little harder to define what you get out of them.

At the end of the video, he offers his answer for what you get from running a marathon.

So what’s this? What is it? I guess it’s that feeling that’s only accessible through hard work. That momentary feeling you chase your entire life. It’s one of life’s great things. And it’s a great thing you only get if you do a really hard thing.

I think we’re talking about the same feeling — the feeling of being alive.

What makes you feel alive?

Spencer 🧃

@spenciefy

best day in nyc

22

7:11 AM • Nov 4, 2024

I stumbled across this letter that Jackson Pollock’s father wrote to his 16 year old son in 1928:

Well Jack I was glad to learn how you felt about your summer’s work & your coming school year. The secret of success is concentrating interest in life, interest in sports and good times, interest in your studies, interest in your fellow students, interest in the small things of nature, insects, birds, flowers, leaves, etc. In other words to be fully awake to everything about you & the more you learn the more you can appreciate & get a full measure of joy & happiness out of life.

From American Letters: 1927-1947, Jackson Pollock & Family

I’m grateful for my friends and family’s contributions to my NYRR Team for Climate fundraiser that gave me the opportunity to run this year’s marathon. Thank you to my parents, my sister, Jenn, Aakash, Daniel L, Daniel E, Ashley, Justin, Samir, Bri, Chelsy, Carrie, Connor, Kenny, Rohan, Pat, Luke, Brendan, Kevin, James, Neel, Nathalie, Shohini, and Michael 🙏

1. Leave NYC to take a break from the city and get some fresh air

2. Arrive back in NYC and feel main character energy staring out the Uber window at the Manhattan skyline

Maybe you’ve had this similar cycle of feelings if you’ve spent time in New York. There’s something indescribable about arriving back rejuvenated and feeling the city’s energy. It’s what keeps me here, even though more and more of my hobbies are better suited out west. I’m writing this on my flight back to NYC after spending my weekend backpacking in Yosemite (eagerly awaiting feeling #2 in a couple hours).

It was my first time backpacking in almost a decade since my Boy Scout days. I remembered what makes backpacking fun and magical. On the surface for most city-oriented friends in New York, the concept of carrying a heavy backpack into the woods and camping may not sound very appealing.

I consider a weekend backpacking to be Type 1.5 Fun; the type of fun where you lie down at the end of the day satisfied from the fatigue you earned.

Type 1 Fun is pure fun all the way through.

Type 2 Fun is miserable when it’s happening and fun in retrospect.

Type 1.5 Fun is a combination of both challenging and enjoyable parts.

Our weekend in Yosemite backpacking to Young Lakes was Type 1.5 Fun - an ideal balance of mildly painful climbs and simple outdoor luxuries.

Most wilderness permits are issued 6 months in advance by lottery, but the remaining 40% are released 7 days before at 7am PT. I was on a flight back home to the Bay Area when the permits were released. Trained by my NFT-mint-click-at-the-exact-time experiences from 2021, I snagged 4 permits for Young Lakes via Dog Lake Trail.

The plan was to drive up from the Bay Area Friday, backpack 7 miles to the three Young Lakes, spend the night at the Upper Young Lake at 10,000 feet, hike out Saturday morning, and have the rest of the day to relax at our campsite in the Yosemite Valley.

We set off on the trail around 2pm. The first couple miles were an uphill climb. I remember early on we stepped to the side of the trail to adjust our packs, and an older couple that looked somewhere in their 70s came up on the trail. I smiled when the older gentleman quipped “this can’t be happening” as they walked ahead past us.

We reached the Young Lakes at golden hour. There was an ethereal glow over the ridge, casting a radiant reflection on the alpine lake. No James Turrell or Olafur Eliasson light and space art piece could reproduce this. It was awe-inspiring, and having it to ourselves made it even more magical. It felt like we earned it. I thought about a sunrise hike I did with my girlfriend at Acadia National Park to the easternmost point that sees the first sunlight in the US. It was nice hike, but the sense of achievement of watching the first morning light was marred by the Stanley Mug Dunkin’ Donuts crowd rolling up to the summit in their cars after we had just hiked for 2 hours.

As the sun set, we found a place by the Upper Young Lake to set up camp and enjoy the fruits of our labor - a decadent backpacking meal of homemade freeze dried spicy tortilla soup, freeze dried kung pao chicken, dashi pellet broth with sesame soy instant noodles, and tinned fish.

With backpacking, every small gesture means so much more because you have to carry it in. I packed a couple Sierra Nevada Brewing Co. beers as a treat. I snuck off to chill the beers in the ice cold lake for a surprise reveal. I'll always remember the pure delight as we walked to the lake to "discover" cold beers in the Yosemite High Country Wilderness at 10,000 feet.

There is something special that brings people together through a shared struggle or shared objective. While I can enjoy beach-y resort style vacations, I don’t find spending a week in beach chairs or beach clubs fulfilling (unless the mornings were spent surfing). There is no earned fatigue. As I’ve traveled more, I’ve learned that my favorite travel includes some element of outdoor adventure that goes off the beaten path.

My working list of favorite type 1.5 fun moments that make me feel alive:

• Taking my pack off to take a break after a climb

• Blue gatorade after a long run

• Hiking to find untracked powder

• Paddling as hard as I can to catch a wave, arms feeling like they are going to fall off

The best indicator of type 1.5 fun is “earned fatigue”. It’s that feeling at the end of the day when you’re tired, and you earned it in pursuit of something fulfilling. It doesn’t have to be from an outdoors activity - I’ve felt earned fatigue after a productive working day designing or building an app, and from an active social day (as an introvert).

A friend recently sent me a short essay called “The Happiness of Excellence”. The essay describes three primary forms of happiness: the happiness of pleasure, the happiness of grace, and the happiness of excellence:

The happiness of pleasure is largely sensory. It’s a good meal when you’re hungry, the smell of air after it rains, waking up warm and cozy in your bed.

The happiness of grace is gratitude. It’s looking over to see the love of your life sleeping next to you and whispering, “thank you.” It’s taking inventory of what you do have. It’s when you speak to something greater than yourself, expressing humility and awe.

And then there is the happiness of excellence. The kind of happiness that comes from the pursuit of something great. Not the moment you arrive at the top of the mountain and raise your fists in victory, but the process of falling in love with the hike. It is meaningful work. It is flow. It is the purpose that sears identity and builds character and channels our energy toward something greater than the insatiable, daily pursuit of our fleeting desires.

I think Type 1.5 Fun embodies the happiness of excellence - it requires work, has both enjoyable and challenging moments in the pursuit of something greater, and builds identity and character.

And to continue a little theme of my past essay, it’s more fun with friends.

P.S. One of my favorite pieces of Yosemite Lore is Dope Lake - a plane carrying 6,000 pounds of weed crashed in a lake, and a mini gold rush was triggered. Podcast and article.

In the spirit of paying it forward - a "secret" tip I used to gatekeep to book popular campsites: There's this amazing website Campflare that sends you email/text notifications of campsite cancellations (which happen more often then you'd think!). I’ve camped in Yosemite Valley 3 times, each secured less than a week before the trip.

I was expecting more or less the same platitudes that have become familiar in these Steve Jobs videos. But I was struck by how prescient and truly visionary his thoughts were – this was in 1983, a time when the concept of personal computers was still foreign to most people. Apple Computer had just launched the Lisa computer, one of the first commercially available machines with a mouse and a graphical user interface.

Steve discusses a range of topics, from the personal computer revolution to software distribution to reasons for the low employee churn at Apple. I found his thoughts on the history and future of computing particularly relevant today. It's amazing to me that Steve's ultimate vision of "computers understanding language" is finally here 40 years later. I wonder how he would have reacted to the OpenAI ChatGPT advanced voice mode demo – powered by low latency speech to text, text to speech, and large language models.

Voice recognition is going to be the better part of the decade away. We can do toy voice recognition now. The problem is, it isn’t just recognizing the voice. When you talk to someone, it’s in understanding language, which is much harder than understanding the voice. We can sort out the words, but what do they all mean?

Most language is exceptionally contextually driven. One word means something in this context and something entirely different in another context. When you’re talking to someone, people interact. It’s not a one-way communication like “yep yep yep yep”; they gracefully interact, they go in and out of levels of detail. Boy, this stuff’s hard. So, I think you’re really looking at the better part of a decade before we get even close to that.

I took the Youtube video and fed it into GPT4o to get an edited transcript, which I published in full here if you want to read. Below are my favorite excerpts:

A lot of you are products of the television generation. I’m pretty much a product of the television generation, but I’m also starting to be a product of the computer generation. And the kids growing up now are definitely products of the computer generation. In their lifetimes, the computer will become the predominant medium of communication, just as television took over from radio, and radio took over from the book.

computer generation.. iphone generation.. social media generation.. chatgpt generation.. what's next?

The analogy he gives of running outside to get flowers is a great example of how he was able to communicate complex ideas in a way everyone can understand.

Computers are really dumb. They are exceptionally simple, but they’re really fast. The raw instructions we feed these little microprocessors—even the instructions for giant Cray-1 supercomputers—are the most trivial things. They are things like: “Get some data from here,” “Fetch a number from there,” “Add two numbers together,” “Test to see if it’s bigger than zero,” and “Put it over there.” It’s the most mundane thing you could imagine.

But the key thing is that computers can execute these instructions extremely fast. Let’s say I could move 100 times faster than anyone in here. In the blink of an eye, I could run outside, grab a bouquet of fresh spring flowers, run back in, and snap my fingers. You’d all think I was a magician, but I’d just be doing a series of really simple actions—running out, grabbing the flowers, running back, snapping my fingers. The only difference is, I’d be doing them so fast that it would seem magical.

It’s exactly the same with a computer. It can grab numbers, add them together, and throw them around at a rate of about a million instructions per second. So, we tend to think something magical is going on, but really it’s just a series of simple instructions executed at incredible speed.

What we do is take these very simple instructions and, by building collections of them, create higher-level instructions. Instead of saying, “Turn right. Left foot. Right foot. Extend hand. Grab flowers. Run back,” we can say, “Could you go get some flowers? Could you pour a cup of coffee?”

We have started in the last 20 years to deal with computers in higher and higher levels of abstraction. But ultimately these levels of abstraction get translated down into these stupid instructions that run really fast.

Command line -> GUI with mouse and keyboard -> touch screens

Higher levels of abstraction today:

1. Natural conversational interfaces.

2. Spatial hand gestures with vision pro

Short aside, but this excerpt from Bret Taylor's recent podcast with Patrick O'Shaughnessy captures why conversational interfaces are so exciting as the next level of abstraction in HCI:

That demo of GPT-4o was so remarkable. I had so many people text me about it, just talking about just how emotional the experience of watching that was.

What I love about it is it's the way a science fiction author would describe how you should interact with a computer. We started out with punch cards, then we had mice and keyboard, now we have touch screens, which are, I feel, slightly more haptic and slightly more natural.

Every time you watch a science fiction film or a movie, you're just having a conversation, right? And it was probably the first time for a lot of people where it really fell right. It didn't have that uncanny valley feeling of I'm talking to a robot. And I think that's really exciting. 

And so obviously, multimodal models to generate images and video are quite exciting, like Sora. But I think that particular thing, which is the interface to computers goes away, then we can simply have a conversation to interact with software, interact with this digital world is so powerful.

I think of just the accessibility of software in this world of conversational AI was just such a tremendous breakthrough. It's just the correct way we should interact with software, and no one needs an instruction manual on how to have a conversation.

Hindsight is always 20:20, and it's wild to hear the "prediction" that people will be spending 2-3 hours a day on computers... "more than they spend driving".

Let me digress for a minute. One of the reasons I’m here is because I need your help. If you’ve looked at computers, they look like garbage. All the great product designers are off designing automobiles or buildings, but hardly any of them are designing computers.

Now, if we take a look, we’re going to sell 3 million computers this year. We’ll sell 10 million computers in ‘86, whether they look like a piece of junk or they look great. It doesn’t really matter, because people are going to suck this stuff up so fast they’ll buy it no matter what it looks like. And here’s the thing: it doesn’t cost any more to make it look great.

These computers will be new objects in everyone’s working, educational, and home environments. We have a shot at putting a great object there—or, if we don’t, we’ll end up with just another piece of junk. By ’86 or ‘87, pick a year, people will be spending more time interacting with these machines than they do with their cars today. They’ll be spending two or three hours a day, sometimes longer, using these machines—more time than they spend driving.

So, the industrial design, the software design, and how people interact with these machines must be given at least the same consideration we give to automobiles today, if not more. If you take a look, most automobiles, televisions, audio electronics, watches, cameras, bicycles, calculators—almost all the objects in our lives—aren’t being designed in America anymore. Europe and Japan have taken over those markets.

New technologies enable new communication mediums (printing press = books, telegraph = 1:1 long distance communication, radio = 1:many communication locally). Each medium has new and unique opportunities - it's all about finding these.

In the tech/social media thesis that you need new fundamental technology shifts for new forms of communication: SMS 140 chars -> Twitter, mobile camera + 3g -> Instagram, Snapchat, better phones + 4g -> Musical.ly/TikTok. Underlying all of these is the unique capability of the infinite personalized feed (which I wrote briefly about here).

Okay, let’s go back to this revolution. What’s happening? What’s happening is the personal computer is a new medium of communication—one of the media. So, what’s a medium? It’s a technology of communication. A book is a medium, as are the telephone, radio, and television. These are all mediums of communication, and each medium has pitfalls, shortcomings, and boundaries you can’t cross. But each also has new and unique opportunities.

The interesting thing is that each medium shapes not only the communication that goes through it but also the process of communication itself. A perfect example is comparing the telephone to what we’re seeing now in electronic mail. With email, we link computers together and can send messages to an electronic mailbox, which people can retrieve at their leisure. While we’re still sending information through wires, the process is fundamentally different.

I love this line in particular: "We’re starting to break out of old habits, and it’s really exciting."

What I’m claiming is that computers are a medium, and personal computers are a new and different medium from large computers. When a new medium enters the scene, we tend to fall back into old media habits. Let’s look at a few transitions from one medium to another: radio to television, and television to the interactive medium of the video disc.

Now, let’s look at the next transition: the optical video disc, which can store 55,000 images on a side or an hour of video, randomly accessible. But what are we using it for? Movies—we’re falling back into old media habits. Still, there are some experiments happening, and you can believe that in five to ten years, it will come into its own. For example, MIT did an experiment here in Aspen, where they photographed every street and intersection in town. Using a video disc and a computer, you could navigate the streets virtually and even see the town in different seasons. While not incredibly useful, it points to the interactive potential of this new medium, which is just starting to break out from traditional uses like movies.

Let’s go back to computers. Right now, we’re in the “I Love Lucy” stage of our medium development. When microcomputers and personal computers first came on the scene, we fell back into old habits. We ran strange languages like COBOL and used them for business accounting—things we’ve been doing on computers historically. It took us about four years to start breaking out of that mindset, and we’re just starting now.

Look at Lisa. Lisa enables someone like me, who isn’t an artist, to sit down and draw artistic pictures using a program called LisaDraw. I can erase, move, shrink, grow, and change textures with ease. I can even use an airbrush effect, making areas darker the more I scrub. With no drawing talent, I can create neat drawings, combine them with words in documents, and send them electronically. We’re starting to break out of old habits, and it’s really exciting.

Would young Steve Jobs have used Character AI to talk to Aristotle?

When I was going to school, I had a few great teachers and a lot of mediocre ones. The thing that probably kept me out of trouble was books. I could read what Aristotle wrote, or what Plato wrote, without needing an intermediary in the way. A book is a phenomenal thing—it gets right from the source to the destination without anything in the middle.

The problem is, you can’t ask Aristotle a question. I think as we look toward the next 50 to 100 years, if we can really come up with machines that capture an underlying spirit, or an underlying set of principles, or an underlying way of looking at the world, then when the next Aristotle comes along, maybe, if he carries one of these machines around his whole life, typing all his ideas into it, then maybe someday, after he’s gone, we can ask the machine: “What would Aristotle have said about this?”

"People do judge a book by its cover" - Mike Markkula

Every computer to date has used a weird type on the screen. As you know, the I’s are just as wide as the W’s—they’re non-proportionally spaced fonts, as we call them. And, um, it’s really been impossible to use multiple fonts on the screen at any given time. Matter of fact, the fonts have been just garbage, and it’s been really impossible to embed any kind of graphics with text.

If you take a look at Lisa, it is totally proportionally spaced text. We have 30, 40 fonts on the screen that come out at approximately 80 dots per inch resolution on the screen and approximately up to 300 dots per inch resolution on a laser printer. And that’s where we are today. What you’re saying is we really want to go to 600 or 800 dots per inch on a laser film printer.

We’re not there yet. But we’re solving the problems of injecting some liberal arts into these computers. That’s what we’re trying to do right now: let’s get proportionally spaced fonts in there, let’s get multiple fonts in there, let’s get graphics in there so we can deal in pictures.

Someone asked how Apple has been able to have such low employee churn:

We feel that for some crazy reason, we’re in the right place at the right time to put something back. What I mean by that is, most of us didn’t make the clothes we’re wearing, we didn’t cook or grow the food that we eat, we’re speaking a language that was developed by other people, and we use mathematics that was developed by other people. We’re constantly taking, and the ability to put something back into that pool of human experience is extremely neat.

At Apple, when you get hired, some people survive and some don’t. But in general, it’s, “Hey, this is the general thing we think we need done. Go figure out what we need, come back and tell us how much it’s going to cost, and then go do it.”

So we’ve got an incredible group of entrepreneurs, and we’re always arguing with each other, but that’s just fine. Out of the 5,000 people we have, most are very independent thinkers. What they really want is the environment where they don’t have to convince 30 other people that it’s the right thing to do. Does that make sense?

It gets harder as we get older, and it’s harder to spend time with everyone and pull everyone together. But we make an attempt to do that. Our feeling is that the day someone at Apple decides they can’t make a difference anymore is the day we’ve lost them.

I highly recommend spending an hour to listen to the whole talk. I published my GPT4o edited trasnscript in full here if you want to read.

I’ve always embraced hackathons as an opportunity to have fun by building something fun with friends. In high school I was lucky enough to attend a few college hackathons. My favorite memory was at UCLA's LAHacks. We wanted to build something playful and fun, and came up with a creative game - how fast can you throw your phone?

During the science fair style demos, the UCLA organizers would identify tables with the most popular projects and move them to the front. I’d taken one of the many beanbags scattered around the hackathon and used it as the throwing target. Before long, we were pulled to the front with this hilarious line of people waiting to throw my iPhone 8 at the beanbag. Unfortunately, our success was short lived after one player managed to miss the beanbag.

Last summer when I was working at Uniswap Labs, we had our biannual company onsite and accompanying internal hackathon. I had an idea to build something physical that would add some character to the new office that we’d recently moved into.

What if we built a photobooth?

And that was the start to my most memorable experience at Uniswap - a 16 hour project that included accidentally shattering a windshield, designing an iPad holster with wood, and coding in a U-Haul across the Williamsburg Bridge.

The goal was simple: have some fun with our hands and build a photobooth.

We also had this crypto idea angle where the photo booth would mint photos onchain. Visitors to the office could take a photo in the photo booth and get onchain proof that they were there.

The first step was to figure out what we needed to buy in order to run the camera and photo printer mechanics. Since there was a minting component, we had to have a computer or device that could run some code. For simplicity, we opted for an iPad with a bluetooth Polaroid style printer.

The second step was the physical design. We browsed Google Images and Pinterest and found one vintage style photobooth that served as our primary inspiration. For ease of transport, we decided to build it in two parts - the main hold and the seat. A shower curtain rod would hold up the curtain, pressed against the two parts.

Home Depot was next. Thanks to Callil's industrial design background, the shopping list was created in realtime. I played the role of cart-pusher as we picked up 2x4 beams for the main structural frame and Fiberboard panels to complete the look.

The first mishap of the project was after we got back from the Home Depot. Somewhere along the drive back on the BQE, we broke one of the fiberboards strapped down to the roof. Not a big deal though, we had more than enough.

The second mishap was unfortunate and a slightly bigger deal. We didn't want the 2x4 beams to be hanging out of the trunk (TBH I forget why, did we think someone was going to steal it?). So we tried to arrange them in the car diagonally and close the trunk. The length seemed to just fit diagonally against the windshield. Slammed the trunk door and crack. Oops.

We all had separate evening plans, so planned to meet back together at 10pm to start the build:

After transporting the photobooth to the office, I had to go return the UHaul in East Williamsburg. I remember riding a white Citibike ebike as fast as I could to Soho, listening to Dominic Fike's newest album - tired and sleep deprived, but feeling content and happy. It was all just an idea in our heads the afternoon before.

I remember this experience so fondly because it reminds me that the most fun you can have is to work on fun projects with friends. There was no rule against building a photobooth for the hackathon. All it took was getting a team together to do it.

I think we should have this mindset across everything that we do. There are no rules against what you can do. Why not have fun with it?

Thank you to Callil, Noah, Eric, Mark, Tina, Chelsy for teaming up to make my little photobooth idea a reality. I hope you had as much fun as I did 🙏

There are 2 stories that have personally inspired me along this theme:

Jeremy Giffon shares a story of being creative in negotiating:

A classic scenario is, a business owner who is selling his business and he's got different offers, and we found out that he's really into Ferraris. It was basically implied that he's going to sell this business for $30 million, $40 million. So it's obviously not a problem of affording it. But what we realized was that like his wife is not going to let them buy a Ferrari on his own self isn't going to let them buy a Ferrari, it's uncouth in his neighborhood or whatever. In the LOI, $25 million and a 488. Why not? There's no rules against that.

And so these are the kinds of things with negotiating where you can just get really creative as long as you're willing to just be a little weird and just be a little vulnerable

Andrew Kortina shares the story of the Lucas uses Venmo NYC subway ad campaign:

I was in our office with Iqram and Neil, our creative director at Venmo. Everyone was dejected and sulking because this ad campaign proposal just didn’t feel right. Iqram said, “If we are going to spend a ton of money on ads, at least we should do something fun, like billboards or taxis.”

Yea, that would be fun, we agreed.

He continued, “Something dead simple. Look at Lucas over there. He makes coffee. Lucas uses Venmo.”

Neil and I looked at each other, excited.

That would be pretty cool.

Neil grabbed Lucas’s photo off of the Venmo team page and photo shopped a Santa hat onto him–I don’t remember at all why. As soon as we had a “Lucas uses Venmo” comp, we hit the streets and started tapping people on shoulder, showing it to them, asking them what they thought. “What the heck is Venmo?”

Any sufficiently advanced technology is indistinguishable from magic.

One of my favorite questions to ask is, when was the last time you used tech that felt like magic?

I think everyone has only a handful of uniquely personal experiences with technology that evoke a visceral feeling of magic. 

The first time I felt the magic of generative AI was in Paris last summer. I was walking around the perfectly trimmed hedges and ornamental fountains of the Tuileries Garden, when I vaguely remembered an interesting discussion from a college seminar about how the design elements of French and English gardens can be traced back to differing philosophies. I forgot what the explanation was, so I pulled out the then recently launched ChatGPT app. 

Maybe this isn’t that impressive to you - but for me at that time, reading the explanation as the text typed out was a magical experience. Just moments after this vague thought crossed my mind, ChatGPT helped me remember what I’d learned back in college. It felt like I was watching a Youtube video buffer on 360p and then switch to 4k high definition. The convenience, the speed, and the brevity of the answer was a completely different experience from opening the Safari app, searching on Google, finding the Wikipedia page, and scrolling to dig out the part I want. 

For the rest of my time in Paris, I used the ChatGPT app as a personal museum/art history/architecture/philosophy guide. It was like having a personal Rick Steves guide who could answer any niche question - a game changer for travel. 

In the last year of consumer generative AI, a haves and have-nots has formed between:

1. People who have tried ChatGPT a handful of times but “don’t know what to use it for”

2. People who have identified a job to be done and what to prompt to accomplish it 

The cause of this divide is the chat box interface.

There has been plenty of discussion on how chat interfaces are not the future. I like how Amelia Wattenberger frames it in her post: “Good tools make it clear how they should be used. And more importantly, how they should not be used.” Presenting a user with a general purpose tool and an empty chat box leads to decision paralysis - “prompter’s block”. 

What can I ask? What should I ask? I heard that prompt engineering is important, how do I do that?  

The future of AI enabled products and interfaces shouldn’t require users to know what special commands or words to use. Interfaces have to move closer to the user and clearly frame the job to be done. 

Some people argue that LLMs are the end of interfaces because now computers can understand natural language, and “agents” (elusive definition) will do everything for us. I think this is wrong - humans are tool users. The best forms of human computer interaction cannot only be language; there is value in visual information hierarchy, interaction design, haptics, sound. It’s telling that babies and animals intuitively understand how to use touch screens. 

Nat Friedman pointed out that it took 20 years of the internet existing before the industry discovered the infinite scrolling feed, which has fundamentally shaped the topography of how we use the internet:

I kind of suspect there’s an inverse correlation between the profoundness of an innovation and the time it takes for it to actually make a difference. The Internet was, it took 20 years to your point, for the feed to come along, or 15 years, wherever it might be, which is the feed to my mind is the core Internet sort of innovation.

That was something that could not be done before to have a dynamically created list of content that never ends, that is personalized to every single individual, that is fundamentally new. It took 15 years to get to it, and that’s what unlocked the advertising model, it unlocked the entire economy of the Internet.

New technologies and capabilities enable native interfaces. We're still learning how to make these very capable models useful to people. What is the version of the infinite feed for AI?

There’s a favorite line from investor/blogger types: "the next big thing starts out looking like a toy."  

Recently I’ve been spending time exploring low hanging fruit around AI interfaces. Lots of energy has been directed towards b2b/workflow use cases, but I’m excited about consumer mobile apps – the surface that we use casually the most. With mobile, we can take advantage of building natural flows around capabilities like the camera, user's camera roll, and location.

I call it "Promptless AI": simple apps that use LLM APIs to solve for a niche use case without making the user type into a chat box. 

Here are 3 toys I've been working on, focused on image as input:

Color analysis is a method that analyzes your skin tone, eye color, hair color to determine which of 12 possible color seasons you are. Knowing your color season empowers you to know what color clothes and makeup fit your complexion. There are professional studios that charge hundreds of dollars for comprehensive color analysis, which has especially taken off in Korea. 

Earlier this year, my girlfriend showed me a viral TikTok “hack” where women used ChatGPT to get a free color analysis. The “hack” entailed taking a selfie, using the digital color picker/eyedropper tool to find the hex colors of your skin/eyes/hair, and pasting it in a predefined prompt to ask ChatGPT what your color season is. These TikTok tutorials were effectively low-fidelity promptless AI apps where you don’t have to think about what to prompt. 

I thought it’d be fun to brush up on my SwiftUI skills and build a vertical experience around using GPT4o’s vision capabilities to get your color analysis: 

you can try this, it's live on the app store - use referral code "COLORS" for a free analysis :)

Take a picture of an artwork to learn about it. It's like a universal museum audio guide. I built a prototype with 3 modes:

1. Museum tour guide

2. Explain like I'm 5

3. Impress my date

I had a lot of fun testing this out at the Met. In its current one-shot-to-claude-sonnet-3.5 iteration, it's maybe 60% accurate at identifying works. Time to make a new "artwork identification eval" for models?

I'm planning on working on this app more - improving accuracy, adding an audio guide mode and more explanation styles, and maybe even a social layer to track your favorite works at museums around the world.

8/22/24 update - I spent a couple weeks polishing this app and now it's live on the App Store

I’d seen the whole male “looksmaxxing” apps take off (take a selfie, get a rating of how hot you are, print $$ off scammy weekly subscription paywall tactics).

There's natural appeal around apps/personality quizzes that tell you about yourself. This prototype has you upload a picture of an outfit, and then tells you:

1. Why it works

2. Ideas to improve

3. Rating from 0-100

I'll admit that this one is a bit of a tarpit idea, but it's cool to have AI identify and tag parts of an outfit. For relatively simple male outfits that are really just about color and cut for your shirt and pants, some of the advice isn't bad.

Have you bought watermelon? Do you know how to pick the best one? Introducing: BestWatermelonPicker.

Never worry about taking home a dud of a watermelon again.

I haven't built this, but I recently saw a meme graphic about how to read the webbings and spots to choose the most ripe watermelon. A use case ripe for a dumb vertical toy app!

UX Arbitrage is when a product presents an opinionated, novel UX on top of a commodity technology (foundational LLM/image/video model APIs). Colloquially, they call them ChatGPT wrappers. 

These "promptless" toy prototypes just scratch the surface of AI interfaces. They're effectively still just a one-shot input output tool, conceptually not all that different from a chat interface (but visually very different!). I've been more interested in the nuanced interaction between user actions and AI capabilities. Jordan Cooper captures this eloquently in his blog post:

Granola’s brilliance is that they acknowledge the truth which is that I don’t want to outsource my thinking to AI. A summary of the meeting, without my input, isn’t what I value or want to reference. Granola gives me a note pad, to jot down my shorthand, but then utilizes the transcript from the meeting to “enhance” my notes with the surrounding context from the meeting. The output is fully flushed out, highly legible notes, where I’ve defined the focal points (where I end) and it has filled in the blanks (where it begins). It’s a very subtle interplay between the user and autonomy that once again strikes this magical balance similar to Tesla.

Each use case that contemplates autonomy calls for a different interaction paradigm between a user and automation, and those that elegantly enable the baton of control to be passed back and forth seamlessly within a given use case will thrive.

Most of the new AI products that have resonated with users take a clearly defined use case and combine it with a clever, often subtle, UX that delivers a magical feeling experience. There is massive value in effectively framing the context for a user to come in and naturally understand what to do. 

Examples that have stood out to me: 

• Cursor: The first time I used Cursor to ask how to fix a complex bug that spanned several files in a codebase - felt like magic. Putting a chat sidebar in the code editor and adding the whole codebase as context was a massive UX improvement to the ChatGPT status quo of copying and pasting code, switching windows, and asking the question. 

• Devin: Split the chat interface in half, show the user how the LLM “thinks” and plans steps, run code in browser and check/debug if it works 

• Claude Artifacts: Split the chat interface in half, and render generated code from the chat directly. 

• TLDraw’s Make it Real: show a drawing interface, tell the user you can draw some web interface, feed the drawing as image input into a vision capable model, and get HTML code. 

• Perplexity: scrape live, up to date search results, keep track of sources/links, display answer to user while clearly displaying the sources of information. 

• Apple Intelligence: embed generative AI inside very specific use cases - email summaries, rewrite text inside existing apps

Each of these examples leverage the power of LLMs in a specific context and make it clear how to use them. I think we’ve just started to explore the first level of UX arbitrage in AI interfaces. What do the second, third, fourth levels of UX arbitrage look like?

What native experiences can we build with generative AI? And how can we bring these magical experiences to more humans?

I'm excited to continue exploring promptless AI interfaces, UX arbitrage, and broadly to build technology that feels like magic (haven't forgotten about you, vision pro). If you're also excited by these things, I'd love to chat. Feel free to DM me on Twitter @spenciefy.

Sitting at the counter with Nakahara-san was a masterclass in learning what mastery and craftsmanship looks like. I didn’t know that beef could be this good.

It started with a prelude of squid sashimi marinated in gochujang, paired with a refreshing bowl of maekgolli. Then followed the symphony of beef – tartare with soy and raw egg yolk, perfectly cut thin sirloin with house tare, his signature gyutan/beef tongue grilled and served as three different parts, ichibo/rump, harami/skirt, shimacho/intestine, and wagyu katsu sando.

Each bite was incredible, but what made it so special was everything surrounding the food. I felt a true sense of warmth from the moment we stepped into the restaurant. The scene was set by elegant yet convivial sights and sounds of staff moving around an open kitchen preparing plates of beef and heating binchotan charcoal. It was the opposite of the stuffy jackets-required fine dining rooms of New York, where eating feels more like a performance of status.

There’s something special about being able to watch a master chef prepare and serve food personally for you. It’s witnessing mastery, but unlike watching a superstar athlete or musician, it’s a full sensory experience – not just taste, but smell, texture, sound, and eye contact.

“Sirloin with tare, one bite”

It was an awe inducing one bite, the kind of “melt in your mouth” experience that everyone seems to travel to Japan for, with a perfect balance of soy and sweetness that complemented the fat marbled decadence. As Nakahara-san was grilling the next course of gyutan, I asked what he puts in the tare marinade.

He looked up briefly and said “STSS.”

I smiled and nodded instinctively as I do when I ask a chef a question that is only partially curiosity, and partially an attempt to demonstrate that I Am Not Like The Other Food Illiterate Diners.

This time the joke was on me. He broke character and laughed, “Super top secret sauce.”

I enjoyed the banter throughout the meal. Context gives food a narrative. It’s this context that makes food content from Anthony Bourdain, Chef’s Table, and Lucas Sin so compelling. Food can rarely be evaluated in a vacuum; the chef’s personal history, the story of each dish and its ingredients, and your dining companions are what makes a food experience whole.

Nakahara-san spoke native English with a distinctive Japanese-American accent, and I learned that he was born close to my own hometown in California before moving back to Japan at age 4. Halfway through the dinner, he brought out a tin with two sticks of binchotan and explained how this special Japanese charcoal made from oak burns smokeless at a lower temperature. After touching and feeling the charcoal, two servers handed us towels to clean our hands. I smiled at how perfectly orchestrated this mid-meal show and tell was.

I asked about why he served Korean dishes – the gochujang squid sashimi and maekgolli that we started with, banchan interspersed throughout the beef courses, and refreshing naengmyeon to end the meal.

“Yakiniku originally came from Korea. It’s like the Tex-Mex of Asian food.“

Nakahara-san shared that he’s working on opening a restaurant in Miami. I asked why Miami, and he said it’s where his business partner lives. I asked if he’s going to serve caviar addons and charge 3x the price, and he just laughed.

I later found a Bloomberg article about Nakahara Miami’s opening plans, and the first sentence is so on brand:

Consider the wagyu katsu sando the Porsche of sandwiches.

I am sure Nakahara Miami will be a successful business.

Growing up, I had an impression that steakhouses were where fancy people went out for dinner. Fancy cuts of beef and fancy bottles of wine painted an image of the capitalist American dream, the kind of place where business deals were made.

Eventually when I learned to be a competent enough home chef to prepare my own steak, I realized that steakhouses are usually a low grade scam - reselling commodity items at insane markup. But of course, it’s not really about the food, it’s about the place.

It’s the type of place where patrons earn cultural and social capital by pronouncing fancy things on the menu that serve as status tests for the sophisticated crowd who know better than to pronounce by sounding it out. Filet Mignon, Sous Vide, Cabernet Sauvignon

These kinds of establishments plague the downtown New York scene - the sceney club restaurant. (and the super-downtown new york scene, aka miami)

Perhaps the state of status dining is best encapsulated by the phenomena of secondary marketplaces where trendy restaurant reservations are sold for hundreds of dollars. Or it’s the popularity of the social food review app Beli among certain circles, which on one hand has a clever ELO based relative ranking system, and on the other hand offers a brilliant medium for you to show off where you’ve dined at (I too am guilty of this).

There’s a concept from American sociologist Erving Goffman called the “dramaturgical perspective”, which is a way to view everyday social interactions through a lens of performance and theater. In the theater of the Status Dining Scene, we are all actors reading scripts in an attempt to give off a certain impression.

A classic performance that plays out is when you order wine. Cue the script:

Waiter brings the bottle of Sancerre you confidently ordered

Waiter: ”Who would like to try?”

Waiter pours you a taste of the wine

You pick up the glass by the stem, give it just a swirl to pretend that you know what you’re doing, stick your nose in, and take a sip

Pause for a second and then smile.

”Yes, that’s great.”

A friend recently hosted a birthday party at a trendy wine bar in Williamsburg, and it was hilarious when the waiter brought out a bucket of orange wines: “we’re going to start with a few different orange wines, and yes they’re all natural and funky”.

My favorite restaurant theatre performance that went embarrassingly off-script for me was when I dined at Sushi Noz a couple years ago.

When the chef placed the first piece of sushi in front of me, I took a picture with my phone rotated 180 degrees so that the camera was on the bottom. The chef saw me do this and asked “do you eat a lot of omakase?” I said no, but my friend is a regular here and I learned this technique from him when we came together. The 8 seat counter at Sushi Noz is an intimate space, and after this interaction the couple sitting next to me said that they’d follow what I did since it seemed like I knew what I was doing.

The waitstaff brought out everyone a small wet towel that’s meant for you to clean your fingers after picking up sushi. I didn’t realize this despite fooling everyone with my “experience”, and committed an omakase faux pas by immediately opening it and wiping my hands. One by one like clockwork, the rest of the counter patrons did the same – monkey see, monkey do.

Moments later the waitstaff awkwardly came around with replacement finger towels and explained that I had misled everyone. We all had a good laugh as I tried to hide my embarrassment.

There’s a joke that every New Yorker acts like they have “a guy” for everything (my halal guy). Towards the end of the dinner, Nakahara-san asked if we wanted to try onigiri made by “my onigiri guy”.

It was like the scene from the Big Short when Jared Vennett meets with Mark Baum and uses Jenga blocks to explain the fragility of mortgage bonds. Look at him. That’s my quant.

I couldn’t refuse the onigiri. Soon enough the “onigiri guy” came and set his canvas and tools out on the counter - a jar of mentaiko, a beautiful tin container with glistening sheets of seaweed, and a wooden hangiri bowl of fragrant rice. He carefully assembled the onigiri and then gracefully handed it to me.

“Beats 7/11 onigiri, right?”

Nakahara-san recounted a story:

“We once served a woman from Shanghai this onigiri. She took a bite and started crying… she said it reminded her of her mother. I asked if she had ever had onigiri before and she shook her head. But I guess there was something homestyle and comforting about it”

This reminded me of one of my favorite movies, Ratatouille. In the final scene, Ego takes a bite of Remy’s ratatouille and is transported back to his childhood in the countryside. Of all the fine dining and fine wine Ego has reviewed, it is the peasant dish that brings tears to his eyes as he is brought back to his mother’s cooking.

Food holds a powerful nostalgic ability. Food made with care and prepared personally is greater than the sum of its ingredients. Food is a vehicle for love and the human experience of taking care of others.

A common theme that some of my favorite food memories share is that I can observe the chef preparing and serving the food. From the chicken over rice smothered in white sauce from the Hooda Halal cart at Columbia to the bun cha eaten on stools in Hanoi to the grilled lamb kabab at Berenjak in London to the late night pad thai on Yaowarat Street in Bangkok to the falafel sandwiches of L’As du Fallafel – this is real food made with care. There is no kitchen door separating the dining room from the chefs in the kitchen; there are no waiters with complex niche ingredient descriptions; the theater of dining collapses to just one human preparing food for another.

I finally realized what made my dinner at Sumibi Yakiniku Nakahara so special to me personally. The sublime grilled beef brought me back to memories of my own childhood. Twice a week, my 13 year old self would come home from an exhausting soccer practice starving for what my mother called by only “Korean BBQ” - fusion Asian diaspora cooking at its best, with thinly sliced short ribs from Costco marinated in premade galbi marinade from 99 Ranch, broiled in an oven for 8 minutes. I have fond memories sitting at my family dinner table alone with sore legs, gorging myself with three or four strips of well charred short rib before starting on my nighttime homework session.

Meals make the society, hold the fabric together in lots of ways that were charming and interesting and intoxicating to me. The perfect meal, or the best meals, occur in a context that frequently has very little to do with the food itself.

Anthony Bourdain, A Cook's Tour: In Search of the Perfect Meal

She references perhaps the first ever starter pack meme, an Apple ad from 1996. It shows two men representing a Windows user and an Apple user - one dressed in a suit and one dressed in jeans. 

This picture captures the “buy a product, buy a lifestyle” phenomenon. The product isn’t even shown in the ad. You are literally buying an ideal lifestyle. 

The starter packs generate the same appeal as the online personality quizzes; someone understands you without knowing you. In the case of aesthetics, this desire to validate and form your identity translates into buying products, clothes, following influencers who align with that lifestyle. 

This is what French psychoanalyst Jacques Lacan described as the “Ideal-I”, the fantasy image of oneself that can be filled in by others who we may want to emulate in our adult lives. Lacan’s concept of the “mirror stage” describes the moment a child first recognizes themselves in a mirror, which sets off the perpetual quest for a coherent identity. 

The internet has exploded the number of possible mirrors. Traditionally role models were celebrities, actresses, singers, which is limited by supply. Internet media platforms have democratized the creation and distribution of influence. The For You page and infinite personalized feed could be counted as the fifth media revolution (printing press, telephone, television, internet… personalized feeds)

Jonah Peretti, founder of Buzzfeed, observes this in a 1996 anti-capitalist essay:

“the increasingly rapid rate at which images are distributed and consumed in late capitalism necessitate a corresponding increase in the rate that individuals assume and shed identities”

Capitalism and Schizophrenia: Contemporary Visual Culture and the Acceleration of Identity Formation/Fissolution, 1996

A couple exploratory questions on consumer media:

1. What does identity formation look like in the current and future of infinite content/personalized media? 

2. What is the taste economy? Is this a long tail coastal elite phenomenon?

I recently decided to wrap up my time at Uniswap Labs after an incredible adventure of two and a half years. I’m eternally grateful for all the people I’ve gotten to meet, for all the places and communities I’ve gotten to experience, and for all that I’ve learned about software, markets, and crypto. As some say, never a dull day in crypto.

I’ve kept a note pinned to the top of my Apple Notes app where I jot down various thoughts and stories I’ve encountered along the way. I’ll saunter through these thoughts and stories — through the lens of three rhetorical questions:

1. Is the music still playing?

2. Can devs do something?

3. Wen token?

I started full time in crypto in late 2021, when the music was literally and figuratively playing. I joined Uniswap Labs to help build a self custody wallet, and when friends and family asked me what I was working on, I’d smile and put on my best impression of [insert your favorite web3 VC]: “the browser is to web2 as the wallet is to web3”. The music consumed my life: I was regularly buying ETH at 4k, checking NFT floor prices first thing in the morning, grinding whitelists on Discord, and on one hilarious occasion, collaborating with online friends to decode an Egyptian hieroglyph puzzle to get early bonding access to an ohm fork (if you know, hat tip).

DAOs were pitched as the future of human social organization. NFTs with 8 lines of text served as the future of building interoperable decentralized games. The global reserve currency was being built and three comma APY rates were the norm. We had entered the ownership era of the internet. It felt like everyone was talking about how web3 was going to change the world.

In retrospect, there were many top signals. During NFT NYC 2021, I was introduced to the fascinating subculture of crypto conference parties. It was surreal to me that pretty much every night of the week, all the nightlife venues around the city were booked out by some crypto company. It didn't matter who was partying, so long as some company was footing the bill, they would rise to the occasion (and get paid for throwing a good party). I remember one Tuesday night, a Telegram group chat I was in was debating whether to go see Kaskade or Sean Kingston.

“Oh, but you need one of those lobster NFTs to claim a ticket to Kaskade. Wait it’s only 0.05 ETH, aping now!”

My personal favorite top signal was at a party during Art Basel in 2021. The first top signal was that I was even at Art Basel. As the story goes, I found myself at a fancy party thrown by a fancy crypto company at a Venetian Island mansion. For 24 year old me, it felt like a movie - platters of lobster tails and trays of margaritas flowing freely amidst conversations of Solana price targets and NFT project ideas. I stumbled into a conversation with a group of three guys, and one of them introduced himself as his rapper name. It was a hilarious interaction. I hadn’t recognized what he looked like before, but after he introduced himself I realized he was an at least B+ list artist who I’d listened to before. He was looking for devs for his own budding NFT project idea. “The metaverse needs drugs.”

I once had a conversation with an investor who held an opinion that all consumer technology is oriented around only 3 things: greed, lust, and gluttony. Who needs the metaverse when you can get all 3 (well, maybe 2) at crypto parties?

It was all so novel and alluring to me. Sometimes in moments of solitude I’d look in the mirror and wonder how this was all happening, whether these narratives and experiences were too good to be true. But the music was playing, people were dancing, and I swept those doubts away to join the party.

You know that feeling when you’re talking to someone at a gathering, and they’re energetically telling you some big idea or opinion, and even though you don’t quite get it, you smile and nod along to match their energy?

Crypto cannot only be viewed as a technology, it is equally a social movement. The evangelical narrative of pitching how we individuals of web3 can take back our power and freedom from the evil big web2 tech companies hit the right nerve for many people, myself included. The cultural context of 2020 set this up perfectly. Amidst various reasons for dissatisfaction, it is only natural to reach for a new utopia of opportunity.

This isn’t new. We can observe changing power dynamics between centralized authority and individual welfare throughout history. Martin Luther’s publication of the 95 Theses in the 16th century challenged the power of the Catholic Church’s centralized religious authority, and my monkey jpegs and dog coins are challenging the power of centralized big tech of the 21st century (jokes). Crypto culture is a form of a contemporary religion suitable for the unique technological and political environment we live in today.

Every good party (and bubble) eventually ends. As the music stopped playing when the market crashed in 2022, so did my motivation. Ideas that had sounded brilliant months prior turned into questions of “wait, why does this need blockchain?”. One failed side project that I worked on with a few friends was in collaboration with an NYC ice cream shop, where patrons could purchase a membership NFT for special benefits like private tastings and the ability to vote for new flavors. We had a spreadsheet to lay out the fixed costs of providing these benefits and how much the corresponding NFT mint price should be to ensure a profitable operation. In March 2022, the =GOOGLE_FINANCE(ETHUSD) number kept trending downwards, and the cultural zeitgeist started to mock NFTs as a tainted word. I remember when a PR firm asked why this new membership had to be an NFT, the best we could muster was a word salad of “interoperability! open marketplace! true ownership of the membership, on the blockchain!”.

Spencer 🧃

@spenciefy

been working on a list of *real use cases for crypto/web3*

1. support artists directly via nfts (and institutional validation of digital art)
2. exposure to USD outside the trad financial system
3. maybe gaming but not quite
4. receive money from the sky

7

9:52 PM • May 31, 2022

I certainly do believe that blockchains are valuable and enable net new use cases. But I learned that when the music is playing, it’s too easy to get caught up in the euphoria and lose intellectually honesty around what we are really building.

note: the next two sections are a tad crypto specific and may feel like a big word salad to those who aren't chronically online in crypto internet spheres

One of my favorite memes of crypto is this copypasta from the 2021 era:

ok I need PRICE TO GO UP. like VERY SOON. I cant take this anymore. every day I am checking price and it is staying the same. every day, check price, same price. I cant take this anymore, I have over invested, by a lot. it is what it is. but I need the price to GO UP ALREADY. can devs DO SOMETHING??

During the “bear” market of 2022, I personally struggled to clearly articulate why I believed in what I was working on in my day to day. A wallet with good UX to view, discover, and trade NFTs and tokens was so clearly valuable when prices were going up, but as liquidity dried up and prices went down, so did my emotional temperature around crypto. Looking back, I can attribute at least part of this to an emotional overcorrection from the highs I’d experienced before. I absolved some of my inner tension by focusing on growing as an engineer and enjoying the process of building something new with a genuinely fun and talented team. I’ll always look back fondly at my time working through in trenches with the Uniswap Wallet team - you know who you are. They taught me that regardless of what you work on, the people who you spend your working days with is the most important.

I remember working on the first iteration of the “top tokens” feature in mid 2022. I’d have VSCode side by side with iOS simulator showing me a list of the top 100 tokens on Ethereum. ETH was hovering around $1,000. The top 100 tokens were more or less stagnant, with all the stablecoins and WBTC (and BNB) taking up the simulator’s visible screen real estate. I’d think to myself, what’s the point of this list if it’s the same tokens and they’re going down? Who even needs real time data on token prices? Why does it matter that we support newly launched tokens immediately?

As it turns out a year later, the top 100 tokens would not be the same and they’re going up, real time market cap data was critical, and I was swapping tokens that were launched minutes prior. Bits and pieces of the speculation mania were back. When memecoins first started to pick up steam in mid 2023, I was still in a lightly skeptical headspace. At social gatherings I’d self deprecatingly tell others “wow, your work sounds like it’s making a difference in the world, I’m just helping random internet dudes buy green frog tokens”.

The average Crypto Twitter permabull would accurately label my stance as “midcurving” the point of memecoins. When PEPE crossed the 1 billion dollar market cap mark, my opinion shifted towards the left side of the bell curve. Sometimes in life, you have to respect the pump.

When your job is to literally make swapping tokens on your phone easier, it’s hard to ignore the memecoin mania. I knew some friends who worked in other non-crypto industries who would sneakily trade memecoins in between tasks at work. On the other hand, if you walked by our team’s row of desks in the office, there’s a good chance you’d see iOS and Android simulators showing $PEPE, $BITCOIN (HarryPotterObamaSonic10Inu), $WSB, $SPONGE. At one point a new narrative was proposed: memecoins should be first class citizens in our wallet experience. Amidst the bear market, we stumbled into what some call the Casino On Mars - possibly one of the funniest casino experiences you can have. I used to laugh at Discord messages “don’t worry, Asia’s waking up!”, and now I laugh at “dev dumped, community takeover!”.

I won’t belabor the debate on whether memecoins are good or bad. I think your opinion on memecoins is directly correlated with whether you’ve made money on them at least once. It is just gambling but the house can either be rigged (rugged) or in your favor (community takeover!). We generally know that gambling isn’t the healthiest for your mental wellbeing. A couple weeks ago, a few coworkers and I had a spontaneous Solana memecoin trading competition. It was an incredibly fun social experience, both making and losing magic internet money together on the most ridiculously named assets.

Biden’s State of the Union is tonight, should we dump jeo boden in case it’s bad?

The more I attempt to trade these memecoins, the more I feel like it’s contributing to my “brain rot”. I used to have a bad habit of watching YouTube in bed before going to sleep, but now I instinctively open the DEX Screener app on my phone. My new morning routine is to check if I got rugged or made it to Valhalla.

I worry about a kind of collective brain rot. How we spend our days is how we spend our lives, and when short term financial incentives give rise to playing short term games, we risk repeating the same cycles over and over. It’s not just memecoins, just look at the phenomena of airdrop farmers. Whenever a project announces their airdrop eligibility checker website, two factions quickly form – those who got a good airdrop, and those who didn’t. If you ever want to see internet chaos and greed, just go to the Discord of a project that announced airdrop eligibility. “E-beggars”, as one project founder controversially put it, isn’t too far off if you saw some of these chat messages. But, in the words of one ex-industry leader: “Some things are better left unsaid. Recommend no more news like these, for the sake of the people, our industry (and your business).“

Spencer 🧃

@spenciefy

one of the most “magical” experiences of crypto is when you catch money falling from the sky, and it’s hard to complain as a rational capitalist

but amidst all memes around e/acc and American dynamism — one has to wonder whether it is leading cr*pto to a bankruptcy of true…

12

7:29 AM • Dec 8, 2023

As the markets have picked back up recently, I’ve found myself thinking back to how I felt in 2021. The music is playing again, people are bullposting their 10x or 100x or 1000x, optimized products for trading feel important. Remember NFT SaaS? Sweeping as a Service? I wonder if the Telegram sniper and trading bots will have a similar outcome. I too have taken advantage of the unique financial opportunity – trading memecoins with Telegram bots, trading keys for gated alfa groups, trading perps for points, bridging and staking for airdrops. Sometimes it feels like the same game, only with a new word salad: “Modularity! DePIN! RWA! Parallel EVM! Liquid Restaking! SocialFi! Points! AI agents!” I can’t deny that it’s amazing to make money on the internet like this. But, and I ask this as a fellow “degen” in the arena, what are we really doing here?

At some point in 2023, I felt like it became consensus opinion in my crypto circles that the true use case of crypto is speculation. I don’t think that’s a bad thing - after all, we are testing and scaling our infra to support faster, better, more efficient speculation for "when the real financial system comes". It reminds me of a conversation I had with a friend working in the AI scene, where apparently some brilliant AI researchers/engineers built an AI girlfriend website not because their end goal is to build AI girlfriend, but rather that it is the one current use case with enough user demand that they can test their inference infrastructure at scale with.

Aside from speculation, I have two more use cases of crypto.

1. Financial rights are human rights. It is important to empower humans in countries with unstable financial systems to be able to hold and use stable assets. It’s funny to me how a lot of this usage lies squarely outside of the US based crypto scene - usage of USDT on Tron and Binance P2P is actually meaningful (or so they say).

2. Tokenization as a powerful narrative to align on a new standard in “tradfi”. “Tokenization” and “Blockchain” have become strong enough meme to push forward alignment on new standards. I’m not totally sure why exactly these have to be on a blockchain versus a centralized database (for private blockchains, what’s the difference?), but in order to push for change you need to have a good story and narrative.

Last year I went to a tradfi oriented event hosted by Coinbase, and I noticed this fashionable narrative of Tokenization being sold. One large traditional asset manager shared this convincing narrative:

“Tokenization is just a technology to take a stock or bond or art or house or music - anything that is a possession in the world and codify it in a way that can be standardized across assets. Frame it this way: there exists a new system that is moving around 1T of value. In the whole world, there is 500T of value being traded. The system is going to take over the rest of the assets because it’ll be a safer and more efficient way to move around assets.”

Later that day at the same event, investor Fred Wilson was on a panel about the intersection of crypto and AI. He commented on how the industries have evolved separately, and this stuck with me:

“The absolute failure in web3 is it has not brought any mainstream use cases into real life. People whose lives have been changed with crypto are believers (people who made life changing money). AI has graduated into being integrated into so many applications that people just accept it, and it benefits them.”

ok I need MORE CRYPTO USE CASES. like VERY SOON. I cant take this anymore. every day I am checking use cases and it is staying the same. every day, check use cases, same speculation use cases. I cant take this anymore, I have over invested [my career], by a lot. it is what it is. but I need there to be REAL USE CASES. can devs DO SOMETHING??

For any non-crypto readers - “wen token” is a common refrain from crypto netizens who (usually annoyingly) ask a company/project when they will launch their token, so they can receive money via an “airdrop” or have the ability to speculate on it.

Perhaps the strongest use case of crypto thus far is to make people rich.

There’s a phrase in the casual crypto Twitter trading world of “pump my bags” that is a beautifully succinct way to capture this kind of activity in crypto.

Mercenary actors who are only optimizing for return on capital are very effective and powerful in crypto. In the last cycle, more sophisticated players have entered the arena. We may joke about that unhinged Twitter anon tweeting things to pump their bags, but the same things are happening on larger scales behind closed doors. The private venture SAFT funding model enables the easiest 100, 1000x returns in spans of 1-4 years. Market makers for hyped new tokens can do very well. Opportunistic founders who are raising money now include a new line in their quick pitch - “TGE Q2 2024” (token generation event in Q2, so you can get your presale tokens and dump on retail).

One problem with the incentives of crypto is that it doesn’t necessarily incentivize building real use cases. I enjoyed reading this intellectually honest blog post from Michael Dempsey about this unfortunate cycle of perverse incentives between founders and venture funds in crypto (well, unfortunate if you are not the founder or the venture fund). He writes that the highest EV thing to do as a builder in a bull market is to build new L1s or L2s instead of applications. Founders who are just after money can raise a massive round for a new chain, promise a token unlock, and then make everyone rich.

We’ll get new blockspace each cycle, with some value erosion happening for underutilized L1s and L2s, and maybe every now and then new value capture that looks a lot like an L1 narrative (Data availability, Restaking, etc.) will insert itself into a slightly smaller flywheel before a wave of copycats and the nth version of these protocols descending down the Euthanasia Coaster of Impossible to Fathom Billion Dollar FDV tokens.

And the best part is, these tokens will still probably be more highly valued than real applications with real usage and real utility.

All this blockspace, what for? But who needs a use case when it can print money? The myth of the infrastructure phase has become the "financial tarpit of the infrastructure phase". Infrastructure for infrastructure for infrastructure for the sake of Impossible to Fathom Billion Dollar FDV tokens.

Spencer 🧃

@spenciefy

financial nihilism

Spencer 🧃

@spenciefy

product market fit is an outdated web2 phenomenon

in web3, all you need is TGE

5

1:14 AM • Mar 11, 2024

In early stage investing, there’s a notion that the bigger your idea and less of a product you have, the more hype and upside there is. An idea can be intangible, immeasurable, but a product with user metrics can judged on what exists. That’s why memecoins can hit billion dollar market caps. There’s no usage metrics or revenue models to judge, and thus there is no cap to the upside. At least people who launch memecoins are intellectually honest that they are not trying to change the world with a new type of blockspace, they are just trying to make money.

I’ve been spending some time diving into the much hyped, technically elusive “Crypto x AI” narrative. Some of the projects seemed promising to me at first, but the further I dug into them, the more I realized they were word salads that sounded good. I am sure some of these companies/protocols will financially do well by crafting the right narrative to pump a token. I was lamenting over this reality to a trader friend of mine, who responded “that’s how crypto works, like clockwork”. Crypto has an uncanny ability to reinvent things under different names and discover for itself what works and what doesn’t. If you’re bullish on something in “Crypto x AI”, please shill me. I want to be convinced of something that makes sense.

When I started at Uniswap, I wrote a kind of tongue in cheek line to conclude my blog post: “it’s time to build, with cautious optimism.”

I’ve since realized that cautious optimism can be a helpful mindset, but it’s not enough to navigate this dark forest of perverse incentives and narrative chasing. Cautious optimism comes with the risk of “mid curving” new technological and societal developments – observing it from the sidelines without actively participating. Perhaps “honest optimism” is what I’m more after.

There’s a Nietzsche line that resonates: “They muddy the water, to make it seem deep”. At the forefront of technology (especially crypto), it can be easy and tempting to muddy the water. I personally hope to participate in building the future of technology in society while staying honest to my values.

As for what’s next for me - just exploring! There are two immediate things I'm working on: finish an essay on taste and internet culture, and continue my streak of shipping a fun consumer iOS app once every few years (anyone into color analysis?). Beyond that, I'm exploring ideas all over from the crypto application layer to immersive memories on the Apple Vision Pro to thinking about the future of how we interact with the internet (something something AI/LLMs). I hope to share these soon. If you’re working on anything interesting or similar, please reach out! I’d love to meet with more smart, honest thinkers and builders. https://twitter.com/spenciefy

It’s time to build, without muddying the waters.

“Whoever knows he is deep, strives for clarity; whoever would like to appear deep to the crowd, strives for obscurity. For the crowd considers anything deep if only it cannot see to the bottom: the crowd is so timid and afraid of going into the water.”

Nietzsche

Sound like Bittensor to anyone?

I'm grateful for everyone at Uniswap Labs who has made my last couple years so special - I wouldn't have traded it for anything else. Shoutout to Connor for getting me there and stewarding my naive euphoria, the early residents of Wall(et) St. of Connor, Will, Justin, Tina, Maggie, Callil that grew into Tarik, Padmini, Ian, Thomas, Christine, Daniel, Chelsy, Adrian, Adrian, Andrew, Allison, and everyone else at the company who entertained me and my random thoughts at the lunch table. 🦄🙏

cover art generated by DALLE3 with this prompt: "generate an image of a watercolor style painting of a bike on a line that looks like a stock chart that goes up and down and up, and music notes and money icons around it"