Here are my favorites with tldr notes, tools, and takes.

• Read on X

• Read on Linkedin

2025 LLM Year in Review by Andrej Karpathy

• Reinforcement Learning from Verifiable Rewards (RLVR): New major training stage added to the LLM production stack after pre-training → SFT → RLHF. Uses automatically checkable rewards (math/code/etc.) to induce reasoning strategies.

• Ghosts vs. Animals: LLMs (unlike humans) are optimized for text + puzzles, not survival. Capabilities spike in verifiable domains but remain uneven elsewhere. Benchmarks became unreliable “benchmaxxing”

• Cursor / New Layer of LLM Apps: Cursor popularized the LLM app pattern: bundling context, multi-call orchestration, vertical GUI, autonomy slider. Are there green pastures for apps? Yes. Labs are generalists. Apps are specialists in verticals by using private data and feedback loops.

• Claude Code / AI That Lives on Your Computer: First convincing demo of what an LLM agent looks like. Runs locally with your environment/data instead of cloud. Analogy: a little spirit/ghost that "lives" on your computer.

• Vibe Coding: Natural language → code crossed usability threshold. Non-programmers can build apps; programmers can build far more. Code becomes ephemeral, free, disposable. “Vibe coding will terraform software and alter job descriptions.”

• LLM GUI: Text is efficient for machines but not for humans; we prefer visual/spatial formats. Who is “actually going to build the LLM GUI?” Nano Banana hints at this by combining text + images + world knowledge

Standing Out in 2026 by Lulu Cheng Meservey

• Everything is fake now: Fake content by fake influencers with fake engagement from fake followers, launching fake products with fake testimonials. Real has never been more precious.

• For comms, 2024 was going direct and 2025 was winning attention. 2026 will be about doing real things.

• Doing real things means: Putting in real effort, Showing real evidence, Real world events and artifacts, Showing up as real humans, Forming real relationships

• “Once you are Real you can’t become unreal again. It lasts for always.”

The GPT-9 Test by Michael Bloch (Quiet Capital)

• What happens to your business when GPT-9 ships?

• Does your biz depend on AI being bad at something?

• Do you have networks effects where the product improves as more people use it?

• Does your biz require physical presence that can’t be automated away?

• Does better AI make your product more valuable, or less?

• Most biz today are “arbitraging a temporary capability gap”

• Build something that lasts.

The Last Moat Standing by fintechjunkie

• If anyone (kid in dorm room) can build your product in a weekend, what's actually defensible?

• Last real moat: An opinionated perspective on the solution

• Building is now easy and fast. Having an informed opinion is hard and takes time.

• Copying opinionated teams is like hitting a moving target.

The disappearing middle of software work by Karri (Linear)

• Middle of software = opening the codebase, booting up the environment, and writing the code

• For a long time, this has been the most important work and where most time was spent.

• This middle is disappearing/thinning thanks to coding agents.

• Understanding the problem, gathering context, and directing agent work become the most important work.

The End of Reusable Software by Sherwood

• Code is now free. No longer need to use existing software. Claude can create from scratch.

• Why create reusable programs? Why not just write one-off for every scenario? Coding agents already do this.

Observability's Past, Present, and Future by Sherwood

• Observability emerged to tame cloud and microservice complexity: distributed tracing + a new reliability mindset that actually worked at first.

• Today we over-collect telemetry and obsess over dashboards, but the real bottleneck is humans making sense of the data, not generating more of it.

• AI is about to flood the world with wayyy more (and messier) software, so we’ll need a new kind of observability that helps us reason about and operate this infinite codebase.

• h/t 1 in every 5 founders I meet in the Bay Area is building an observability platform for agents

Founding of Claude Code + Cowork by Boris Cherny

• Initially launched Claude Code to Anthropic team to dogfood

• Started with Sonnet 3.5 before model was good at agentic coding

• Couple months later, non-eng (research, data sci, design) started using CC daily

• Now people are using CC to “control their oven, recover wedding photos from a busted hard drive, analyze their DNA and medical records, haggle with customer support.”

• Realized they needed to “make it easier for people that want to use the Claude agent for things that are not coding” → Introducing Claude Cowork

How I use Claude Code by Boris Cherny, creator of Claude Code

• Run 1-5 Claudes locally, run 5-10 Claudes on the web (claude.ai)

• Opus 4.5 for everything

• Team shared and updated claude.md

• Start most sessions in plan mode

• Create slash commands for repeat work

• Create subagents for automating common workflows

• PostToolUse hook to clean up code

• Use /permissions > dangerous skip permissions

• Allow Claude Code to use all your tools (MCP) for you

• Verify and/or ralph long-running tasks

• Give Claude a way to verify its work

Scaling long-running autonomous coding by Wilson Lin (Cursor) + Leerob summary

• Cursor ran ”hundreds of concurrent agents on a single project, coordinating their work, and watching them write over a million lines of code and trillions of tokens.”

• Single agents are good for focused tasks but slow for complex projects

• Flat structure of agents failed bc agents became risk-averse and avoided difficult tasks

• Separating into planner and worker roles and judging agents solved coordination problems and allowed cursor to scale to very large projects

• Lessons deploying trillions of tokens on long-running tasks: Model choice (GPT-5.2) matters, removing complexity (unnecessary roles) but the prompts matter most

Coding agents need product agents by Jordan

• Coding is cheap. Decision-making is still expensive.

• More pressure on the part teams have always struggled with most: deciding what to build, why it matters, and staying aligned.

• Coding agents help teams ship faster. Product agents (Async) can help teams ship the right thing.

Shipping at Inference-Speed by Peter Steinberger

• You can now ship code now at a speed that seems unreal. Now limited by inference time and hard thinking.

• Important decisions have become languages, ecosystem, and dependencies.

• It’s getting harder and harder to trust benchmarks. Try multiple models/tools to understand.

Notes on AI Apps in 2026 by Anish Acharya (a16z)

• Thinking tools vs Making tools: Many execution tools exist, but more exploration tools are needed.

• Software eats all the “service” functions in the organization: Agents will replace human service functions (legal, finance, HR)

• Compounding AI apps: Apps that benefit from multi-modal data, proprietary datasets, networks, and ecosystems (e.g., thick apps) will compound.

• Humans discover “the rest” of AI: UX/UI will improve and more consumers will create with AI.

• Notes for (incumbent) CEOs: Collapse customer-facing roles, software-first everywhere, and price boldly. For most enterprise tasks, AGI is here.

LLMs vs. Marketplaces by Dan Hockenmaier

• LLMs are on a collision course with marketplaces

• Collision = User → AI interface → DoorDash → Sandwich delivery

• This is a problem because marketplaces pay back CAC from repeat transactions and would have to spend more per tx because orders are coming from ChatGPT rather than their app

• Marketplace defensibility to LLMs comes from:

• Difficulty of supply aggregation: hotels (easy for LLM) vs airbnb (hard for LLM)

• Degree of management: search and tx (easy for LLM) vs risk and service (hard for LLM)

• Nature of customer engagement: low frequency and high consideration/research (good for llm) vs high frequency and low consideration (bad for llm)

• What marketplaces should do? Do things LLMs wont or can’t.

Consumer AI predictions by Eugenia (Wabi)

• Screenless AI devices will flop: voice is good secondary interface but bad primary, hard to fight our addition to feeds/screens

• “Always listening” devices won’t work either: most things dont matter to record and things that do you wont dare record. granola is good.

• Mini-apps will unlock UGC personal software: full apps are hard to make and heavy to onboard and use, ai coding + mini-apps = UGC software

• By 2030 there will be two big general purpose AI chatbots: today we have cGPT and lots of niche bots, tomorrow cGPT-like assistant and AI friend will be the big ones

• Performance marketing for apps is dead: saturated channels and copycats will push margins to zero. paid acquisition is a boost, not a biz model

• The fastest consumer product to reach $1B ARR will be an AI webcam girl: dropping video generation cost will result in a hyper-personalized 24/7 super OnlyFans

• Whoever solves AI discovery wins: normal people use text input for chat and search. consumer AI winners will unlock hidden beyond search/chat use-cases

Human data will be a $1 trillion/year market by Ali Ansari

• All functions (digital and physical) will be automated.

• Automation pushes humans towards higher-value creative work.

• Frontier AI requires structured human data to learn.

• A lot of time and money will be spent on “expert human data creation or structured human judgment”

21 Lessons From 14 Years at Google by Addy Osmani

• The best engineers are obsessed with solving user problems.** — “User obsession means spending time in support tickets, talking to users, watching users struggle, asking “why” until you hit bedrock.”

• Bias towards action. Ship. You can edit a bad page, but you can’t edit a blank one. — “First do it, then do it right, then do it better.“

• Your code doesn’t advocate for you. People do. — “In large organizations, decisions get made in meetings you’re not invited to, using summaries you didn’t write, by people who have five minutes and twelve priorities.”

• The best code is the code you never had to write.** — “The problem isn’t that engineers can’t write code or use AI to do so. It’s that we’re so good at writing it that we forget to ask whether we should.”

• Focus on what you can control. Ignore what you can’t. — “Dwelling on these creates anxiety without agency.”

• Writing forces clarity. The fastest way to learn something better is to try teaching it. — “The act of making something legible to someone else makes it more legible to me.”

• The work that makes other work possible is priceless — and invisible. — “Glue work - documentation, onboarding, cross-team coordination, process improvement - is vital.”

• When a measure becomes a target, it stops measuring. — “The goal is insight, not surveillance.”

• Admitting what you don’t know creates more safety than pretending you do. — “When a leader admits uncertainty, it signals that the room is safe for others to do the same.“

• Your network outlasts every job you’ll ever have. — “Your job isn’t forever, but your network is. Approach it with curiosity and generosity, not transactional hustle.”

• Most performance wins come from removing work, not adding cleverness. — “Before you optimize, question whether the work should exist at all.”

2025 Year in review by Paul Stamatiou

• Rewind lessons: “prioritize task-based workflows over pure recall, and explore using lightweight visual models for data classification.”

• Limitless pendant lessons: AI wearables have real uses-cases but some “take privacy extraordinarily seriously and lean introverted” and arent excited to wear recording devices

• Who to work with? “exceptionally talented team, at the forefront of AI, with leadership and a CEO who genuinely care about quality, and as little organizational friction as possible between us and an outstandingly well-crafted product.”

• How to be successful? “if successful, will make the rest of my career look like a footnote.”

• Lessons exploring: Designers who code aren't a nice-to-have anymore: they're the norm. People “who weren't already working closely with AI were thinking about leaving their companies.”

• Sesame (hiring) is building lifelike personal agents via software + hardware

• Restraint with coding agents: “Just because you can, doesn't mean you should.” The hard part now is restraint.

New AI tools

• Scott AI: agnetic workspace for eng alignment

• Claude Cowork: Claude Code for non-technical tasks.

• Ralph Wiggum: self-referential AI development loops in Claude Code.

• Worktrunk: git worktree manager for running AI agents in parallel

• CallMe: plugin that allows claude code to call you on your phone.

• Async: slack-based product agent that learns about your company's product, customers, codebase, and team from existing work.

• Universal Commerce Protocol (UCP) — enable commerce inside Google AI products like Gemini

Good AI takes

• Solve problems that won't get solved with the next model update — Maor Shlom (Base44)

• Claude Code is the Gutenberg press moment of software — Sergey Karayev

• Figma is a bottleneck + 1 month of Figma in 5 days. — Ryan Lu (Cursor)

• Claude Code built all of Claude Cowork — Boris Cherny (Anthropic)

• the era of humans writing code is over. — Ryan Dahl (Nodejs)

• Bros building Claude Code setups — John Palmer (Area)

• Curse of LLM interfaces — David Holz (Midjourney)

• Personal software is incredible — Issac

I'm a big fan of anyone who wants to build. Anyone who wants to make more than they take has my respect. That's the main thing you should aim for. To make more than you take. Be a net contributor to society. If you want to create something valuable financially, you don't pursue that. It's best to pursue providing useful products and services.

If you do that, then money will come as a natural consequence of that, rather than pursuing money directly. You can't pursue happiness directly. You pursue things that lead to happiness — fulfilling work, study, friends, loved ones

It sounds very obvious, but generally, if somebody is trying to make a company work, they should expect to grind super hard and accept that there's a meaningful chance of failure. But just be focused on having the output be worth more than the input.

Are you a value creator? That's what really matters. Making more than you take.

In our house, we often say, "What's the goal of life?" The answer is simple. Be helpful. But how do you be helpful? Well, I think the best way to be helpful is to make more than you take.

Making more than you take is obvious from a career perspective. You either join a company or start one to create a useful product or service. If you do that well, you're probably making more than you take.

But it applies to more than just your work. You can make more than you take in all aspects of your life.

Eat at a restaurant a lot? Compliment the chef or leave a glowing review.

Live in a neighborhood you love? Be helpful to your neighbors. Join the HOA board.

Use a great software product at work? Give the team feedback on how to make the product better.

Lurk on a social network every day? Try posting. Follow more people. Reply, like, and boost their content.

Not everyone needs to build rockets.

Every time you consume something, whether content, food, products, or services, you have an opportunity to make more than you take. Being helpful in these small moments compounds. It's the easiest way to make more than you take and be a net contributor to society.

One thing I’ve noticed is that AI applications are a lot like marketplaces. There’s a clear demand side, a supply side, and an interface connecting the two.

The demand side are the end-users of AI applications. For Cursor, its developers. For Granola, notetakers. For Mixy, DJs. For Particle, newsreaders.

The supply-side for AI applications are LLM tokens. These tokens are good at different things, complete tasks at varying speeds, and come in various price points, like freelancers on Upwork. There’s an ever-increasing fragmentation of tokens, much like books or collectibles before Amazon and eBay aggregated them into a marketplace. Tokens and the foundational model companies behind them want to multi-tenant, meaning listed on different and competing applications, and the end users don’t mind, just like drivers on ride-sharing and delivery marketplaces. 

And just like the most successful marketplaces, the most exciting opportunities for AI applications are in non-consumption, not in competing with human labor. They unlock the liquidity of LLM tokens to enable new experiences and expand the demand for labor that was previously inaccessible. It’s the designer generating code with AI, the nurse hiring an AI notetaker, or a student learning from an AI tutor. 

There’s endless talk and fear about AI eating jobs. Creative destruction isn’t new. AI will overtake many jobs, but I think the most exciting opportunity and the thing I’m most optimistic about is that AI is making nearly every job more accessible and infinitely expanding the market for those jobs, just like marketplaces made products and services one-click away. 

If you’re building something new in AI or just trying to make sense of it all, I’d love to hear from you.

More people should have questions they live by. It’s a good way to direct one’s life, and it positively impacts the world by making others more curious. Imagine if every person had a question. The world would undoubtedly be a more curious and interesting place.

Thiel isn’t the only one with a question. Other famous examples include:

Can machines think?

— Alan Turing

What did you get done this week?

— Elon Musk

When I’m 80 years old, will I regret not having tried this?

— Jeff Bezos

If today were the last day of my life, would I want to do what I am about to do today?

— Steve Jobs

What’s the ONE thing I can do such that by doing it, everything else will be easier or unnecessary?

— Gary Keller

I try to live my life with no assumptions untested, so my question would probably be “What assumption haven’t you tested?”

What’s your question?

For whatever it is that you’re creating, a product, song, book, podcast, token, the number of people who want that thing today and in the future dictates your success more than anything else.

Here are some timeless quotes and links on this subject.

Personally, I’ll take the third position—I’ll assert that market is the most important factor in a startup’s success or failure. Why? In a great market—a market with lots of real potential customers—the market pulls product out of the startup. The market needs to be fulfilled and the market will be fulfilled, by the first viable product that comes along. — Marc Andreessen (a16z)

Markets trump all. I want it to be the reverse but to use a surf analogy is the rider skill matters but it is the second order bit after the wave and where the wave is going. — Garry Tan (Y Combinator)

As I look for patterns in successful startups, the more I believe the market, inclusive of timing, is more important than the initial startup idea. People get so enamored with the idea — even putting it up on a pedestal as the be-all-end-all — that they don’t step back and spend enough time assessing the market. … Take any famous entrepreneur. I bet they picked a great market (and timing!), found an opening in the market (an initial idea), and then built a suite of offerings to service that market over many years. … Pick a market, not an idea. — David Cummings (Atlanta Ventures)

I think market is often the most important variable to consider when thinking about an endeavor … For us we think we try to think about the overall size of the market and not in some abstract terms but in terms of how many people are paying for this thing and how much are they paying per year for this thing and then try to think about if you are successful how many of those people can you convert over to your product and how much are they going to pay you and it’s really important to think about both the overall size of the market but also the address ability of the market and how hard it is to reach those people and sort of capture that spend. — Mike Vernal (Sequoia Capital)

We have always focused on the market. The size of the market, the dynamics of the market, the nature of competition because our objective was to build big companies. If you don’t attack a big market it’s highly unlikely you’re ever going to build a big company. — Don Valentine (Sequoia Capital)

When starting a new company, you want to get to monopoly. Monopolies mean you have a large share of a market. How do you get to a large share of a market? You start with a really small market and you take over that whole market. And then, over time, you find ways to expand that market in concentric circles. The thing that is always a big mistake is going after a giant market on day one, because that's typically evidence that you somehow haven't defined the categories correctly, and it normally means that there's going to be too much competition in one way or another. And so, I think almost all the successful companies in Silicon Valley had some model of starting with small markets and expanding. — Peter Thiel (Founders Fund)

I think a much better mental model to have is that you’re serving some market, and then there’s the percentage of the market that you’re serving. And whatever percentage you are not serving, you just haven’t built the go-to-market functions and organization that’s brought the product to those market segments ... What we did not do, but what I wish we did, is six months after launch, we should’ve mapped out the concentric circles of our market.” — Patrick Collison (Stripe)

The only only advice I have about trying to start something big is pick pick a market where it seems like there's some version of the future where it could be big if it works but other than that it's like one dumb foot in front of the other for a long time. — Sam Altman (OpenAI)

You also really want to take the time to think about how the market is going to evolve. You need a market that's going to be big in 10 years. Most investors are obsessed with the market size today, and they don't think at all about how the market is going to evolve. In fact, I think this is one of the biggest systemic mistakes that investors make. They think about the growth of the start-up itself, they don't think about the growth of the market. I care much more about the growth rate of the market than its current size, and I also care if there's any reason it's going to top out. You should think about this. I prefer to invest in a company that's going after a small, but rapidly growing market, than a big, but slow-growing market. — Sam Altman (OpenAI)

Notice I don’t talk about the idea. I think ideas are almost irrelevant … The more important thing is that you pick a large space that you’re knowledgeable and passionate about. And then you will figure out what the right thing to do within that space is.” — Naval Ravikant (Angel List)

Simply put, a company’s choice of markets and M&A is four times more important than outperforming in its markets. This finding comes as something of a surprise, since many management teams focus on gaining share organically through superior execution and often factor that goal into their business plans. … Startups can also learn a lesson from this. Riding market growth in a fast growing market is a lot easier than trying to take market share in a slow growth market. — Jeremy Liew (Lightspeed)

I thought it would be hilarious if there was a TBPN mini app on Farcaster that would allow Farcaster users to watch the segment about Farcaster ... on Farcaster. LOL. I spun up a quick app, added the manifest, updated the actions, deployed it to Vercel, and posted it to Farcaster.

A few days later, my notifications started blowing up, and I saw that John and Jordi announced the mini app live on the show. 

John: And Big news, you can watch TBPN on Farcaster. Someone built a Farcaster mini app ...

Jordi: Which is a very cool novel experience you have on Farcaster. People can create sub-apps within the app, and they live in the feed. I think it's Jayme Hoffman who built this. 

Dan: Are you talking about the pirate feed of TBPN?

Jordi: It's authorized. Fully authorized. 

That day, thousands of Farcasters watched the TBPN livestream through the mini app. Dozens still use it daily. Wild for just a joke. 

Fast forward to last week, coinbrad.eth reached out asking if i could make TBPN work inside Coinbase Wallet. I thought "TBPN on TBA" was such a good meme.

Introducing TBPN on TBA

TBPN on TBA is a mini app for watching the TBPN's daily show on TBA, Coinbase's new app.

TBPN will be livestreaming from A new day one, Coinbase's launch event today, July 16th, at 3 pm PT.

You can watch it through the TBPN mini app on TBA.

And since TBPN on TBA is technically a derivative work, we are legally required to Drink Tap Water.

Support this post on Paragraph to get some $TAPWATER🚰

TBPN

@tbpn

Here's a quick reminder of the official TBPN Derivative Works Clause in our Terms of Service.

420

11:39 AM • Jul 14, 2025

I've spent a lot of time thinking about this. I've started multiple companies and pivoted countless times. Last year, I took some time off to reflect deeply on what to work on next. Recently, I started a new company and chose an idea I'm truly proud of.

I wanted to share the lessons that impacted me while they were still fresh.

1. Pick something you want.

The easiest way to pick a good idea is to pick something you want. If you want it, there's a good chance others will, too, and you'll understand exactly what to build. This is how many iconic companies started.

The best way to come up with startup ideas is to ask yourself the question: what do you wish someone would make for you?
— Paul Graham

The very best startup ideas tend to have three things in common: they're something the founders themselves want, that they themselves can build, and that few others realize are worth doing. Microsoft, Apple, Yahoo, Google, and Facebook all began this way.
— Paul Graham

In the great tradition of startups, we started Apple because we wanted the product ourselves. There is 1 person or 2 people who are both the creator and the market so that they know the marketing requirements intuitively because they're the market themselves. They can have the engineering-marketing meeting in their head and arrive at what the product ought to be.
— Steve Jobs

2. Solve a problem.

No pain, no gain. If there's no problem, there's no opportunity. Don't chase ideas. Notice real problems and solve them.

The way to get startup ideas is not to try to think of startup ideas. It's to look for ("notice") problems, preferably problems you have yourself.
— Paul Graham

You start a company because there's a very important problem to solve that's not getting solved.
— Peter Thiel

The reason I started Tesla is because I think it's important to accelerate the adoption of electric vehicles.
— Elon Musk

I always say, "Make sure you're actually solving a problem." Now there are different ways that can be wrong. It can be solved already, really well, in which case you're not solving a problem. The idea might be great, but you're not solving a problem. It can be a really neat idea with not a lot of people that need that problem to be solved, in which case, it's not really a problem for the world. Maybe we could broaden this by saying, Solve a problem for the world.

— Kevin Systrom

3. Pick a large market.

In hindsight, this might be the simplest and most important advice: Don't pick an idea. Pick a market, ideally one that's large, growing, and that you're part of.

"Notice I don't talk about the idea. I think ideas are almost irrelevant… The more important thing is that you pick a large space that you're knowledgeable and passionate about. And then you will figure out what the right thing to do within that space is."
— Naval

The only only advice I have about trying to start something big is pick pick a market where it seems like there's some version of the future where it could be big if it works but other than that it's like one dumb foot in front of the other for a long time.
— Sam Altman

As I look for patterns in successful startups, the more I believe the market, inclusive of timing, is more important than the initial startup idea. People get so enamored with the idea — even putting it up on a pedestal as the be-all-end-all — that they don't step back and spend enough time assessing the market. … Take any famous entrepreneur. I bet they picked a great market (and timing!), found an opening in the market (an initial idea), and then built a suite of offerings to service that market over many years.
— David Cummings

We have always focused on the market. The size of the market, the dynamics of the market, the nature of competition because our objective was to build big companies. If you don't attack a big market it's highly unlikely you're ever going to build a big company.
— Don Valentine

Most people think first of what they want to express or make, then find the audience for their idea. You must work the opposite angle, thinking first of the public. You need to keep your focus on their changing needs, the trends that are washing through them. Beginning with their demand, you create the appropriate supply.
— 50 Cent via Sama

An entrepreneur's job is to have a hypothesis on what a good market is going to be.
— Chris Dixon

4. Be obsessed.

Startups are impossibly difficult. To succeed, you need to be obsessed to persevere through all the difficult things you must do to make it work. If you're not obsessed, don't do it.

I'm convinced that about half of what separates the successful entrepreneurs from the non-successful ones is pure perseverance. It is so hard. You pour so much of your life into this thing, and there are such rough moments in time that most people give up. I don't blame them. It's really tough, and it consumes your life. I mean, if you've got a family and you're in the early days of a company, I can't imagine how one could do it. I'm sure it's been done, but it's rough because it's pretty much an 18-hour-a-day job, 7 days a week for a while. So unless you have a lot of passion about this, you're not going to survive. You're going to give it up." … "You've got to have an idea or a problem or a wrong that you want to right that you're passionate about. Otherwise, you're not going to have the perseverance to stick it through. And I think that's half the battle right there.
— Steve Jobs

I came up with 40 new business ideas—everything from creating software to setting up hospital chains, since my wife's father is a doctor and has a hospital. Then I had about 25 success measures that I used to decide which idea to pursue. One success measure was that I should fall in love with a particular business for the next 50 years at least. Very often, people get excited for the first few years, and then, after they see the reality, they get tired of the business. I wanted to choose one that I would feel more and more excited about as the years passed.
— Masayoshi Son

I don't like mercenaries, and I don't like mercenary cultures. The missionary is building the product, building the service, because they love the customer, because they love the product, because they love the service. The mercenary is building the product or the service so that they can flip the company and make money. One of the great paradoxes of entrepreneurship is that the missionaries usually end up making more money than the mercenaries anyways.
— Jeff Bezos

5. Be different.

Thanks to AI, it's easier than ever to create—making it harder than ever to stand out. In a world full of noise, don't pick a better idea; pick one that's different.

Tolstoy opens Anna Karenina by observing: "All happy families are alike; each unhappy family is unhappy in its own way." Business is the opposite. All happy companies are different: each one earns a monopoly by solving a unique problem. All failed companies are the same: they failed to escape competition.
— Peter Thiel, Zero to One

Focus on being different rather than better. "I've learned over time that you need to have differentiation. A lot of storytelling is about differentiation. Why are you different than anybody else? You don't want to compete head-to-head. You want to tell the story about why your company is unique. I think sometimes we forget that. We just talk about why we're better than the competition.
— Alfred Lin

6. Be narrow.

There's no focus that's too narrow to start. The best startups begin by serving laughably small groups of people. As my friend Shane Mac says, "Think smaller to go bigger."

Focus on the smallest possible problem you could solve. There's a resistance to focusing from a fear of being trivial. If you get to be #1 in your category, but your category is too small, then you can broaden your scope. Focusing on a small niche has many advantages.
— Ev Williams

Better to make a few users love you than a lot ambivalent. Ideally you want to make large numbers of users love you, but you can't expect to hit that right away. Initially you have to choose between satisfying all the needs of a subset of potential users, or satisfying a subset of the needs of all potential users. Take the first. It's easier to expand userwise than satisfactionwise.
— Paul Graham

I would definitely advise a small startup company to be as narrow and as focused as is possible to be. If you look at the original Amazon business plan, there was no hint of anything other than books in it… I wanted to build an online bookstore, and that was it. But the online bookstore worked better than they thought it would. So Amazon launched music, and that worked better than they thought. Then video, and that worked too. So Jeff sent an email to customers: "I picked about 1,000 customers and I said, besides the things we sell today - books, music, and video - what would you like to see us sell? And the list came back incredibly long-tailed… So it's been one foot in front of the other.
— Jeff Bezos

7. Be you.

Play the hand you're dealt, not the one you wish you had. Pick an idea you're uniquely positioned to solve—one that your entire life has led you to.

You want to look for the intersection of what you're good at, what you enjoy, where you can create value for the world. In my experience if you don't find something at the intersection of those three it's hard to really have impact. Most people just fall into what they work and don't give it much thought. It's worth upfront thought about what you're going to send most of your waking time doing— Sam Altman

In my opinion, the best predictor of whether a startup will achieve product/market fit is whether there is what David Lee calls "founder/market fit". Founder/market fit means the founders have a deep understanding of the market they are entering, and are people who "personify their product, business and ultimately their company.
— Chris Dixon

Understand you are one of a kind. Your character traits are a kind of chemical mix that will never be repeated in history. There are ideas unique to you, a spell of rhythm and perspective that are your strengths, not your weaknesses. You must not be afraid of your uniqueness and you must care less and less what people think of you.
— 50 Cent, The 50th Law

Your time is limited, so don't waste it living someone else's life. Don't be trapped by dogma - which is living with the results of other people's thinking. Don't let the noise of other's opinions drown out your own inner voice. And most important, have the courage to follow your heart and intuition. They somehow already know what you truly want to become. Everything else is secondary.
— Steve Jobs

8. Be useful.

Pick an idea where you can be helpful to other people. That is the whole point of startups. Your success is determined by the number of people you help and how much you help them.

When one serves for the sake of service—for the satisfaction of doing that which one believes to be right—then money abundantly takes care of itself. … Concretely, what I most realized about business in that year-and I have been learning more each year without finding it necessary to change my first conclusions-is this: 1. That finance is given a place ahead of work and therefore tends to kill the work and destroy the fundamental of service. 2. That thinking first of money instead of work brings on fear of failure and this fear blocks every avenue of business-it makes a man afraid of competition, of changing his methods, or of doing anything which might change his condition. 3. That the way is clear for any one who thinks first of service-of doing the work in the best possible way.
— Henry Ford, My Life and Work

Whatever the thing is that you're trying to create what would be the utility delta compared to the current state of the art times how many people it would effect. Having something that makes a big difference but effects small to moderate number of people is great as is something that makes even a small difference but effects a vast number of people. The area under the curve would be roughly similar for those two things. So it's really about trying to be useful.
— Elon Musk

How strange is the lot of us mortals! Each of us is here for a brief sojourn; for what purpose he knows not, though he sometimes thinks he senses it. But without deeper reflection one knows from daily life that one exists for other people — first of all for those upon whose smiles and well-being our own happiness is wholly dependent, and then for the many, unknown to us, to whose destinies we are bound by the ties of sympathy. A hundred times every day I remind myself that my inner and outer life are based on the labors of other men, living and dead, and that I must exert myself in order to give in the same measure as I have received and am still receiving. … A man's value to the community depends primarily on how far his feelings, thoughts, and actions are directed towards promoting the good of his fellows.
— Albert Einstein

9. Think in first principles.

Don't think in analogies by picking an idea that is "X for Y" or because someone says, "A is hard and B is easier." Instead, think in first principles and choose an idea based on fundamentals—an important problem for a clear customer in a growing market that you're born to solve.

The paradox of teaching entrepreneurship is that such a formula (for innovation) cannot exist; because every innovation is new and unique, no authority can prescribe in concrete terms how to be more innovative. Indeed, the single most powerful pattern I have noticed is that successful people find value in unexpected places, and they do this by thinking about business from first principles instead of formulas.
— Peter Thiel, Zero to One

It's mentally easier to reason by analogy rather than first principles. But first principles is kind of a physics way of looking at the world…what that really means is that you boil things down to the most fundamental truths…and then reason up from there…that takes a lot more mental energy.
— Elon Musk

10. Live in the future.

Good ideas are gaps in the world, and it's hard to see them if you don't live in the future. Live in the future by using the newest technology, spending time with smart people, and observing what's missing that you wish existed.

Paul Buchheit says that people at the leading edge of a rapidly changing field "live in the future." Combine that with Pirsig and you get: "Live in the future, then build what's missing." That describes the way many if not most of the biggest startups got started. Neither Apple nor Yahoo nor Google nor Facebook were even supposed to be companies at first. They grew out of things their founders built because there seemed a gap in the world.
— Paul Graham

An inflection is an event that creates the potential for radical change in how people think, feel and act. An insights is a non-obvious truth about how one or more inflections can be harnessed to radically change human capacities and behaviors. If you have an insight your initial idea doesn't have to be right.
— Mike Maples, Pattern Breakers

11. Bet with conviction.

Pick an idea you believe in so deeply that you'd bet everything on it—even if no one else thinks it's a good idea yet. If you don't feel a "hell yes," then it's a no.

We said okay they're going to try a lot of things and we've just got to pick one and really concentrate and that's how we can we can win here. Most of the world still does not understand the value of like a fairly extreme level of conviction on one bet. That's why I'm so excited for startups right now it is because the world is still sleeping on all this to such an astonishing degree.
— Sam Altman

And here is the prime condition of success, the great secret-concentrate your energy, thought and capital exclusively upon the business in which you are engaged. Having begun on one line, resolve to fight it out on that line, to lead in it, adopt every improvement, have the best machinery, and know the most about it. The concerns which fail are those which have scattered their capital, which means that they have scattered their brains also. They have investments in this, or that, or the other, here, there and everywhere. "Don't put all your eggs in one basket" is all wrong. I tell you "put all your eggs in one basket, and then watch that basket." Look round you and take notice; men who do that do not often fail. It is easy to watch and carry the one basket. It is trying to carry too many baskets that breaks most eggs in this country.
— Andrew Carnegie

The first thing I heard when I got in the business, not from my mentor, was bulls make money, bears make money, and pigs get slaughtered. I'm here to tell you I was a pig. And I strongly believe the only way to make long-term returns in our business that are superior is by being a pig. I think diversification and all the stuff they're teaching at business school today is probably the most misguided concept everywhere. And if you look at all the great investors that are as different as Warren Buffett, Carl Icahn, Ken Langone, they tend to be very, very concentrated bets. They see something, they bet it, and they bet the ranch on it. And that's kind of the way my philosophy evolved, which was if you see – only maybe one or two times a year do you see something that really, really excites you… The mistake I'd say 98% of money managers and individuals make is they feel like they got to be playing in a bunch of stuff. And if you really see it, put all your eggs in one basket and then watch the basket very carefully.
— Stanley Druckenmiller

12. Choose good quests.

Startups are hard, and life is short. If you're going to spend the best years of your life working on one, pick an idea that matters—one that leaves the world better than you found it.

What's your quest? Quests tend to manifest as an objective we center our lives around. Your quest might be to reach a specific milestone: to become a senator, to publish a book, to make a million dollars. But not all quests have an end state. You might be on a quest to maximize your net worth, or to bench-press more weight than anyone else at the gym. Maybe you're just on a quest to have the most fun possible before you die. A far cry from refrigeration or running water, but a nice life. As you've probably already intuited, not all quests are created equal. In the most simple terms possible: a good quest makes the future better than our world today, while a bad quest doesn't improve the world much at all, or even makes it worse.
— Trae Stephens & Markie Wagner

13. Take your time.

A successful startup takes years. It's better to spend an extra month choosing the right idea than to waste years on the wrong one. Don't rush the pick.

If you want to do great work, you must work on important problems. Great scientists made time to carefully think about the most important problems. Average scientists spent almost all of their time working without carefully considering the importance of problems.
— Michael McGuiness

Picking the right thing to work on is the most important element of productivity and usually almost ignored. How do you think about it more? Leave enough time in my schedule to think about what to work on. Reading books, hanging out with interesting people, and spending time in nature.
— Sam Altman

I am willing to take as much time as needed between projects to find my next thing. But I always want it to be a project that, if successful, will make the rest of my career look like a footnote.
— Sam Altman

Startup founders often treat fundraising as a solo sport. They guess when to raise money, how much, and how to go about it. They craft their pitch in isolation and reach out cold to investors, which leads to wasted time, ignored emails, and missed opportunities.

Fundraising, when done right, is a team sport.

You learn the market conditions from other founders and investors. You get feedback on your pitch, investor leads, warm intros, and recommendations on investors as potential partners. You vent to trusted founders who remind you that the struggle is normal. You negotiate and close with support from your fundraising friends, investors, and lawyers. Almost nothing in fundraising is truly solo.

I always tell founders to lean into this team aspect of fundraising. Make “fundraising friends.” At any given time, there are always founders raising or who have just finished. Find them. Create a group.

Being part of a small group of founders fundraising at the same time is one of the best things you can do. Your pitch will improve, you’ll 10x your investor pipeline, and you’ll feel less alone during one of the hardest parts of building a company.

Here are some additional fundraising links I frequently share.

Run a process.

• Andrew Farah (Density) has a great video on running a process.

• Travis Kalanick (Uber) on running a process.

• Watch all the YC videos on fundraising.

• Paul Graham on how to raise money.

• How to get intros from friends?

Keep your deck simple.

• The Sequoia pitch deck template and investment memo on YouTube are great inspirations for creating a simple and concise pitch deck.

• Vinod Khosla gives a great talk on pitching the way VCs think.

• Kevin Hale on designing a better deck — What are the 5 to 7 most important ideas people should know about your startup?

• How to pitch your seed stage startup via Michael Seibel — "A lot of people think that they need to bring a bunch of energy and pizzazz and sizzle and shark-tankiness to a pitch. You don't. You actually stand out by being concise and easy to understand.”

• YC Series A Memo Template

Send investor updates.

• Send investors updates every month. Here’s a simple template from Dan Romero.

Sharing a few "co-founder lessons" here in case they are helpful to others.

Don’t co-found a company with a remote co-founder

Co-founding a company and building a magical product is like making love, having a child, and raising it. It's hard to do this successfully from a distance.

It's hard because, in the early days, startups are all about learning and moving fast. Working remotely, especially from a distance, reduces your ability to communicate, share information, and collaborate. You can't turn around in your chair and say what if we did this, you can't vent about what's not working on the way to lunch. You can't see the frustration, confusion, or excitement in each other, which is important information that fuels additional conversations and leads to better decisions. In other words, working remotely slows you down when you should be trying to figure out ways to move faster.

Perhaps you work remotely with your co-founder right now and are doubting this. Well, ask yourself: If there were two versions of your team, one fully remote and one in-person, which would you bet on? If you've worked together in person and remotely, which time did the startup move faster?

I'm not saying that all remote work is bad. Working remotely might be better for you if you're an individual contributor. You have a specific task where you're heads-down researching, designing, coding, etc. Working remotely might be better for you because you don't have distractions from other colleagues or the bureaucracy of the company.

If you're trying to create something new with partners, whether as parents, co-founders, a small team within a big company, or a group of executives, you should do it in person if you take what you're creating seriously.

We tried hard to make it work over the years, and tried recently to start remotely and work towards in-person. Unfortunately, it became clear that this was too challenging, both in the short term and the long term.

Be aligned on your worldview

Co-founders should have a shared vision of the world and excitement for a group of customers, behaviors, and problems. These beliefs form your mission. A lack of alignment here creates issues and limits your options.

For example, I have been a crypto holder, user, and believer for years. In the past, I worked with co-founders who didn't own or use crypto and weren't interested in it. As a result, I set aside my interests to make something work with people I believe in and enjoy working with. This is a mistake. Life is too short not to explore your interests.

If you're interested in AI coding, work with people who use AI coding tools and believe AI coding will be a big deal. If you're creating YouTube videos, work with people who watch a lot of YouTube and think that YouTube will continue to be an important platform in the future.

Being aligned doesn't mean everyone on the team must fit the persona of the person you're trying to create for. For example, if you're building software for nurses, it's unlikely and probably not optimal that everyone on the team will be a nurse. But it's crucial that everyone on the early team is excited about serving nurses and believes that's important work.

Partner with people who push you

After doing a couple of startups and watching dozens of friends start companies, you realize that building magical products and enduring companies is nearly impossible. If you partner with other people to achieve the impossible, they should push you, and you should never have to pull them along. The journey is too difficult for you to have to pull them and also push yourself.

My friend, Mike Rosengarten, taught me that the ideal scenario is when the person moves so fast that they push you to the point where you feel uncomfortable and have to slow them down. If you're not slowing them down and don't feel pushed, you probably don't have a partner. You have an employee.

First appeared on Farcaster

The good: I felt more focused and got more done. My mind was quieter, and I was more present with family, friends, and nature.

The bad: I felt more alone. My thoughts were siloed and unsharpened, and I felt like I was learning slower. I missed my internet people and learning in public.

I think the system that works well for me is using social media only on my desktop when working to learn faster and be helpful, and keeping social apps off my phone to be more present in person.

It's good to be back!

First appeared on Farcaster and X

We pivoted dozens of times but failed to achieve PMF because we didn’t find and fall in love with a good customer and problem; we wasted too much time on low-frequency and unimportant problems and found that remote work sucks for startups.

We started the company in the fall of 2020 and went through YC W21 to help people relocate during the pandemic by connecting them with local experts. We pivoted to crypto in 2022. Here’s what we built and learned.

What did we build?

• Newzip — an online marketplace for neighborhood advice. Connect to local experts through on-demand calls and livestreams to answer questions like, “What neighborhood should I live in?” or “What school should I send my kids to?”

• We thought consumers would pay for local expertise, and we could build the next Airbnb by helping in-demand local experts monetize their knowledge. We acquired hundreds of local experts and influencers in dozens of cities and sold thousands of dollars in local advice. We ended up selling Newzip to friends at Realeaux Inc., who are doing great things with it.

What did we learn?

• Solve a frequent problem: The number of times a consumer has pain high enough to want to pay for this type of advice (moving cities, changing schools) is extremely low. And the amount of money they were willing to pay was low. This made it difficult and costly for both us and the supply side to find, acquire, and keep customers. Low frequency can work if it’s high margin (Airbnb), and low margin can definitely work if it’s high frequency, but low frequency + low margin is very difficult.

• Demand is the game: In a two-sided marketplace like Newzip, acquiring both supply and demand is critical. We got pretty good at finding and acquiring high-quality suppliers, but without strong demand, suppliers weren’t satisfied and churned. Driving demand is key to retaining the supply side.

• Don't be afraid to change your idea: We hesitated to pivot due to our attachment to our initial “YC idea” and sunk cost bias, resulting in a lot of wasted time. We should have pivoted months sooner when we lacked conviction.

• Take big swings vs. incremental ones: We wasted months exploring two slight pivots in this direction (On-demand local tours, Tiktok for locals). When we decided to pivot, we should've committed fully to exploring big, bold options rather than incremental SISPy ones related to our initial idea.

What did we build?

• Offload — Marketplace for excess building supplies. Save money buying doors and windows from wholesale building suppliers’ excess materials.

• While renovating my house, I discovered you could buy doors and windows cheaply from the back of the warehouse. We thought we could help suppliers sell their excess to home improvement buyers and professionals looking to save. We launched a simple MVP and helped one supplier sell $10K in doors.

What did we learn?

• Understand user incentives: Small, old-school building suppliers had no incentive to sell their excess material. They didn't want to cannibalize their primary sales and found it to be too much extra work. Local liquidators ended up being a better solution.

• No-code tools can take you far: We weren’t sure if these suppliers wanted to sell their excess building materials, and we didn’t want to waste time building a product until we saw some proof. We built the entire MVP using Airtable in a day. I'm still shocked at how far you can get with Airtable, Retool, and now with all the GenAI tools.

What did we build?

• Maindrop.xyz — Venmo for sending and receiving NFTs. Send NFT gifts to anyone with a wallet, email, or phone.

• Sending and receiving NFTs is a bad experience, so we thought a solution could be a wedge into the large greeting card and gift card markets. We processed thousands of transactions and provided users with fun tools for creating and sending NFT gifts.

What did we learn?

• Beware of fake vs. real trends: While I’m still a big believer of NFTs, the number of active NFT users at the time was very small. The number of them actively sending and receiving NFTs was even smaller. If you're solving problems around an emerging trend, make sure it's a real one.

• Find a frequent customer: We couldn't find a user type who frequently wanted to send NFTs. We explored NFT communities, traders, and crypto marketers. We learned that the largest greeting card companies get around low frequency by primarily serving retailers and getting good at marketing a few days a year (Valentine’s Day, Christmas, Mother’s Day).

• Avoid incremental solutions: Even though the experience was subpar, you could send NFTs using OpenSea, Metamask, Rainbow, Coinbase, etc. Our product was fun to use, but it wasn’t 10x better or different enough, making it difficult for even our most loyal fans to develop a habit around it.

• Set test deadlines: Towards the end of this product, we got a lot smarter and set deadlines for our tests. This simple change forced us to face reality and find ground truth much faster. I think this is part of what makes Y Combinator's 3-month sprint so powerful.

What did we build?

• Launchcaster.xyz — Crypto Product Hunt. Launch and discover the best crypto projects by simply mentioning @launch on Farcaster.

• Launchcaster was never meant to be a big thing; it started as a side project. Over time, we thought it could evolve into a Steam or App Store Connect for crypto apps. We grew Launchcaster to thousands of crypto projects and over 14K crypto builders. It continues to be a valuable public good.

What did we learn?

• Solve a venture-scale problem: Hacker News, Product Hunt, and Techmeme were not venture-scale. We quickly learned this when we started focusing on Launchcaster. The number of people interested in consuming this type of content frequently is very small, and finding an aligned business model for the community can be tricky. The best business model for these communities is often to have them owned by someone else (e.g., YC runs HN, Hyper runs PH).

• Kill your darlings: I’m so glad we launched this side project. Building the community and brand gave us a small edge in crypto and changed my life. The mistake we made was revisiting it multiple times after concluding we didn’t see PMF or a venture-scale path. We never got new information that changed the opportunity. It was more that we struggled to find PMF in a difficult crypto market, and Launchcaster was kinda working on a small scale. That “kinda working” state distracted us much more than it should’ve.

• It’s hard to find PMF on customers who haven’t found it themselves: We repeatedly tried to serve crypto projects with Launchcaster by offering App Store Connect-like CRM and analytics tools. We struggled to find a winning path due to the lack of crypto projects with real PMF. Every direction we explored felt like a house of cards.

What did we build?

• Launcher.xyz — a platform for crypto mini-apps. Create and share a crypto mini-app on top of any smart contract by pasting a transaction link or entering custom functions.

• We thought the lack of things to do in crypto was a problem. To address this, we released two apps. The first was a progressive web app and token-curated registry where users bet on apps they wanted to be built. This violated securities laws and had some other flaws, so we pivoted to making it easy to build and remix apps on existing smart contracts. We thought this could become a Dune-like platform for crypto mini-apps.

What did we learn?

• Listen to what people do/pay vs. say: We spent weeks talking to hundreds of users and demoing the product. Everyone raved about it and told us all the different ways they wanted to use it. Come launch, we couldn’t get a soul to use it. Take what users say with a grain of salt. I like the advice that a potential customer’s opinion is not valuable unless they’re willing to pay you (financial capital, time capital, social capital), and there are lots of great lessons on this in The Mom Test.

• Avoid nice-to-have: The product didn’t solve a major problem for the users we talked to. They already had their interface, and they used Etherscan write features to fill in the gaps. While our product was cool, it was more of a nice-to-have than a 10x improvement on Etherscan.

I wrote a full post on Frameboard lessons here.

What did we build?

• Frameboard.com — A crypto Pinterest where you can earn on your curation. Curate anything in web2 or web3 via URL, image upload, or browser extension, and earn when users collect it.

• We thought crypto could empower higher-quality content curation and that users would support great curators by collecting (paying for) it. Since launching in March, hundreds of users curated over 6,500 posts to 800 boards, with thousands more consuming those boards on the web app, Farcaster, and XMTP.

What did we learn?

• Avoid SISPing: The first mistake we made with Frameboard was approaching it from "solution in search of a problem." At this point, we were out of ideas and desperate to figure out what to build next. We let impatience guide us, brainstorming crypto and web3 socical ideas instead of identifying a problem and reasoning from first principles. No surprise—we ended up with mostly bad ideas and a few that just sounded good.

• Solve the right problem for the right user: While cool and fun to build, Frameboard was the wrong product for today’s crypto audience. Active crypto users prefer text-based content, are time-constrained, and mostly driven by speculation. Frameboard was more visual, time-intensive, and lacked speculative elements. We didn't solve a real problem the market wanted solved and built something people didn’t need. Don’t make that mistake—be brutally honest about who your users are, what their behaviors are, and what they truly want.

• Don't forget about demand: Like many web3 ideas, Frameboard sounded great from the supply side (i.e., the curator) but was weak on the demand side (i.e., the collector). Every creator wants more money and control—that’s a given. The real question is: why will the demand-side users spend their time and/or money in a meaningful and frequent way? We struggled to answer that. Without demand, supply-side users—whether content creators, platform developers, marketplace sellers, or service providers—won’t be satisfied and will churn. As I mentioned above, demand is the game and key to providing a magical experience for the supply side.

• Take your time pivoting: In hindsight, we didn't spend enough time between pivots. We were eager to build and ship (always be launching), and didn’t like waiting around. Looking back, I wish we had taken a little more time to research, understand users and their problem, and develop a strong hypothesis that we were going after a big, important market and be missionaries about it. I think in crypto, more so than other industries, teams really need to be grounded in truth and mission-driven to navigate all the skepticism and market swings.

Fall in love with a customer and problem

Ultimately, we failed to reach PMF because we didn't find and fall in love with a good customer and problem. I never forgot the advice from Michael Seibel as we were pivoting: "Who can you give a shit about for a long time?" We struggled to get excited about serving real estate agents and infrequent movers. When we pivoted to crypto, we enjoyed serving crypto users and builders. But over time, we realized most current crypto users just wanted to passively own digital assets, while the active ones wanted to speculate. We weren't drawn to the speculative use cases. We loved (and still love) helping crypto builders, but few have PMF, making them not the best customers. The experience drove home the importance of falling in love with a problem and serving good, rich customers.

Make sure the team fits the pivot

I got into crypto in 2014 and NFTs in 2017. When we decided to pivot to crypto, I was the only one on the team who actively used and owned crypto and NFTs. The team was excited about it because I was, not because they believed in or used it. In hindsight, the team probably wasn't the best fit for crypto. I think this hindered our ability to find PMF and ultimately led us to shut down instead of continuing to pivot into crypto. If you're pivoting to a new problem space, make sure the current team is a good fit for it.

Remote work sucks for startups

Our team was fully remote. We got pretty good at it, but I have mixed feelings about remote work. It can work well when everyone knows what they're doing (post-PMF), allows you to recruit more and better talent, and works ok during COVID. But I question if it's the best path for finding PMF and building a great, enduring company. Our team tried to address this by meeting regularly, working out of the same office or Airbnb, and living in the same city for a few months. When we did, it was magical, and you could feel how much faster we moved. Looking back, I wish we met monthly and pushed my co-founder or myself to move to the same city. I'm certain this would have increased our odds. If you're pre-pmf and working remotely, I recommend stopping and doing whatever it takes to be in the same room as much as possible.

Build in public

When we pivoted to crypto, we got frustrated with the fragmented audience. I wanted to control our destiny and committed to building our own audience. I posted on social media daily for years and built an audience across Farcaster, newsletters, X, and Telegram. This audience gave us such an advantage when acquiring users, getting feedback, fundraising, recruiting, and more. I can’t recommend this enough — it’s not hard, and you already have more content than you think.

Always be launching

For years, we shipped weekly and often daily. We became good at announcing those launches, no matter how small. This habit allowed us to gain traction quickly in everything we did. It made the team feel like we constantly had momentum, which made us ship even faster. It helped us build trust with users and gave us endless feedback.

Get good at the marathon

We got really good at sprinting and making a ton of progress in a short period with limited resources. However, startups are a marathon. No matter how good you are at sprinting, you will lose if you sprint in the wrong direction. Before you sprint, make sure you’re heading toward something important that you care about and can have conviction for a long time.

Of everything we built, I’m most proud of the team and community. Huge thanks to everyone for your support over the past four years.

I want to start another company in crypto. I've never been more determined to build great products and make an impact — I made so many silly mistakes over the past four years. I'm excited to get back up and build on all these lessons, but I'm taking my time. Startups are a grind and, if you’re lucky, a 10-20 year journey. Whatever I work on next, I want it to be mission-driven and commit for a decade and beyond.

In the meantime, I just want to be helpful. Don’t hesitate to reach out if there’s anything I can help with.

Farcaster: @Jayme

X: @JaymeHoffman

The original idea for Frameboard was a Pinterest-like app and protocol where users could earn from their curation. We thought crypto could empower higher-quality content curation and that users would support great curators by collecting (paying for) it.

I came to this idea after thinking a lot about:

• What products don’t give users skin in the game (i.e., Pinterest)?

• What could grow quickly on new web3 social primitives (i.e., frames, bots, NFTs, XMTP)?

Since launching in March, hundreds of users curated over 6,500 posts to 800 boards, with thousands more consuming those boards on the web app, Farcaster, and XMTP. We saw Frameboards for travel, food, fashion, art, train stations, mood boards, NFTs, erc20s, and more.

We experimented with a ton of features, including Farcaster frames and cast actions, browser extension, OpenFrames and XMTP notifications, NFT curation, collectible boards, AI curation, and more.

By May, it was clear the product wasn’t working, and our core assumptions were wrong. 99% of users churned, no one wanted to collect boards or the NFTs curated on them, and we grew skeptical of people collecting content as a big frequent consumer behavior. I remember walking around a rainy London one afternoon, thinking, “Damn, we built the wrong product for crypto.” In that moment, I felt like I trully and painfully understood what product-market fit is—and what it isn’t.

1. Avoid SISPing

The first mistake we made with Frameboard was approaching it from "solution in search of a problem." At this point, we were out of ideas and desperate to figure out what to build next. We let impatience guide us, brainstorming crypto and web3 socical ideas instead of identifying a problem and reasoning from first principles. No surprise—we ended up with mostly bad ideas and a few that just sounded good.

2. Solve the right problem for the right user

While cool and fun to build, Frameboard was the wrong product for today’s crypto audience. Active crypto users prefer text-based content, are time-constrained, and mostly driven by speculation. Frameboard was more visual, time-intensive, and lacked speculative elements. We didn't solve a real problem the market wanted solved and built something people didn’t need. Don’t make that mistake—be brutally honest about who your users are, what their behaviors are, and what they truly want.

3. Don't forget about demand

Like many Web3 ideas, Frameboard sounded great from the supply side (i.e., the curator) but was weak on the demand side (i.e., the collector). Every creator wants more money and control—that’s a given. The real question is: why will the demand-side users spend their time and/or money in a meaningful and frequent way? We struggled to answer that. Without demand, supply-side users—whether content creators, platform developers, marketplace sellers, or service providers—won’t be satisfied and will churn. Demand is the game and key to providing a magical experience for the supply side.

4. Take your time pivoting

In hindsight, we didn't spend enough time between pivots. We were eager to build and ship (always be launching), and didn’t like waiting around. Looking back, I wish we had taken a little more time to research, understand users and their problem, and develop a strong hypothesis that we were going after a big, important market and be missionaries about it. I think in crypto, more so than other industries, teams really need to be grounded in truth and mission-driven to navigate all the skepticism and market swings.

For those interested in curating or building around curation, here are some tools and ideas worth exploring:

Curators

• Paragraph is the best place on the internet for creating, monetizing, and owning curated content—whether it’s text, links, or visuals. The team is world-class and the product keeps getting better.

• Farcaster channels are currently the most underexplored tool for community curation. Rounds.wtf and Fiids are early glimpses of what's possible.

• Zora is great for visual and art curation. With no fees, a mobile app and a more accessible experience, the opportunities here feel endless.

• For non-crypto curation, Arena, Cosmos, Mix, and Crate are all giving Pinterest and LTK a run for their money with AI and monetization tools.

Developers

• Side projects: I recommend exploring Farcaster channels. Curating text, links, nfts or even memecoins is so much more fun and interesting when it's done as a community. I'm certain entire apps and website will be built around single channels. I also think Zora protocol is underexplored for curation. Mint.Fun was amazing when it first launched, but as low-cost NFTs grew L2s, it became difficult to filter signal from noise. Daylight is making a lot of progress here. Similarly, there could be opportunities in curating memecoins on platforms like pump.fun, rug.fun, and other emerging products.

• Venture-backed projects: I think commerce is the biggest opportunity in curation. Great curation—whether it’s architecture, fashion, places, food, or finance—leads to a purchase. The challenge is figuring out how to reward creators for driving those purchases. Ads hijacking the attention creators work hard for, without their permission, feels wrong, and platforms like LTK redirecting users to affiliate links in clunky mobile web views is a terrible experience. A lot needs to happen before crypto can solve these issues.

A huge thank you to all the Frameboarders who experimented with the product and gave thoughtful feedback, including Kugusha, Andreaboi, Jane, Steen, Steph, VM, Chic, Adam, Oxb, Pol, Linda, Justin, twolf, 0xshash, David, Lisa, Beecurious, the teams at Farcaster, Open Frames, XMTP, and Neynar, and many others. This was a fun experiment, and I’m proud of the product and the community we built.

🖼🛹✌

Jayme

We launched Launchcaster two years ago because we didn’t have a great place to share and discover crypto projects with a community that cared. Since then, our side project has evolved into a public good, fostering thousands of launches and attracting over 15,000 crypto builders.

We wanted to find a long-term home for Launchcaster where it could grow and continue as a public good for crypto builders without needing to turn into some short-term speculative experiment. Just as YC runs Hacker News and Hyper runs Product Hunt, Orange DAO running Launchcaster is a perfect fit!

The community has been a big part of Launchcaster from the start, including @jo telling us to first launch on Farcaster, early feedback and launches from @jomessin, @ravi, @jpren, @bli.eth, @robrecht, @timigod.eth, @sayangel, @berk, @huh, @maurerbot, @chaseadam17, and many others. They’ve invested in more Launchcaster companies than any other investor, including Neynar, Frens, Events.xyz, Quidli, Perl, and more! This awesome group of people deeply cares about finding and empowering crypto builders at the earliest stages with software, community support, and programs like Fellowship.

We now have the full force of Orange DAO behind Launchcaster, including the community creating and hunting down more launches and Berk Ozer as a head scout to continue to grow into the best place for crypto builders to launch and discover the best new projects.

Live, laugh, @launch!

— Jayme and Mike

I'd also like to thank Jay, Jason, Emily, Anna, Tamzen, and the rest of the CSX team for organizing a great program and group of founders.

• Achieving PMF is like holding a winning hand. - Jason Rosenthal (a16z)

  • It gives you optionality (e.g., fundraise, get acquired, acquire other companies, build a massive generational company).

• Design your product for the dinner table. - Ben Rubin (HNT, Towns)

  • How do you want people to talk about your product at the dinner table?

  • Start backward: What does the user say at the dinner table? Map the product funnel to that conversation.

• Web3 is the hardest place to be an entrepreneur. - Mo Shaikh (Aptos Labs)

  • The market is “neck-whipping” fast, and you get to “be really close to the customer”

  • A crowded market (in web3) doesn't matter. There are barely 1m users. 5B users are not using your product.

• Build a great community by genuinely giving a fuck. - Jing Wang (Optimism Foundation)

  • How does Optimism have such a great community? “We genuinely give a fuck” - Jing

  • What are vibes? People are having fun. Vibes are the outcome of a bunch of early decisions

  • Vitalik on why unicorn shirts “keep out a particular type of toxicity “

• Network tokens are the most interesting token type - Eddy Lazzarin (a16z)

  • There are four token types: Memecoins, Stablecoin, Arcade tokens, Network token

  • Network tokens are the most interesting: You don't design a token; you design a protocol

• Find people who love what you are doing - Hilmar Veigar (CCP Games)

  • Finding a few people who love what you're doing is better than millions who just like it. Avoid ambivalence.

  • Start there and then diffuse into the mainstream. The best things started as niches: Metalica, Apple, Eve

• More cryptography is better - Valeria Nikolaenko (a16z)

  • “Richer the toolkit, the better. The more tools we have, the more free we are.”

  • Our privacy toolkit is very poor — Data is leaked all the time. With MPC, you can shift it around

• The new Five Forces: the source of competitive advantage has changed. - Scott Kominers (a16z)

  • The source gets stronger when you have interoperability

  • Value capture in traditional form is more difficult

  • Network effect forms around protocol and digital assets

  • Embeddedness becomes a source of competitive advantage

  • Community cohesion: Users want to engage in your ecosystem; shared ownership is a powerful loop

• Regulation should reflect democratic values. - Brian and Michele (a16z)

  • Shouldn't: create new powers, censorship

  • Should: support innovative tech, freedom, right to transaction, promote competition

• A brand is a collection of beliefs people have about your company. - Steven Ebert

  • A brand is not something you set and forget. The reality is that a brand is like a living org that evolves

  • Brand building is the practice of shaping a brand.

  • Minimum Viable Brand is why you exist, how to tell your story in a compelling way, and how you differentiate from competitors

• Fundraising is answering questions about a business. - Arianna Simpson (a16z)

  • Why is this a big market?

  • Why are you best positioned to tackle that problem and win?

  • How deeply have the founders explored the idea maze?

  • How good is the founder market fit

  • Assuming everything goes well, what will you be doing in 5 years?

• All decisions in the beginning compound more than you think. - Alex Blania (Worldcoin)

  • The first eight people should be complete rock stars, work day and night, and be a perfect cultural fit.

  • There is a massive difference between top-tier and non-top-tier investors, so optimizing Series A leads matters more than founders think. Investors who have not been founders should not advise founders on how to run a company.

• View life as art and paint pictures. - Chris Lyons (a16z)

  • Your job is to paint pictures. Maximize what each moment can be. You can make 1 + 1 = 2 or 1 + 1 = 11.

  • What is the next best situation? "There is always a move. If something is a little boring, you need to shake it up. There is always a move."

• Think smaller to go bigger. - Shane Mac (XMTP)

  • Mistake Shane made with the previous startup: went horizontal instead of vertical. Thought dog walking was too small: "were not just a vertical app were bigger and better."

  • Think smaller to go bigger. We didn't pick a small vertical to nail the experience, and we couldn't have picked a small enough community.

  • Key insight: Do not go broad; Go less. If that doesn't work, switch to a vertical.

• Communities are the most powerful force. - Chris Dixon (a16z)

  • The moat is your community.

  • 1K True Fans: you need 1K super fired about contributing. Think, “How do I get 1K people to love me?”

  • It's surprising how small of a community you need (Wikipedia is basically 20K people).

• The best company in every credible category.

  • You can't control the category, but you can control the best team.

  • Category is the hardest thing about the investor's job. Not overly strict about the category because predicting the future is difficult. Very strict on the team.

  • Example: You can be unsure of decentralized social, but you can be sure that Farcaster is the best team.

That's a wrap! Thank you, a16z Crypto and London 🍻

• Got into Crypto through AI

  • Grew up in rural Germany

  • Theoretical physics before → LLM → Worldcoin

• Being recruited by Sam Altman

  • Got a two-pager email from Sam and Max

    • The best startup ideas sound like horrible ideas.

    • It will likely fail, but if it works, it will change the world.

  • The interview

    • Explain physics model

    • Sam liked that he disagreed with a lot

    • A lot of questions about childhood

• How to piss off investors 4.5 years ago

  • Bitcoin failed

  • AGI will happen

  • That's what got a16z excited

• Why Tools for Humanity

  • What will be the things we need on a global scale that the government will not be able to do?

  • economically reach everyone

  • global form of proof of personhood

• Worldcoin Pre-launch

  • Spend eight months of research

  • Talked to a lot of AI and crypto experts

  • Crypto (tokens) allows you to incentivize alignment.

    • Crypto has been around and is still a niche thing

    • What would be the incentive mechanism?

  • Paypal inspired GTM

    • Email $20 invite links to kickstart and scale the network

    • It is a network; if it doesn't scale, it doesn't work

  • Interviewed the Facebook Web of Trust team

    • Facebook is already struggling with that

    • Ignored because AI will become increasingly strong and break any web of trust

  • Spent $50M and lots of prototypes to get to Orb

• Worldcoin fully-realized

  • Proof of personhood is important in an AGI world

  • The majority of the population will have a WorldID

  • Worldcoin will be the most owned digital asset and the largest financial network.

• Twitter bots

  • This is a super early warm-up of how the internet will be

  • Almost everything you do will be enhanced by AI

  • Cool for most things, difficult for other things like social media

    • It will be hard to distinguish between human vs. AI

    • AI will interfere with elections

• Why Worldchain?

  • Worldapp now at 10M users

  • On many days, 80% of OP mainnet

  • "We just need the blockspace."

  • We could not compete anymore with other apps

• On government interaction

  • A lot of founders and startups shy away from gov; Worldcoin has taken a different approach

  • They didn't have the capability/function to interact with the government at first - now starting to turn

• Growing as a CEO

  • Coaches can help. Alex coach setup:

    • 1 VP of door dash, one scientist, and one exec coach

    • 2hrs a week, every week, told me how I sucked this week

  • Talking to other founders won't get you far because they don't know what they are doing. Look for successful people in the things you're trying to learn

  • Different learnings each year

    • Year 1: hardware

    • Year 2: scale team

    • Now: policy

  • Matt Majorie blog post - turn the ship around

    • you don't come to the CEO with a problem

    • you come with a problem, solution, and decisions - raise your hand if you disagree

• Wish he had known earlier

  • All decisions in the beginning compound more than you think.

  • Examples

    • The first eight people should be complete rock stars, work day and night, and be a perfect cultural fit.

    • There is a massive difference between top-tier and non-top-tier investors, so optimizing on series a lead matters more than founders think.

    • Those who have not been a founder should not advise founders on how to run a company - Khosla.

• Path to Web3

  • Started in music

  • Producer for Jermaine Dupri

• On Kevin Hart

  • He built an empire by thinking of his career as an entrepreneur

• Relationships

  • Always look for long-term relationships

• It's not who you know; it's who knows you

  • The more you can create a halo effect, the more you can naturally generate success

• View life as art

  • Your job is to paint pictures

  • Maximize what each moment can be

  • You can make 1 + 1 = 2 or 1 + 1 = 11

  • What is the next best situation?

    • "There is always a move. If something is a little boring, you need to shake it up. There is always a move."

• Creators

  • 85% of all streaming revenue goes to centralized systems - it is impossible to have a full-time career by just making music

  • Read, Write, Own — The only word cdixon uses more than Cryptohain or crypto is creators.

  • If you ask kids what they want to be when they grow up, "I want to be a TikTok star."

• Hide the wires in crypto

  • Life is like a restaurant — If you're in a restaurant, how can your application be valuable and be used by each person

  • We (web3) have to do our job so well we don't have to talk about it ... "I never went to an MP3 conference."

• IQ + EQ

  • Realize they're regular people.

  • Making a connection — Always figure out the things that have nothing to do with the profession. Find the commonality.

  • It can lead you into a familiarity of trust and then lean back and talk about the business

• Leadership

  • What do you call a leader with no followers? Someone taking a walk.

  • You have to be a leader, be a star, sell the dream

• Shane background

  • Appreciate moments where perspective changes

  • Arc: grew up in a small town, went to western Illinois

• Whole career got rugged

  • Zuck/Jack - do whatever you want with our social data

  • Then Twitter API changed, and Facebook changed

  • Whole career was built by being the best expert

• Building on blockchain

  • 2019 defi summer — Why can't I message a wallet?

  • To not have to trust me is to be able to trust us

  • Spent so long not being able to trust web2

• Best way to kill a startup

  • Zaarly: raised 35M, magical demos

  • Problem: went horizontal instead of vertical

  • Thought dog walking was too small "were not just a vertical app were bigger and better."

  • Think smaller to go bigger

    • We didn't pick a small vertical to nail experience

    • We couldn't have picked a small enough community

  • Key insight

    • Not going broad, Going less

    • If that doesn't work, switch to a vertical

• On VCs

  • Don't want to get shot? Just commit suicide.

  • If you don't have user growth, focus on that

  • Be honest about bad things about confidence in the mission and rituals.

• On Language

  • Language almost killed startup - blog post

  • Change the question and the environment to get different outcomes. Don't ask binary questions; ask how.

  • To employee: "On a scale of 1-5, how much do you wish I was there?"

  • Your company's culture is its language and rituals.

• Vulnerability

  • Doubt is the greatest way to let someone help you

• Building Toys

  • Everything starts looking like a toy

  • You have to be confident in building something everyone thinks is a joke.

• On writing

  • Be more present, listen to things I cant here

  • Thinking about writing: connecting disconnected dots

  • Process: Sit at a hotel and write. Then, resist publishing. Edit for the next few days

  • Impact vs. output: What's the smallest show-and-tell

  • Every time I publish I edit publically for four years

• Being an imposter

  • Don't worry about imposter syndrome because everyone thinks about themselves.

  • Insignificance theory via Tim Ferriss

• Ask to web3

  • Just build something usable. The bar is so low.

  • I think we've gotten lost in all the words and jargon

Here are some links that came up this week.

• FIT21 passed!

• Farcaster raised $150M

• Polymarket raises $70M

• Collectible boards

  • You can now collect a board on Base to unlock it.

  

• Cast action settings

  • Choose the board you curate to with the Frameboard cast action. This also allows you to collaborate on curating casts!

  

This is not bad, but if you're building crypto stuff for consumers, I think you need to be aware of it. If you run the cryptonative GTM, you need to cater to crypto professionals (traders, developers, artists, creators) first or plan on abstracting away the crypto aspects after finding your first 100-1K happy users. If you GTM elsewhere, this doesn't matter.

The new consumer projects going to market with cryptonatives that have the most staying power right now have a prosumer angle.

Some examples executing this well …

• Hypersub: NFT subscriptions for crypto artists

• Paragraph: publishing tools for crypto creators

• Farcaster: protocol + community for crypto builders

• Dexscreener: realtime dex analytics for crypto traders

• Bounycaster: labor marketplace for crypto companies

• OnceUpon: blockchain explorer for superchain builders

First appeared on Farcaster.

Farcaster

Jayme Hoffman on Warpcast

Some thoughts on consumer crypto. Curious if anyone else feels this right now? https://imagedelivery.net/BXluQx4ige9GuW0Ia56BHw/196c0ea9-2400-45a9-6f56-7d7833cad100/original

• What is a brand?

  • Something you affiliate with

  • A collection of beliefs people have about your company.

  • People = customers, partners, team
    

    

• A brand is not something you set and forget

  • The reality is that a brand is like a living org that evolves

    • product evolves

    • bring in new users

    • grow to more employees

• What is brand building?

  • The practice of shaping a brand

  • Tools for shaping a brand ...

    

• Minimum Viable brand

  • Why you exist

  • Tell your story in a compelling way

    • Not just about marketing

    • Hiring is difficult when you do not have an interesting and compelling story

    • A big part of fundraising

  • Differentiate you from competitors

    • Classic example is AMM

    • The way you tell your story is what sets you apart

    

• Two areas are useful

  • Brand story

    • The language you represent

  • Visual identity

    • What do you look like

    • It builds goodwill with the public

• Zora does a good job at brand

  • Define the problem

    • Did a good job of defining the problem

    • Established a meme of what the new future looks like

    • Creating language became super useful

      • we see this problem

      • we see new things because nfts

      • we are going to be part of that movement

  • Explain how you solve it

    • Creating free and valuable

  • Prove in the product

    • Simple 3-step product

    

• Elements of brand story to consider

  • Why

    • problem

    • how we solve

  • Key benefit

    • what are the reasons someone would want ot use it

  • Tone of voice

    • thoughtful around how you’re communicating

• On creating your visual identity

  • Think about your audience

    • If shitcoins identity you will have will be very different than institutional investors

  • Look at what competitors are doing

  • Define the vibe you're going for

• A good visual identity

  • Connected to the brand story

  • Designed for flexibility

  • Opinionated and differentiated

  • Party does a good job ...

    

• Additional thoughts on brand to consider

  • Be careful of brand stories already taken

  • Watch out for confusing nomenclature

    • Aave = could be talking about 1-7 different things

    • Good = Opensea vs. Seaport

  • Dont handcuff yourself to the current meta

    • bad: gm

    • don't handcuff to current vibes

  • Refine brand story as the product evolves

  • Visual identity is extra important for B2C

  • The more saturated, the more differentiated you need to be

• On why, how, what

  • If early, focus on why

  • If competitive, focus on how

  • Very few examples where "what" is the most important thing

Thank you, a16z Crypto, for making me feel a little less homesick 🙏🌮

• Arianna intro

  • Many VCs are afraid to tell you what they think, not Arianna.

• Fundraising is answering questions about a business

  • Two important questions to answer before deck

    • Why is this a big market?

    • Why is your team best positioned to win?

      • Why are you best positioned to tackle that problem?

      • How deeply have the founders explored the idea maze?

      • How good is the founder market fit

      

  • Question Arianna asks every founder:

    • Assuming everything goes well, what will you be doing in 5 years?

• How to get ready to pitch?

  • Understand market

  • Practice with angels and friends beforehand

  • Tailor the deck

    • have teaser deck

    • the goal is to get a meeting

    • hold a little info back and then reveal all

• Core tenants of the pitch

  • Tell a story

  • Have a team slide

    • a lot of it is do we think this is the right team to build this thing

  • Never lie

  • DYR

    

• On negotiations

  • The best investors will not play games

  • The best strategy is to be authentic

  • Trying to eliminate the reasons why people would say no

  • The name of the game is momentum: You will be better off if you get more term sheets

• Post-raise relationship

  • From pitching to portfolio

    

  • Never surprise your investors

  • Engage early and often

  • Use their network for

  • It's a marathon

    

• Other helpful fundraising tips

  • Generate momentum by time boxing conversations

  • On team slide: logos > 10 years of experience

  • Party round = nobody has (real) skin in the game

• Memorable fundraise: Anton at Talos

  • FMF was perfect

  • Charismatic

  • MIT smart

  • Technical but can sell

  • Plow through any wall to make it happen

Here are some links that came up this week.

• From a16z Crypto Marketing & Comms

  • On communicating in crypto, and beyond with Matt Abrahams & Sonal Chokshi

  • On marketing web3 with Amanda Cassatt, Kim Milosevich, and Sonal Chokshi + deep dive on launching & marketing your project with Amanda Cassatt

  • On strategy & tactics for social with Ish & editorial with Sonal

• What’s Open Layer?

  • OpenLayer is the first trustworthy data layer empowered by users.

  • Last week, Kevin, Charlie, and the OL team launched the OpenLayer AVS on Eigen Layer.

• More 🐐 Shoes of CSX

  • Tata Flats @aleksijavujicic

  • Air Reads @ishverduzco

• New Frameboard homepage with discovery ✨

  

• Policy vs. Regulatory

  • Policy: Goal for laws and regulation in the future

  • Regulator: Enforcement of the laws and regulations that are here now (i.e. what you need to be aware of and compliant with now)

    

• a16z Crypto Philosophy

  • Regulation should reflect democratic values

    • shouldn't: create new powers, censorship

    • should: support innovative tech, freedom, right to transaction, promote competition

  • Should be appropriate and clear

    • Appropriate: What is the scope of the audience intended to address? is it targeting the right risk?

  • Regulate apps, not protocols

    • “Don't regulate software; regulate businesses”

    

• Policy issues a16z crypto is focused on

  • Launching tokens

    • A regulator path to decentralization

    • How do we create rules that recommend decentralization as a benefit?

  • How to think about DeFi

  • Stablecoins and staking

• What they hear from policymakers?

  • I like blockchain, but not crypto

    • Translate: I like innovation, but I don't like speculation

    • Response: you can't have decentralization without crypto assets, world computer vs. world casino

  • Same activity, same risk, same rules

    • Translate: political tagline, not sure they really mean

    • Response: see it as two questions followed by a statement

  • Crypto is the Wild West

    • Translate: I'm worried about crypto and illicit finance

    • Response: Comes down to data

      • See State of Crypto 2023

        • talks about the amount used in illicit finance

        • definitely a concern, but need to scale appropriately

    

• Remember about politicians

  • Policymakers are extremely busy!

  • They have to have thoughts on every major policy

  • Booked 30min entire day every year

  • Their thought often comes down to "crypto is bitcoin, FTX or defi"

• Where your service is offered is what matters

  • People think where you incorporate is what matters, in fact, what's most important is where you are offering services.

    • If state, you need licenses with the state. The same is true for international

• Top questions a16z hears

  • Sanctions questions

    • if moving any type of value, you have to follow sanctions law

      • whether you are running a business in the US or not

      • you need only a little bit of touch in the US to have liability

    • Think...

      • How would a criminal in XYZ country try to take advantage?

      • What can I do to mitigate?

    • “Im not here to scare you and be the boogeywomen” just here to talk to you about the risk"

  • Tax questions

    • Where you need to pay

    • Do you have reporting obligations

  • Securities questions

    • Love these questions the most: How do we design, launch, and build tokens and token ecosystem

• How to interact with lawmakers

  

  • Reactive

    • Often unpleasant

    • 98% of the time just looking for info in a case

      • Reaching out because they seem to believe you may have info access to

      • Should you get legal counsel? Sometimes.

      • You shouldn't panic if they reach out - They are often confused about where to look

  • Proactive

    • Demos

    • conferences

    • schedule meeting

    • great ops for them to learn and understand

• Legal advice on Crypto Twitter

  • Not who you want to be following when comes to legal and building business

  • Good for humor and excitement

• On crypto in the U.S.

  • No question we're in a fight in the United States

    • Significant legal challenges

    • Anything and everything is part of their regulator's jurisdiction

    • Crypto is going to be around longer than the Chairman of the SEC.

  • There are a lot of good parties

    • I have faith that the courts are going to give us more clarity and info that will cause the agency to go in a different direction or scratch their head that this aggressive approach is needed

  • Tech is here to stay

    • The technology is here to stay. The US will get this right. The question is, will it be in one year or three to five years?

  • a16z Crypto is fighting

    • "Cdixon 150 meetings a year, Sriram, just as many"

    • In those meetings, "you see their lightbulb go off"

• Spectrum of consumer protection/innovation

  • Almost always on a spectrum of consumer protection or innovation

  • Strong protection = anti-innovation

  • Strong innovation = anti-consumer protection

  • Embracing decentralization can do both

• Remember, regulation is needed

  • Protect consumers from the casino

    

• Where to find Brian and Michelle online?

  • Brian Quintenz (Head of Policy) / Twitter

  • Michele Korver (Head of Regulatory) / Twitter

Here are some links that came up this week.

• Frens

  • Frens is a map-based crypto messaging app. The app and screenshots of it came up a ton this week due to Farcon and Far(away)Con. The Frens founders are great ❤

• EigenLayer

  • EigenLayer came up a couple of times before and after their controversial airdrop.

• Paragraph + Mirror * Kiosk

  • Paragraph announced that it acquired Mirror's assets and raised $5M from USV. The previous Mirror team announced they're working on Kiosk, which looks super cool!

• Leveraging Farcaster for Growth

  • I gave dozens of builders tips on how to leverage Farcaster for growth. tl;dr be consistent, be authentic, and post from founder vs. brand account (more here)

  • Here are the tools I use for alerts, feeds, and search.

• a16z decentralization and regulation links

  • On decentralization, an overview by Miles Jennings

  • On decentralization for builders: principles, models, how by Miles Jennings

  • On launching compliant tokens by Miles Jennings

  • On designing effective governance by Porter Smith

  • On crypto regulations, illicit finance, and more with Michele Korver, Jai Ramaswamy, Sonal Chokshi

This week was a record-breaking week! We shipped a couple of new features and lots of small tweaks.

• Edit an image curated from a URL or uploaded.

  

• Frameboard cast action - You can now curate to Frameboard directly from Farcaster

  

• Miss out on Farcon? No worries, there's a Frameboard for that.

  

👋 See you next week!

• The a16z Crypto research team

  • Releases major papers multiple times a year

  • Collaborates with a16z Crypto portCos

    • Help them succeed

      • office hours - what are the hurdles?

      • another set of eyes on a design

    • Sometimes deeply influences product/protocol

  • Helping portCos in an immature field is a competitive advantage — “18mo advanced notice on important papers to write by talking to founders.”

• Scott Komiers - Farcaster / Twitter

  • Excited about consumer crypto

  • Market design

    • economics - how does the world work

    • market design - how should the world work

  • Design incentives in the system to shift or change behavior

  • Dropping tokens

    • Think really hard about the users that are most valuable and the ones that are going to build with you — “Who are going to Invest to grow the ecosystem?“ - Scott

• You cant black-box the blockchain

  • computer in the sky, anyone can use and share

• The source of competitive advantage has changed

  • Web2

    • Platform lock-in

      • Hold everything in this enclosure

      • Become extractive over time (ads, charge more)

  • Web3

    • Break out of it

      • Web3 breaks a lot of it

      • Data to be individually owned

• Five forces of web3

  • The source gets stronger when you have interoperability

  • Value capture in traditional form is more difficult

  • Network effect forms around Protocol and digital assets

  • Embeddedness becomes a source of competitive advantage

    • difficult for anyone to disintermediate

    • value capture = small fees

  • Community cohesion

    • Want to engage in your ecosystem

    • Shared ownership is a powerful loop

• On Farcaster

  • Farcaster gives users a very powerful form of ownership

  • Users own their identity

    • They can use it and any value built up across any client

    • This is absent in web2

  • Places limits on how extractive any one client can be

    • Users can switch to Supercast

    • Users have more power

  • Farcaster did very well curating the early supply of users

• Reason to do additional work for creators

  • They have to do a lot of work to invest

  • They have a higher activation cost

  • Early platforms have to subsidize creators

• On cryptography

  • There’s always something exciting in cryptography

    • “Cryptography is resolving this physical paradox in the universe — kinda mind-blowing cryptography thinks about this” - Valeria Nikolaenko

  • 10 years is the median time for cryptography papers to become a reality

    • Rely on hardware to become practical

• Snarks

  • Close to crossing the threshold to where they’re much more useful

  • Crossing a performance and usability threshold

• More cryptography is better

  • “Richer the toolkit, the better. The more tools we have, the more free we are.” - Valeria

  • Our privacy toolkit is very poor

    • Data is leaked all the time

    • MPC you can shift it around

• How to nerd swipe cryptographers

  • Ask, “when will quantum computers arrive”

    • Very many decades away from being here — Very far from a cryptographically relevant quantum computer

    • Joke in cryptography: “Quantum computers are always ten years away”

    • Progress is very slow now

      • Once scaleable tech, it can grow exponentially fast

      • Maybe it will take another military conflict to motivate people to work on it - the US and China being a little afraid of each other helps

      • Thinking that problems are simple is good because it will attract more people to work on them

• Breaking down a system (Titles.xyz examples)

  • What are the …

  • Components

  • Flows

  • Objectives with incentives

  • Ways we acquire users

  • Puzzles thinking over

• Note: Titles is cool, and Soren and Parker are great ✨

  

• Marquee users

  • Good for pulling in other users and creators

    • I.E., if Optimism is using, another chain will want to

  • At scale, marquee users get a lot from the platform

  • Have a different strategy for recruiting different types of users

• On fees

  • Easier to wrap a transaction fee when a transaction is already happening

  • Airbnb adds value at the time of the transaction (fee model) vs. the moment of access (subscription model)

  • Fees on spam

    • Minimal fee for anti-spam

    • Can kick back the first time there's a mint

    • Pay to create a model

• Fiids

  • Fiids is like a Hootsuite or Tweetdeck for Farcaster and web3 social. IMO, it’s the most powerful desktop client. DM Tiago for an invite.

• Farcon

  • Farcon, a multi-day conference in Venice Beach, CA, is like Woodstock for Farcaster. The event is sold out, but you can find a list of side events here, including Far(away)Con, our London satellite event.

• Spindl

  • Sindl is a suite of attribution products for Web3 companies. Its founder, Antonio García Martínez, writes an incredible blog on Web3 attribution and growth (h/t Eddy Lazzarin).

• Story Protocol

  • Story Protocol is building a protocol and universal ledger for IP. They have a bold and beautiful vision outlined here (h/t Jason Rosenthal).

• Gurley on ILTB

  • There’s a great new Invest Like the Best podcast with Bill Gurley and Michael Mauboussin (h/t maks).

• Reboard and follow boards - Farcaster / Twitter

  

• Shoes of CSX - Shoes of CSX is a work-in-progress Frameboard highlighting the great shoes and founders in a16z Crypto’s CSX.

  

• We made great progress on the protocol and onchain side of Frameboard — excited to share more soon 🖼🛹⛓

• Token = balance in a crypto program

• Four types of tokens

  • Memecoins

  • Stablecoin

  • Arcade token

  • Network token

• Memecoins

  • Explicitly no purpose

    • If purpose, then more interesting than a memecoin

  • Gambling and purely speculative

  • Could be decentralized, could be a scam

    • Scam = deep asymetry of info in the memecoin

  • Memecoins have been around for a while

    • Look at CoinMarketCap in 2013

      • #40: Memecoin with $5,709 market cap

• Stablecoins

  • Prixe fixed to a target asset

  • Backed by collateral

  • Regulation likely

• Arcade tokens

  • Price dampened

  • Limited transferability and use

  • Continuous issuance and redemption

  • Wouldnt be reasonable to buy them and expect price appreciation

  • Not a security because price dampened and issuance

  • Example: SLP from Axie

• Network tokens

  • Integral to decentralized protocol

  • Potentially high volatility

  • Complete economic model

  • The most interesting: You don't design a token; you design a protocol

• Faucets and Sinks

  • Fundamental power = issuance

  • Always consider the sink

  • Protocol as a Marketplace

    • Supply = parties creating good

      • Network capacity, blockspace

    • Demand = people pay to consume that

      • People pay for compute

    

• Bad reasons for tokens

  • Incentives

    • Pay for this is terrible; printing money and throwing it somewhere

    • The infinite ad budget you didn't pay for is not protocol design

    • goal = pmf

      • make a product people want

  • Voting

    • who wants to vote?

    • voting is a last resort

  • Payments

    • Talking about network tokens

    • Evidence

      • People do not want to pay in protocol tokens

      • People don't like paying with a volatile asset

    • Payment is not a direct sink

  • Good reasons for tokens

    • Align the network

    • Support network effects

      • Pay for network effects - Incentivize things that lead to network effects

      • Solve cord start - token issuance can help

    • Securing a protocol

      • Pay for security

        • ETH proof of stake pays for security

        • People should be paid to take risk

      • What is your protocol for selling?

        • Is this good and high quality = verified

        • Pay to ensure quality

• Incentives changing over time

  • What about the thing you want to change over time?

  • Manual programs

    • You want to be able to walk away from the system and it still works

    • Initially, use manual token allocations aggressively.

      • You're trying to find PMF

• Path to decentralization

  • “Decentralization is the regulator strategy”

  • End users don't care about decentralization

    • They don't care about monetary policy

    • They care about banks don't steal money, hold value, high-quality products

  • High-quality products

    • Built on the most solid foundation

    • if network token, needs to be decentralized

  • Where to start?

    

  • Centralized but functional

    • Points

      • Blackbird doing a great job

        • $Fly = offchain (good idea)

        • Nobody is going around buying fly

        • People are excited about it

    • What about testnet?

      • “If transfer entire balance, was it just mainnet all along?”

        • people can still be burnt

        • markets can still happen

    • How does it look?

      • Did people lose money

    • Avoid risk

      • Tweet bad, and the price goes to 0?

  • A simple test

    • if the team disappears, what happens to the token price?

      • if goes to 0, you’re not decentralized

  • Memecoin decentralization

    • Bottom left - decentralized but no functionality

      

    • Why not go right?

      • Legal risk

        • The minute you benefit, you are moving way back up the decentralized path

        • You can become a security again

      • If totally decentralized, how are you going to find PMF?

        • Easier when centralized because you’re in control and you can feel pull of the market

• When to consider token design?

  • The earlier, the better

  • Design a protocol first, then see where a token's role is necessary—it may not work without one.

  • Design: Centralized app over here, Decentralized protocol over here

    • Imagine if Coinbase made Bitcoin

    • Separately, they have a business with a client of that protocol

    • Compliments together - “I have a protocol I like but I can think of a company i want to be a client or vice versa

  • Does value go to equity or token?

    • When we invest, we prefer both to preserve alignment

• How to distribute tokens?

  • Automatic programs

  • Manual programs

    • Wish people thought more about them

    • Airdrop

      • We think of it as something to give to users

      • Uniswap example

        • Anyone that swapped

        • Those were primarily developers

      • Discount to regular people

        • Think more about long-term stakeholders who are going to build on it

        • Those are the people who are most valuable

      • Retros

      • Grants

      • Prizes

      • Partnerships

• Note: this was one of the most dense and actionable talks. I recommend watching it when it hits YouTube.

• Security in crypto

  • Who are the parties that want your crypto

  • What are they doing to try and get it

  • How can you address it

• Types of threats

  • Independent actors

  • Criminal organizations

  • Lazarus - mid-level nation states

  • Sand Eagle - The scary one

  • Criminal orgs and Lazarus are the ones to care about

• Motivations

  • Hacktivism = defacing websites, leaking tools and emaik

  • Financial crimes = steal crypto

  • Espionage = ip theft, steal secrets

• Targets = anyone that has crypto

  • You

  • Your org

  • Your customers

• Attacks against accounts

  • Mostly

    • Sim swap

    • Password guessing

  • TTP - people

    • phishing and bribery

  • TTPs - dependencies

    • find someone the target depends on, hack them

• In order of likelihood

  • someone gets phished

  • someone gets sim swapped

  • someone password guessed

  • smart contract exploit

  • disgruntled employee

  • highly motivated

• How to avoid?

  • For you

    • avoid using sms for auth

    • use uniqu passwords

    • harden account recovery

    • MFA = most important

  • For your biz

    • MFA

    • SSO to force auth across services

    • Keep track of dependencies and due diligence

  • For your customer

    • How do you make sure the password isn't guessed

    • Forced OTP

    • Notify user on new logins

• Audits in crypto

  • Almost everyone who gets hacked thinks they cant be

  • You need your system reviewed

  • How?

    • Find someone good

    • Reserve some of their time

    • Get the report and fix the issues

• Crowdsource audits

  • Dont have as many examples of code getting hacked

  • If the code isn't up to snuff, you will get 100s of comments

  • After you do it, you need to get another audit

• Hilmar

  • Student of history

  • Expert on economic primitives

  • Talked to oil industry - “omg that is well game-designed”

• Eve Online

  • Players work on projects that take 100s of people … for a year not months

  • Mission: Make virtual worlds more meaningful than real life

• How is crypto like virtual worlds

  • Tokens equally not real

  • Jobs aren't real

  • The economy is equally not real

  • What is meaningful?

• Kant on rules for happiness

  • Something to do

    • The game

  • Someone to love

    • Friendships in the game

    • Real-life friends have never been tested

  • Something to hope for

    • The next expansion

• Real life has failed people

  • 7B people, only 200m people like what they do

  • Can we solve that?

    • Give people agency

    • Give people meaningful social networks

• Why decentralization

  • Today: user agreement

    • Nothing belongs to you

    • You must adhere to the rules of CCP

  • Eve is like Bitcoin

    • Derives its value from social consensus

    • Lindy trust properties

• Layers

  • is there a way to replicate it in a game simulation?

  • is there a way to design it so that it's better than real life?

    

• On collective suffering

  • Eve = Where nobody dies, but great loss can be experienced

  • Something magical about the collective suffering

    • Underestimated part of biology

    • People seek this out

      • adventures together

      • marathons

  • Crypto is similar

    • Winters = collective suffering

    • Such bonds → What winter did you join?

• Building in crypto

  • Very similar to when we said were going to make a database game

    • People were like, “Weird.” .. “elves in Iceland are going to use a database to make a game”

  • Comparison

    • Slow: databases 20 years ago were slow

    • Transferability: happening already, claim they don't want it, but people are already buying/selling

    • Financial: People have more money in Even than in a bank account

      • Average Eve savings = $1300 vs America savings = $500

      • Should be able to pull on that in case of an emergency

        • “Why would I disallow that?”

        • What is the moral principle of making that wrong?

  • Challenges

    • Secrets

      • You need to be able to have secrets → ZK

      • As soon as we announced (ZK) all the people working on this came out because they needed to test their ideas

    • Agency

      • You need to limit agency (i.e., moving from one solar system to another) → Every action has a cost

    • Blockspace

      • “There is a fight for blockspace anywhere and anything”

      • What hits in block matters a lot if a spaceship does or doesn't survive

      • Theres a queue to get into the frame

        • the queue is unpredictable - “When is my message going to get in”

        • never-ending warfare

          • similar to crypto

      • Time dilation

        • Similar to real-life

        • If a lot going on, time dilates

        • Time dilation = allows you to strategize more

        • Real-time, only so much you can do

  • In some ways, crypto is a big MMO (Solona vs. ETH)

• If you were starting as an entrepreneur today?

  • Find people who love what you are doing

    • Better than millions that just like what you’re doing

    • “We like it, and if you like it, come join”

    • Avoid ambivalence

  • Start there and then diffuse into mainstream

    • main things started as niches: metalica, apple

• On getting inspiration

  • Iceland is small

    • People in your social network are weird

    • You can’t specialize in anything

    • the initial team were artists, i wasn't a game dev

  • Deconstruct history and evolution

    • economists = good at analyzing the economy

    • physics = good at emergent behavior

• Scaling large groups

  • Getting 10k people to do anything is very hard

    • sporting events, concerts, protests

  • Concert = single player replicated 10k times - eddy

  • Burning Man took a long time to scale up

  • Military takes decades

    • Doctrine → Train people

    • Takes 100s of years to train military

  • Eve took a decade to get to 10k

• On design

  • I like to think of design as in “de-sign”

    • Take everything away

    • Figure out what is essential

• What’s Blackbird and $Fly?

  • Blackbird is a crypto startup that connects restaurants and customers.

  • $FLY is like miles on your favorite airline, but instead of one airline to redeem from, the points are good at all the restaurants in the world.

• Computer vs. Casino

  • What’s the difference?

    • Computer: build new networks

    • Casino: speculation and money-making

  • Tokens’ true purpose?

    • Tools that enable community-owned networks.

    • Community ownership doesn’t work unless communities have a way to own.

• All things tokens from a16z

  • On tokens as a new digital primitive By Chris Dixon

  • On designing internet-native economies: a guide to crypto tokens by Patrick Rivera

  • On token design mental models, capabilities, and emerging design spaces by Eddy Lazzarin [watch time: 32 minutes]

  • On a novel framework for reputation-based systems by Jad Esber and Scott Kominers

• Sign in with Farcaster - Farcaster / Launchcaster

  

• Right-click on any image to curate to frameboard

  

• See quality dot and image size on image selector in curate from URL and browser extension.

  

• See the Chrome extension on the home page

• URL support in bios 🔗