Original title: Death to Stagnation
Original article by: Joel John and Siddharth
Original translation: TechFlow
Hello!
A few days ago, I noticed a problem with the way we write. Most of the time, our articles are about technical topics written for founders and investors. This is fine. We skip the drama, politics and scams, and nerd out for weeks on topics that can be read in minutes. But we are not in academia, we are in the free market, investing and building with founders. Therefore, it is very important to keep up with what is happening around us.
If I were to confide in my therapist about my experiences in crypto over the past six months, I would sum it up like this:
A 13-year-old launched a meme coin and then dumped it. Scotty Pippen of the Chicago Bulls somehow predicted the price of Bitcoin after having a dream about Satoshi Nakamoto . Speaking of which, Jack Dorsey might be Satoshi Nakamoto. A Frenchman made $28 million betting on Trump’s election. By the way, Trump launched a meme coin a few days before he took office. Meme coin . Two weeks ago, the president of Argentina tried something similar and lost $4 billion in value. Now his opponents are trying to oust him. People lost money on Su Zhu’s exchange . Oh, and Hawk Tuah girl also launched a token that turned out to be rugged . Dave Portnoy also launched a token that was also “rugged.” Cz talked about his dog Broccoli , and it looks like that token isn’t “rugged” yet — at least not yet.
Meme coin now has the power to bring down a president, which is truly a far-reaching social effect.
It’s not that good things aren’t happening in our industry. In October, the UAE clarified that cryptocurrencies are not taxed. In the US, banks can now hold custody of Bitcoin. Someone made $25,000 by tricking an agent into transferring money. Oh, and Marc Andreessen sent a proxy $50,000, driving a coin’s market cap to $1 billion. There are rumors that the US government is developing a strategic cryptocurrency reserve. OpenSea may eventually issue tokens to bail out NFT traders who lost a lot of money in 2022. FalconX acquires Arbelos. Coinbase clarifies that the SEC has dropped all lawsuits against it. Last week, Bybit survived the largest hack in human history.
The point is, we are a resilient bunch. But much of our collective psyche is entangled in bad news — the kind that is price-driven, laced with gossip and a lot of embarrassment. You can't help but wonder, "Am I really in the company of these clowns? Am I in a circus? Am I a monkey in this game?" It's all so exhausting. Especially considering that the human brain can only process 10 bytes of information per second when thinking.
How do I navigate everything as the fraud frenzy is live streamed in 4K UHD ?
If you don’t consciously set boundaries, working in crypto is like throwing your brain cells into a vortex of headlines, spinning at the speed of light until the heat of the sun evaporates all your cherished memories.
A gradual descent into madness marked by reckonings and an endless but often meaningless stream of news. It's like circling Dante's Inferno, jumping from one sin to another.
That’s why we, as a current affairs media, tend to distance ourselves from the daily drama.
But given the state of the market and the feedback we hear from founders, I think it’s time to do a “vibe check” on the culture, so to speak, in response to the brief “ vibecession ” we’re experiencing.
The “Memesis” Era
David Perell's article on Peter Thiel's investment philosophy was one of the important enlightenments in my career.
One of the themes is mimesis. The concept of mimesis, defined by Rene Gerard, revolves around the human tendency to imitate and compete with others.
Think about the career choices you made when you were 17: you looked at what your smartest peers were doing, or at an adult who had the lifestyle you aspired to, and chose their career path. As humans, we are wired to imitate and compete with our peers because blazing new trails requires a high cognitive load. We like the safety of “strength in numbers.”
This also applies to startups .
Put enough smart people in a room together and you’ll see them emulate and compete with each other. Label an accelerator or investment fund like Sequoia or YCombinator as a “top dog” and we’ll see a bunch of smart people rush to join — not just for the financial resources it unlocks, but for the status it brings.
YC knows this, and claims that its acceptance rate is lower than Harvard. Status isn’t a commodity that has a price tag, but it’s definitely something that’s implicit in people’s minds. That’s why young, driven, and ambitious 20-somethings pack their bags, head to San Francisco, pay the high rent, and hope to move up the status ladder.
Mimesis drives us to be the best at what we do to satisfy our desire for status.
Image from Luke Burgis' blog
When I was 15, I used to wonder why so many Indian VCs formed their worldview based solely on what A16z partners tweeted on Twitter.
After a decade in venture capital, I realized that they are just copying what the "big money" is doing. Why innovate if you can "copy" the best? Is it an advantage? Not necessarily. But it makes money. That's why there are a lot of "me-too-many" products on the market. Many people imitate and iterate on a concept.
Just like Facebook wasn’t the first social media platform. Instagram wasn’t the first media sharing platform, Spotify was definitely not the first service to allow users to stream media.
Iteration benefits consumers. Initially, the market will be crowded, but over time, the market will determine who survives. Therefore, you need multiple founders and multiple VCs working on a problem, often with similar solutions.
Keith Gill is the Soros of meme stocks. Murad is the Soros of meme coins. I’m trying to be the Soros of newsletters.
This is a common occurrence in liquid markets. George Soros rose to fame by short the pound and crushing the Bank of England. Keith Gill — known as “ Roaring Kitty ” — is famous for sparking the GameStop frenzy. Both are skilled at pulling large swaths of the market into trades they’ve entered. Soros convinced traders to short the pound, putting pressure on the Bank of England . GameStop investors pushed the stock higher until Robinhood stepped in to limit short short.
What Roaring Kitty is doing is just a modern version of Soros’s “reflexivity theory” — both are designers of self-reinforcing loops. A big deal attracts attention, people follow suit, the price goes up, more people pile in, and suddenly the asset hits a new high.
In Soros's time, there was no Twitter for endless rants. In fact, he left the market to study philosophy and write books. But today, you can pull people into trading by simply posting a ranking of meme coins. What Soros called reflexivity is what we now call meme coin mania.
It’s often argued that financial nihilism is why regular people invest in meme coins. The idea is that this generation finds themselves at 30 without stable jobs, partners, or houses, and so are betting on random codes on Twitter in the hope of cracking the financial system that’s bankrupting them (and keeping them bankrupt). But I think that’s a weak argument.
The real reason is "mimesis", which means imitation.
That’s right, the same thing that determines your career choices, which startups YC invests in, and how Keith Gill makes his money is also the reason you lose a ton of money on a bunch of bullshit tokens called TrumpShibuInuWallDoesnotExistcoin.
Image source: Murad’s tweet
Let me explain what happens when the Internet breaks down the barriers to entry and information in financial markets.
Things usually go like this:
You see a 17-year-old kid on TikTok or Instagram telling you about the “road to wealth” while sharing the Meme coin he traded the day before. The social media marketing manager at the office shows off an NFT worth $150,000 on LinkedIn. You see the bills piling up and think you’ve found another way out. A friend shares a ticker in a WhatsApp group. A new coin is launched on the exchange, and its icon is a dog in a hat. This is your chance, you think. You invest $100 first and watch it go up to $117. You think, what if I invest $1,000? Or $10,000? Before you know it, your credit card is maxed out and you’re in it.
Note: It should be noted that Murad is mentioned here out of respect for his huge impact on the Meme coin market, not as a dig. He defined the Meme coin category today, just like Keith Gill once defined Meme stocks.
Pump It
The emergence of Bitcoin in 2009 completely overturned the way capital operates on the Internet.
You can provide "labor" through "proof of work" (PoW, the thing that secures the network) and get rewarded with Bitcoin for it. Since the value of the asset will rise further in the future due to demand and deflationary pressure, the future value of current labor is higher.
People are therefore motivated to provide computing power to the network and hold Bitcoin. But with the advent of ICOs, asset issuance is decoupled from proof of labor, and you can mint tokens without labor.
Between March 2017 and 2018, ICOs raised approximately $28 billion.
VCs claim this is the future of “finance,” enabling individuals to coordinate capital and resources to launch new networks. This sounds reasonable until you realize that VCs often invest at low valuations (e.g., $10 million) and then raise money within a few months at a much higher valuation (e.g., $100 million).
In the traditional venture capital world, this kind of madness happened during the Internet bubble 18 years ago, and cryptocurrency has reproduced this madness.
Between 2018 and 2023, the token issuance market matures. We no longer have the ICO craze, but VCs are still keen to invest at low valuations in an attempt to go public at high valuations. This is arbitrage.
Image: This video by Meltdem Demirors does a great job explaining how capital allocation in crypto starts to free fall the moment a token goes public
There is nothing wrong with raising money at low valuations, but reselling it to retail investors at a high premium without real progress is extremely predatory. Chamath used a similar strategy with his SPACs (special purpose acquisition companies) during the COVID market in 2020. Currently, his SPACs are down 42% on average.
Now, everyone can become their own Chamath or venture capital fund with one click through PumpFun. PumpFun is either the most innovative financial product of the last century or the most predatory platform. The truth is probably somewhere in between. Meme coins are to the market what pornography is to the media. Just as pornography will not disappear, meme assets will exist as long as greed and speculative desire exist. And many memes will drive meaningful innovation, just as pornography has driven technological progress.
While discussing the ethics of this is beyond the scope of this newsletter, I would like to share two interesting figures.
PumpFun cumulative revenue chart. Their total revenue last year was $500 million.
Another chart shows the number of new assets issued on Pump over the past few months.
The second chart is strikingly similar to previous hype cycles.
I can overlay it with a chart of ICOs or NFT issuance and it looks exactly the same. I’m dealing with (or adjusting to) two parallel realities: one is that PumpFun is probably one of the most profitable startups ever; the other is that it’s a stark demonstration of how crypto works in its rawest form. Dave Portnoy wondered with interest in a tweet what a “rug” is.
When we say that blockchain can coordinate capital, we don't say who the capital is coordinated for. It can be used for cancer research or urban planning, but marginalized people on the Internet tend to be indifferent to these. Everyone cares about profits and avoiding losses. We are adults, with bills to pay and dreams to pursue. As a result, everyone in the market is betting on the most speculative investments, and capital and attention are drained away from what really matters. This is the current state of the market.
The real impact of PumpFun is that it has transformed cryptocurrency from a niche to a mass tool. When influencers, presidents, and countries issue tokens and watch prices plummet, we are not creating wealth through cryptocurrency like Bitcoin did in 2009, we are destroying it. But just as personal expression on the Internet cannot be defined by its creators, market outcomes cannot be determined by tools. The creators of TCP/IP cannot decide what I write in this newsletter today.
This frenzy is the price of releasing the genie in the bottle.
Figure: This chart was inspired by a conversation with CK from ThirdPrime as a joke. I promise you, I have 0.0042 ETH in my cold wallet. I am an Ethereum person.
Without a willingness to be offended, there is no real free speech. Without providing the tools for the free market , there is no way to prevent most of those tools from fueling greed and speculation. Especially in an era when regulators are “dozing off”, the consequences of launching a “rug pull” asset are minimal. Free speech works because there are consequences for saying the wrong thing. How does a free market work without consequences? This is the big question that cryptocurrencies are trying to find answers to. But as with most things, the market will eventually find its own solution.
The meme market has many similarities to blogging and personal expression on the Internet.
In the early days, blogging was a niche activity. You could start one, but not everyone would read it. I love seeing old WordPress blogs online because they remind me of a time when people wrote for expression, not for influence. The same was true of Meme assets a few years ago. Doge is special precisely because the founders didn't launch it just to make it a "meme." As Meme asset tools develop, everyone will be able to launch a Meme coin, just like everyone can start a Facebook page.
In social networks, attention ultimately focuses on a small number of creators.
In meme assets, capital also ends up concentrating in a few key names. The challenge is that people lose money as they mature.
Image: This tweet from Wassielawyer captures what many of us are thinking
So, where do we go from here?
Are we "finished" as the Zoomers say? Is there a light at the end of the tunnel? Do I need to find another industry to build my career in? What do we do!?
I would be lying if I didn’t admit to thinking about these questions many times over the past two quarters. It’s not because I’ve lost faith in the potential and future of cryptocurrencies, but because of the way attention is being allocated. The only antidote I’ve found is to keep my reality rooted in human interaction rather than the noisy rhetoric on Twitter.
Every time I see my newsfeed flooded with the latest scam, I talk to one of our portfolio founders for a mind massage to restore my confidence . This is perhaps the greatest privilege of working at Decentralised.co – we get perspective from founders beyond the newsfeed.
So if I were to zoom out and look at what’s exciting and what’s going to define the next decade, I’d lay it out like this: Think of it as a blueprint.
Respect the Pump
You just listened to me complain for ten minutes about scams and lack of substance in cryptocurrencies. If you’re still here, buckle up and get ready for a dose of optimism and the logic behind it.
First, let's look at the apps that are actually making money, as this will give us an idea of whether these are worth our time. TokenTerminal doesn't track all the apps we'd like to see, it's slightly biased towards EVM (Ethereum Virtual Machine), but has the best data on price, revenue, and profitability. So, I started my investigation by looking at the top ten categories by revenue.
The revenue data you see below is monthly revenue for each of the last ten years. The data is granular because we want to play the skeptic. We want to see growth because - hey, what's the point of staying in a stagnant industry? We want "hockey stick" growth.
When you think about revenue, you'll see that there are three phases of industry growth.
The first phase I call the “Dark Ages.”
Before the rise of Ethereum smart contracts. Before 2018, blockchain transaction fees were the industry’s main economic output. Sure, you can consider miner income and related businesses (like Coindesk), but that doesn’t apply to marginalized people. Developers didn’t directly benefit from it.
Then between 2018 and 2022, there will be an explosion of applications and use cases.
Revenue is no longer limited to transaction fees, but extends to on-chain businesses. You can see "blockchain layer-1" fees gradually decline, making way for businesses such as exchanges and lending. I think this is the "enlightenment era" of cryptocurrency - a magical period when people question the limitations of blockchain, go against the dogma of "savior" Satoshi Nakamoto, and create alternative business models.
It was during this “Age of Enlightenment” that models like Play-to-Earn (Axie Infinity) and yield farming took off. Of course, some chapters ended in failure, like the political alliances in medieval Europe. But it paved the way for people to question possibilities and explore new paths.
After 2022, a new category has emerged at an astonishing rate.
Guess what I call it? This is the “industrial age” of cryptocurrencies. Just as humans discovered the ease of machine manufacturing and transportation, crypto practitioners have discovered the scalable income possibilities without human intervention.
The "alchemy" of this era is stablecoin.
Everyone wants yield and fast money movement. Fintechs still rely on banks, which in turn rely on governments to dictate the rules for money movement. Stablecoins abstract away the rules that fintechs have historically faced, giving rise to a new generation of businesses. Tether and Circle have made over $5 billion in a single year.
According to The Economist, last year stablecoins transferred $2.76 trillion in value, about two-fifths of all transactions on the blockchain, up from just one-fifth in 2020. The same article states that stablecoin transactions accounted for 4% of Turkey’s GDP in March 2024. This article is not specifically about stablecoins, so I won’t go into details.
Now that we have established that this industry can make money, and that it is in the billions, it is worth discussing how "sticky" these revenues are. Cryptocurrency revenues may be seasonal, but at scale, it can be life-changing. Think of OpenSea, PumpFun, or the Play-to-Earn craze. Of course, some will say that NFTs and Meme assets have created a huge "bubble" and ruined many people's lives.
It’s beyond the scope of this article to discuss whether markets are zero-sum games (and I don’t want to send you into a deeper existential crisis), but one way to think about it is that revenues during these cycles are high enough that companies can make more money than they would otherwise make in their lifetime.
A mental model that applies to blockchain-native applications is that product maturity will vary.
Depending on the market cycle, the speculative premium on a product will increase. Take stablecoins, which have grown to the point where large businesses can use them for remittances. That’s why Stripe made a $1 billion acquisition in the space last year.
On-chain analysis tools sold to governments (such as Chainalysis) and smart contract auditing companies (such as Quantstamp) have also reached mature revenue scales. These businesses have considerable cash flow and sufficient profit margins to attract the attention of traditional capital.
Figure: For reference only, not as an absolute basis, many of the positioning here are subjective
If you stack these problems, you get a matrix. On one end you have centralization and decentralization, and on the other end you have seasonality. Extremely seasonal apps like FriendTech may be centralized, but have a hard time bringing value to the ecosystem because they rarely pass value on to edge users. Uniswap, while mostly decentralized, has a hard time returning value to shareholders. That’s why last year they started adding fee switches on the front end. Since then, Uniswap labs has generated nearly $103 million in fees.
The business needs of entities — revenue, profit margins, and control over capital allocation — often conflict with our desire to decentralize and deliver value to users.
A few companies have found a balance — being both innovative and decentralized while delivering value to users and having plenty of room to grow.
My personal favorite is Layer3.
If you're hearing about Layer3 for the first time, think of it as an ad network that aggregates one of the largest user bases in the crypto space. They help new users discover crypto-native products and reward early adopters with USD or crypto. So, if you're a new app or a newsletter like us, you don't run ads on Google to attract untargeted eyeballs, but instead go to Layer3 to get a hand-picked group of crypto-native users. Unlike Google, Layer3 gives most of the value generated back to users in the form of incentives.
I've written in the past about how they've aggregated users . On average, about 60,000 users trade on their product every day. Last quarter alone, they gave back about $1.4 million to users. That adds up to $5.8 million for the year.
In the advertising world, $5.8 million may seem like a pittance. Kanye West spent more on his Super Bowl ad. But the point is, $5.8 million is interesting because it’s an early example of a product giving value back to its users. It’s an open ad network where consumers (i.e. early adopters) are compensated in real dollars through a global pipeline of funding.
Anu of Working Theorys has a great framework for zero-sum vs. positive-sum products. She points out that some products reduce usage of other products. For example, you either use Google Docs or Notion. Companies rarely switch between the two because each product suite has a lock-in effect and you need to manually migrate entire teams. But some products are positive-sum — you can use Perplexity, Claude, and ChatGPT to address different use cases on the same day. These products won’t cannibalize each other’s market share until user preferences are as clear as Google vs. Bing.
Predatory products are often zero-sum because they make (ordinary) users worse off. Markets are not zero-sum games if there is an underlying economic reason. Someone who bought Tesla stock in 2018 did not need others to lose money to profit from the subsequent rise. You could say that the stock movement is correlated to the output of Tesla the company. But in meme markets, the game is increasingly like negative-sum.
The reason is simple:
Users need others to invest capital in order for asset prices to rise. This is fair in itself - all capital markets work this way. But users also need this "atmosphere" to last a long time. From the beginning, you expect the frenzy to continue. This is also possible because the "greater fool theory" will make people buy assets. But as prices rise, people revalue their "net worth" based on thin liquidity pools. For example, there is a token related to Congo with an FDV (fully diluted valuation) of $1 billion, but there is less than $5 million in the trading pool. People "revalue" their net worth based on this illusory capital, and when the game inevitably ends, they are usually in a worse situation.
Products like Layer3 are positive-sum because they don’t extract value directly from users. If you use their tools and stick around long enough, you can make thousands of dollars just as an early adopter. As crypto crosses the user chasm, we’ll see more and more products optimized for positive sum because that’s how you build critical mass of users. The more users Layer3 has, the better equipped the team is to negotiate better deals for its users.
Marketers also prefer positive-sum games because they know that Layer3 brings in more skilled and experienced users — they’ve already used multiple related products.
One way to think about the current state of cryptocurrency revenue is through what I call the “marginal adjacent user” lens. In emerging fields, founders are often better suited to solving niche problems. In the 1970s, Jobs and Wozniak built products for hobbyists who were familiar with computer technology but needed portable and affordable home computers - niche, technical, expensive, and targeted. In contrast, Jobs in the 1990s was obsessed with bringing technology to the masses - less technical, more universal, price-sensitive, and extremely user-friendly.
If you were building in crypto before 2021, building products for on-chain active users was viable because the revenue per customer was high enough. In a curious market, selling novelty is effective. But as users lose money and lose interest in the following years, it becomes important to expand the market. A few chains and applications (mainly on Solana) have done well and captured the next wave of on-chain users. The risk is that these products may repeat the mistakes of DeFi in 2019 - just slightly faster, slightly cheaper, and slightly better experience, but essentially doing the same thing.
According to a recent report by Mary Meeker, there are about 3 billion people on the Internet today. The most optimistic estimates put the number of monthly cryptocurrency transactions at between 30 and 60 million. That’s generous. But think about it, there are multiple products competing for those users. Essentially, this is “industry attrition” — the market is not growing enough to accommodate multiple players.
As a result, teams compete on either (i) pricing, (ii) incentives, or (iii) features until they die out due to insufficient profit margins—a veritable “race to the bottom.”
What is the alternative? Build something that goes mainstream, because mainstream markets have both moats and profit margins. The “popularity contests” we have in crypto about who has the highest TPS (transactions per second) and who is more “aligned” are useful for information flow, but they don’t pay the bills. Layer3 attracts me because they are not chasing existing users, but expanding the market and capturing a portion of the value from it.
The theme of "expanding the market" is not limited to Layer 3. OpenRank is a product built on Farcaster. Since user activity on Farcaster is often relayed on an open network (blockchain), it is easy to evaluate which users are valuable and which communities are natural. OpenRank helps identify the right users and incentivizes them directly with tokens, NFTs, or early access rights. This means that any developer can target any user on the social network.
Image: Karmalabs mapping Farcaster users in their filter
Layer3 and OpenRank are two different approaches to advertising in the blockchain era.
One is to curate the protocol, users, and incentives, and the other is to allow the market to identify users and directly target them on the open web. While it remains to be seen how Farcaster evolves and which advertising model will last, it is certain that blockchain is changing the way the value of content is transferred on the web. Of course, this market is still niche at present, but they have the potential to grow exponentially.
One example I've witnessed firsthand is stablecoins. In 2019, the total market value of stablecoins was about $1 billion, and it was hard to imagine that five years later their market value would reach $204 billion. But today, the total market value of stablecoins is that much.
The key question, then, is can the market of cryptocurrency-related users be just as large? Can they grow to hundreds of millions of users in the next five years? What will it take to build such a world?
We are already seeing some early clues in the parts of the industry that interact directly with hardware networks.
For example, Frodobots and Proto Use points or tokens (like USDC) to incentivize users to map geospatial data. Frodobots ships physical robots to users who ride them around town and upload data to create the world’s largest urban navigation dataset. Proto incentivizes users to help map dense urban networks with their phones. What attracts me to these models is that they can capture data from third parties trustlessly (via device sensors) while incentivizing users with a global network of capital.
UpRock is using the crowd to provide data for website monitoring, which is a variation.
UpRock's SaaS platform, Prism, provides an alternative uptime monitoring system with comparable accuracy to DePIN (Decentralized Physical Infrastructure Network). Their network consists of nearly 2.7 million devices worldwide and forms the backbone of UpRock - the core consumer product that powers Prism. When developers need insights, they can tap into UpRock's user base - many of which earn rewards by running mobile and desktop apps to collect data. Can this be done on a fiat rail? Of course it can.
But try making millions of small payments every day in over 190 countries and see what happens. UpRock uses blockchain to accelerate payments and maintain a verifiable, public record of past payments. They connect it all with their core token (UPT). As of writing, the team burns UPT when they earn external revenue through Prism.
This isn’t to say we’re in the midst of a radical wave of innovation that’s about to make tokens Iron Man-like.
No, we are not there yet. The best example of this gap between innovation and token performance is agentic tokens. Let me start with a set of charts.
After the hype cycle, we can safely ask ourselves: Is there anything valuable lurking in this space?
The most compelling intersection of cryptocurrency and AI lies not in the agents themselves, but in decentralized training and computation networks — training and serving AI models on a globally distributed network of computers and GPUs incentivized by tokens.
You can see an early iteration of this via Pond’s Model Factory . They are incentivizing the creation of a model with machine learning capabilities that simulates how judges rank open source contributions.
So, you can collect data (like UpRrock does), and you can find people to build models (like Pond does). But where do you run these models in a world where energy and computing power are scarce? Shlok explained this in a previous post , and I’ll share his original words:
Who are these new AI-native marketplaces? io.net is one of the early leaders in enterprise-grade GPU supply aggregation, with over 300,000 verified GPUs on their network. They claim to deliver 90% cost savings over centralized incumbents and over $25,000 in daily revenue ($9M annual revenue). Similarly, Aethir aggregates over 40,000 GPUs ( including over 4,000 H100s) to serve AI and cloud computing use cases. Previously, we discussed how Prime Intellect is creating a large-scale decentralized training framework. In addition to these efforts, they also offer a GPU marketplace where users can rent H100s on demand. Gensyn is another project betting big on decentralized training, taking a similar training framework and GPU marketplace approach.
In other words, cryptocurrencies have evolved to the point where we can acquire data, models, fine-tune models, and assemble the physical infrastructure needed to run AI models. Meanwhile, the average person in crypto is still asking “when will it go up?” — completely unaware that they are sitting on a pile of uranium that can take us much further. Some developers are realizing this. Gud.Tech and Nomy Teams such as are working hard to build trading agents that can accept user input, understand context, and execute trades.
what does that mean?
Chatbots have been around since at least 2015. Getting information from a bot itself is not new. What attracted me to Gud and Nomy was how they abstracted the complexity of purchasing assets across chains. Nomy provides a simple chat box where you type "buy 50 po on base with my eth" and the agent automatically completes the transaction without the user having to sign multiple times. Similarly, Gud is developing a trading product that almost always provides users with the best price by optimizing liquidity sources. These products blur the line between how we consume information (Twitter, communications, etc.) and how we execute trades, and it's all built on advances in AI.
Why do I mention these two teams in particular? Because they are the essence of everything we have discussed so far.
They are targeting cryptocurrency-related users rather than competing for the existing small user base.
They are not zero-sum because their growth depends on the continued engagement of their users.
They package critical infrastructure (like gas-free transactions) into a single product that users can use.
They are building on the cutting edge of cryptocurrency and AI in a way that is relevant to the average internet user.
I highlighted Gud and Nomy because they were built by teams that started deep in infrastructure before moving to the application layer. This reveals a simple truth: the age of applications has arrived. If cash flow, revenue, moats, and user retention really matter, then applications will take on this mission. Infrastructure has matured to the point where arguing about TPS (transactions per second) doesn't make much sense anymore. Our reluctance to abandon these rules of thumb is just another sign of stagnation.
To evolve, we need to change the way we talk. Perhaps, even develop a shared language of optimism.
Cathedrals, not trenches
In the mid-1940s, as the threat of World War II receded, the aviation community was presented with a new challenge - to build an airplane that mimicked birds, pigeons to be exact (this comes from Where's My Flying Car , a great book, by the way). The challenge was to build an airplane that could land in a backyard, without the need for a long runway. Despite huge improvements in the efficiency of airplanes in terms of passenger capacity and fuel efficiency, we still don't have retail-owned flying airplanes today.
Does this mean that the airline industry has failed? No. In 1961, John F. Kennedy said:
We choose to go to the moon. We choose to go to the moon and do the other things in this decade not because they are easy but because they are hard, because this goal will organize and measure the best of our energies and skills, because this challenge is one we are willing to accept, one we are unwilling to postpone, and one we are determined to win, as well as other challenges.
Fueled by this optimism, decades later, U.S. investment and energy during the Cold War laid the foundation for today’s Silicon Valley. Even more interesting, in 1903, the media openly argued that we wouldn’t have flying machines for another million years. A million years!? It sounds like they were counting on evolution to help us. But we didn’t wait blindly.
We made it happen. In 1969, we landed on the moon.
But between 1903 and 1969, attempts at flight failed multiple times. Cryptocurrency feels a lot like that — we’re too focused on perfecting the “engine” instead of making it actually serve the purpose of “moving passengers.” Kennedy’s political will injected new energy and direction into aviation — mobilizing people, resources, and policies toward a mission. It wasn’t a one-time event, it was a pattern.
In David Perell’s article on Peter Thiel, he highlights how medieval Europe came together to build cathedrals during times of destruction and danger. These buildings were symbols of hope, and took hundreds of years and enormous resources to build. They required the coordination of finances, security, talent, and labor—far beyond the everyday efforts.
The average worker who built a cathedral could only hope to live to see its completion. Today, the equivalent is spending 4-5 years scaling a consumer application while much of the industry obsesses over the technical details.
Today, with Memecoin , we can see humans grappling with how to harness money when assets can be issued and traded as fast as texting. We’ve been through this before. In the 1400s, we adapted to the effects of interest rates on capital. In the 1700s, we discovered the power and chaos of trading company shares. The South Sea Bubble was so severe that the Bubble Act of 1720 prohibited the creation of joint-stock companies without a royal charter.
Bubbles are a feature, even a necessity. Economists view them negatively, but bubbles are how capital markets identify and evolve into new structures. Most bubbles start with a surge of energy, attention, and obsession with an emerging field. This excitement pushes capital into anything investable, driving up prices. Lately, I’ve seen this obsession with proxy projects. Without the combination of financial incentives (token prices) and hype, we wouldn’t have so many developers exploring this space.
We call them “bubbles” because they burst. But falling prices aren’t the only outcome. Bubbles fuel radical innovation. Amazon didn’t become “useless” in 2004, it laid the foundation for today’s AWS, which powers the modern web. Could investors in 1998 have known Jeff Bezos would succeed? Probably not. That’s the second aspect of bubbles — they create enough variations of an experiment that eventually a winner emerges.
But how do we conduct enough experiments to see a winner?
That’s why we need to focus on building cathedrals instead of dancing in the trenches. My argument is that meme assets are not inherently bad. They are great testing grounds for financial innovation. But they are not the long-term, revenue-generating, PMF-finding game we should be playing. They are test tubes, not laboratories. To build our cathedrals, we need a new language.
We are already seeing the beginnings of this new language in projects like Kaito and Hype. I don’t own their tokens, but they are category leaders in their own right. Their token prices have also found stable footing, unlike many VC-backed “star projects” that have nothing to show for it after 36 months of talks. As people realize that the core metagame of crypto has evolved, more people will gather around this new language.
Figure: To end stagnation, one must face the "monster" within
In 2023, when I wrote “Is Crypto Failed?” , I thought we were heading towards a market where founders and VCs both recognized the importance of revenue and PMF. That was my post-FTX PTSD, hoping the industry would converge to rational players. In 2025, I fully realized that this was no longer a reality, and hope was no longer a strategy.
Next, high-motivated people will create a parallel game and build their own “cathedrals.” Instead of chasing fame on crypto Twitter, they will listen and build for marginalized people on the internet. They will make money through actual product revenue, rather than relying on the “blessings” of exchanges. The impetus for this comes from the crisis of faith I mentioned at the beginning of the article.
People will start asking, “Why am I here?” and switch up the game they play.
The split between revenue-generating tools and non-revenue-generating tools will be a huge watershed for crypto. Eventually, we will no longer talk about “working in crypto,” just like no one says “I work on the internet” or “work on mobile apps” today. They talk about what the product does. This is the language we should use more often.
When I started writing this post, I was trying to figure out what really bothered me. Was it fraud? Was it meme assets? Not exactly. I think the real pain came from recognizing the gap between effort and impact in crypto — especially compared to AI. Sure, we have stablecoins, but there are so many other innovations that people barely know about or don’t talk about. It made me feel stagnant.
In an ever-evolving market, stagnation means death.
If you stagnate, you die. But if you end stagnation, you survive. That’s the irony behind “end stagnation” — to preserve life, you have to be willing to end a core part of you. That’s the price of evolution, and the price of staying relevant. Crypto is at that crossroads. It must choose to end certain parts of itself, to give those parts that can evolve and dominate a chance to grow beyond the early stages.
Admire the Cathedral,
This post was very much inspired by The End of the Bubble and Stagnation. If you haven’t read it yet, I absolutely implore you to do so.
