The infrastructure builders in the crypto industry are falling into collective confusion—outside of DeFi, the market ceiling is only $200-300 million, and more people are shifting to AI. However, injecting crypto’s capital superconductor into AI-driven growth products might be the real way forward. This article is adapted from Kydo’s piece “Why crypto is not that fun anymore,” edited and translated by Dongqu.
(Background: U.S. CFTC Chair Selig has set the focus: weeks of internal discussions on perpetual crypto contracts, ending the turf war with the SEC)
(Additional context: Is Wall Street getting anxious? JPMorgan, Goldman Sachs, and other giants are considering suing OCC to protest the loosening of crypto banking licenses)
Yesterday, a tweet of mine resonated with many. It expressed a widely understood sense of helplessness: many of us, working in this industry, no longer find happiness.
If you’re not working on stablecoins, and you’re not extremely passionate about financial markets, chances are you no longer find any joy in crypto.
Worse still, you see AI racing forward at lightning speed and high enthusiasm, while your daily work seems stagnant. I understand this feeling because I’ve been there before—that’s why I posted that tweet.
And the fact that so many people resonate with it (I’ve received over 60 DMs so far) shows that many share this sentiment, just not spoken aloud.
In the last crypto cycle, building products for crypto developers was straightforward. Data metrics mattered more than revenue, partner brands mattered more than revenue, community atmosphere mattered more than revenue. You were just a cost item, and people were willing to spend money.
But this model has been outdated for about 18 months.
Today, more and more people agree: cryptocurrencies are only useful for finance. Haseeb from Dragonfly, Kyle from Multicoin, Toly from Solana—they’ve all said it. Most of you probably think so too, just not willing to admit it publicly.
I understand why people arrive at this conclusion. Most tokens are meme coins; they own nothing and can’t promise anything.
That’s why, in native crypto applications, only two types of on-chain earning are truly viable: trading and lending.
But many of us are not in DeFi. We’re building infrastructure to support new scenarios beyond DeFi.
And a harsh truth is: based on our analysis, the total addressable market for this work is roughly $200-300 million a year, spread across hundreds of institutions. The most successful ones earn only a few tens of millions of dollars. After years of development, this has become the market’s ceiling.
If you want to build a venture-scale company, the choice is clear. If you want to serve crypto developers, the market is very small. So, either you suit up and sell your infrastructure to traditional finance, or, like many others, pack up and shift into AI.
Many are turning to AI because they understand their strengths; they know their team’s expertise; they also realize their competitive edge isn’t in long, complex B2B sales in traditional finance. That’s why so many feel lost now.
So, you look toward the only vibrant field: the intersection of crypto and AI, trying to find a foothold. But the options aren’t easy either.
First path: traditional business + AI + issuing tokens. Tokens that serve no real purpose, just like most tokens. Essentially, you’re creating a normal product and forcibly overlaying a layer of financialization. It’s disheartening because you’ve been playing this game for five years.
Second path: decentralized AI infrastructure. Privacy, security, verifiability—classic evangelist routes, old-school approaches. But I guess most of you don’t want to do another long-term, faith-based movement without immediate feedback or income.
Third path: providing stablecoin infrastructure for intelligent agents. From a business perspective, this is interesting, but the competition is fierce. Circle, Stripe, and all major stablecoin players are aggressively attacking this space. For a startup without a clear breakthrough, competing head-to-head would be discouraging.
So, the real options on the table today are basically these:
Beyond that, there’s not much choice. That’s why most of you will probably leave and shift to AI. Honestly, I completely understand.
That’s why crypto is no longer as exciting.
These are my observations. It’s the conclusion I’ve reached after six months of feeling, analyzing, and repeatedly validating. If you just want to know why the crypto industry feels stagnant, now you know. You can close this article and go for a walk, relax.
But I want to tell you, I am more excited about what I’m doing now than ever before. The last time I felt this way was probably when I first understood what crypto was and what it could do.
So, if you want to hear what I truly believe, keep reading. Fair warning: from here on, I’ll start promoting my ideas.
Over the past six months, I’ve been trying to answer one question: how to find a sufficiently large market, with a clear, crypto-rooted business model, that is product-oriented and non-financial, so that both insiders and outsiders can use it?
I’ve always emphasized: cryptocurrencies are superconductors of capital, driven by capital to grow. The problem in the past was that we kept injecting into things that don’t grow—like pouring gasoline on ice, hoping it will burn brighter.
AI makes it unprecedentedly easy to create “growth-enabled, useful new products.” Tasks that once required 50 people can now be done by one. Building real products, real businesses, costs are plummeting. These things grow rapidly, generate revenue, have real users, and real feedback loops. They need fuel to accelerate. And crypto, historically, is the best fuel mechanism for this purpose.
This is the only truly interesting question I see now: how to leverage crypto’s superpower—the instant, global, programmable capital formation—in genuinely growing ventures?
We believe the answer is: Agent Companies. To achieve this, tokens must be able to “own” things. Coincidentally, we’ve been working on this for the past five years. Now, we aim to turn it into a product and realize this goal.
