I published "Crypto is Dead" in December. It got far more attention than I was expecting. People said it labeled something they already felt but couldn't properly articulate. For a little while, I was pretty proud of it.
Then shortly after Christmas I got sick and had nothing to do but lie there.
The ensuing days leading up to New Years were quiet for me. I couldn't work. I couldn't really do anything except think. And what I kept thinking about was: if I was right, what comes next? The article made the case that the era of crypto as an insular identity was over. I believed that. Still do. But I had been so focused on what was ending that I hadn't fully reckoned with what I was supposed to do next. What the industry was supposed to do next.
I had some answers but a lot more questions. Three days on the couch with the flu and a thesis I believed in but hadn't finished thinking through.
I decided I needed to go earn some certainty.
Over the next several weeks, I started using AI the way I had once gotten native to crypto. Daily. Seriously. And I also spent a lot of time on calls with people in DeFi that I respected. I asked them where they were seeing real demand, where they weren't, what was working for their users, and what was failing.
I was asking about customers primarily. About what users actually want to do and what's stopping them.
A pattern kept showing up. Teams building DeFi products for non-crypto native companies/users kept telling me the same thing: fintechs and institutions are getting serious. Not dipping a toe. Serious. They want to own more of their financial stack. They want a larger share of the economics that currently route through banks and intermediaries to route through them. They want to offer financial products directly to their users without asking those users to go create accounts somewhere else or navigate infrastructure they didn't build. The pitch wasn't "DeFi is cool." The pitch was "we can help you cut out the middlemen and control more of what you offer."
Then the public announcements started filling in the same picture at scale. Coinbase crossed a billion in on-chain loans through Morpho because the permissionless rails just worked better than keeping it siloed. Kraken launched DeFi Earn. And in late March, Whop launched Treasury: 21 million users, yield powered by Aave. The user gets a better product, the DeFi layer looks different than before.
The list got longer and soon enough this wasn't a prediction anymore. It was already happening.
Separately, I was spending enough time with AI to understand something that changed the math on all of this.
The way companies will be built over the next decade is different from the last one. Speed is different. Capital requirements are different. The kinds of founders who can now start companies and the kinds of businesses they can build have expanded dramatically. And as I watched that unfold, I kept thinking: the financial infrastructure underneath all of this is not ready for it. The way capital moves, the way companies raise money, the institutional stack around it, it was built for a different era. It's going to feel increasingly insufficient.
AI didn't create the demand I was seeing in those conversations. That demand was already here. What AI did was make me believe the demand will compound in ways that make this feel inevitable, not just interesting.
While all this was sharpening, DeFi kept looking worse.
AI had the narrative. Crypto felt tired. Yields had compressed. The speculation that had driven on-chain activity for years was running out of gas. But the part that shook me the most was watching people who had real conviction in DeFi, people who had been there from the beginning and understood what the technology could do, quietly step back. They didn't think the risk-reward justified staying on-chain.
Watching them lose conviction made me question mine.
I kept thinking about what they were actually saying.
They were right that the products in many ways had failed them. The yields were subsidized and then they weren't. The use cases that emerged were mostly circular, speculation funding more speculation. Everything had been built for people already inside the system, and most people never came inside it.
I think they were wrong about the direction though.
The things I said were dying in "Crypto is Dead" are still mostly dying. But the underlying system, the part that survives, is even more important than I appreciated at the time. While many of the existing products were disappointing, the need underneath them was getting stronger, not weaker. The world was moving toward exactly the kind of financial infrastructure DeFi was built to provide, at the exact moment DeFi looked least convincing. That gap between where the world was heading and where the products were is not an argument against DeFi. It's the argument for why the work isn't done.
Here's what I kept coming back to.
Traditional finance packages access inside institutions. The counterparties are chosen for you. The constraints are chosen for you. If you fit the packaging, it works. If you don't, your options narrow fast and the ones that remain get worse. The problem isn't always pricing, a lot of the time it's access.
On-chain finance changes that. Once capital becomes part of the software stack, it becomes designable. A company can define what they need, what constraints make sense for them, and find a counterparty on the other side who shares that criteria. The deal is structured in code. The constraints are enforced programmatically. What used to require months of institutional negotiation and human process can start to happen directly.
DeFi makes counterparty discovery and matching programmable. That's what it's for.
A fintech platform's treasury should be able to hold on-chain dollars, keep some liquid, deploy some under tighter rules, and move between those states without rebuilding from scratch through five separate providers. Legacy finance can approximate that as a sequence. DeFi starts to make it a system.
As my conviction solidified, the question it forced was: where do I want to be when this plays out?
Four years as an investor taught me something about where value actually accrues. You see enough companies succeed and fail that patterns stop feeling like coincidences. You build a view that isn't just about individual companies but about the aggregate, what the compounding of multiple founders and businesses and capital flows adds up to.
What that view showed me was that the startup layer in crypto was compressing and that incumbents in fintech and traditional finance were going to move on-chain. The interesting thing to me wasn't competing with them; it was understanding what infrastructure layer they would all route through.
At some point my thesis stopped pointing toward another fund and started pointing toward a specific place in the stack. The layer that issues trusted dollar infrastructure. The layer that connects institutional capital to on-chain markets. The trust boundary, where the guarantee gets made and the deal becomes real.
That boundary is where the lender side and the borrower side actually find each other. Enterprises and fintechs who want control over their balance sheet, how their users access finance, and the underlying economics. Founders and on-chain businesses who want access to capital without conforming to institutional packaging that wasn't designed for them.
The opportunity is building the system that allows them to transact directly.
@paxoslabs sits at that boundary. It helps companies leverage stablecoins and connect them to lending markets through an integrated stack. Trusted dollar infrastructure meeting on-chain deployment. That's the bet made real.
That is why I joined Paxos Labs as Head of DeFi.
DeFi doesn't win just because the thesis is clean. It wins when the products get good enough to deserve the role the world needs them to play.
And the need is clearer to me now than it's ever been.
The next era of finance will be defined by the systems that make counterparties easier to discover, exchange easier to structure, and trust boundaries easier to use.
That intersection is where I want to build and why I'm betting on DeFi again at Paxos Labs.