Uniswap Is Building The Liquidity Network For Everything | Hayden Adams

Uh, is it? It might be. Thrilled to be here though.

It's been good. Uniswap has been having a crazy year. We've been shipping things left and right. There's been a huge amount of attention and traction. Recently the Robinhood launch in particular has gotten a huge amount of usage. But yeah, it's been a great year. I'm having a lot of fun.

Yeah, I think that it's so funny. Sometimes I think that a lot of crypto moves on these very short cycles. And I think a thing that has been really motivating for me is that ideas or plans that were laid like two, three years ago are coming to fruition now and starting to have an impact. And so I think that there's just this long-term strategic thinking and when that stuff pays off, that's some of the most motivating things. And so for example, Uniswap V4 hooks. Uniswap V4 was initially, we started working on it almost maybe more like even four years ago and we spent years developing it. And we had all these ideas around how we could make the market structure of Uniswap just much more flexible and customizable and programmable. And when we released it, we didn't have any of the hooks themselves built because we just spent the whole time building the platform. And a lot of people were, there was a lot of skepticism, like what can you do with hooks, or is it actually that big of an innovation, will it actually speed up development, how will routing work. There was all these conversations and questions and I think people didn't quite understand the power and impact. And it's only even now, it's been on the market for like a year, whatever, it's only really now that we're starting to see hooks have their really big moment where people are starting to recognize, oh, I can create a completely customized market structure and still have it work within the entire Uniswap ecosystem. You know, I could do that, it speeds up the development time. Building a new AMM from scratch could take years. And so being able to build something in a matter of days and being able to have it be extremely customizable, that's really powerful, but it took a long time for people to even kind of grok the core thing here. And so that's just an example, one of many examples, but just having these sort of long-term vision and then having it sort of start to pay off like years out when along the way there's all sorts of conversations like, was this a bad decision? Should we have just built a more opinionated singular implementation of an AMM that doesn't give people as much flexibility, but adds some improvements on top or something. But it just feels really good to kind of see that payoff. There's a gratifying feeling to be like, I wasn't completely crazy two years ago when I said...

Yeah. And there's also just this interesting dynamic where when you're working in crypto, you have to make a lot of very fast decisions and you have to sort of balance your short-term decisions against your long-term. And sometimes you have to move really fast, like with the launch pad we did recently, we actually moved very quickly from should we build this to let's build it to it shipped was a matter of weeks, not years. And so you sort of have to balance when you're moving really, really fast with when you're moving kind of almost really slow. And it's really fun for me, that's a lot of the fun of it.

Yeah. So I think that, and this has kind of been at its core, I think you've described us as a liquidity network. And I think that people sometimes, because of how it shows up in so many different expressions, people often think of us as like they're an exchange or they're a decentralized exchange. And they think of what is the role of an exchange and that's what that is but it's decentralized. But I think Uniswap is a lot more than that. It's more like an Ethereum, right? It's like a platform. It's a network. Its value comes a lot from its network effects and its ecosystem and the things building on top of it. And I think that even the word decentralized exchange has actually sometimes almost done us harm because when we talk to exchanges sometimes they're like, oh, but you're an exchange. You're trying to compete with us. And I'm like, well, no, we're like a, you could also think of us as like a coalition of market makers or you could think of it as, there's so many different ways to think about it, but at its core, it makes it really easy to create a market. Sometimes people under-reflect on the name automated market making. A big part of how we've created this liquidity network is by automating and enabling the automation of market making itself. And that is a really big part of it. And exchanges, when you talk to a centralized exchange for example and you say we're an exchange, they're like, oh okay, we compete with you. If you say, actually we're like a liquidity market making system, they're like, oh, we integrate with market makers, and it's a very different kind of vibe. And so I think that even as we've recently been talking to more folks in tradfi or more folks in fintech, the positioning and the framing has become increasingly important to helping people understand what Uniswap can do for them. And so that's just been a really interesting learning for us and something we've been working on a lot.

Yeah. So I think that we are a lot more B2B now than we ever, I think that there's this really interesting dynamic where Uniswap, it's a network, it's a protocol. And so I think that we can build, we certainly build several applications on top of Uniswap and they have value, but I think that the value of our applications is always going to be small relative to the overall value of everything that is built around Uniswap and that's always kind of the goal of it. And so with Uniswap, that's been a big part of even our shift in focus from just building our own apps on top of Uniswap to kind of empowering everyone else that is building apps. And so that comes in the form of supporting hook developers, it comes in the form of providing our trading API to other frontends, the same API that we use for our front end which just makes it easier, you also don't even need our API, you can build your own API. We've built out a developer platform, we've built out documentation, we have a team of four deployed engineers that will help other teams do their Uniswap integrations. And so I think another thing is some of this shift was not just because of Uniswap. Part of it was also because of the timing, we spent so much time building the core platform with Uniswap before that we didn't have time to help people building on top of Uniswap V4. And so now that it's out there and in the world, we've been able to shift our priority and focus to empowering everyone else that's building on top. And so I think a lot of it is talking to asset issuers, even people that are creating tokens, whether they're memecoins or crypto-native tokens or real world assets being brought on chain or stablecoins. The user base of Uniswap is very broad, it's anything that could benefit from liquidity, which is pretty much any asset on earth. And I think one of the interesting challenges we often face, there's oftentimes there'll be an internal push, people will be like, well, which assets are we prioritizing? What are the most important ones? And I'm like, it's a really tough conversation. It's like, we want every asset.

Yeah. It's sometimes I try to compare against YouTube. It's like, what type of video is blowing up on YouTube? I mean, it's a lot of, it's like, because YouTube you have like weird memes to people selling protein supplements to high quality educational content to news to people's own businesses to video game streamers. There's a lot of variety to the content on a social platform or on an internet platform. And I think people are not used to the world of finance working more like the internet. And so I think that part of what Uniswap is doing and part of what DeFi is doing is trying to create this sort of infinite, user-owned, user-generated landscape but in a world where it kind of works for financial markets, which is just such a novel idea. And that's why even, what is the point of automated market making? Well, automated market making is a critical component of every exchange system, or market making is a critical component, but the model from tradfi just does not scale to the scale of the number of assets and things that happen and are going to happen on top of blockchain rails. And so yeah.

Yeah, I think that's a really interesting question. It's true. I haven't talked about it a lot to be honest. It feels like every three to six months, one of the reasons I like what I do is because every three to six months I have to almost completely reinvent what I do and how the company, it's a constant evolution. It changes, the people change, the priorities change, the needs change. And obviously not every aspect of everything. There's the through line and I think a lot of it is grounded in this vision of making markets and exchange, lowering their barrier to entry and creating value for people. Everything has to be downstream of that vision and the impact that we want to have on the world. So that's the starting point. In terms of how you map that, when I think about the flywheel of Uniswap, I think that's another important part that people have to understand when you're running a company. And so the flywheel that I've been explaining, especially internally, and part of it is constantly evolving and improving how you talk about things, but the way I've been talking about the flywheel of this liquidity network that we're building around is that it's basically two things. There's growing liquidity and then there's unlocking distribution for that liquidity, all the different volume sources. And when we talk about the value prop of an AMM, a lot of it, what I've been trying to explain to the team and then ultimately to our customers is putting your assets on Uniswap, you're not getting listed on an exchange. It's almost like getting listed on every exchange on Earth at once, right? That's the distribution of Uniswap. It's integrated into literally thousands of frontends around the world through whether it's through our API or through direct integrations or data partners or whatever it is. There's this mass global distribution that happens when you put your tokens on Uniswap. And when you talk to an asset issuer, and I'll come back to the team stuff, but it's all kind of downstream of this, but when you talk to an asset issuer, they might not understand every nuance of what makes automated market making cool, and we're going to get into that, and that's really cool, but they understand the need for distribution for their assets. And so it's like, oh, you put your asset in Uniswap, it now shows up in all of these places, all of these surfaces around the world. And so there's kind of the distribution side and then the liquidity side is we make it really easy to just take an asset and create liquidity in an automated way. And the more distribution you have, the more value flows to that liquidity, the more benefit of the liquidity. The more liquidity you have, the more valuable our distribution points are becoming, things like our API, which has grown massively. So I think helping people internally understand the strategy is a really key part to running the company. Understanding it yourself. And then I'd say that generally you want teams mapped to the organization's priorities. And so I'd say that one thing I'll say is when we were a little bit more B2C and where we were a couple years ago, it was a little bit more top down and kind of structured. We've become a lot flatter. I think part of this is also, and we've become a little bit smaller. I think we peaked out at about 160 people.

Yeah. And it's definitely, there's an AI component. There's, I think there's just broader learnings. I think a lot of people are even kind of using AI as a convenient excuse of it. But the reality is also that they overhired and were bloated. And we've aimed to be, there's been maybe a few different eras of Uniswap. In the early days we were like five people doing this thing and then we've kind of grown into this B2C thing that was like 150 people and now we're like this B2B thing that is like 120-ish people. But as we've grown we also went from a period where it was extremely flat and small to starting to hire in a lot more leaders from outside of crypto and had a bit more of a top down structure. And I think now it's, I do have a lot more reports than I did say a year or two ago. I have maybe 10 to 12 reports versus before I had like five or six. And I think I used to feel like the more reports, the more work. But I've actually found that in some ways the more reports often, you can do more things in a shared group or more people are on the same page and they're able to work together to figure out things. And so I found like a flatter structure for Uniswap has worked very well. It also lets different people have ownership over different important focus areas or priorities. And I've definitely found that the more reports I've had, almost the easier it's run itself. And it's kind of ironic that I have to learn this as a person that creates decentralized systems for a living. But it turns out that decentralization to some degree, or you don't want consensus decision-making, right? That's the failure mode. But there's something nice about having a broader network of people working on something as long as they're coordinating well with each other. And so I've found that we've definitely shifted also, I think that we've removed any sort of layers that felt like, if a layer is strictly only for management then it's not, you kind of want every leader to be in the weeds of the content that is rolling up to them. A lot of these are basic learnings that people have from running startups but they are still key. And then I think the other thing is that a lot of it is just going to be downstream of people want to be working on work they want to see how their work has an impact on the organization. And so I think that the more we've, I think that one of the biggest positive changes that came out of Uniswap is that increasingly people are able to tie their work to a broader impact on the Uniswap ecosystem. And that's been a really motivating thing internally.

Yeah. So I think that there's a lot of different, and I have to be careful with specifically tossing out names because in some ways we're talking to everyone on Earth right now. But I think that the way I think about it is definitely trading platforms, whether they're frontends or backend systems. I think that Uniswap as a liquidity source tying into any various trading platforms around the world. I think that idea of 24/7 on-chain deep liquidity in literally, at this point, there's technically millions of assets on Uniswap but definitely thousands or tens of thousands that trade every day. Having that plugged into anywhere that people want to access and exchange value is useful. But I think that there's also starting to become these other interesting integration points, things like payment networks. So payment networks, they want to start accepting payment in crypto or more assets. Uniswap has increasingly a valuable liquidity source for them. And so I've had a little bit of this vision of just, people in crypto talk a lot about what is the best decentralized money. And I think the best decentralized money is basically people being able to use literally any asset they want as money. And so sort of having liquidity, as long as you have a low-cost liquidity network that can cover every asset, you can have a system where someone can pay in any asset or even a portfolio of assets, and then have the recipient receive any asset or portfolio of assets. And so I really like this pay anything, receive anything vision. We're starting a little bit around Uniswap's, particularly stablecoins and payment networks. We haven't talked about this a lot publicly, but we've been talking to them behind the scenes.

So this is at the liquidity network layer. So here we're talking about things like Curve and things like other AMMs. It's something like 60 to 70% now of EVM, I think it might even be 70, and then it's something like 40 to 50% of all stable swaps across non-EVM as well is just Uniswap protocol. And so I think that massive growth on the stablecoin side has been helpful for positioning us with payment networks who are starting to, they're not quite ready to accept payments in every token on earth but they're maybe starting to think about how do we accept payments in stablecoins, oh there's all these stablecoins, how do we choose. And so we're positioning Uniswap as a liquidity source for that. That's another example. I think that there's some interesting opportunities here with what's happening around RWAs and tokenized stocks and saying, especially outside the US, giving access to US stocks to non-US folks and having it trade in a way that is 24/7, fractional, nights and weekends. That's actually really powerful and ultimately we can bring that at a very low cost. And so I think that having integrations into whether it's international brokerages or whatever it is, that's an interesting integration point we've been thinking about a bit more recently. I think there's actually also, people sometimes think about it as US stocks outside of the US but I think people don't realize that there's an entire, it's like every country's stocks in every other country. It's not just the, and I think that partially obviously the US stock market has been the biggest one and that's sort of how people think about it, people often think about the obvious one. But part of it you have to think about is every, there's like the piping never connects very well between any of the countries' stock exchanges or systems. And so I think that decentralized networks and things like Uniswap offer an interesting opportunity to really have country A integrate country B's tokenized assets.

Yeah. And it's just so interesting that the way tradfi systems work, it doesn't even connect. And there was an interesting thing we built recently and I'll kind of connect it to this system in a second, but we built this hook that we're calling dual pool with the Spark team, collaborated with the Spark team on it, where you can have assets that are passively sitting in Uniswap liquidity pools also earning yield on a lending protocol. Pretty cool design, pretty useful. It really lets you reduce the opportunity cost of market making because you can also be earning yield. But I think that some of the things like that, doing things like that in a traditional setting, if you wanted market makers in a traditional exchange to also be able to have them earning lending yield at the exact same time, it's very hard to make it all flow and pipe together. And so I think that being able to unlock things like this are just really cool efficiency improvements. And so an example of this, the reason I thought of that and why my brain went there, by the way, was that you can't just take your treasuries and deposit them to an exchange and then swap them. You can't take your treasuries and put them in a centralized, or in a brokerage, a tradfi brokerage. Well, you can in certain ones but the point is that the way the conversions have to happen often on the deposit or they have to happen in a separate system and then you have to move, there's this seamless way that it flows in Uniswap. Obviously brokerages accept treasuries, I misframed it there. But the point here is that the piping is just seamless in crypto and that lets you really have a world where what does it look like when you can convert anything to anything and have that happen on the fly inside trading systems or inside frontends or inside payments networks. And so I think that that's a lot of what will unlock the next stage of value from here.

I, okay. So I think the way I would think about it is, it's just about abstraction layers. I think the underlying network is certainly getting more complicated, the same way that when people are programming, back in the day, I had a family member who worked on computers, who has now 60 years of computer science experience or something. And used to program using punch cards where you're physically programming your computer using a physical, and right, and over time computing has gotten increasingly complex and complicated, the chips, the internals, the firmware etc. And yet, computers haven't gotten harder to use because we've increasingly improved the abstraction layers on top, whether it's better programming languages or, obviously now, vibe coding. And so now I have other family members who never programmed in their life who now basically can write code using Claude, right? And so I think that the point I'm sort of making here is that ultimately the system certainly becomes more complicated. I don't think that the products or end-user experience should be, in fact it should become dramatically easier over time. And I think that a lot of what has been happening, a lot of what has made DeFi sometimes challenging for end users is that we're still building and defining the internal and the underlying and making more radical changes. And we're figuring out the front ends as we do it. But I think what we're starting to see is those abstraction layers become a step easier to use on top of it. And so I think of it as, definitely by the way agree with Paul on DeFi is becoming infrastructure and it's going to be increasingly integrated into products around the world and into systems around the world. And I think that what that ultimately will mean is that way more people can use it and make it way easier. Even on the side of, you mentioned maker and taker side, I think that there's a lot of things that we've started to experiment with that will make the act of LPing or providing liquidity a much better experience. But I think it took building V4 which increases complexity for us to be able to start building these things. And so we have various hooks that we're building right now that will actually dramatically simplify, starting with stablecoin stable LPing, so that you don't have to think as much about what you do but it still is optimized, improving the returns relative to what you might get today. So that's a more technically complex system but the end user product of providing liquidity there will be a much easier experience. And so that's something that is not rolled out yet, but is only possible because we have V4 hooks and now we're able to build these things on top of them. So that's kind of how I think about it there.

Yeah. It all kind of, ironically is all the same thing and it's just a different framing of it each time. So let's talk, go. In the very beginning of time it was like order books, this will never, like before even launching Uniswap it was like AMM will never compete with order books, it's just a less efficient system, market makers need to be able to update their orders and leaking your strategy on chain will leak value, whatever it is. That was the very first. And then eventually it was like, oh, actually RFQ systems will do this and that's kind of the same thing, it's more actively managed off-chain. And then it was like, Prop AMMs will do this, right? Prop AMMs are just another expression of the same thing where it's kind of like Prop AMMs are sort of like RFQ but adapted to be a little bit more programmable on chain to try to learn from AMMs. And by the way, in reverse, I think AMMs are going to learn from Prop AMMs. But the way that I think that usually, I think that there's two things. I think that a lot of, there's probably two things. One, it's very hard for people to, AMM reasoning is very complicated. I've been having internal mental breakthroughs about my own internal understanding of it in the past few months and I've been working on Uniswap consecutively for nine and a half years, almost 10 years now. And I'm still kind of informing my own mental model constantly. So I think that it makes sense that when you sort of hear about this new thing, it's very easy to think, oh, this thing is just going to outperform this other thing. But I think that there's a lot of, one thing is that it's often misunderstanding why AMMs were successful in the first place. And two is kind of underestimating the ability for AMMs to continue to improve and adapt from here. And so those are the two. So I'll start to kind of parse it out a bit.