Zero Just Made Blockchains 100x Better | Bryan Pellegrino & Raz Zarick

The big unlock with Xero is that everything is part of the same system. It's all owned by the same chain. There is no separate entity that's deploying a layer 2 or owning the DA or owning any part of the stack. The whole stack is one unified blockchain.

Nothing said on Empire is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only, and the views expressed by anyone on the show are are solely their opinions, not financial advice or necessarily the views of Blockworks. Our hosts, guests, and the Blockworks team may hold positions in the companies, funds, or projects discussed.

Welcome back, everyone, to Empire. I'm really excited to have this pod. I was with Brian and the LayerZero team last week in New York where they announced Zero. I've known about this coming for a while. When I first heard the pitch, I was blown away. I mean, I've been in the space long enough so that my, my, you know, what really excites me is the bar is higher now. But when I had that conversation, Brian came to me, I was genuinely blown away as someone that is now trying to build and bring businesses on-chain. I think it just opened up my mind. So, Raz, Brian, welcome to the show. I'll let you guys introduce yourselves for the context, just for the audience, and then we can go from there. Before we get started, though, full disclosure, I am an investor in LayerZero. So of course, a lot of this discussion will be just objective and gloves off, really, because I think it's— but, you know, obviously I'm biased. So anyways, guys, welcome.

Thanks so much for having us. Quick intro. Brian Pellegrino, co-founder and CEO of LayerZero Labs. And Raz?

Yeah, Raz, co-founder and CTO. Of LayerZero Labs.

Awesome. Well guys, why don't we just get started? I mean, you announced Zero last week. I think the reception of that was pretty amazing. I was in the room. We were with Citadel, Citadel were all your existing investors, which is probably the most like non-crypto set of investors, you know, Sequoia and PayPal and a bunch of other folks. Give us the origin story and the thesis behind building Zero. You know, what was the initial conception and what was like, going into Zero, like what gap did you see focusing on from the messaging layer that you said, okay, we need to build a, we need to build like the more foundational piece here, which is Zero. Like walk us through that journey.

Yeah, I think one thing, obviously we sat in the middle and continue to sit in the middle of all the different chains, all the different architectures, right? So we, I don't know that there's another group in the world who's gone as deep in every different VM and architecture as we have. Because when we build a new endpoint on Solana, we have to rewrite it from scratch with the SVM and we have you have to figure out how you deal with dynamic dispatch and memory and all of these different things that go into it. And then when you build on TON, you have to figure out how TON deals with sharding and what that looks like. And so we had gone really, really deep in that. Again, I think we are deployed on the most non-EVMs out of any org that exists. We found crits in a huge amount of these systems. We've gone very deep. We've worked very closely with many of these teams. So we understand the trade-offs that people have made. But really the origin story of like how this happened was, was Raz was really generally just frustrated with this space in general. I, I lean a little bit more pragmatist sometimes. I have been in the space since 2011. I've seen lots of evolutions. He came in a little bit after that, but really, really, I think both of us like believed deeply in the early vision of Ethereum. And so he was complaining like every day, every single day about the state of the world and the Ethereum basically saying layer 2s inherit the security of the layer 1. And he was just like, that's just like fundamentally not true. And like, he didn't understand why this basically is being propagated. It's not like layer 2s can exist. They can be just like, didn't in the state of the world, like inherit the security. So he was like complaining every day. and I was, dude, chill out. Like, all the layer 2s are like our biggest customers where all of our volume comes from. Like, settle down. Um, but ultimately it just came to me being like, you know, like, all right, like, what, what would you do? Like, just how do you solve this problem? What would it actually look like? Uh, and it was interesting because back then, I mean, ZK was still like really nascent. Uh, but he had this really early architecture that he went and he whiteboarded. They really said like, hey, Who knows when this technology is going to be ready, what it's going to look like. But like, as soon as it's done, this is the way that you solve this, right? Like the underlying beacon chain should own all of these things. It should be 100% one system. You have trustless community, like drew out this whole thing. But it was like, it was, you know, some futuristic, like 20 years out, like who knows what this is sort of going to look like. And so once he did that, I think we're very early on, we were just like, oh, like that's that, that is like, like clearly correct, right? That is 100% what the world is going to look like. And then just like, I think we had conviction that, okay, like we have to build this, but we no clue how long it's going to take, no clue what that's going to look like. And then it was just like, just like grinding away piece by piece. It took 2 and a half years, right? So it was, uh, I think substantially faster than the rest of the world thought that it would happen, uh, for solving a bunch of these problems, but still like a, a very meaningfully long time of just being 100% stealth in the lab, like solving all of these bottlenecks along the way.

Raz, before I go to you, I mean, like, just, just for zooming out a little bit in context, like you built what I think, like objectively, I think the most widely adopted messaging protocol, like across crypto, right? You start off saying, look, there's gonna be a plethora of chains, let's connect them all. And there needs to be a common, like connective tissue, right? So I actually, in my journey building Inversion, your guys are one of the first ones I called because from my perspective, I care about onboarding users. And, and, you know, so I, I care less of if I'm going to tap into different ecosystems. I got to make sure that the end customer, the user, doesn't see degradation, right? And, and I think you guys have been super helpful in that journey. We've seen other folks like Circle launch ARK, and you could argue that that may or may not be the right move because you're basically undercutting your customer. In many ways. In this case, like, talk us through like the decision to kind of pivot to launching your own L1. Like, you're, you're— isn't that competing against like some of the chains that you serve, as you said?

Yeah, so, uh, one thing that we've learned that I think is very interesting is, I mean, we work super closely with a bunch of chains, or a couple of chains that we're just— have gone very, very deep with. We've done a ton with Arbitrum, we've done a ton with like all of these different environments that we've worked in developing native solutions for developers. There's a bunch of chains are close with. There's a bunch of chains that we're not close with at all. And at the end of the day, we thought early on that our relationship was with chains and like that's who we served and that that was the customer. We're just totally wrong. Like in, in isolation, chains don't want Layer 0. They don't care. What they want are the assets, right? And so like our customers are USDT Zero and WBTC and Athena and EtherFi and PYUSD from PayPal and all of these things. And ultimately what the chains want is they want the assets. They want USDT Zero, they want PYUSD, they want, uh, Athena, they want all of these things, right? And so what we, I think, learned over time is at the end of the day, it's fine. There are some chains even prior, just they just outright don't like us that much, and it still didn't matter because they want the assets.

Like, when the asset goes into the chain, not when it leaves. So it's awesome.

Yeah, yeah, exactly, exactly, right? Like, uh, You actually just make it easier to move across chains.

So the barrier to exit is lower when Layer 0 works. And that fundamentally is sort of, you know, why they wouldn't like you, right?

Yep. And so our customers, we've realized over the many years, is the asset issuer. Our job is to make their product as compelling and as seamless as humanly possible that, you know, they want to drive AUM, they want to be everywhere, they want to do these things. And at the end of the day, they are the pull to the chain. So like the chains don't care about us. They want the assets. And so we just build and do what's best for the assets. And then we're, you know, we're going to be in the chains anyways.

Yeah.

I heard you say, like, when you announced Zero, that it moves the industry's roadmap forward by at least a decade. Um, I think that's an extraordinary claim. Um, like, walk us through, Raz or Brian, like, walk us through, like, what is this, like, generational leap? Like, what is— how does this improvement really show up?

Yeah, I mean, the, the problem that I guess when to go back to what you were saying with Brian, uh, about like why are we doing like doing this is decentralization has become like a, a footnote and now everybody's focusing on expanding in terms of how fast it could be, uh, what kind of throughput they can do, uh, and how cheap, right? And Ethereum has was for a while going after the L2, like having other people expand and have them scale that way, right? Which was a more centralized approach. And we felt we, like, they're getting, like, the person who's fighting the most for decentralization has kind of put that down and said, at the time, and said, let's have a more centralized approach to expansion so that we can compete. And so what we realized is the only way to do that, to make an app, like to make a blockchain, I guess, competitive with Solanas and everything like that in the world is it has to, and still have those features is we have to build, I guess, a ZK approach, right? The only way to have it is to have these validators do less amount of work. And so this design that we drew, I guess, 2.5 years ago was end state. Imagine we have ZK that can do unlimited scale. And if we could do that, then we can have validators that are lighter than what Ethereum has. So imagine something smaller than a Raspberry Pi, a watch, something that's just incredibly easy. Anybody could just pick it up and run it. Um, and the work is not being replicated. Like that was the big unlock that we saw, uh, when we were as Layer 0, one of the things Brian left out is like seeing all these other chains and seeing what was wrong with them. We also saw what was special about ZK, which was elimination of replication where you could generate this proof. One big, uh, actor could build a proof and then everybody else would not have to replicate that work. And like, that's the expensive part of a blockchain. And so if we could remove that, like that's Solana's problem, that's every blockchain's same problem over and over and over again. So if you remove the replication, um, there's other bottlenecks still. And so we went through and solved all of them. We listed them out and said, okay, this is, this is where every, every part of the stack is broken, right? There's a storage problem with NPT. There is a scheduling problem within the VM. So how do you parallelize all the transactions? There is a network problem where how you disperse all this data and still retrieve it. Uh, verifiably. And then obviously the ZK problem. So we limited to those 4 issues that exist and, uh, we solved all of them beyond our wildest dreams. Like our original goal was let's make something as fast as Solana. Uh, so we're like 2,000 TPS, 2,000 to 4,000 TPS and as decentralized as Ethereum. So something that can actually compete, uh, and with the, with the principles we cared about. And then when we did QMDB, we just blew that away. We're like 2 million TPS, like we could achieve now. That was like, well, that's a new benchmark. Let's try to, if we can do 2 million TPS, if the bottleneck now for storage is 2 million TPS, uh, that's the ceiling, then what can we do, uh, with everything else? Like, let's just start shooting for that instead of 2K TPS. So then we went after that, just the scheduling. And Faffo, we were able to do, uh, well, 1.1 million TPS on EVM and even 2 million TPS if it wasn't an EVM.

Um, and so then just keep walking down that with that new goal of, okay, just to be clear, at the time of Faffo, like common wisdom that was like the The EVM is slow, right? Like the EVM is just, the EVM is the thing, it's slow. It's like, that's the problem. You know, the EVM itself, that's why you need SVM, that's why you need Move. There's all, you need these other approaches 'cause the EVM is just inherently bloated. It's bad, it's slow. And so this was like proving a million TPS on like a, on a single node with all of the overhead of the EVM. So that, that alone was like a very massive breakthrough. Sorry, go on.

Yeah, so, and then the ZK side, we, we thought it wasn't there yet, but we thought everybody in the world is working on ZK. There's like 15 ZK teams. We're just going to evaluate them all, pick one off the shelf, plug it in. We solve these other things and we're good to go. And then we realized they weren't really, they're actually trying to solve Ethereum's problems, which is like 15 TPS. And the goal is for them is drive price down, not become high throughput and do as much TPS as possible there. And so they're all trying to solve this one problem and not the problem that we want. And, uh, and then they're also reaching this like ceiling of what they could do. This technology is older. This math is older. Their approaches are like all, they're all just precompiles now to the point to try to get edge out every little, um, performance gain they can get. Uh, and so that's where we, you know, started looking at other, other new research that was coming out and that's where we found Jolt and we're like, this is greenfield, this actually scales, this is where we have to go, but this is obviously just research. How do we make a production team and build that? And we know what we want to do is 2 million TPS, so let's drive towards a terahertz cluster, a GPU-like proving cluster. And that's what we're building.

Before we go to your 4, what I think 4 proprietary systems like QM, BD, Faffo, SVID, and Jolt, I want to just go back to like, say you're a trad fi allocator coming in and say you talk about what is this heterogeneous architecture, right? I think most people would understand, like, there's always a trade-off. And in crypto, we, we've talked about the scalability trilemma, you know, whether you believe it to be true or not in today's world, different story. But, you know, you've kept hearing that. Um, so if you're like a— put yourself in the seat of a trad fi allocator who's like never used a blockchain but cares about performance, What, uh, why does like, I guess, separating execution from verification matter?

So there's two things. I want Raz to give like a slightly more technical side of this, but there's, I want to just double back on you. You had asked, well, why does this bring the industry forward 10 years? And I just, we started diving into the specifics. I want to be very clear there. Like ultimately Ethereum laid out a lean, lean Ethereum roadmap and they said 2035, they hope to hit sort of 10,000 transactions per second and like 10 megabytes per second, um, of throughput, right? And so, uh, that was like the goal for 2035, right? And so when you look at the Zero side, and again, the, the, the reason is all of the rest of the world, and that's one of the benefits of, of being able to start from scratch, is just go back to first principles. Like, the core insight was assume ZK is solved, which nobody thought at the time. Everyone's taking ZK and they're retrofitting it to existing systems. And how do I, how do I use this? Go back to the start of the world. ZK solved, what do you do? And what you do is ultimately this. And then once you do that, That's where— so now instead of, you know, now you're at 10 gigabytes per second and millions of transactions per second. So when you're talking about the performance trade-off, right, most blockchains, the bottleneck is twofold. There's two real constraints in a blockchain. One is that if you have a million nodes, every single node in the network has to download all of the transactions and run the same computation, right? So if your node is a MacBook, you can only do as much compute as that MacBook can do. In all 1 million MacBooks, download all of the transactions, and they all run the compute. So you have a huge inefficiency factor. You're doing this computation a million times, a million times more expensive to do a simple set of instructions. Um, and then the second thing is, once you start to scale up the transactions per second, you really get bottlenecked by bandwidth. And so, like, actually being able to download all of the data and transactions from a lot of, a lot of transactions per second, a lot of throughput becomes like you can't pipe it over the internet, right? You actually can't do it. And so this is like Solana is rolling their own fiber. You have DoubleZero, you have like all of these things to try to solve for that. So, uh, if you want maximum decentralization and you want all the nodes there, you have a huge replication factor and you're bound by, by the weakest link, right? So your, your smallest node in the network, if you want the person to be able to be homestaking, that is the most powerful the network will ever be because that person needs to be able to download everything around the computation. Or you scale up the transactions for a second, you get bottlenecked by the internet. So Zero does two things, and before we go into the breakthroughs, one is it still has this beacon chain. And so if you think about most blockchains, there's this chain that runs, and again, they all do the work and they all download all of the transactions. ZK is really— all ZK is for the most part is compression. That's like one of the largest unlocks. And so now you have the ability to say, if you have Ethereum and you have a bunch of validators on beacon chain, they're just fully saturated. They're doing as much as they can possibly do of data that's coming down from the EVM and just running these transactions. They're at absolute capacity. Now with ZK, you compress all of the compute, and with SVID and ZK, you compress all of the data. And so now you go from like the pipe being totally full to the pipe being like almost nothing, right? You can support so much more above, and that really is like the really big unlock. So what is the trade-off? The trade-off is you still have all these really light nodes. They can be even lighter now because they, they're not at capacity. They don't need to do this. So you, you can actually be much more decentralized. But if you were going to run just that chain and they were going to run like every blockchain, they would be constrained. You'd be constrained at 20, 50, some small number of transactions per second because these are like Raspberry Pis or little MacBooks or nothing. What Zero does, and that's why we describe as a heterogeneous architecture, is allow you to have really big nodes. We call them block producers. That can do a lot of the really heavy lifting, right? And so they can run— this is like the servers that run Nicey today. This is— anybody can run these systems or large or sequencer level. They take much more stake, but permissionless, anybody can run them and they can do all of the work and they can do millions of transactions per second. And then they just generate the proof and all the verifiers need in the network is a transaction commitment, you know, a slice of the data and the proof, right? And so then they can validate it and then they do everything just like it would be in a normal network. And so if you had no block producer, this would look like a normal blockchain in many ways in terms of the constraints. The whole point is you get to have these really big nodes, or like these two tiers of nodes. And Ethereum does this a little way, right? There's proposer-builder separation. The nodes who actually build the blocks and the searchers and everything, those are like way bigger and larger and much more computationally heavy than like the tiny nodes that run the network. This is just taken to the, the absolute extreme. And again, they had a vision for this kind of 10 years out But that vision is really only because you're retrofitting the system to about 10K transactions per second and about 10 megabytes per second. So this is just taking it to the absolute extreme, saying rebuild the whole thing from first principles. And this is, you know, this is what a system will look like 50 years from now. Let's solve the way to it.

So I'm curious on that. Like, as I was hearing you say that, someone might say, wait a minute, I've heard this idea before. Eigen, this idea of like having every node do the same work. Other people have tried to do it before. You know, the forcing full replication is like something a lot of people moved away a long time ago. Where do you diverge in like versus like someone else like Eigen DA? I think they use like a Razor code widget.

Everybody's just doing a piece of the stack, right? So everybody's trying to modularize and L2 just does like execution or Eigen's just doing DA. We built the whole solution, the whole stack that works. Together, right? As a single blockchain. It is what I think closer to what Lean Ethereum wants to be in their like 5 to 10 year roadmap. Obviously with the far less throughput and all of that, but that we built the whole system and we built it like Brian said, from first principles and built it at the question, you know, make questioning every decision we make is like, is this, does this need to be here? Does this not be here? And so we've made like, Even our validators, they're just verifying blocks, verifying the state transitions with the proof. They're not, like any of the logic for like how votes and all that stuff happens actually happens in its own zone. So we've made it so light that they don't do almost anything. And which allows us to have multiple zones that are like 2 million TPS if we want, many, which is crazy.

Just talk about though, like very specifically, like a system built on a different DA, right? You have a DA layer, you're rolling up to Celestia. Are you rolling up to like some other DA? Like, what is the difference between that and between Zero and getting a holistic system?

Yeah, so that modular style, right? You don't inherit, there's no secure, those validators of Eigen or Celestia don't secure anything, right? They're a place that you can point to and look at your data. Um, but you don't, it's not a holistic system, just like an L2 doesn't inherit the security of the layer 1. Uh, it's a sovereign blockchain that kind of posts its data there. Um, you know, the DA has no requirement to do, uh, to secure your data or do anything. Like, they, they are not, uh, going to be slashed by your network for doing something malicious to you. They're, they're, they're an external resource. Um, and so the big unlock with Zero is that everything is part of the same system. It's all owned by the same chain. There is no separate entity that's deploying a layer 2 or owning the DA or owning any part of the stack. The whole stack is one unified blockchain.

I want to go back to this TPS. You guys, I think, went out saying, hey, we can achieve 2 million TPS, which is massive. To put it in perspective, Ethereum does 20 to 30. You said 15 earlier. Solana does what, 3,000 TPS? What's the asterisk here? Like, is that like theoretical across all zones, per zone, under what conditions? Because I think, you know, I think crypto as a whole— not to discount that, of course— but I think it's, it's reasonably normal to assume, like, be skeptical when someone says everyone can manufacture TPS in a vacuum. It all comes with certain conditions. I'm curious what those are, if any, in your case.

Yeah, so there's two sides, right? There's like what can you actually scale to? Where does, where does the system break, um, cost overall? And then the other side is, is there actual demand, right? Being, being like scalable to 2 million TPS, uh, it doesn't mean you're gonna provision at 2 million TPS in terms of like Prover networks and all of these things that are doing it, because you're just like wasting a ton of resources if the network has 100K TPS, right? Uh, so the point is that the system itself can scale to many millions of transactions per second. We're like Ethereum layer 1, like, can never scale, right? Solana is not going to do— again, this is why they're, they're struggling at the bandwidth layer. There's just a bunch of constraints on the actual network. This network, the pipe, has become so large that you can easily, easily now have the capacity up to millions of transactions per second. But you have to actually be able to fill in the other side. And that's why we've talked a bunch about this. This is why we've aligned with a bunch of the partners we have. This is why we're doing the things that we're doing. Uh, it's not enough in 2026 to like just build the technology. Nobody cares about just like a, an incremental— and not that this is incremental at all, but there's so many been so many incremental layer ones, and it's just like, it doesn't matter. The demand is what matters. And the one of the driving forces for us is that, again, almost 15+ years now, right, 2011 forward, for me, for that entire time, crypto has been a pool of capital that sits over here, sometimes very big, you know, up to trillions of dollars. And the rest of the world has lived over here, and they have never intersected. You had exchanges bridges and a couple other things that allow people to go back and forth. But for the first time ever in the last kind of 18 months, you are having like a real, real convergence of these two systems. And what happened is we're very happy in this land to be at like 1,000 TPS, 2,000 TPS. Like, we've done it. Uh, the systems are cheap. You're paying a penny per transaction, or maybe 50 cents per transaction, or whatever the number is. Um, But now you have this like behemoth of like, guess what? Like the new— if Nicey comes on, like Nicey is like 2 million transactions per second, right? Like modern systems, it's just like a joke compared to what we have. And so what happened is everybody started making these like last mile trade-offs to try to get like— we waited our whole lives, we built the system, we've proven the value, and now the rest of the world has sort of been forced into, into acknowledging the power of like global permissionless systems, right? And like Tether is one of the largest holders of US debt in the world and, uh, the largest non-sovereign holder of gold in the world. And like, this is just from like permissionless access to these things. And so now when you have this convergence happening, everybody's now saying, oh well, just like, you know, they'll just have their own centralized blockchain or they'll do this. And there's all these really centralizing trade-offs that are made at the last mile that like really kill a bunch of the underlying value of the whole system itself. And so like you have to be able to solve for the demand side and you have to be able to say, how can you bring that entire existing world into an environment that is still like neutral and decentralized rail and can meet the demand? Because they're not going to wait for you for 30 years to catch up with the demand. They're just going to replumb the back end of their existing systems with some permissioned stuff and like, that's it. That's it. That's the end of it. So like, we, we should— we are so fortunate that we have, you know, $14 trillion BlackRock saying every day like that public chains, like there actually is real appetite for using public chains and for driving this. And they like, there is that everybody has seen the value of what the system can do. And we have a chance to do that, but we have to actually be able to like provide the supply side from the technology in a way that upholds the, you know, the principles of the space. That's the only thing that really makes it valuable.

Also, the asterisk that you wanted, there's no asterisk on the network, the proving. Proving is just a cost function. QMDB, there's no asterisk because QMDB's best case is its worst case. So they could always just do that. And the asterisk would be kid state contention. So imagine if the only app on the chain was after the same piece of state and everybody was doing transactions on that state, then yes, that is a, that is not a Web3 problem. That's not a blockchain problem. That's just like everything problem, uh, any, any type of systems problem. Um, so we are limited by that. So that is the asterisk, but imagine there are many applications and there's not, and the state contention is not all on one thing. Uh, then there is, if 2 million is, 2 million is kind of just where we stopped. That's what QMDB can do on a, on a, a modern SSD. Now, if we choose, moved QMDB to run on, uh, you know, a RAM drive. That the TPS could go even higher. It's just like, why at that point? So 2 million TPS is not even necessarily our ceiling, but we think it's high enough where we can talk to it.

So before we round out that, I want to go to kind of the institutional side of the house and demand side. I think about that constantly, right? And it's where I want to go next. Rounding out the technical architecture kind of chapter here, it's like, okay, like validators download a very small subset of the data, the block data. Are there any decentralization trade-offs? Can this realistically be run by anyone? Or eventually, in practice, does it centralize around well-resourced node operators? Or is that just not going to happen?

So we built it— that is a concern with every blockchain, including Ethereum, right? You have all these other things.

Vitalik says anyone can run a Raspberry Pi. But let's just assume that's not in practice. Most people are lazy and are not going to do it. And so you have—

it's also scary, right? Because you can get slashed. There's like a bunch of things that make it like, man, there's a lot of incentives not to do this and just put it in a pool.

Um, so many people were skeptical of Solana because like higher hardware requirements. I came to the realization of the big pivot saying, wait a minute, if I, to your point, are on demand, it's like, well, what's the ROI of someone that is provisioning resources in the same way that you have proof of work? Like you eventually do the calculus of I'm going to spend all this in compute and bandwidth. Well, am I going to get compensated for that? And the calculus for that was very much so, yes, for Solana, but you still have centralization.

Yeah.

Yeah.

So we try to attack that with the, like, how do we attack this problem? Right. And we see that Ethereum has it and everybody else has continued to have it. So our approach is we've built and are building a client that's easy to download and run without like a terminal. Like you don't need to be a, we want to make it as accessible as possible. So anybody who cares about decentralization wants to participate in the network can run it. We've removed slashing at the validator level, so only the block producer, but the block validators don't do it. So we want it to be easy, much easier to do and participate in the network. We wanted to just make it accessible in that way. So like, those are the approaches. Things like governance, right? So right now it's kind of opaque, from my perspective at least, like how are decisions made on these blockchains? Things seem to be kind of be done in a, in private chats or whatever. It's just not open. So we want to, we put this governance also part on-chain. So it's within the validator network that like within your app, you know, proposals to add new zones, add, change configurations, add, increase the TPS for a zone. All of that will pop up in your, in your application that's running the client, the validator, and you can participate in it. Or you can delegate your decision to somebody you think is an expert that represents your, your beliefs on what, where the change should go. And so it allows, we're trying to make it very easy and active to, to, to be an active participant if you care about it. Right now it's just scary. Like even like if you're not, if you don't have technical background and you don't, and you're not like really interested into this stuff and you're okay with like the risk, like if you lose internet at home, like if a snowstorm happens or something, you're just getting slashed. Like that's crazy. And that, that's definitely a barrier to entry and where it makes everybody wants to run it in the cloud and running it in the cloud puts you at risk.

US salvages and stuff like that.

Yeah, it's brutal.

So there's, there's two, two sides. One is extremely— anybody will be able to run a verifier. Like, the goal of the Beacon Chain, the goal of like validators, is to be hundreds of thousands to millions, right? As small as possible, super accessible, anybody can run. And then when you talk about— so you talk about Solana, like, big node requirements, much more limiting factor in terms of who can actually run them, who can afford to run them, etc. That will be 100%, um, the case for block producers, right? You're just going to need bigger servers, or so you're going to have a smaller set of block producers. But all the block producers do is boost, right? They are not— they cannot affect the security of the underlying chain. They cannot censor the beacon chain. They cannot do anything. Like, the beacon chain is the chain of the thing itself. These are just providing a boost to the network by creating this transactional throughput and generating the proof down and doing that, right? So you will have a smaller set of these big nodes. The whole point is, you know, imagine you can have these big nodes in Solana and then you can still have a million-node decentralized network that actually is like moving the state of the chain forward. And so that's the delta, is like, that is why we call it a heterogeneous network. That is the unlock, is that you can have a lot of the work put off, uh, but everybody, everything, just like you said, proof of work backs out to cost of energy. Everything is going to back— there has to be economic incentive because it is so cheap to run the chain because you can do millions of transactions per second at a fraction of the cost of you would do it with every node doing this. And in a traditional sort of full replication setup, um, the economics are just much easier. They make much more sense, uh, than they do in most structures.

Coinbase crypto-backed loans powered by Morpho empower users to borrow up to $5 million using Bitcoin or $1 $1 million using ETH as collateral. Loans originate in seconds with no additional credit checks and competitive rates typically between 4 to 8%. Use USDC for down payments, debt refinancing, unexpected bills, and more. Repayment anytime with no fixed deadlines. Over $1 billion in loans have been opened through Coinbase to date. With the Coinbase app, you can manage crypto-backed loans with ease. To learn more, click the link in the show notes. As always, investments in blockchain technology involve risk. Terms and conditions apply. The Blockworks Digital Asset Summit is back in New York, March 24th through 26th. We'll have top speakers from leading asset managers, financial institutions, DeFi protocols, crypto companies, and policymakers all under one roof. Think BlackRock, Coinbase, Robinhood, and more. Follow the link in today's show notes to buy tickets and make sure to use code EMPIRE200 for $200 off at checkout.

When I first heard it— and look, I don't pretend to be technical— but, um, when I first came to crypto, zero-knowledge proofs, even explained in sort of layman's terms like Zoco, I heard him describe it, and I was early investor in Zcash. It's very fascinating. Um, we're at a time in crypto where I guess a lot of people would say the— you've heard people say, look, the innovation between Solana and Ethereum was very much 0 to 1. And then Solana, you had other like parallelization and you had Zui, you know, a number of other different protocols with similar architectures and Monad. But you could argue it was very marginal relative to what perhaps Solana innovative versus Ethereum layer 1. I think if Vitalik were to start Ethereum today, he would very much use kind of similar to your architecture, like Solana's architecture or some variation thereof. But it's pretty interesting to see you guys say, wait a minute, let's just forget about what Solana and some of these other high-throughput blockchains say. We don't accept the premise that we've figured out and solved innovation on the X, you know, as an L1, and let's just build it with some open source stuff, right? I think you used like a couple of different systems. I think Jolt was, you know, I think that was the AC team was very involved in that. I want to make it more of a discussion, not so much on the technicality of these systems, but more on what does this mean for a space that obviously is open source and you guys are building something pretty unique here. Like, how do you think about the defensibility and moats of the system that you're building?

Yeah, I think there's two. I mean, I think one thing is we're, we're pretty fortunate in that, well, Jolt is definitely on the ZK piece for like the ZKVM. That, that is something that like originated Justin Thaler, Michael Zhu, like they, uh, were the origin of that work. And we've sort of taken that and together productionized it in full. And then we build out our own prover and again, moving to this terahertz proving cluster and like all of these things, right? So. That is one, but every other component of the stack— so QMDB, we invented, we originated, like we built from scratch. Faffo, same thing. SVID, same thing. And then the proving, the GPU prover on the Jolt side, Jolt Pro, 100% ourselves. So like, they're like part of the defensibility is in, is in the innovation itself, like, and just knowing that. But then also like the world is changing a little bit. It's very interesting. There's been a lot of, a lot of discussion. So like Optimism just had this happen yesterday or whatever, where Base, uh, basically like forked, uh, the OP Stack code and are like running it on their own thing. And there's a lot of conversation to be had about that. Um, it was the same thing with LayerZero ultimately, right? LayerZero came out, we had other people sort of copy the architecture. There is— it is very hard to retrofit an existing system to do this. This is why again ETH is like, in 10 years we'll be able to get like, uh, you 1/1000th and sort of 1/200th of the way there. Um, so it's very hard to retrofit an existing system. So we are first. We have in it, we have created from scratch 100% of the underlying infrastructure. And one thing that we have been very good on the Layer 0 side is like, you know, we have $100 billion plus of assets built on top of us, hundreds of billions of dollars moving across us. Like we understand how to bring real usage, how to bring something to market, how to actually I think people get really trapped in the noise. I talk about a lot of people get stuck in, in POC hell. People get stuck in wanting to have a, a big name that's not actually going to move any messages. I remember at the time that the huge— there was a huge governance forum in LayerZero for, for Uniswap, um, and we went through this battle with everybody publicly. Everyone wants Uniswap to be built on top of them. At the end of it, we realized like, that is going to send one message ever. It's literally going to be turn on the fee switch. That's it. That's a whole thing. Like, all of this, everything would have just been for like a single point of name brand, whereas like we go and we build out and like help build USDT Zero. And now you move $70 billion in, you know, 9 months and grow AUM by $10 billion and are driving like hundreds of millions of dollars to the bottom line. We build out stuff with Athena, we build out stuff with EtherFi, with Sync Pools, all these things. And like that drives like real value and real usage. And so I, I think people are going to be very surprised. Obviously, we stood on stage with the announced partners of, of, you know, uh, DTCC and the Intercontinental Exchange and Citadel and Tether and ARK and Google and Nexon and on and on, right? All of these groups across the board have already announced that GTE is coming and GTE is building there. Like, yeah, I think people are going to be really, really surprised of how focused we are on having like real meaningful use cases very early on in the life of it. And so it is again the technology is not enough. I think the moat is actually creating— like, ETH is just— ETH is amazing at the decentralization side. Again, there have been centralizing forces over time. I would say that has, has waned some, but in general, that is still the thing they have. But what they have is they're, they're early and they got a bunch of sticky stuff and it has like stayed there. And it's just very hard to replicate that for most new things unless you are net new. And like you said, Solana was like this almost zero-to-one moment, and they were able to bootstrap a bunch of really sticky stuff. Like, you have to be able to bootstrap something that cannot exist in the world, right? Like, at 1/10,000th of a penny per transaction, at something that can serve the institutionals where nobody else can serve them now for building some of this stuff, uh, you just have to focus on like, how do you build out those things and take them to market? Because nobody else can do it. And that's like the feedback from a lot of the institutions was We didn't think that this was possible and now it is. And that's like a huge unlock for them.

Um, yeah, I think what you talked about is as much as we have like a $2, $3 trillion asset class, you still have seen very little of TradFi or the real world come on and use decentralized, um, you know, public, public infrastructure, decentralized blockchains. I think it goes back to this idea that like just the scalability, the throughput was not there. And as much as Solana's, you know, you have some primitives and you have the tokenization, like Figure is an interesting example for me. And I think when I've heard you guys announce or disclose it to me, I was curious of like, from my side, I think about, okay, if I go and acquire a business or try to bring a business that we through Immersion want to bring on-chain, like what's the footprint of that business on-chain? And certainly like Citadel, and DTCC or ISAF, you know, that, that's a very different animal than like Figure bringing HELOCs on-chain. Different value creation, different, you know, both interesting in their own right. What would you say to someone that's listening to this and say, guys, like, I hear you talk about 2 million TPS, maybe the upper bound is 4 or 8, what does it actually matter? Like, in practice, like, are you just saying, look, I take the strong position that like we're replacing trad fi, everything's going to come on chain and we need this type of TPS. And, and, and we finally, when we went and talked to Citadel and showed them what is possible, they just finally are coming and saying, okay, yeah, we want to build it. We want to build a zone. Is that real? Like, or, or where do we see most of the builders come? Like, I guess the question is like, where do you see really the industry growing and what type of demand, right? Is it the Citadels and the DTCC? Sees, or is it just more figure type of implementations that are more lightweight? Perhaps could go to Solana or somewhere else, you know, it's totally fine.

Yep. Yeah. So I think there's like 3 main portions. So one, like, I, I very strongly, and I, I think it's very hard to disagree with, uh, this at this point, like, markets are moving from 7/5 to 24/7. Like, that is happening broadly, right? And it has been proven outright that like permissionless global access to markets is just an incredibly powerful force. Tether, Tether and Hyperliquid are literally like the two most profitable per employee or per headcount companies on the planet, right? And they are providing permissionless access to dollars and permissionless access to like trading through a bunch of different instruments, right? And so like, this is happening. No, like, there it is too large to ignore when they're like creating more profit than Goldman Sachs and from a bunch of other, you know, a bunch of other organizations, uh, that, that is happening. And like, I will come and I will take that stance. And so really the question is like, there's only a couple of reasons that institutions care right now, still today. And there's, there's really just a couple. One is fear of being disrupted. And I think fear of being disrupted is like a very powerful motivator, uh, and it's not always like a pure fear. Sometimes Um, you know, Jobs' maxim is like, you, you have to be willing to like disrupt yourself or somebody else is going to do it. Like just seeing that there is something that could be a disrupting force in either one, wanting to get ahead of it, or two, realizing that in a changing landscape, maybe you can go from 6th to 1st or 2nd, right? Like that you can actually capture a bunch more, uh, in this like changing world order regime, regime, et cetera. So that's one is fear, fear of being disrupted. Second is just pure P&L, and I would say every asset issuer in the world views it purely through this. So this is, I have some asset, um, you know, high-yield, um, fund from Apollo, whatever this is. And like crypto as a whole, just as a pool of capital, is bigger than like all of high-yield credit. I think high-yield is like, high-yield credit's like $1.5 trillion or something. Like crypto is bigger than that in general. Like to be able to launch a Biddle or one of these things and have like billions to tens of billions of dollars of inflow is like extremely attractive, right? So asset issuers just want AUM. They're AUM-driven beasts. That's all they care is just P&L, and they view it largely as distribution. And then you now have the banks behind them, so the merchant banks and all these other banks who do services for these groups, who say, hey, I want to be able to provide our asset issuers, our customers, a way to issue their asset. They want to have it out there. They want to chase the AUM. We need to create a structure to do that. And so you have A lot of these huge institutions are saying, we must have a strategy for, for this, for giving distribution to our customers, because we risk losing our customers, right? Uh, and then I think the third is this sort of like appetite of, we want to own what the next chapter is. So if the world is moving to global permissionless markets in 24/7 markets, and they're trading across all asset classes and all of that, I want to be like front and center in owning that. So there's sort of a race of who is going to be the one who tokenizes all of the equities and provides trading rails for that, who is going to be the one who establishes the end platform for where all of these things trade and what it looks like. Stablecoins have been in this, you know, death fight for a very long time. There's just like the incentive is far, far too massive, and like that's not lost on organizations. Uh, you know, Tether making $10 to $15 billion a year, he's just like not going to be ignored. And then seeing what like the rest of the world's markets actually look like. Um, and so I view those as the three major drivers. All of the institutions that we talk to are driven for the most part by one of those three things. Uh, but I, I think the general thing that I was told forever is like institutions like this are— they move so slow. It's like multi- multi-year sales cycle. Like, everyone is so slow, it's so hard to do anything. And like, we started talking to groups not that long ago. Like, people moved heaven and earth to be like involved here because they are driven by one of these forces and are moving way faster than people think. And like, that's not to say immediately on day one you're going to see 100% of all of these businesses on-chain, but like, they're, they are like very very actively pursuing out of self-interest what that looks like.

Yeah, I mean, look, businesses are rational economic beings, and Citadel is certainly one of the smartest guys in the room. And so they see an opportunity to run markets 24/7, 365. I mean, I just saw— no, they paid like a billion dollars to get retail flow. Like, they'll move if they— the money is there. If they can all of a sudden— imagine what a market 24/7, 365 does. Does. Like, like, yeah, certainly there's like after hours and there are Asian markets, but like increasing 2 days of the week is not trivial. I mean, shit, there's a lot of value that I think the human brain can't comprehend, like exponential concepts. This is very much one of them. And it's like, you know, you basically expand— what is that, you know, 30%? You could theoretically have 30% just more active days. In the year trading days. And then add on top of that the assets that are coming, come on and the composability of that and this new primitives, you know, just let alone like that is a huge component.

But now you're talking about like we just barely got T+2 to T+1 and that was like a massive, like imagine going from like T+1 to like real-time net settlement, right? Like people have no concept as like what that actually means and what that unlocks from like a capital perspective for every single instance, you know, for all of these things and knock-on effects of some of them are just unbelievably large and people just like don't understand the real, again, rational— this is not somebody— nobody is benevolent. They're not doing it because they love permissionless systems and they're like decentralization max. There are like people in these orgs who actually care and like do care about decentralized systems and like actually care about that, but they are purely in general acting as out of self-interest, something that's better for their business, something that's better for their customers. And you just like, you cannot ignore the effects of what that will actually do.

Yeah, I want to make it as— I want to go to like, I think, I think you said, look, we're not going to launch until later this year in the fall. Um, um, I'm curious, like, take us into the future. But before we go there, I mean, I do want to underscore, like, when I made an investment in you guys, When I— again, I'm not technical enough, but I do make bets on teams. And I think you guys, in my estimation, have grown to like, what, a 150, 160-plus team. What was interesting that I've heard from you, Max, and others in the team is like, you kind of forced everyone to be in the office, like in Vancouver. So yeah, I'm sorry about— for OPSEC reasons, but in Canada. And now I think you have a satellite office in in New York and I was talking to Michelle and the event and stuff, but I want to like make it be, take a breather on the technical and like on Xero and just kind of focus on that because I think about it constantly at building Inversion, but also I think it was very opinionated. And, and from what I've heard in engineers is, A, you have very little turnover. Um, I think you hire exceptionally well and you hire your acceptance rate similar to RAP is very, very Wow. So I've relied on you guys from like a, as I want to bring in a few engineers, but yeah, well, like, what, walk us through that. Like, uh, I, I'd be very curious to think of other folks would benefit from hearing like how opinionated you guys have been on that front.

Yeah, I, I will, I will say Raz, Raz has been the driver of, of a lot of that, which is like amazing. It's been, uh, he's been the strong force, but, uh, we started, we were 100% remote when we were early on. We were, you know, 7 people and we went and we sort of, all lived in the Bahamas for 5 months and we grew from 7 to 25 together and we shipped more in that couple of months than we ever could have done. And when we left that group, these are 25 people from all over the world, everywhere. And we said, we can't lose this. That matters more than anything. Every single person, either you move to Vancouver or like you won't be here anymore. And this was like, we just said like, this is what we're going to do. And like you said, extremely opinionated stance.

And all of them thought they could go back to being remote too.

Yeah, we had, we had, we had hired them and said we're remote, a remote company. And so that was like, they're in— every single one of them packed up their lives like nobody left, right? Everybody came. And so then it started to— I actually love it now because it is just an embedded litmus test, right? Like 85+ percent of the company is not from here. Most of them are from the States or from Europe. They're from other places. 85% of the company has packed up their lives, moved here, and chosen to do this. And like that level of conviction early on, like they are bought in to the mission. They're, they are totally restructuring their life to come and work on this one thing. And like when people get here and the culture is intense and like all of those things, but like they have chosen, consciously chosen to like take that leap. And I think the type of people that do that are like the type of people that we want here. And so yes, of course, I think that means one, we've been able to— much harder to hire in general, so much easier to hire senior talent X and Y and Z. But once you find somebody who is willing to do that, like, this is really, really something as we care about, they care about, and like That just does an unbelievable amount for the organization as a whole.

Raz, I want to go over to you because I think you guys have managed to hire some of the leading ZK engineers, especially for this endeavor. Uh, like walk us through that. I mean, like, I think you didn't announce who they were or they sort of just for a variety of reasons, but what was the ultimate driving factor for them to come join you guys when they could be doing a million things?

I was told by people who recommended us these people is you'll never get them. But I guess what was, but the real thing that drives people with pitch, like the pitch is we're trying to change the world. We're trying to build this new internet. We're trying to build something decentralized, permissionless, censorship resistant, that view, that pitch of like what this looks like. And we have a real path to do that, right? Like the original goal of Ethereum. But like that they weren't headed, they weren't going to get there fast enough and like centralized chains were going to win and that this was the chance, this was the shot to do something meaningful. That vision and like being able to pitch that to them and then see, right? Because they heard from people they knew like that we have the fucking amazing engineering team. We have the part of it is like we have this crazy culture of we spent a lot of time like myself and Isaac, our chief architect, spent a lot of time Um, going through this interview process of how we ask, how we figure out who is a good engineer. Like we, we engineered our engineering process over months so that we know, like based on how long it takes you to answer this question, whether you're an A, B, or C. Uh, and we filter out engineers so fast, uh, early, cuz we're so good at that at this point that, uh, by the time you hit to the third or fourth round, you are exceptional. Uh, and it's just about how do we convince you to move your life, uh, to come here. and we'll, we'll bring them here and show them what it's like. But like, people who are exceptional and like great at something want to be around other people who are great at something, right? You want to— you don't want to be the guy carrying everything. You want to be like, um, surrounded by people who are like, who are better than you. And that's what you hear constantly is like, we bring in A's and they're like, oh my God, I'm a B here. And they're, they don't know that they love it. But, uh, it is, it is a shock for them. And then they tell their other A friends like, you got to come here. Like, this is like lightning in a bottle.

Um, and so yeah, it's been like referrals internally of like engineers saying, hey, I'm bringing in people I worked with. Because I think, I think that's good hire, right?

Yeah.

So every A you get, they bring in like a few other A's and it kind of just goes out, right? Uh, not all A's know each other. A lot of them are like lone wolves too, where they're the, because the thing about it, most companies are built around like a few A's and then B's that surround them. The A's are carrying it, right?

So they don't know too many other A's, but the ones that do would bring And one thing that's really helpful too is we, we, we've gotten the chance to work with a bunch of like really, really brilliant academics and people on the other side. And like, they have gotten so bought in and have like formed this relationship over time where like every incredible, like one out of X PhD who's coming out of their labs or wrapping up their like almost all of them come and talk to us very early on in that lifecycle. That's because we've pulled people out of there. I think our research team, uh, at one point I think we were like 13 people with 14 PhDs, you know, it was just like we were, we were just pulling in all of these brilliant people from all of these different labs, from Stanford, from, from all over the world. Uh, and I think that too, while some of them like didn't know each other directly, like they were bound by like the leading expert in this thing is saying like, this is the group that you like. If this is something you care, if you care about ZK, if you care about this, like this is the group you want to go sort of work with. And that has been a really strong driving factor as well as like having that stamp of approval from these people who sort of like everybody are looking up to as they navigate that.

Yeah, it's interesting that I heard it from when I was like looking at a layer like Zcash early on, it's like there's like a handful of cryptographers and everyone knows who they are, but like they really understand this tech. Like you can count them with two hands at most. They all kind of operated in their own silo. They're at Stanford and then they're in Germany. Like is what you've done just kind of centralize that and put everyone in one room and align towards this objective. To your point, everyone in person or just collaborating more closely. And I guess someone could say there's been plenty of examples in crypto where people raise a bunch of money and just run like a research lab and they don't really put stuff in a production. I mean, God knows the Ethereum Foundation spends a lot of money, but for a variety of reasons hasn't really— it's just constrained, right? Because they're first, and I guess if they were to rebuild it from scratch, you guys just found yourself in a position to re-architect it from first principles and use a lot of stuff out there.

We had an objective. We knew what we were going to build.

Yeah.

Yeah. Yeah.

It's not enough to just, it's not enough to just be a researcher here. You have to be both researcher and engineer. Like there is no pure research position. Every single one of our incredible, like incredible researchers are writing, like are writing code. Everybody must be both. It is only people at the intersection of the two. Um, so yeah, I mean, very like very opinionated, very, very focused on what we're looking for on that side. Um, but yes, I, I said the other day, actually one of the ETH-ZK researchers reached out like, hey, in this podcast you said you, uh, you know, you've assembled the best ZK team in the world. I was like, yeah, like, I, I am extremely, extremely confident in that statement, and like, I will plant the flag and I will say that. And I said, but does it matter? Like, it's on us to prove that statement. And we have until the fall, and we're going to publish everything along the way. And like, I, I, there is zero doubt in my mind that this is the single best collection of ZK talent on, on the planet. Um, and that's like a very bold statement, and I rarely make like pretty absolutes. Extremely confident in that statement.

Uh, as an aside, I want to get back to just kind of the, the later this year, what happens and tying it all together. I'm curious, uh, AI, like how are you internalizing AI in the organization? How are your engineers like using it? Like does that influence who you hire and for what? Um, but walk us through that. It's something I think about constantly and I think others as well.

So different parts of the organization use it more frequently than others. We're, uh, we're not where I want to be, but we will be. in terms of how we're using it. Um, but, uh, it is amplified, especially our best engineers and our best people. They become, they were 10x and they had another 10x there where the things that they finish in a day, it's like, what the fuck? Uh, but, um, besides that, uh, it's trying to have that trickle down to every single engineer. Um, I think some people in general, and generally, um, veteran engineers are like, like, you know, I'm, uh, I know better and I don't need to depend on it. Uh, but I think that's just not the— that's not where things are going. Like, everybody needs to be using it, and they need to be more of a conductor than, uh, like the person who's writing everything. Um, and being better at prompting and, and having multiple agents working on things and reviewing. And like, not just because like you could have AI do everything for you and they'll just create junk. You actually have to verify that that stuff is incredibly good and, and tweak it and give it feedback and iterate with it to make great work.

Yeah, I mean, obviously Reza and I, we spent a bunch of our prior life in AI on like counterfactual optimization, published with Noam Brown, did all of this. So it's very interesting to see the application side. My intuition was that it would raise the floor across the organization and you'd say raise the floor 50% or whatever and every engineer will get better. I think it has raised the floor some. But like Raz says, it is actually the ceiling on the absolute best engineers, like night and day difference. Like, uh, I remember when Isaac was really diving into ZK and Isaac is like, he's brilliant. He's unbelievable. He's incredible at a bunch of different fields, but like really new to ZK. He said, listen, as I'm diving in like early side of Jolt, like this would take me 3 years working side by side with Justin Taylor every single day to be able to do this. And in 3 months, like just with relentless AI and sort of how he is diving into this, he then found like 6 errors in a 90-page math theory. You know, like he, like he, he's the speed that he is able to grow and the speed that a bunch of our best engineers have been able to grow, uh, is just like night and day. We had incredible, incredible researchers who are churning out progress on the GPU side or on other side on like actual in production that like never would have happened before. So again, I think the best, it has elevated a ton and we've seen a huge, huge progress. But across the org, it is something you just have to, is changing constantly and you have to like get as much as humanly, humanly possible out of it. So we always, we always want to do more.

Yeah, I definitely agree with that. I was thinking about that. Like you've always seen like the 10x engineer, like it's even Steve Jobs to say like, The difference between a, a B engineer and an A engineer is like definitely not linear. Like you get like, you know, a 10x output. I think with AI, you have even more discrepancy where the 10x engineer is doing 100x output, but the 1x, 1.5 engineer maybe does 5. Uh, and it really comes down to like, do you have field knowledge? Do you have like the ability to architect and think in the system? And like, I think what you guys, from what I'm hearing, just everything I've heard is You kind of envision the end state and work backwards and have used not just your team, but tools like AI to just really drive towards that, um, which is really impressive. Uh, would you say, like, did in, in this building of Xero, if you were operating a couple years back, I guess, without AI, like, do you think you would have been able to get to where you are in the speed and the quality, or Did AI not come into—

who's the most helpful in the ZK ramp up? So like studying and becoming like really good at ZK ourselves so we can evaluate all the ZKVMs and come to the conclusion that Jolt was the best and become really great at Jolt. Um, it was, that was like the supercharging part. Um, but it didn't like, it wasn't really like impactful on QDB or any of the other stuff like that stuff was happening kind of pre all that. It was really just probably the last 6 months. It probably had a big impact on how fast we can get the GPU tuned and all that other stuff.

Maybe when I think I heard you guys pitch it to me once, it was like, do you think that you guys were able to see the future more than others? Because you come at it from, to your point, Brian, you said earlier, which is from your standpoint, building Layer 0 as this cross-chain communication layer, you kind of saw the limitations of some of the other chains and came to the conclusion, okay, if we were to build one, A, we have to build it. We understand who our customer is, and this is how we would build it. Having— because again, the reason why I went to you first when I thought about launching my own is like, from your vantage point, you actually have the most amount of insight into this disparate set of systems that are increasingly more connected, but you understand where it breaks, right?

Yeah, I mean, I think it's made a— again, huge— you talk about like right to win and what that looks like. A huge amount of early insights of why, why are these systems like, where do they break down? What are the trade-offs between this approach to sharding versus this approach to sharding? What is the breakout, uh, like the breakage in this scaling system? Like Aptos sharding does a million TPS with a Move, uh, VM. Um, so like rip out the EVM overhead, we hit a million TPS. But when you break it down, it was still like like a bunch of nodes that were really large supercomputers that were co-located in like a single data center. So like, why? Like, why, why can that not be— why was that only hitting like 33,000 transactions per second per node? It's like, oh, it's actually like the physical storage layer of that. And so like every one of these things, you start to see like, what are the constraints of the system? Where are they making trade-offs? And where do they break as they're scaling along one of these vectors? And again, one of the first realizations outside of the core insight of like ZK being the unlock for like the completely new architecture was like, hey, it is actually like the physical storage layer. And then you start to rethink that layer and it's like, okay, well, actually, like, why do you need tries? Like, why, why is that the thing, right? Like, why does— why do we need MPT? Why do we need to base this on, on, you know, hardware that was like 15 years ago when everybody has SSDs now? And like, then you get to like, of course it's a log-based database and like, of course this like solves all of the things. And so So part of it is seeing what is the constraining factor of many other systems. And then part of it is just being willing to say, I think people are so— I mean, I love, again, I've spent 15 years of my life here. So it's like hard to neg this space a little bit, but for the most part, like a lot of the people here for a very long time were sort of like the rejects or like people who like didn't make it in other, you know, it wasn't like the best and brightest in the entire world. Maybe it was like the most principled or the people who cared a lot about this thing, but it wasn't like the best engineers on the planet from every other discipline because crypto has always had this like sort of bad aura over it around different ways. It's not always been super cool to work on unless you were like chasing money in some way. And so I think there was just like a huge amount of hero worship on just the way that these things were done, like must be the right way, or like, of course, because like this thing has MPT, we must have MPT, even though there trying to transition to Verkle and then seeing now new approaches like QMDB. Like there's just so many decisions that people were not willing to rethink either because rebuilding the whole thing was too difficult or because like people don't get past that, like that, that is the way it's done. And so that's like a fixed requirement. And I think like questioning the requirements matters a lot culturally.

It's from Brian and I from the beginning have always just, we don't trust experts and we challenge each other on everything. So we're always just like. Why? Why, why, why is that the right approach? Um, which has served us well in the company.

Why? Just basically on end. Yeah.

That's what we do to each other. And we just argue and argue and get to ground truth. Uh, and like, that's part of QMDV gets solved, right? It's like, why did we need MVPT? Is there a vibe? Is this like a vibe property that you have to have it? The, uh, the order matter or anything like that? Like, uh, we start going through all of these questions and realize, oh, There's no reason for MPT. Like, that's, that is just because they did it before and everybody just thinks that's how you're supposed to do it in the future. Um, but we came to that every, every step of the way is why, why is that? Why is that the approach? Can we do something different?

Yeah, I think it's important to put it in context. Like, I feel like it's important. Like, I always tell my team, look, if you're going to have a strong opinion, fine, just document it. Like, and be willing to change your mind, right? Because I think a lot of times in the crypto, like, you've seen this, like, tribalism, this, diehard loyalty because you made money with a particular token. And I think it just is very— it's very weird to see someone come into the space with intellectual curiosity and that curve just degrades over time for a variety of reasons. Where if you truly understand the space, we're still very, very early. So you— it's sort of like the worst assumption you can make is that we've solved it all or that you've like stayed constant to like a particular set of parameters because they've worked to some extent. But the reality is, to your point, Brian, like, we haven't really— the work hasn't really begun yet. Like, we haven't really brought on the real institutions or the real workflows on-chain. Um, so anyways, pretty amazing. I think, like, just to kind of wrap up the, the— towards, like, closing out the discussion, I do want to bring it back to— I think you mentioned launching 3 zones, right? And one of them was an EVM, the other was privacy payments, And then the last one was trading. Kind of walk us like high level, like these three interesting on the privacy side. My first, I just go, how is this compliant? Because we've seen iterations of this not be compliant. And, you know, so, so anyways, if you could just spend a moment talking about these three zones and what that looks like.

Yeah. Yeah. I think there's— we just view crypto really a bunch of different things you can do. Uh, but the 3 things that have product-market fit more than anything else is like general purpose smart contract zone. You want people to be able to come and deploy, but the things that we think the chain itself, Xero, needs to like have an opinion and internalize and like build the underlying infrastructure for is global markets. Again, it is just this very, very strong thesis that the world moves from 7-5 to 24-7, moves from local markets to global markets. Has permissionless access, trades across many different asset classes, all of those things. And then payments, obviously payments. We did $138 billion last year across LayerZero. You know, at one point, I think in November, we did $38 billion, which was like 5.5 times Western Union, 2.5 times Wise, like what, literally moving more assets than some of the largest payments companies on the entire planet. And so like that is just something that we, we know extremely well. We have the partners around, like we have these things, and it's just like, how do you— if you believe— and I, I think like you talked about opinion, I think it's very important to be opinionated. I think there's like only a couple of things that we are super opinionated on. We are super opinionated on the fact that markets are moving in this direction, like that shift is happening. It is— there is a move from traditional markets to these global permissionless markets. Payments, again, you need these, especially now. Like, this has been so hyper-accelerated from, like, Agentix stuff. Like, the difference between paying $0.30 per transaction and, like, 1/10,000th of a penny per transaction is, like, a very, very— you know, you're talking, like, many, many orders of magnitude and what that can actually, like, look like. And even that, when we talk to some of our, like, large payment partners, they're like, listen, long-term, we don't even think, like, doesn't matter, 2 million TPS, 10 million TPS. Like, it won't actually matter. Like, you have to actually, like, side channel a bunch of the— there's like a bunch of optimizations you even need to do past that to present like the long tail scale that, that, that is going to require. So like our, our stake in the ground is that you must— like the world is changing in this way and we are going to build for that. The things that are invariants for us is like permissionless, decentralized, censorship resistance. Like care very deeply, and I will come across and all of the code and all of the way that the systems are built. So those are like the invariants for how you build. And then you talk about the privacy side, like, yeah, it is a sort of like you need in order to bring any of this institutional stuff on chain, you need to be able to have some form of privacy. Like you must. It is, it is a like hard requirement for a bunch of this stuff, whether it's corporate payments, well, it's anything. And obviously we believe from like a fundamental human right, like people should have a right to privacy anyway. So it is like tricky and nuanced into like what the end state of that looks like. But there is like both a clear requirement from needing to bring the markets over and then also just like a clear imperative broadly. And Balaji talks about this a lot is like ZK being the, the counter to all things AI and sort of like what AI exposes, like ZK gives you back from a bunch of these things. And so I think that's, that is a interesting decision point that we have not talked a lot about publicly yet, that we certainly will closer to launch. But, but there is a, you know, is a whole very nuanced conversation around that side.

Yeah, I would argue that's whether people are public or not, but I think that's one of the reasons that is holding back a lot of institutions for coming on chains. Like, hey, I don't want my counter— my competitor to look at. And, and it's, um, it's a real thing. And I think it was sort of— it's weird that we've come to a state of the world where privacy requiring privacy is seen as a red flag of doing something nefarious or obfuscating when the reality is privacy is just a right to be private and be left in peace. And the beauty is the technology allows it to be, have a semblance of privacy where you can be compliant. I think other iterations of it didn't require it. And there's like little issues. Like Monero was like a, a laundry mat that you put all the coins and it would just like shake it enough so that you could bring— take out the coins back. And but there's ways that you could kind of— it wasn't truly private. And then you had Zcash in the early days where you'd say, hey, we're gonna— you could shield your transaction, but it wasn't by default. And anything— whenever something's not by default, then, you know, if only 2% of the network shields, and you can always kind of bring it back. And anytime you shielded something, you'd get a call from the IRS or the government. Um, but I think— and then there's Tornado and other instances, like— but I do think that one of the things that most people don't appreciate is that you can have privacy and you can be regulatorily compliant. Like, we've had that for a while now. Like, you could, you could go back to Coinbase and say, hey, here's actually— like, you can, you can open up the door if, if it's required by regular regulators or just to be compliant, but you could be private by default. And, uh, I think that's one of the areas that excites me the most, candidly, just having like a payment, a truly private payments network that is high throughput. Um, you know, it doesn't have a lot of the bottlenecks or limitations that historical chains of privacy-focused chains have had. But it's good, it's exciting to see you guys do that. Hopefully maybe at a later time we can talk more about that as it gets closer to launch.

Definitely.

Um, what else, guys? I mean, I think we've covered a lot of ground, hopefully useful for just people that are less technical. But I do, I do think we've done a good job just going through some of the breakthroughs here that are important to recognize. Uh, I think the fact that Citadel, some of these larger institutions, are building zones is very instructive. And I don't think this is like death by a thousand pilots. I think it's very real. What is, uh, walk us through like what are the next 12, 18 months look like from here? You've announced Xero last week with very good partners. What does the future look like here?

Yeah, I mean, I think the last 10 days— how many days have been? 9 days. Like, I've spoken to many of the, you know, we were even internally 24 people out of, you know, 165 were even aware of this until like right before the launch. So just like extreme stealth. There's probably 8 or 9 external conversations that we had. And so post that, I've had conversations with like a huge number of the largest institutions on the planet. We've had conversations with like a huge number of the largest sort of like current crypto DeFi protocols, asset issuers, etc., right? So interest has been like incredible. Our only focus right now is launch. Like again, there is— there is technology alone doesn't matter. You have to be able to do both things and you have to be able to pull this all together. And you have to be able to like generate real economic value and utility. Like that is the only thing that matters. And so our only focus is how do we build the systems in these zones to the spec that sort of allows for that? Like how do we actually get in all of the utility and all of the value to make it maximally useful for the broadest number of use cases? and to be able to deliver value. Same thing we've done with LayerZero, same thing we've done in the past. And so like, we're just laser, laser, laser focused when it comes to the Zero side. Obviously on LayerZero, we're still doing all the things we've been doing, but just laser focused on that. And so, uh, you know, what you should expect to see from us is, is for a launch to not just be like, here's a chain, hope people use it, right? It is, it is like real things, partners, usage, like all of it. Like, we're very focused on solving both sides of the equation.

Yeah, that's good to see. If, for instance, someone like us, you know, we're focused on the other side of the demand, right? Like bringing— making businesses more efficient with this tech. We, we've talked a lot about, like, for me, it's like, okay, I've historically thought that I could buy this type of business because that's sort of the limitation, which is, which is still like a massive universe of companies that, similar to Figure, you could bring on board on-chain with the existing tech and feel pretty good over the next 5 years that, you know, it can absorb that. And you and I have talked about that at length of like, okay, there's a lot of wood to chop even with existing tech. But then it begs the question, okay, what else could we be buying? Because these zones allow for just something quite more powerful. How do you, as a builder, like, what does this mean for other chains? Like, you know, of course you built the messaging protocol for other chains, But do you take the position that, look, this is the dominant architecture, like everything eventually just doesn't get as much scale and therefore everyone just should build zones? Or is there still a place for other chains to exist that are application specific or just something else?

Raz, you want to go?

Sure. I mean, Without crushing your— I think, no, yeah, I'm not trying to crush them. I think like long-term, this is the approach that everything will kind of go towards. Like there's just no reason to have replication of this compute everywhere. I think that a lot of the chains will probably have to wait to see the ZK mature, right? We're getting our stuff formally verified. They'll need to see that stuff be in the wild long enough because they have a lot riding on their current approach. But I think like this is like we say, it's the last blockchain. It's the last blockchain architecture for sure. This is where things are going. Going. Like, how do you compete on price when you're spending 10,000 times more on compute than we are? Just because you're duplicating it, right? It doesn't make any sense. And so I think like they'll go this way.

I understand that. So I think about it like the printing press, like before the printing press, you had a block and then you had to copy that. Like the assembly line was super inefficient. Then you had the printing press. And then so the proliferation of books just kind of did that.

Yeah.

Anyways, when I hear you say that, that's how I internalize it in my lizard brain. It's like, okay, you basically—

Good analogy. Yeah.

Yeah.

Uh, you can't reason by analogy, but you know, the human brain needs to comprehend it through analogies.

Sure. Uh, but for sure. And so we think this is where things will go. I mean, Ethereum's, uh, like we came up with this and started working on it 2 and a half years ago thinking that The reason why we did it is we thought no one was going to build this and we wanted it to exist. We wanted, you know, something to compete with the high, the fast chains, but be as decentralized as Ethereum. But Ethereum is since kind of their lean roadmap that's over the next 5 to 10 years is headed in this direction. So I think like you're seeing them go there. I think others will slowly start to follow. This is the way things will go. And like in terms of like unlock, one of the analogies that Isaac just used is he says, currently we have like edge network, like for like phones, right? And like what we're bringing is 5G and like the things that you could build on 5G versus edge, right? You couldn't like back then you wouldn't even thought of, right? And so if we just put out 5G phones when you already had edge, you'd be like, we, I can't tell you what's going to be there, but we've unlocked something that couldn't have been done before, right? You could like viably put things that aren't finance related, things that you want to be decentralized. You could put a new DNS there. A new internet. Like, the whole internet could run off of something that's decentralized and not controlled by, uh, the state or some actor, right? A single actor. Um, and that's a future we want to build.

If you're, if you're an investor listening to this, I think where I've, um, you know, I think it serves you well to update your priors. Um, if a category in crypto has not worked before, when you look at something like zero immediately your brain should be like, okay, what has failed before with the current architecture and infrastructure? And, and why did that fail? And if it was a dependency or a technological constraint, to your point, if you were operating on edge, well, maybe Uber would have gone off the ground.

Mm-hmm.

YouTube wouldn't have gone off the ground. And because you have 3G, or in this case, 5G, to your point, then it serves to revisit, you know, a bunch of categories in crypto that have just not been able to get off the ground for that particular constraint. Um, and so I think if you're, if you're listening to this and thinking, okay, I think that's where your brain should be going. It's like, what are some of those use cases that just haven't worked? Because to be an example, like prediction markets kind of work, uh, like Augur, for instance, or some of the other prediction markets like Gnosis.

They do—

I was an investor in them. I don't think they work because you didn't have a stablecoin to settle. And if you don't have a, if you have a volatile instrument like ETH, it just, it just degrades the user experience. Right. But if you have a stablecoin and you have a more performant, you know, chain, then yeah, you know, you could, you could see how, you know, Polymarket would work under that environment. Do you guys have an opinion on beyond like these 3 zones of the type of maybe financial workflows or type of market that really gets off the ground when you go to this like 24/7, 365? Like, is it just what exists in Wall Street, like, on steroids? Or are there new primitives, markets, or type of instruments that get off the ground, uh, that, like, more of the— what Chris Dixon talks about, like, skeuomorphic. Okay, like, we understand that, but, like, what are the non-skeuomorphic things that I think you will— yeah, we'll be able to see in Zero? Just kind of, if you guys have an opinion on that.

Yeah, I mean, I think it's— I think there's a bunch of— there's a bunch of things, right? And so, uh, I mean, when you talk about— you talk about prediction markets, there's just the, you know, the article that the Fed wrote about like cow sheep predictions being like lower margin of error than, you know, like the Fed prediction thing. There's all these crazy things of just like information sets that basically get leaked through these alternative markets that get built around it, right? And some of that is like access, some of that is ability to like create these markets, uh, anywhere, uh, and some of that is like viability of actually having them. And so I think, I think you're just going to see like payments is just a natural— I mean, payments has been talked about like ad nauseam everywhere in terms of like what that looks like. And there's all of these pieces to solve right there. There are no really great— there's no really great like FX settlement on-chain. There's no really great like way to actually go from stablecoin to merchant directly and deal with settlement windows. There's like all of these things that are happening that people are trying to solve. So a bunch of stuff again, AgentX side, like let's take that out of it because the area of what that could look like is just like an insanity at the rate that things are moving in terms of like what that actually looks like around. But I think you're talking about ideally every market that involves anything that can live on chain. And so again, I think, I think metals have been really like the 0 to 1, uh, of, of metals on-chain over the last little bit, of like Tether going from like nothing to like the largest non-sovereign holder of gold in the world, from Hyperliquid doing like many billions dollars a day of silver, you know, all of these things like happen like that, right? And so you're talking about like the more that these markets are open and the more that there is like a convergence of how these things trade, like capital just wants to be there's like a couple of sides. You want to be able to be, you know, whether that's protected in a way and in like the way that markets are structured now with Prime Brokers or like whatever that might look like. But it's also like you, you don't— I don't want to— nobody wants to have to like deposit to somehow move money into here to be able to trade this market and then go over there to be able to trade that market and then like withdraw it to a bank and send a wire to trade this market over there. Right. So like the more of a convergence that you have on these things, I think is like a massive compounding knock-on effect of just like what can actually be served because liquidity is just like there and available and it is like much easier to take these incremental positions in a way that just like makes zero sense to do when they're really, really disparate by platform.

Yeah, I think we're going to be sitting here maybe 5 or 10 years' time and said, Gosh, yeah, it was pretty obvious. Like the financial system was like predates the internet and all these current workflows, T+3, T+5, all the friction that exists in the system, A, was not built for the world of AI. And if you think that, you know, Agentic Commerce or whatever, like, you know, AI can't interact as easily in the current financial system. And so it's like sort of the direction of travel is very clear. The nuance is like, how do you actually implement it? Right. And the thing that I have always felt when people talk about like, for instance, real world assets, it's like, okay, well, yeah, like you're bringing a real instrument on chain. It's not like digitally first. And, and that I think is where friction continues to exist, will probably continue to exist. Like you have to trust someone when you're tokenizing real world assets, like someone needs to verify that. Someone needs to inspect it. So there's, I guess, like certain businesses that like continue to survive in this world, like an Oracle, like Moody's and S&P, phenomenal businesses because they sort of put their stamp of approval to things.

So, um, uh, an intercontinental exchange for a $2 billion check in a Polymarket, largely driven, uh, you know, because of like data. Like, that's like their express— like the data itself is is valuable and fascinating, and they do— they have a huge business through data, right? Like, absolutely. Uh, most of these things do not just disappear overnight.

One perhaps parting question that I just obsess about is, like, in this technological, like, disruption, creative disruption of AI, certainly what you guys are doing, like, who stands Who in the old guard stands to benefit the most? Because you could argue there's like a whole slew of companies that just die naturally because that's sort of the effect technology has. But there's some other businesses just become more durable or just continue to compound because they have a moat. Uh, goes back to the Morgan Housel concept, like there's just things, just certain things never change even though technology comes and revamps the experience. I think you can argue the financial system is just due for a massive, massive disruption within the financial sector. Are there like legacy firms that you'd say, hey, these guys are like, they're going to use Xero, they're going to be made vastly more efficient? Um, or does it allow for like a hyper liquid to just crush Wall Street as we know it? Like, what, where do you stand in that, like, you know, opinion?

So I think a lot of it I think we'll know— a little play on words— I think we'll have a lot more clarity after clarity, right? Like, like, a lot of this is like, how much— like, where, where does the world move in regards to our— like, in a world where it is just open systems and anybody can sort of like compete in any of these environments, whereas right now it is like deeply entrenched in many ways. Um, versus a world where like there's, you know, more like stablecoins can't pay yield and this and that. And it's like, you know, this marginal change in the system, um, there's like just a natural push and pull. And I think that is like sort of going to define how quickly that disruption comes. I think again you will have groups like a Tether or like another group who sort of like take something to— I mean, prediction markets are sort of a form of this anyways, of like taking a possible vector of a intersection of things to, to an extreme and maybe getting so much adoption early on that like the world reshapes itself around these things. Like the world has kind of now started to reshape itself around stablecoins, right? Like it was a small idea, and arguably like a lot of the stablecoin issuers got chased for many years Uh, but now the world is totally reshaping itself around stablecoins itself and around the technology. Um, and so I think, I think it really depends on how some of that stuff develops. And on the application side, like, certainly there are parts of TradFi that are like much more exposed than others. Um, but, but I don't think it's just like an immediate switch overnight. I think you're going to see, like, our view has always been like, and people, there's a natural question of of, hey, you guys say you care about like decentralized neutral systems and yet you're working with like all of these huge large TradFi players. And it's like, that is like a totally reasonable thing to ask. Our view is what you don't want is the way the world was headed where like, again, you're making all these centralizing things and everybody is moving to their own permission chains because nothing can service what they actually need. Like what you want, even like we, we have stablecoins. Stablecoins are just fully centralized things, right? Like, it is permissionless access, but they, they have the ability to seize funds, they have the ability to enact enforcement, they can do all of these things. There's a central issuer who is holding the collateral. There's like all of these things that happen, um, and yet like we still would way rather have them built on neutral permissionless rails than anything else, right? What you want is the rails themselves to be neutral, decentralized, permissionless. And if you put all of these things in that environment, overall, like, the freest option is going to— like, you want them to be able to compete on that playing field versus it just being like, oh great, new backend, plug in the new database, we all have an upgraded system in TradFi, and like nothing changes. Or that you do it in a way that is super centralized and we end up with just like an even worse system that we have now for the end consumer in many ways, slightly more efficient for the banks. Their P&L goes up, but like everything else is worse. You want to have something that is as credibly neutral as the internet itself, that has this multi-jurisdictional decentralization, that has censorship, that has these things built in. If someone builds a centralized system on top of that, like, that's totally fine. That's actually so much of a better outcome than a centralized system like existing in a vacuum, because the adjacency to those rails and the natural competitive forces that are going to happen there are going to drive disruption and are going to drive better systems in like more embedded value and long term. So like we should fight at all costs to keep the rails as free as possible. And then what gets built on top of those and like again, stablecoins being a prime example now, I think it is like only a positive factor in the, in the long run to bringing things to those rails and allowing competition and freer options to exist over time.

Yeah, I think that's incredibly well said. Like, and that's some point that is constantly lost in the space, which is let's decentralize everything. And you're like, well, no, because a stablecoin, like, a technology is incredibly centralizing force. And I think in many ways what you guys have done here with Xero is a reset to introducing not only credible neutrality, decentralization, but also throughput in the execution layer. Because to your point, like a lot of people were saying, I'm interacting in Ethereum. So I don't know, you're interacting in a layer 2 that doesn't have fraud proofs. It introduces way more security. You're better off not working within crypto because you have a bunch of degradation the minute you go to a decentralized system. So the question is, when you're moving billions of dollars, yes, it matters to have a credible execution environment. And the biggest bottleneck has been just the execution. Like, you know, you just couldn't have. But I'm incredibly excited and fired up what you guys are doing. I think, you know, people are going to want to have a dollar. Dollar. You're trusting the US government and their ability to service that. And then you're trusting a counterparty like Circle and their auditor to make sure that that represents dollars in the bank account. It's like turtles all the way down. There is still going to be trust, but I guess like you do want to have at least the transparency that these systems provide, even when you have privacy, but you still have a transparent system. So I guess a lot of times, I mean, we've been in this space long enough, and I think people get lost when they want to be perfect in the execution. It's like, no, marginal improvement is still a marginal improvement, and this is a really fucking big marginal improvement. Because go— anyone that comes from, from TradFi, you could tell them, hey, imagine a system where you have perfectly auditable collateral 24/7, 365. Wouldn't that be a wonderful financial system? Wouldn't you have avoided the global financial crisis in many ways? Wouldn't you at least be able to price risk better in real time? You know, move money faster? Like, that's like— again, I think like people are just lost in that. And that don't really— to your point, I think a lot of people that came into crypto didn't come from TradFi initially and just build in a vacuum of like, look, the technology is really cool, zero-knowledge proofs is amazing, like it's a hard problem to solve, you're going to attract A players like you guys have done, but they lack the practicality of saying, okay, what does it actually mean in practice and trying to build for a perfect state and you never do that. But I think what you guys have built here is— I'm really fired up. I think most people are going to be. Surprised that in a year or two you're going to see this amount of volume, this amount of traction. But you know, you're always going to be discounted, right? So I was like, oh, you know, these pilots— how many times have we seen these financial institutions, you know, from Facebook and others announce these partnerships? But I'm rooting for you guys. I don't know if there's anything that— Roz, you've been quiet for a minute. Um, is there anything that you guys haven't talked about either publicly or like in you know, a lot of the podcasts you guys have been on, um, that you guys want to just underscore in this discussion?

Uh, I guess the one thing that, uh, to, to, to add to Brian's thing on clarity, it's like regardless of when that happens, uh, the institutions that will do well though, I think are the ones that embrace and try to do it with the technology. Like we found some partners that are, like he said, moves incredibly fast. Because like they want, they, they see a value in it. They want to do something with it. Uh, and the ones that stay behind and are conservative and don't want to embrace the new technology, like this is just for financial, from a financial system, blockchains are just better in the same way that you go from pen and paper ledger to a centralized computer, uh, database. Like a blockchain is just that next leap. And so if you stay behind and use the pen and paper, uh, someone's gonna eat your lunch. Um, so it, it's, uh, It was incredible to have to work with those teams and see them be so excited about it. And like, you know, they're clearly leaders in each of their respective fields and being like bought in and wanting to do something means I think that they're going to be a formidable force to the other existing like DeFi applications and all that because they're taking it seriously. And the ones that don't will be left behind.

No. Brian, anything else to add?

No, I think this, uh, I think, I think it's been great. I liked it. It's, we say a lot, it's not enough. This is what we say internally all the time. Like we, we talked about, um, we built LayerZero, like nobody knew who we were, zero adoption, zero anything. We had to grind our way from zero market share to 80+ percent market share to, you know, hundreds of billion dollars, all of these things. And we did it. By really focusing on just like, not, not POCs, not marketing announcements, nothing. Find the people who actually want to build something and decide what you can build together to meaningfully make their business better. And those things that we've done with those people have generated the huge bulk of our business today, right? The huge bulk of the value. And when we went to TradFi, and I still think even just from purely interoperability, I think people are going to be shocked um, at some of the stuff that is, you know, even just like Fidelity just announced that they're running DVNs and they're like doing stuff. Like, I think people are going to be shocked at the amount of institutional, uh, involvement that we've had on that side. And obviously a bunch of the zero partnerships sort of like highlight this, but it was the same thing. I told the team like, listen, it doesn't matter that we have, you know, super majority market share in crypto. It doesn't matter. It's day one again over there and we have to go and we have to like do exactly what we did for. Find out who is serious, build real things, and meaningfully make their business better. Must do one of those things. It has to be real, and that is the only thing we focus on. We'll put all of our time and effort— and it is just like repeat and replay again here— it is like, it is not enough to just build the technology in a vacuum. It is not like a logo doesn't matter. None of that stuff is going to make your business. You have to go and you have to build and solve for real use cases, drive real economic value, And that is what we're going to do. And like, again, like you said, people write it off. A lot of skeptics. It's fine. We're going to do the exact same thing that we've done. That's how the team approaches it. That's how we approach it internally. And we're going to build the things that drive real meaningful volume and sort of shift the way that the world works. And like that, that is the only goal.

Amen. So it's a timeless kind of advice that applies to crypto and everything else. Just go. Build something useful. And then, uh, so, but no, it's great to see you guys do this. I was blown away when I heard it, you know, uh, early on, and, uh, I'm, I'm incredibly excited to see what's possible here. And I think it's going to be much faster than what folks, uh, probably on the sidelines are looking at. But, um, thanks for coming on, guys. This has been a great discussion. Really enjoyed it. And, uh, I guess, uh, where can people find you or learn more about Xero?

Yeah, well, first of all, thank you for having us. Obviously you can follow LayerZero Core is still the account for everything. So Twitter, obviously we're on a bunch of different social channels. Myself and Raz are both there, but we will have a continuous stream of technical partnership and sure, go-to-market from now until the fall when we launch on mainnet. And so there'll be a lot to see.

Right. We'll definitely have to have you guys on before that happens or when that's happening. So appreciate— you guys are super busy. Appreciate you guys coming on. Roz, Brian, thanks so much, guys. See you next time.