On Ep. 22 of Tokenized, Simon Taylor, Head of Content & Strategy @ Sardine, and Cuy Sheffield, Head of Crypto @ Visa, are joined by Bryan Pellegrino, CEO @ LayerZero to discuss multi-chain stablecoins, interoperability, compliance changes and more!
On Ep. 22 of Tokenized, Simon Taylor, Head of Content & Strategy @ Sardine, and Cuy Sheffield, Head of Crypto @ Visa, are joined by Bryan Pellegrino, CEO @ LayerZero to discuss multi-chain stablecoins, interoperability, compliance changes and more!
Timestamps:
This episode is brought to you by Visa
A world leader in digital payments, Visa is bridging the gap between traditional financial institutions and innovative blockchain networks, helping players in the payments ecosystem navigate the ever-evolving world of tokenized fiat currencies with confidence and ease. Learn more at visa.com/crypto.
Tokenized is also presented by Avalanche.
With Avalanche’s purpose-built Layer 1s, institutions can tailor digital asset strategies to their exact needs—while still tapping into the power of public blockchain innovation, developer communities, and seamless interoperability. Learn more at avax.network
***
We’d also like to remind you that the views or opinions of our contributors today are their own and do not necessarily reflect those of the companies they are representing. Nothing we say should be taken as tax, financial, investment or legal advice, do your own research!
Music by Henry McLean
Speaker 1 00:00
Our approach was layer zero is not a blockchain. It's an immutable set of smart contracts libraries really like a TCP IP library that lives on your computer on every single blockchain. Then any issuer, any asset issuer, can basically deploy and run those rails themselves. So if you look at some of the recent assets that went in production over us, like bitgo, wbtc, right? So they move wbtcs Like a ten billion asset. Rather than delegating that security out to other third parties, they run a part of that security themselves. Same thing with usdt Zero, same thing with many of these others.
Sy Taylor 00:42
You welcome to tokenized. The show focused on stable coins and the institutional adoption of real world assets. And we got a great show today. My goodness, we are interviewing Brian Pellegrino, the CEO and co founder of layer zero. I am, of course, your host, Simon Taylor, author at FinTech brain food, head of strategy at sardine. And joining me, as always, is Mr. Kai Sheffield, head of crypto visa. What's up? Kai? A
Cuy Sheffield 01:10
lot is going on. I'm excited to get into this topic. I think interoperability, cross chain bridging. How do we demystify this world for our
Sy Taylor 01:19
audience? Let's get into it. Brian, welcome to the show. Thanks for joining us. Thank you so much for having me before we jump in with you. I just need to remind our audience of something tokenized is also brought to you by avalanche. Major banks, FinTech challengers and industry leaders are using avalanche to create new business models on a fully customizable blockchain infrastructure. Think of it as more than a blockchain. Think of it as an entire network built for financial institutions to innovate with purpose built layer ones. Institutions can tailor digital asset strategies to their exact needs while still tapping into the power of other public blockchain innovation developer communities and seamless interoperability join the institutions shaping the future of finance on avalanche, and you can learn more at avax dot network. I just need to remind everybody that the views and opinions of contributors today are their own, and they don't necessarily reflect those of the companies they're representing. And of course, please don't take the tax, financial investment or legal advice ideas from this. Do your own research. All right? Brian, thank you so much for being with us. Let's just start with some super basic stuff, right? There's a lot of talk about interoperability between blockchains, between everything. It's just that buzzword that you see everywhere. So can you define interoperability between blockchains and other things, and the various approaches that are kind of out there?
Speaker 1 02:57
Yeah, sure. My mental model is, I mean, it used to be actually quite difficult for two computers to talk to each other like so, surprisingly now, when everything is so seamless, like that, wasn't actually a super easy thing. And so you used to have these isolated, distributed execution environments. You had Stanford, you had DARPA, you had, you know, all of these different environments. If you wanted to run some data or program you have, you literally would take a floppy disk and, like, fly to Stanford and put in a computer and do it, and then we invent this cool protocol, TCPIP and now everything can talk to each other, and, like, life is good. That's how we talk now. That's how we do everything today. And it's just like, in the background, you don't ever think about it. Even computer science students barely learn about the networking stack anymore. So OSI model is like, oh, that's just fully abstracted a couple years ago. So 2021 when we really start working on this fully, and everybody in the space was talking about bridging, that was the thing. And typically these were through third party providers who would lock up your asset here, and then they would, in some fashion, mint you a representation, some wrapped asset in another chain, and you sort of carried that provider risk. And we had four or $5 billion worth of hacks through this structure, because the person responsible for doing this had some lapse of security. And so our original insight back then was, listen, bridging is just a special case. Value transfer in general is just a special case of generic messaging. What we care about is making a protocol for purely generic messaging. So think about it like you think about a packet on the internet. It can be any data. It can be anything, and if that happens to be value, great. That's just a array of bytes. So you basically include in this so that was the original vision. Recreate TCPIP across these distributed execution environments, which is really what a blockchain is. And so when you get to the different models, the one model that almost everybody did in the past was you'll have what I call a middle chain. I will make a chain. I'll have a set of validators and some stake in that chain, and everything will flow through that as a central hub. So hub and spoke model, and that's a really nice dream, to connect everything and have everything. Thing flow through that. But it's really, really hard. In reality, when you're going to connect hundreds of billions or trillions of dollars of assets to have all of that secured by one central hub, if anything happens to that, they can take everything right? So you can imagine, every bank, every ledger that exists in the world, has the ability to be corrupted by that thing being corrupted, even for a matter of blocks. And this can be a client issue. This can be an actual security like 51% attack. This can be a whole array of different things. So our stance was always there is zero chance in the world that the future financial system is going to delegate 100% of its security to some some third party validators. It's just very unlikely to happen. So our approach was layer zero is not a blockchain. It's an immutable set of smart contracts libraries really like a TCP IP library that lives on your computer on every single blockchain. Then any issuer, any asset issuer, can basically deploy and run those rails themselves. So if you look at some of the recent assets that went in production over us, like bit go wbtc, right? So they move wbtc is like a ten billion asset. Rather than delegating that security out to other third parties, they run a part of that security themselves. Same thing with usdt Zero, same thing with many of these others Ondo. It allows them to actually own their own security, and if you want to backstop that by including some other groups, sure, we have 45 groups who run security now. Google Cloud runs it. You have ZK lite clients with a full technology stack of providers. And if you're a small asset and you just want it to work great, you delegate to any of the external parties who do this professionally, and they can have stake, and they can have all these things you would want. But if you're a bank, if you're a financial institution, if you're any of these things you can imagine. This allows you to create your own cctp circles, methodology for moving yourself any institution seamlessly. You own that, and you don't need to actually again. Our contracts now have been in production for three plus years. There's $240 billion worth of asset issuers building on top of them, but $100 billion of actual assets deployed and $75 billion transferred over the rails, immutable. We cannot change them. Nobody should ever need to trust us or have any delegated authority to us. It is a protocol, in the truest sense, a piece of technology that exists, and it allows asset issuers to easily and credibly have full control over those rails if they want them or delegate any piece that they want. A
Cuy Sheffield 07:28
ton of interesting concepts there. I want to take a step all the way back, even getting to like, what is the problem that ultimately this technology and stack solves? And so I like to start with one of my biggest pet peeves that I've had for years now, when I talk to institutions about blockchain technology and crypto, is whenever I hear anyone use the term the blockchain, we're just going to put it on the blockchain. And I'm like, Oh, the blockchain. Like, it's If only there were just like this, one thing that everyone agreed upon, everything would be super easy, but it's not the blockchain. And so there are actually many different blockchains. I think one question is, how many blockchains will ultimately be successful? I think today we see dozens of blockchains exist. They're probably significant activity on maybe half a dozen. I don't know how many you would say have meaningful activity. And so if I understand you correctly, why does this matter? If you're an asset issuer, whether you want to create a stable coin or a tokenized security, a tokenized bond money market, whatever it is one of the first steps. And questions is, which blockchain do I issue it on? And it feels like the answer today is, it's hard to choose one. If you choose one blockchain, you're kind of missing out. What if you choose wrong? What if the other blockchains different properties you scale and get adoption of that use case. And so it seems like asset issuers are saying, we want to issue multi chain. But then, if you're issuing multi chain, how do you do that in a secure way? It seems like it's more work, more complicated, to issue multi chain. And then once you have an asset that's issued multi chain, how do you enable a consumer who's got USDC auto Ethereum to be able to move that to USDC on Solana, and so is that the right way to think about it. And so is the framework that our audience should start with. Is anyone who's in the position to issue assets probably wants to issue multi chain. And then how does layer zero specifically make it either easier to issue multi chain in the first place, or once assets have been issued, multi chain make them be able to move between each other. Is that the right way to think about
Speaker 1 09:47
Yeah. I mean, in my mind, every asset issuer has a singular goal, which largely is, is usage, is TVL, is AUM, like everybody wants to, just like a business, a business wants to expand its customer base. And wants to expand what it's able to do. And so if you talk to from the largest you know, from the black rocks and Apollos, who are issuing on chain now, every single one of them wants more they want more exposure for the underlying asset. They want more customers to be able to access this instrument in a way that's compliant with what they're doing. And so, yeah, if you issue only on Solana, well, you're missing out on a couple 100 million dollars worth of latent capital that's living on Ethereum. If you issue only on Ethereum, you might be missing out on, you know, something on a different chain. And so most of them want to be multi chain, because that just means more usage, and the more usage should generate more revenue, more you know, everything that sort of they're looking to do. And so then the question is, how do you one make that very easy to issue? So we have this standard we call the oft. This is where that 100 billion dollars is largely issued across this standard. You can think about it like ERC 20 on Ethereum. It's a single standard. You use it, and inside the token itself, it allows messaging to any of these other chains. And you choose what you enable. You choose the security model, you choose all of these things as an asset issuer. So on, that standard has made it very, very easy for a lot of people at scale. So going multi chain itself now is that part is quite easy from the first step, but what you really want to enable for most of them, especially as we get to more real world adoption, here is the end. Consumer needs to not feel it right. Like you said, if you have USDC on arbitrum, and your merchant that you're spending with, or thing you want to do lives on Solana, the prior method was you're going to write a transaction here. You're going to wait for 10 minutes. You're going to have to claim your transaction over there, which means, like, now you're to find out how to get Solana when you probably never done it before. You're just a user. You're not thinking about this. You're going to go to a centralized exchange, send it to that wallet, claim it. The goal of layer zero is like, listen, the user is going to click one time. There's going to be a bunch of embedded fees, post some of the new EIPs. It's actually the users can pay directly in stable coins. They won't even need to have underlying gas right now. You need to have gas on the chain you're on, so on arbitrarimon, Ethereum, on whatever. But one transaction, whole flow is executed. The user never sees it can jump across seven blockchains. It can execute 10 different smart contracts. So the user never sees that. They're billed one upfront cost of what is the gas going to cost on all of these they click it. Everything happens in the background, and it's done. It's dead seamless. And I think that's what you need to be able to do and even more right now the latency for that median latency is is, you know, you can have outliers, probably 1012, seconds, but median time is about 29 seconds, at least between L twos or some of the major networks. Some of the networks are a little bit faster, some are a little bit slower, but that's median time as a whole. Obviously, that needs to go down into like some single second, sub second. Ideally, there's a lot of work that needs to be done there to really have this really seamless, abstracted consumer experience. But the goal is, just like when you do anything today, this platform that we're using to record this is talking to servers all over the world, each hyper optimized to do a specific thing. One might be optimized for storage or database, and you know, you know, you have all of these different pieces that are coming together to make this seamless experience. Again, a blockchain is very easy to make it this abstract concept. It really is just a it's a distributed execution environment, a bunch of computers that all just are doing the same thing, and those computers are tuned to do something specific based on how the VM is. And so you really need to orchestrate that really nicely. I
Sy Taylor 13:23
really like that practical example of, I'm a consumer. I want to buy a thing, I want to hit the Checkout button, and I want it to just work. And that's kind of what I expect today with cards and a visa auth, if it takes longer than half a second, you know, people in behind the scenes are like jumping up and down with their hair on fire, like this has got to happen quickly. And with blockchains, you don't want to see all of the wires, you don't want to see all of the hops as a consumer, you just want it to happen, and it should. So you kind of help pick that happen there. And the wallet experience gets better. But from an asset issuer standpoint, it struck me that was it yesterday. As we record this, the CEO of Bank of America said, if and when regulatory clarity comes, they will, of course, move in to seek this opportunity. An organization of that size doesn't want to be issuing assets on every single network and ledger, so having a single standard that they can do that with, makes a ton of ton of sense. Could you contrast of T and your standard approach of messaging to some of the other sort of cross chain solutions that are in the market? There's axelar, there's wormhole, there's others. Help me. Help me, like, differentiate layer zero from some of the other approaches in
Speaker 1 14:42
the market, sure. So as I described in the beginning, both axelar and wormhole have taken the approach of, we're going to be this hub and spoke model. So wormhole, they have 19 validators, and they say, Hey, these are our trusted 19 parties that we've assembled. You can think about it as a proof of authority system. And. Everybody who interacts with this is going to rely on that set of 19. So it doesn't matter if you're JP Morgan trying to transact with Goldman Sachs, or if you're some user on Ethereum trying to transact a meme coin or an NFT, or any of these things, everybody is going to rely on the same set, same thing with XLR. XLR is a permissionless set, so anybody can be one of those validators, but they are the hub and spoke model, so the issuer doesn't actually run their own security. The issuer does. You know they can join and be one of 75 or one of 19, or they can't be one of 19 in the wormhole case, but they could let them into that if you're a big enough issuer, but that's really it. They have a hub and spoke model. Originally they were issuing wrapped assets, so it was like, Hey, we're going to lock it here. We're going to have $5 billion locked in this chain. We're going to have 5 billion minted over that. But that really changes the relationship for the user with the asset. So if you think when a user uses uniswap or uses any product, right, they're using some smart contracts, some code, so there's some embedded risk there, but it's been battle tested for a long time, and the user clicks the button, it happens, and as soon as the transaction is done, they never think about it again. There's no carried risk by the user. They have their asset, they go on their way. When you have a wrapped asset. What that means is, if at any point in time the user doesn't have a real asset, they have an IOU on another chain, right? And so if that contract on the destination chain gets broken or hacked, that asset is worth nothing. If the underlying contract where all the assets are held originally gets broken or hacked, that iru now has no backing, it's worth nothing. If the middle chain, the consensus mechanism that is controlling this an axillary wormhole case, if that ever gets hacked, like I said, even for a matter of a couple blocks, they can arbitrarily write the most malicious state to every application everywhere. So really was just a fundamental difference. Our belief was that all the largest institutions in the world are never going to hand over their security to a third party validator set. Period. They will want ownership, and they will want to run it themselves, and if they don't, then great, you can still delegate it. You have all these groups, actually, in the layer zero model, we have some issuers, so Ondo, for example, Ondo uses axlr through layer zero as one of their validators. So layer zero as a framework, allows you to use one of these networks that they've created to provide your security while not necessarily using their framework, or you're using only that. So for us, we're totally agnostic. You can fit anything in at that validator level. What
Cuy Sheffield 17:28
does it actually mean as an asset issuer to provide your own security? Because I get the idea that if you're outsourcing it to someone else, that's another layer of risk. I also get the idea that the early concepts of bridging, where you lock on one chain and you issue on another chain. It is in IOU, and particularly with a stable coin or an asset that's held off chain. Now it's like this whole other abstraction. So you've got Fiat in a bank account, and then you've got the native stable coin issued on one chain, and then that's locked there, and then you've got an IOU on a second chain for the stable coin on the first and you can start to see how like each one of those layers is additional risk. So those concepts make sense in your model as an asset issuer, saying, I want to provide the security myself. What does that mean? Does it mean I'm actually redeeming or burning an asset on one chain and minting it on the other one. So it's never locked one place and then an IOU another. Is it? You're actually removing the asset. It gets eliminated from Ethereum and it gets minted on Solana, is that the way to think about the mechanism? Yep.
Speaker 1 18:38
So that's the way oft works in general. So all ofts work in that structure. The difference is, if you looked at Circle pre cctp, what they would have to do is you would have to say, I'm going to redeem this, USDC, and you would burn it in some smart contract Ethereum. You would wait some amount of time. They would have some centralized process for going through and saying, Okay, we're going to process these. And then you would have to say, Okay, I've redeemed. Now I want you to issue me new on Solana, that might take like, hours or days or like, you know, this very long time period to actually do that process, and a very manual process. And so they went through and they created cctp, which is, hey, we're going to you already trust us circle. We have your stable coins and banks and treasuries, and you trust us to hold that part. We're already doing this in this manual process. We're going to create an automated attestation, which you will burn this we will sort of be listening for those events, and we will provide the go signal on the other chain to mint it. So they've automated this flow. They own the security. They do all of that. That's exactly what it is in the layer zero case, like circle could do that now on top of layer zero, instead of doing it themselves. So they had to build all of the rails and the infrastructure and all the hard stuff to do that, rather than just say, Here's my sign off, right? Like the easy thing. So new smart contracts, new everything with us. It allows you to one be that level of I must sign off on this, but you're also can backstop. With others. So let's say you're a new asset issuer and you're launching a stable coin, and you say, Great, listen, I definitely want to sign off on everything I do, but I'm also a little bit like I'm new to running infrastructure for all of this stuff. Maybe I don't want to be the only singular point of security. I've seen all these hacks. I saw the buy bit hack. I do everything now you can also with layer zero. You also have Google Cloud sign off. And you also have you pick the groups that you want, and maybe you say, hey, I want these three out of five to sign off, and I must always be one of the three, but I also want to be backstopped for the others. So now the user, or even you as an asset issuer, say, Hey, listen, the only thing that can happen is I must be hacked and compromised, and Google must be hacked and compromised, and one of these other groups must be hacked in compromise, and the coalescence of all of those three things end up being quite a strong security guarantee.
Sy Taylor 20:47
Yeah, it's fascinating. Having worked with financial institutions for a long time, I know they will not want to outsource security, and I know they won't want to give it to a particular place, and they're going to want to kick the tires on it, and it's a thing that you see every time there's a headline like by bit, as you mentioned in the sales pitch internally for moving into digital assets, gets that little bit harder, but really making the case that you can use security models and best practices that you have today, and you can apply those further out, I think is comforting for for everybody to kind of understand, and that you could take the ability to issue assets on to create digital assets, back them with the treasuries you hold back them with the deposits that you have as a financial institution, and bring that liquidity into the space. Is exciting if you are looking to move into stable coins. Because the first question is going to be, is it compliant? Well, it helps that we might get regulations. Is it going to be secure? But I want to move us to stable coins, because they are like the topic de jure at the moment, and you've announced a multi chain version of usdt called usdt Zero, which is with tether, and it's gonna debut on Krakens layer two, i and k I believe it's called and it's launched in a couple of places, but Kraken chose to implement this rather than launch their own stable coin. It feels like everybody's launching a stable coin these days. Everybody's having a go at it. What do you think the rationale for that was, and what do you think's starting to happen as a result of some of the interoperability you have a unique vantage point on the market?
Speaker 1 22:26
Yeah, I think when it comes to stable coins, I mean, you look at basically every major stable coin of production too. Again, circle runs their own, but you're talking Athena, USD. S USD is about $16 billion as an asset total Ondo, PayPal, pyusd, usdt all of these things, right? All of these are solutions that are now, you know, have solutions built on top of layer zero. And so there's a couple of things that are really unique. One is our pitch typically to these stable coins is listen one you you want the maximum velocity of your stable coin. You want, like, maximum utility. You want it to be used in every market in the world that you possibly can. You want it to be as frictionless as humanly possible. You wanted people to be rebalanced easily in the current state of the world, the previous state to even move usdt. You're talking six, 810, 12, bips. If you even had a pathway, a lot of them didn't. You have to go out to a centralized exchange, and you have to wait this time, and then you have to come back out to try to get them, and they're dealing with inventory issues on the back end. Now you look at Kraken chain, Inc, they moved $80 million worth of usdt to usdt zero, and it cost, I think, something like $1.50 right? You can move unlimited. There is no bip, so it's not a for us. Everything is a generic packet. We don't care if you're sending one penny or a billion dollars. You can move $500 million for just the cost of gas. If you're going between Solana and arbitrum, you're going to move $500 million for a fraction of a penny. And so the ability to do that now you think about, okay, now movement between these chains rebalancing becomes totally seamless in the background. And again, with latency constantly decreasing. Now you're talking spreads between chains can be arbed within like subset spread. So now you have much more liquid markets. You have much tighter spreads across all of these different disjointed markets that are happening. And it just becomes an actually, much more useful, net useful tool as a piece of currency than could exist before. And that usually is the thing like that's very exciting to asset issuers, and I think you see it now with usdt, we've seen really large transfers happen for basically zero cost. And the same thing across the board, with PayPal, with Athena, with all of them. I want
Cuy Sheffield 24:35
to unpack this Kraken example a little bit more to make sure I understand it. And so the first piece so Kraken has created their own layer two on top of Ethereum, which is called Inc. I think it's using the optimism op stack. We're seeing this as a trend where you've got many institutions, whether exchanges, even some traditional companies. You talked about Sony in the past. As you know, they're creating their own l2 that they've. Announced. And so there are more and more of these l twos that are coming online if you're the asset issuer tether in this case, part of the question is, okay, are you going to issue usdt directly on every l2 that exists and emerges? I imagine that like that requires some real work. Every time you want to natively issue usdt on a new chain, you have to my understanding is you have to have a smart contract for that chain. You have to make sure it's audited. You have to have the node infrastructure for that chain. You have to have everything for that chain. And I think today, usdt might be on eight chains or so. I think circles on like 16, and so I think circle has gone the farthest, and their other issuers are taking the approach to go even further. And so in the Kraken scenario, is this a usdt is issuing on ink the same way that they issue on Tron, or they issue on Ethereum and they've decided to add ink as another network, or are they using layer zero, basically saying it's just hard for us to keep up and add all these other networks. And so here's a way that you can move from usdt on Ethereum to Inc, which is EVM compatible, without the asset issuer having to do a direct integration and understand all the unique properties of that network is, is that? The way to think about it, it's like the process of getting on a new chain, that's one barrier, and then how to move between chains is the second this is trying to solve both at the same time, where, for the first time, usdt is on ink through this mechanism, and it's on ink, because it can move between Ethereum ink Tron like it is, any chain that usdt is currently on can be bridged. And I don't even know if you use the term bridge, since it's different than the kind of normal bridge idea can go over to ink. Is that the way to think about it? So
Speaker 1 26:56
any chain that layer zero is on is connected in a full mesh. So we're on 122 chains, so most of any interoperability provider. And so all of those, 122 chains are connected to each other. So you're, first of all, you're absolutely right. All of the asset issuers, USDC, usdt, all of them. It is incredibly difficult to go to New chains, to new VMs, even to the same, to EVM chains that are the same. There are huge hurdles along doing that. They have to run all this. They're new ones. Like, every
Cuy Sheffield 27:23
day, every day, constantly. Yeah, it's like, it just happens so much. It's like, how do you keep up? But at the
Speaker 1 27:29
same time, there's a huge incentive for being there first, because if one of these chains takes off, our bear, a chain just launched. There's $3.1 billion got on the chain within the first two days. And so do you want that $3.1 billion is getting paired on these dexes and in these liquidity pools against an asset, and certainly as an asset issuer, you want that as a special as a stable growing provider. You want them to be paired against you. Every single one of them is eth, two, USDC or usdt, right? Bitcoin two usdt or USDC as the pair. And each one of them want it to be them, because each one of them means an extra billion to one and a half billion, effectively half the LP pair in their asset. And that just means more assets in total, more AUM, and that's the business that they're in. And so being able to move quickly to any new network on day one to be the first thing there that is available for everybody to have liquidity and to pair those liquidity that's very hard to unwind later. If you have a billion dollars of assets that are all paired against usdt, and all of the consumers on that chain are already naturally using usdt to trade. Well, if you're USDC and you come there six months later, it's going to be very hard to get everybody to swap over to using USDC now or migrate all of those pairs away. And so really, there's a big advantage to being able to be there fast, to be day one on a lot of these chains, and I think that's something that we provide to a lot of these asset issuers, that is just extremely attractive. I can certainly
Sy Taylor 28:50
see that like if you're in the Aum game, you got to get that aum. But if you are more institutionally focused as well, we talk to a lot of financial institutions, how do you get them comfortable with the compliance side of being multi chain? There's a security side, but compliance is like, it's evil twin,
Speaker 1 29:07
100% so we had to go through this. And again, each asset issuer kind of has their own compliance process, but we via PayPal, pyusd, obviously working with Paxos, we had to go full NY, DFS approval, and you have to get that for every chain that you're going through, and there's like, a very heavy compliance process that goes through it. And so for them, they start on one chain, and then they move to another. So there's Solana and Ethereum, and that was this long process. And they will very likely go to other chains in the future, and they're gonna go through their own process internally to do that. And we are just the tool that they use. Every new chain that they add will add this connected layer between the three. So now you move between Ethereum and Solana pyusd, you're moving with layer zero. You won't know it, and it'll be totally in the background, but you can seamlessly go between them if they add ink as another chain. Now you move in a triangle in any direction you want. And so for us, compliance piece really is on the asset issue, where we are a piece of technology in the underlying. And so they go through their process, and then they just turn it on with us.
Cuy Sheffield 30:03
Is part of the compliance piece that the issuers look at whether or not KYC happens at the interoperability step. Because today, traditionally, again, if you wanted to move from USDC on Ethereum and kind of pre cctp, the circle cross chain Transfer Protocol. If you wanted to move from USDC on Ethereum to USDC on Solana, the most secure way to do that would be through a centralized exchange. So you would go to Coinbase, or you deposit USDC on Ethereum and then you withdraw USDC on Solana. That would take some time, as Coinbase would have to manage inventory. You would have to be KYC with Coinbase, my understanding is now with cctp. That's basically just like a low level protocol, kind of defi infrastructure where you don't have to go and onboard with an exchange. Anyone who's holding USDC in a wallet on Ethereum, self custodial could move that to USDC on Solana without needing to kind of onboard and go through the full process. When you talk to asset issuers today, which approach are they taking? Are they saying the interpretation is someone is holding their existing asset that person is just reformatting which database that assets on, and so that doesn't require a direct onboarding interaction? KYC, or are some of them saying no, if you're going to change which chain you're on, and you're ultimately relying on our trust to do that, because we're the one that's verifying that transaction. It's not just a decentralized blockchain. It's like we are letting you move from USDC Ethereum to USDC on Solana. And if you're someone who shouldn't have that USDC, and we're now saying that you can move to salata like we're then approving that transaction. And so what are you seeing in that split on the issuer side? Are there some that are requiring KYC and some that don't? Or all of them aren't requiring KYC and they're just seeing it as a low layer database transfer
Speaker 1 31:55
on the issuer side, none of them do on a per transfer basis. So they all do this in a way that I'll call reactive fashion. So all of them are very quick to as well. Actually, some are more quick than others, but all of them have a method for responding to OFAC, to sanctions for any of these things. So there's a public list of blacklist. They have the ability to break these addresses in the contract. So as these sanctions list are updated, or is there somebody who's too should not have funds, or should not be able to use the protocol, they will brick them. Those funds will then be frozen or reclaimed. They will not be able to move with that. But it's a constant process of, sort of adding to this list. So you can imagine it is a check before it goes out. Right? None of these people who shouldn't be able to move the asset can move the asset, but there isn't this full KYC step in terms of every asset. So everybody who's not on this list basically can move freely. Now, from layer zero's perspective, everything's an arbitrary packet, which means an asset issuer. They can add full KYC restrictions into the underlying they can add rate limits. They can add anything that you want to add, anything that you can do in a smart contract you can do with layer zero, totally agnostic dos. We are just moving the bites. But in terms of what's actually in production, uh, everything is this, I'll say, kind of reactive. It's proactive if you're already on the list, but if you had moved it and then somebody realized after they would then need to go and add you, and then you wouldn't be able to
Sy Taylor 33:14
move the next top. It's much more institutional focus as I think about it, from a Global Transaction Banking standpoint, it's that way. It's like the flow is happening via me, and I'm sort of watching it and making sure I'm picking anything out. But the end point should be KYC or the same as cash. They're self custodial. It's the way to think about it. The telegram ton network is integrating usdt using layer zero. It's a pretty big network, but they're also interesting because they bridge this like isolated ecosystem back to the crypto world. Were there any technical challenges with this one compared to some of the others? And could you take us underneath that? Yeah, very much so, and remind everybody telegram ton and what that does, because maybe not all of our listeners are familiar.
Speaker 1 34:02
Yep, Telegram obviously is one of the most used messaging apps in the world. There were about 900 million active users, right? And so as an ecosystem, they've launched their own blockchain, the ton blockchain, that's integrated into this messaging app that makes it one of the most compelling ecosystems that could exist, right? You really have distribution downstream into potentially almost a billion users. And so obviously there's very strong desire to be the stable coin of choice within that ecosystem, to have users paying, creators getting paid, all of these things, all of this transactional activity happening, and economic activity happening on that platform to be happening sort of through this asset. And so already usdt was there. So I think it's their third largest. I think you have Ethereum and Tron, which combined for about 136 billion or something. And then I think ton is about a billion and a half or 2 billion. And so they had issued an isolated instance there, but there was no way for them to go so. And part of doing ton the network, we deployed this USD team. Native bridge that allows you to move seamlessly between any of these assets, but in terms of building it any new VM so for us again, one stance that we take, we talk about differences in protocols. Another stance, typically other protocols have, is that their contracts are upgradable, so they can, you know, if they wanted to be malicious, they could be malicious even if they wanted to be benign. A bunch of the hacks that have happened, most of the hacks actually hacks, actually very few hacks of the $5 billion bridge hacks that have happened, happened because it was actually the bridge operator themselves being malicious. It happened oftentimes because somebody pushed code. So wormhole, you mentioned wormhole, they had a three $50 million hack that happened, but then they had a $1.8 billion that was exposed risk that was found by a white hat. And then another time, two weeks later, $1.8 billion at risk that, again, was found through the same issue. And this was not due to anything other than an engineer that was deploying a smart contract forgot to initialize something. In other cases, the case of a nomad hack, it was downstream. They pushed code that they thought was good. It was a bug in the code. And so our stance has always been fully immutable. You will never need to rely on us. We cannot upgrade it. You don't need to trust us. It is a piece of code that exists. But what that means is, when you're dealing with 10s of billions of dollars of assets built on top of it, 10s of billions, hundreds of billions of dollars of economic activity happening, there's a really large sort of burden and getting that right. So it's really like a measure twice, cut once thing. And so we have to go very deep in any new VM that we're deploying, because once it's there, it's there. And so I think we spent almost a year on Telegram, on ton building that we literally rewrote an entirely new programming language, fun c plus, plus on top of their fun C, really, really, really deep technical work I don't think anybody else outside of their core team has done, and we worked very closely with the core team throughout. So it was a very large lift. It took much longer than we expected, but now that we're there, we're super excited to be there. Obviously, you have the usdt moving around pretty actively, and then we have every other large issuer now wants to get to Tai and we're really the only way to do that.
Cuy Sheffield 37:00
I want to switch gears and talk a bit about your business model, and I think it's really interesting your protocol, but this is crypto, and crypto protocols have business models, and they have ways that they make money. And there's this concept of fee switches that we're hearing more and more about, which I want to unpack. So let me just start with how do you make money today as a protocol. Where are the fees charged? Where do they go? Is there a token? Is there governance like, What? What? How do you explain that to someone who's never really looked at a protocol as a business before? So we
Speaker 1 37:35
it's very important to separate out layer zero Labs, which is our equity entity, which is the nd I'm the founder of, and layer zero the protocol, right? So layer zero labs, we view ourselves as you would view consensus relationship to Ethereum, right? Consensus has a very large stake in Ethereum. They're very vested in the future of Ethereum being successful. And they build products that they themselves monetize, that are a creative led ecosystem as a whole. So they build in fura, and they build meta mask, and meta mask makes $400 million a year, and they build all of these things that make Ethereum as an ecosystem have really contributed a lot to its success over time. Layer zero labs, goal is the protocol doesn't care about us. It doesn't know that we exist. It doesn't give us a tie that doesn't know nothing. The protocol doesn't care about us as an entity. We care very deeply about the protocol, and so our goal is to build things that are a creative the protocol. So one thing we've built is this executor I described before. The natural stands for protocol would be message here, wait, and there's some verification, and then the user has to go and claim the message there. And that's just like horrifically bad as an experience, what our executor does. And anybody can run an executor. So permissionless, there's no special privileges for us. Is basically say, Hey, here's what it's going to cost on Solana, you're on Ethereum. You're going to pay us an eth upfront. So maybe you're paying $1 on Ethan a penny on Solana, you're going to pay us $1 it would be $1 in one penny. Maybe you'll pay us $1.05 pennies, and then we're just going to execute the transaction on your behalf. So you as a user, you click, you sign once you trigger their transaction on the other side. So layer zero Labs has done about a period of a billion dollars of that to gas from one chain for gas to another. So you're quoting n squared pathways in real time, and you're dealing with vol and duration risk from the time that you're accepting a quote to paying out. So you're eating losses a bunch of times, because vol has basically made pricing on the destination chain move again. So you're doing all this really complex stuff and managing, honestly, 40, $50 million worth of gas assets across 122 networks running full nodes for all of these. So is
Cuy Sheffield 39:32
it like FX for gas? Almost? Yeah. Is that the way to think about it? It's like you're running an FX desk, but it's you've got native currencies being used for gas on both sides that you have to figure out how to hedge. And
Speaker 1 39:42
now imagine you need to run a full node for every country and every currency, right? That's the other complexity, right? Is we have to run full nodes of all 122 of these networks. We have two jobs. So this is like a really large lift from that side that's laser labs. And laser Labs has a whole long slew of things that it's building around the. Protocol to make the protocol creative. The protocol itself is its own thing. It has a message gets sent. You have groups who are attesting, and they might charge like if you're going to use Google Cloud to validate, then Google Cloud might charge you five cents or 10 cents or 50 cents, or whatever they want to charge. All of them set their own fees independently. If you're running it yourself, you can set it for free. Or you can charge your users, you can do whatever you want. There's execution side, and then the protocol has an underlying fee switch that can be turned on that basically is bounded by up to 200% of what the executor and the DVMs, which are the verifiers, charge in aggregate. And so the reason there is a bound, and I think it's very important, the reason there's a bound is, if it wasn't, then the fee switch could be set to effectively unlimited, and center the entire network, right? So it has to be bounded. And what it also means is, for certain financial institutions, if they want to run all of the infrastructure themselves and do 100% of it on their own, then the fee would be effectively zero, because the verifiers and the executors would be charging zero, right? There isn't an enforced fee. In the underlying technology, there is only a fee if you are using sort of external parties that are aggregating fees within the network. And so right now, there isn't anybody who runs the entire full stack themselves. Everything is run to external of a network, and you can think on a per transaction basis. You know you're talking depending on the chain. It's really based on gas. Is what drives a lot of it. A lot of this is a markup on on the gas that's being spent. But if you're on Ethereum, and Ethereum is charging you three to four bucks, you might be charging 25 to 50 cents that the protocol would capture. So the long term vision of the protocol is, Listen, you're effectively reinventing TCPIP here. If you think this is the way the technology moves in, that all information moves across blockchains, across all of these networks. The protocol itself has the ability to charge a fee on every single message that's ever moved across the internet. Effectively, yeah, and
Sy Taylor 41:53
it gets to stay agnostic. It gets to be a protocol, and it feels like you've designed that first principles. And speaking of first principles, the first principles of podcasts is, I have to thank our sponsors, so I'm just gonna, I'm just gonna do that right now, and we'll come right back to close out this episode, if it's not obvious, is brought to you by our friends at visa, a global leader in payments, visas, tokenized assets platform, VT app uses smart contracts and cryptography to help banks bring fiat currencies on chain. Vtap allows financial institutions to issue Fiat backed tokens, improving financial efficiency and enabling programmable finance. You can check out the links in this episode's description to express your interest in vtap. Alrighty. Thank you to our sponsors, Brian, little change of gear here before we close out, which is, I didn't realize your background. You were a machine learning engineer and AI and crypto has been kind of a big topic for us, but there are a lot of projects just slapping the word crypto on things. But then there's also, like Airbnb for GPUs, and there's kind of agents and that sort of stuff and all kinds of stuff going on. What's your perspective on that? Because, again, as layer zero, you kind of act as this, like Switzerland, type of model where you get to see a little bit of everything. What are you seeing from your vantage point? It's
Speaker 1 43:20
been very interesting to see the convergence of two worlds, right? So I spent a long time. I spent almost eight years in AI, and originally it was, you know, it was just early on. It was fun. I was building models. I sold the models to a bunch of the pro baseball teams. But eventually, like my last academic paper that I published, I published with Noam Brown from Facebook AI research. And Noam Brown is now like the creator of the oh one model at open AI at chatgpt, right? And so this has been, like, the really transformative model of moving things into into actually, sort of transitioning thinking, and having it not just be training time, but inference time compute. And so it's really been interesting to see something I spent so much of my life on now become so large. And then this conversation around the convergence, I've actually been, I've been like the dissenting bear for a long time at the intersection of the two. I think there's a lot of things that don't make sense. I don't think an Airbnb for GPUs model is ever going to be successful against just against economies of scale. You look at running nodes running at like, it's so hard to compete when Google and Amazon and all these big groups have like, a third of your capex cost, and a third, you know, your OPEX cost is just everything is going to be driven down to basically the cost of electricity in a point where it's efficient. You've seen this with every form of mining, with Bitcoin mining, with all types of miners at all. Basically, you know, you're at three cents a kilowatt hour, you're doing great. If you're at six cents, you're okay. And no consumer at 10 to 14 cents is even able to compete at all, like it's just all gets industrialized, basically. And so that model I've never found super interesting. I do find the agentic model very, very compelling. And I think it really comes down to a world of I think deep seek was a very interesting moment. Maybe I'm getting a little bit too technically, but I think deep seek is a very interesting moment. Is it? It shows that you can create something that costs significantly left it really shows there's a heavy commoditization of models. I think there's a lot of discussion now that, hey, maybe the model piece itself is going to get heavily commoditized over time. So llama, deep sea COVID, these show that. And the question is, if all of the models, and all of the fine tuning of the models for what is, you know, your hyper specialized lawyer and accountant and everything that you might be using in the future for these agents that do very specific things, or travel planners or assistant whatever it is, if they all live purely in open AI and chat GPT land, hey, maybe it's just going to be stripe on the back end, and you're just, you're going to find a way to basically do this efficiently. Now, it's kind of a problem in the current world, in micro transactions, they think that's a thing that is kind of a constraining factor in some of the traditional markets now. But if it is unlikely that the end user, the consumer, is going to give the agent that they're using their own assistant who's going to go fulfill something, hey, I want to book this appointment. I want to do this. I want to do that. If that thing doesn't have access to their credit card, to their bank account to be able to spend. What does that actually look like? And I think that is the world these agents needing to go out and fulfill a task, and in order to fill that task, they might need to go talk to a booking agent and this agent and that agent and do all of that. And how do they coordinate payments between themselves?
Sy Taylor 46:17
You start getting into cross communication and cross authentication, and you might need some sort of abstraction layer to do all of that. Listen, I could sit at the Learning Tree for the rest of today trying to learn more from you, because you are at the intersection of so many things, literally as layer zero. But that is all the time I have for today, and I do have to run unfortunately. So Brian, where can people find out more about you and more about layer zero? Yeah,
Speaker 1 46:45
anything layer zero, you can just find at layer zero, underscore core on Twitter or layerzero dot network online. Thank
Sy Taylor 46:53
you so much. And Kai, how about you on x at Kai Sheffield and visa.com/crypto and you'll find me at sy Taylor, or look for fintechbrainfood.com, and with that, we hope you listen again next week. Thank you, and bye for now.