Epicenter - Learn about Crypto, Blockchain, Ethereum, Bitcoin and Distributed Technologies - Antonio Juliano: dYdX – Decentralized Perpetual Exchanges

Episode Date: January 20, 2023

2022 has proven time after time that centralised entities are prone to multiple points of failure, intentional or not. Decentralised finance (DeFi) addresses these, but the user experience has often been lacklustre compared to that from centralised finance (CeFi). Exchanges are the backbone of any market, but high throughput is required in order to ensure that both makers and takers can proficiently use them. Decentralised exchanges have long faced issues due to bandwidth constraints native to blockchains, but recent L2 scaling solutions have provided a considerable increase in terms of throughput. For example, dYdX is currently the largest consumer of STARKs, but growing institutional and retail interest in derivative financial products urges for an even better alternative.We were joined by Antonio Juliano, founder & CEO of dYdX perpetual exchange, to discuss about the upcoming transition to dYdX V4 and what challenges decentralised exchanges have to face amidst technological bottlenecks and upcoming regulations.Topics covered in this episode:Antonio’s background & timing productsdYdX product overviewPrice feed oraclesdYdX user demographics & UXDevelopment timeframes & taking risksOrder book vs. AMMBuilding a sovereign blockchain on CosmosThe challenges of decentralising order booksIBC integrations and composabilityCurrent state of scaling solutionsMigrating from dYdX V3 to V4 & token utilityDeFi vs. CeFiDeFi education & regulatory outreachEpisode links: Antonio Juliano on TwitterdYdX on TwitterdYdX exchangeThis episode is hosted by Brian Fabian Crain & Felix Lutsch. Show notes and listening options: epicenter.tv/479

This is Epicenter, episode 479 with guest Anthony Giuliano. Welcome to Epicenter, the show which talks about technologies, projects, and people driving decentralization and the blockchain revolution. I'm Brian Crane and I'm here with Felix Luch. Today we're going to speak with Antonio Giuliano. He's the founder of D-YDX. D-YDX is the largest crypto-derivatives exchange or decentralized derivatives exchange that have about 1.2 billion trading volume per day. it's also the largest layer two application on Ethereum right now running on Starkware. And I'm sure a lot of people have seen some months ago, they've also announced that they're

going to build their own cosmos chain. So that also got a lot of attention. And so lots of interesting things to touch on. Thanks so much for joining us. Yeah, excited for the conversation. Cool. Well, I mean, let's start at the beginning. How did you get into crypto? And what's the story of how you started at DYDX? Yeah, absolutely. So I got into crypto in 2015 when I interviewed for a job at Coinbase, which was my first job right out of college. I kind of got to Coinbase differently than most people in that I didn't really know anything about Bitcoin or cryptocurrency at the time. It was just one of the like 20 or so companies I applied to senior year of school. And I was kind of just more interested in finding a really high quality startup. But went, I interviewed and all these people seemed really smart. really awesome and really excited about this random thing that I didn't understand in Bitcoin. So I was like, okay, what the hell?

Like, I really want to work with these people. It seems like an awesome opportunity and a great company at least. And they all seem excited about this thing. I really get it. But, you know, I'll take a leap of faith and kind of at least spend a year or two working there. So went out to San Francisco, worked there for a year. I was like the 100th employee or so at Coinbase. And it was a really awesome opportunity just to get a chance.

to have a great front row seat to kind of see what was happening in crypto at the time. So there are a lot of really great people who since then have gone to do really awesome things at Coinbase. We were fortunate to be, you know, by far one of the biggest and kind of only one of the legit companies building in crypto at the time. So a bunch of other really great people in crypto would come and talk to us. Like we had Battalic come and talk really early. Joey Krug, the founder of Auger, come and teach us how to build smart contracts.

And it was a really kind of pivotal time in crypto in terms of that was right around the time back in 2015, early 2016, when Ethereum was launched. And I think it took all of us a while to wrap our heads around what's possible with Ethereum. And I remember kind of learning from Olaf and Fred about, hey, what is gas, what are smart contracts, what is solidity, things like that. And again, I didn't really get it, but I kind of took a little bit of a mention. a leap of faith because there are clearly all these really smart people that were excited about it. And I kept looking into it. And then I think at one moment it just kind of clicked for me, oh, like this is just Bitcoin, but you can build programs on top of it. And it sounds kind of simple now. And I think once people have understood it more on mass, they can explain it more

simply. But it really wasn't obvious to people at the time. But once that kind of clicked for me, I realized this was just a totally new paradigm of computing, where for the first time, you can write these programs that execute totally autonomously, totally deterministically, and without being controlled by anybody. And I was like, ah, aha, this is something that's fundamentally new. Like, there must be something someday that this can be useful for. And I've always been really entrepreneurial. I really wanted to start my company even before, my own company, even before going into Coinbase. And I was pretty transparent with Brian and everybody else on the team about that, and they were super supportive. And then after Coinbase, I,

briefly worked at Uber, but then started D-YDX in mid-2017. And actually, the first thing I started when I tried to start my own crypto company, was not D-Y-D-X. It was a search engine for decentralized apps. And not a lot of people know this, but I built this out in early 2017. And I tell this story because I feel like I learned something important in it. So I tried to build a search engine for decentralized apps using kind of the data that was on the blockchain. to inform kind of the ranking of decentralized apps. But the problem was there were only like 10 decentralized apps in the world in early

2017. So what's the point of a search engine if there's nothing to search for? And the point of this story is I really learned the importance of timing in building products and building companies. Like maybe someday it'll make sense to build a search engine on decentralized apps, but certainly not in early 2017. So I kind of took a pause on that and I thought about, well, okay, what could I build that would actually be useful in crypto right now back in 2017.

And back then, and still to this day, the main thing crypto is used for is trading and speculation. And that was right around the time when the very first decentralized exchanges were just starting to come out, things like ZeroX, things like Khyber. So I took a look at that, and I was like, wow, this actually kind of makes sense. Like you can build an exchange that is just code, and people can use that rather than centralized intermediaries. and that has a ton of benefits in terms of censorship resistance, in terms of transparency, in terms of security of funds, what's coming next after that?

And if you think about finance as kind of a stack where you have to build the things lower in the stack, so things like money, which in a decentralized sense could be things like Ethereum and Bitcoin and stablecoins, and then you build exchanges after that, so things like zero X and now uniswap, kind of the last step in the stack there is more advanced financial problems. And those are things like financial derivatives, margin trading. And that's what we set out to build in DYDX. So that's kind of the journey. And we've been at it since 2017.

Yeah, awesome. That's super cool that you found that niche there. And I guess obviously became one of the big players. We also wanted to just generally for the listeners to get an overview of what are actually the DYDX products right now. Maybe you can like kind of elaborate a little bit on that. Yeah, so we have one main product, and it is a financial derivative, and it's the most popular type of financial derivative that's traded in cryptocurrency, and specifically that's called a

perpetual contract. A perpetual contract is a type of derivative contract that, as per the name, never expires. It's perpetual, kind of goes on forever. And the reason these are so popular in trading cryptocurrency is because they're really accessible to kind of the audience that trades cryptocurrency in this kind of prosumer retail trading audience that's global. And we have built these into decentralized finance. And kind of the way that they work is you can point a perpetual at whatever price you might want to point it at. And specifically, you could make a perpetual for Bitcoin.

You could make a perpetual for Ethereum. You could theoretically make perpetuels that are pointed at pretty much any type of asset that you might want to trade. Right now on DYDX. only perpetuals are available for cryptocurrencies, so kind of just like the top markets on coin market cap. And these perpetual contracts, though they sound advanced, they're actually really critical and integral to the crypto markets, and they're by far the most traded product in crypto. So about two-thirds of crypto volume is traded through perpetual contracts, and only one-third is traded

through a spot, which is kind of like a fancy word for just regular old buying and selling. So these things are really popular in centralized exchanges. They were popularized by Bitmex and now are available on Binance. They had been available on FTX, but RIP for that. And now they are available in Defi as well through DYDX and a couple other decentralized exchanges. Do you rely on like, I guess, some sort of price feed oracles for that? And like how does DYDX solve that? Are you using ChainLink or some other kind of?

Yeah, so perpetuals do rely on price oracles. Even centralized perpetuals rely on price oracles more or less or index prices. And the way that that works on DYDX and pretty much every other legitimate perpetuals trading venue is you take the prices of a given asset on a bunch of different exchanges. And then you kind of take the median of them, which kind of removes outliers. So you might look at the price of Bitcoin on like Coinbase and finance and whatever other exchanges are at the most liquid and have. the most trading volume. So right now we do use chain link price oracles. And then as you alluded to before, the big thing that we're building right now is the next version of DYDX called DYDXV4. And in that version, the price oracles will kind of be built directly into the validators

themselves on our Cosmos chain and there won't be any external price oracles. So chain link now and then eventually our own later. Also, we're going to definitely get back to V4. I think, You kind of alluded to it a little bit already and obviously how the product is very accessible, the perpetuals. But can you like expand a little bit who are the users of the YDX? Yeah, absolutely. So there's two main classes of users that we have. The first class of users is crypto institutions.

And these are hedge funds that are trading crypto. There's a lot of different types of them. I'd say kind of the two main types of crypto institutions that are most actively trading right now are kind of native crypto trading firms, so like crypto hedge funds like a winter mute or something similar to that. And then there are kind of more traditional, call it Wall Street trading firms that have really gotten into crypto trading in a big way. So like a jump trading or Tower Research or others. So that's kind of one big class of users. and most of the actual volume does come from these institutional users,

but that doesn't actually necessarily mean that they're more important than the other class of users. The other class of users, that's big in perpetuals and therefore also on DYDX, is kind of prosumer crypto traders and specifically international crypto traders who make up the bulk of the derivatives trading market, especially in crypto. And these users are just regular old individuals, but they are kind of more sophisticated and demand. kind of more advanced trading products than most crypto traders. It's not your friend that bought their first Bitcoin yesterday. It's kind of somebody that's either for a full-time gig or kind of on the

side likes to basically day trade or kind of trade with more advanced financial products like perpetuals. And these users are actually the most important users, not to say that institutions aren't important, but most exchanges build first and foremost for kind of these prosumer users. because they're the source of most of the, quote, organic taker volume on the platform. And that's kind of the most valuable volume. So you want to make sure the product is really excellent for those users. And they have all the tools that they need to kind of maximize their returns, maximize their risk management and all of that.

And then crypto is actually really different in terms of its market structure than traditional finance. In traditional finance, really almost all of the volume comes from institutions. Sure, there are some prosumer, like retail traders that trade in the traditional markets through things like Robin Hood, but they make up a very small portion of the overall trading volume. In crypto, it's completely the opposite where a lot of the trading volume does come from retail, basically. And then the institutions will follow around wherever most of the retail volume is. So if you're an exchange or an exchange protocol, kind of the name of the game is attracting more of that volume. But we've also really intentionally set DYDX up to be a platform that is really accessible to institutional traders as well. And that's something that I think is pretty novel for a D5 platform because we've really intentionally tried to make the technical trading experience on DYDX similar to a centralized exchange, actually, because that's what the users are used to trading on.

So like there are API endpoints that feel really similar to a centralized exchange API endpoints. And that makes it much easier for a lot of these institutions to come and trade on DYDX as their first foray into DFI. And fast forward to now, we've actually been really successful with that. And pretty much all the top crypto trading firms are trading on DYDX now for a lot of them as their first thing in DFI. And I think that was a really intentional thing we tried to set out and do. Yeah, I mean, I'm somebody who's, you know, I guess, a long-term crypto investors, but not a trader, right? So I've mostly been the sort of, you know, spot volume type user,

but I was, you know, did some, like put on some trades on DIYX before, because I wanted to, you know, try it out before a podcast. And I have to say, it's like so smooth and intuitive and like, yeah, super straightforward. So I think that was really, really nice. user experience, really nice flow. Yeah, I'm glad to hear it. And that's been something that's been an intentional choice as well, like I was mentioning

before, is trying to make the experience, again, feel like just a regular old trading venue. And that's actually pretty hard to do from a technical perspective in Defi, but doing things like being a really early adopter of a lot of these new technologies, like you mentioned it before, but currently we're by far the largest app on Layer 2. and then fast forward to the next thing that we're building, we're going to be building sovereign, our own basically blockchain for DYDX, and all of that is kind of centered around the user experience.

I think you kind of alluded to like a bunch of stuff that maybe it's kind of a larger question here. What do you think are the main things that you did right and that allowed DYDX to get so successful? So I think one of the first things is that we, were early, actually. And that gave us kind of a lot of time to learn and build and iterate. Kind of gave us more shots on goal to build the right thing and build in a really iterative approach. But if you are really early and certainly we were when we were founded in 2017 and I think even

now, DFI is in the really early stages, you have to combine that with a really long-term approach towards what you're building and kind of just the internalization that what you're building really is not going to take off probably for like five to ten. 10 years plus, and I still think that's the case in Defi today. And I think a lot of people don't build with kind of that long-term focus, and that's probably the reason, like, 75% plus of kind of call it new ventures and new markets fail. So I think that was probably one of the first and biggest things. I think this focus that we have on building a great product experience was really useful as well. And then just building a really strong core team and especially a really strong

technical core team. Like we again have been building smart contracts for over five years at this point. I don't code so much personally these days. But I coded a lot of our early smart contracts along with some of our early engineers. We set a lot of the paradigms for how smart contracts are built. And I think it's also this mentality that we have of risk-taking from a product perspective. And I think What I mean by that is, like, we've made a lot of bold choices from a product perspective, and I think we've executed on them pretty quickly as well, especially in crypto terms. So, like we've been talking about, D-Y-D-X is currently on layer two. And I think it's been basically two years now that we've been on layer two.

I know it's gotten a lot more popular recently, but, like, we were on L-2 before people were really even talking about it. And that was a really intentional choice. And we did that, and the previous L-1 contracts are kind of sunset. and that was kind of risky, right? Because we had something that was working on layer one, but we went out and we felt strongly that this new technology could give a better product experience

because it can process way more transactions. And in our previous version, our users were just getting absolutely hammered by gas fees. The latency is way lower. So, Brian, like you mentioned, now our product can feel just really snappy, kind of, again, more like a centralized product. but being that trailblazer

when there hasn't been a really great example of a previous project building on some of these new technologies that's gotten a ton of adoption I think can be scary because a lot of times just in life and in company building especially you feel like you have something to lose.

It's like, okay, we have this thing that's working and we could build on it more iteratively or like, okay, let's just like wait and see what happens with layer two, for example. But instead we've just really put our foot forwards and be like, no, like, we believe this is the best technology and the best product that we can build. And I know we're going to talk about it later,

so I don't get into it too much. But I think this is a similar mentality that's caused us to make a relatively controversial technology choice. In our next version, which is to kind of abandon L2s almost at the time when everybody is like finally hyped on them and move to our own blockchain. But we can talk more about that.

So I think it's that mentality. And then also just the experience and capability of our team to go out and execute on a lot of these things that really have never been done before in crypto. Cool. Thanks so much. So I have kind of like to get the sort of the corollary question on this one. What are the biggest mistakes you made and especially mistakes where you feel like, okay,

here's like a deep lesson that you've learned and something that you're approaching in a different way now? Yeah. I think one of the biggest mistakes that we made was earlier on in the company. And again, we were able to recover from it. because we were so early in the space. But back in 2017, the space just felt different, like ICOs were rampant.

And basically what everybody was doing at the time was publishing white papers, not really building products, but more publishing white papers. And maybe, I don't know if people like in crypto these days remember what white papers even are because luckily they've fallen out of fashion. But basically what white papers were,

were kind of long, like research-based documents, like an academic paper basically, saying what you're going to go out and build. And these things would be, like, ridiculously complicated. And the white papers would basically be like, okay, here's our like 50-page white paper, and we're going to build this whole thing. It's going to be perfect.

And then we're just going to ride off into the sunset with this perfect product. But that is not how product development works. And I think we weren't as guilty of this as some other people were, but we were guilty of it to an extent to, like there was a D-Y-D-X white paper early on. And kind of worse than that, I think we spent basically too long building the first version or two of the DYDX protocol. I think the first one took us about a year or a little over to build, which again is fast.

But we have good developers. And like the problem with the first version was we were just trying to consider like every case. We were trying to build a really general protocol, general meaning. Like you could do a ton of different types of things with it rather than having a really specific use case for the product. and that caused us to build a ton more features. And this is a really common story, not just in crypto, but in product developments in general. Oftentimes you go out,

and you're really excited about something. You build out the early version of the product, but you add way too many features to it, and then it ends up like 90% of those features you don't even need. So you're like, why the heck did I spend so much time building this first version of the product? But the story here is it's really important not to build something kind of for the future, but for right now. And again, I think that kind of resonates with the story,

even the founding story of DYDX that I told earlier. I think too many people in crypto kind of build for too far into the future. And that's not to say that the future is not important or that we shouldn't pay attention to it because we are super early and we need to know or at least have some like guess about what the future could hold. It really impacts the decisions we make right now. But like that being said, like the only thing that matters when you're building products is building the best possible product for right now or maybe like for the life cycle the next year or two of the product. So I think fast forward to now, we have hopefully taken that more to heart and are really trying to build again the best possible product for right now. And I think we've learned our

lesson. Yeah, I think what I wanted to talk about a little bit is kind of the competitor almost to this order book design that you kind of pursued a more traditional look that definitely also had like a big success moment with uniswap. So basically automated market makers or AMMs, probably at that time, right, when you were building the first or second version of the YDX, they really took off because that was kind of something that seemed like needed at that time for in crypto.

Now, I guess, yeah, it seems though that a lot of things like AMMs came about because of the limitations, I guess, of the, of the blockchains to scale for, orderbook. I guess what I wanted to ask is like, what's your general thoughts on kind of AMMs? How do you see them? Is it like kind of something that's just there in passing until everything is like order book base? Or do you think they have like a space in the ecosystem? I definitely do think they have a space long term in the ecosystem. I'm maybe not, maybe proven is too strong of a word. But I think that the success of Uniswap and a lot of the other ones at this

point has proven that they are really useful and they are here to say. But I think they're kind of targeted at different use cases than order books. And I want to kind of take it back to my previous answers in terms of answering the question, why are we building order books rather than AMMs? And I think it really starts with the users that we're targeting. Again, specifically, the crypto institutional traders and the crypto prosumer traders. And these traders really demands, again, like I was mentioning before, some of the more advanced trading features, things like limit orders, things like stop orders, market orders, and all of that. And you can really only get those things on an order book, at least as kind of a first-class citizen.

Also, traders really demand kind of really deep liquidity. And just like a lot of these trading firms are not going to be so much out there, at least like market making on automated market makers, as they are on order books. So it was really, again, kind of deciding what our goal is. And our goal at UIDX is to eventually, meaning far into the future, maybe like five plus years from now, become one of the biggest exchanges or exchange protocols in crypto and really not be satisfied with just kind of being relegated to some niche markets in crypto. And I think that's really informed what we've done in a big way. We're quite literally tackling the biggest markets in crypto.

Like just Bitcoin perpetual trading in crypto is like almost 50% of the entire crypto trading market. So kind of by definition, if you want to be one of the biggest crypto trading firms, you have to play in the biggest markets. And we feel this is the product experience that we need to play in those markets and kind of target the users that at least right now are forming the bulk of the crypto trading volume. I think AMMs can be really useful. Specifically, they have been really useful in a lot of ways on providing liquidity and more long-tail markets. And I think that is really important. And that's something that was uniquely enabled by Uniswap and other AMMs.

So it's definitely something we're tracking long term. And I'm not saying in any way that like order books or AMMs are like objectively better than the other ones. Like I think it's possible at someday like you could even see an AMM on DYDX. We're not currently building that, but I'm certainly open to it long term. I think they just kind of serve different use cases. And again, the use case and specifically the users that we're targeting right now, I think an order book is a much better fit for them. And also like one of the challenges with building.

order books on the decentralized exchange is from a technical perspective. Order books require much, much higher throughput than automated market makers, probably on the order of like 100 to a thousand times more throughput than automated market makers do. Because if you think about what's actually happening on an order book, it's mostly just market makers, which are crypto trading firms, which are placing automated orders to trade on the order book. But it's not humans that are doing this. It's bots. And these bots are placing multiples of orders per second, oftentimes up to thousands of orders per second. And this is just much higher throughput to then kind of an automated market maker needs where with an automated market maker,

you just like click wants to deposit into your LP position. And that's basically it. And that has been, again, one of the things that's driven a lot of the technical decisions that we've made with DYDX. That's the biggest reason why we move to a layer two. And also the biggest reason why on the current version of D-YDX, we operate kind of what's called a hybrid exchange where there's some centralized components to the exchange and some decentralized components. Decentralized components, of course, being the smart contracts that custody user funds and operate the protocol, but decentralized components being the order book. And the main reason for this is because it requires really high throughput. This is actually kind of the main thing that we're solving in DYDX V4, where we're

building the very first decentralized but off-chain orderbook that can serve on the the order of a thousand order places and cancellations per second. And we're really only able to do this because of the technical decision that we're making to build on the cosmos chain. And yeah, we can definitely go a lot more into that, but I'll kind of stop there for now. Let's do it. No, this is perfect. Let's go into that decision. So you mentioned the decentralization of this kind of order book and matching. Were there other reasons for the switch, or was this the main one? I'd say that's probably the main one. I think there are definitely other positive tradeoffs for building your own chain.

Probably the biggest other advantage besides just the order book, which I'll talk more about, is the sovereignty of the chain, which kind of sounds like a fancy word. But basically it means that the entire chain is yours, right? So you can do whatever you want with it. It's like you can have the validators do different jobs. You can have their incentives be different. like their incentives are more aligned with the protocol, right? Because on Ethereum, like, there's no notion that Ethereum validators should care about like uniswop or sushi swap or like whatever else

is like tons of stuff running on the chain, right? But on our own chain, it's like all the validators are DYDX validators. And just as an example, that actually can help you if you're trying to solve certain things like dealing with MEV, for example. But back to the kind of critical portion of the decision, it really was what I was mentioning in terms of the throughput that's needed to serve this product that we felt like we needed to build in a decentralized order book. And kind of the decision as we thought through it, so we effectively wanted to replicate what we have right now more or less and improve on it hopefully, but replicate it for the most part, but in a fully decentralized way.

That was always really core to the ethos of what we're building and was always the plan. So we kind of looked at what we have right now, and we saw, okay, on D-Y-D-X-V-3, there are currently about 1,000 order places and cancellations per second. And actually, if you kind of look into the future with some of our ambitions, that's pretty low for an exchange. Like if you look at something like a Binance, I'm sure their throughput is orders of magnitudes higher than that. But at least for right now, let's try to be satisfied with 1,000 operations per second.

And again, kind of replicating the performance that we have in D-YXV-Rew. IDXV3 right now. We took a look around and we really keep an open mind about what technology we're building on, which for some reason is like kind of controversial in crypto, which I'll understand. But it's like we really don't care like what technology we're building on. We just like want to build a great product. So anyways, we took a look around at all the different chains we could build on. We looked at things like Starknet. We looked at things like optimism, like Solana, obviously like Cosmos, even other chains. And the core question we were asking is, can this given chain support the throughput that we need?

And ideally support this throughput with really low or ideally no gas fees for placing and canceling orders as well. So again, if you think about the use case that we're trying to solve for, it's really at least on the maker side for market makers to be able to cancel and place orders really quickly and cheaply. So we took a look around and we were like, none of them can even close to the market makers to be able to cancel and place orders really quickly and cheaply. support like a thousand places in cancel or like transactions basically per second with low gas fees. I think L2s right now have a lot of promise. And I do think they'll improve a lot long term. But again, to pull back from one of the things I said before, really the thing that we want to do is not build for like five years from now. It's to build for now and like the next year or two. And currently L2s aren't really able to support, you know, that order of transactions per second.

It's more like if you just literally go and look like this publicly available information like the transactions per second of optimism and Starknet is kind of on the double digits rather than like the triple digits right now, suffice to say. And Solana like was a contender and we did sort of seriously consider that. I know there's a lot of like Salana Fudd right now, but for what it's worth my two cents is that Solana is really novel and important technology that is working really well. And it's certainly legit in something that we considered. And they can support a really high amount of transactions on their layer one right now. But I think the whole chain is roughly around 1,000 transactions per second. And we don't have all of the chain to ourselves. So that wasn't really a

huge option either. So we kind of looked at building our own chain. And it's not so much that Cosmos in and of itself is way more scalable. It's actually not. It's probably on the supports on the order of like tens of transactions per second, which is a pretty far cry from a thousand transactions per second. But it's this core concept that we gain in sovereignty that really unlocks building things that are new. And the new thing that we're building is a decentralized but off-chain order book. It's effectively an order book that doesn't exist on chain, but does exist in the network of the validators. It's kind of similar to like, you know, when you send a transaction and it sits in the mempool for a while before being committed to the blockchain.

It's effectively using a mempool like structure to store the order book. But that's cool because the mempool is way higher throughput, like orders of magnitude's higher throughput than consensus of the core chain. And you don't have to pay gas for it, which is sweet. Like you only have to pay gas when things are included on the chain, but that's totally fine and things are only included on the chain for D-Y-D-X-V-4 when a match happens. So it's kind of like this unique product that we had where like the transactions by far the most popular ones are just placing and canceling orders, but only about 1% or less of the

orders that get placed actually match. So if you kind of map that to the throughput that we have with like the mempool, which is like thousands of transactions per seconds, you know, that's totally fine. That satisfies our 1,000 transaction per second goal at least to start and then map the actual matches that happen, which is on the order of like 10 per second to the throughput constraints of the on-chain protocol, which is like on the orders of 10 per seconds, tens per second. That works as well. So that's effectively what we're building in DYDXV4. It's something that's pretty novel. And I don't think at least has been built or at least been used at scale in a production way before. So we're pretty excited about it. We're getting pretty far into the development process

of D-YDXV-4. We have an internal test net version that's up and working and sort of placing orders on it. There's still a good bit more to build, but we're pretty excited about what exists so far and what it'll be able to launch with at some point.

I'm curious, what are the other implications of having battle leaders run this matching engine, you know, aside from now it's no longer one company, but it's like a bunch of different entities, but like, you know, what are the kind of other maybe positive or negative side effects of that?

Yeah, I mean, I guess it depends in terms of what you're comparing it to. So obviously it's like way more decentralized than what exists on DYDXV3 right now, obviously way more decentralized than a centralized exchange. And I think there are a lot of different aspects to it. One aspect is who are the validators and how many of them are there and how decentralized is that network. That network is certainly not going to be as decentralized as like the validator set of Ethereum itself.

But I think it can be much more decentralized than at least from a censorship resistance perspective than kind of the sequencers that exist on like a L2. So I think we can get to a really good level of decentralization to start in something that will only improve and grow over time, hopefully as the protocol continues increasing in volume. There's a lot of considerations, right? So you have to start worrying about MEV in a fairly big way. And that's not unique to DYDX. I think that's a big consideration on any decentralized exchange.

We're still looking a lot more into this from a research perspective. So I won't go too deep into it right now. But I think we do have some pretty interesting novel ideas. And kind of the design space of MEV, if you will, is a lot better and a lot more open when you control the validators themselves. from like a network perspective. So I think that gives us more tools to be able to tackle MEV from like a protocol perspective. So I think that's a pretty big consideration. I think it's also probably maybe obvious but worth stating like when you build this matching engine and you build this order book and the entire exchange in this decentralized way, you get a lot of the benefits of decentralized exchanges in general in terms of things like transparency.

Now the order book is entirely open. everybody can see what's on the order book. It's like provable in terms of security, like the security of the network is proportional to the decentralization of the network now. Everybody can see the rules of the chain. Everybody can see where the funds are. So things like what happened on FTX would just literally be impossible on an open protocol like this. And then a lot of the really great access that you get with decentralized exchange as well in terms of it's much easier to use from certain perspectives. it's much easier to build things on top of it.

And you can kind of build this open network that others can contribute to and can build on top of. So I think it's a really big step forwards. And I think it's also unique in the sense that we're kind of really directly trying to take on from a product perspective, what exists on centralized exchanges. And I think we're trailblazing, like technology that hasn't existed before that hopefully is making that more possible over time. And we're not intending for or even expecting for this to kind of go out and be one of the biggest exchanges in the world on day one. But I think it's a really good kind of building block to move forwards with. And we always have tried to keep a really long-term approach towards what we're building.

And our goal is to, you know, from a protocol perspective, become one of the biggest exchanges more on like a five to ten year time horizon. But this is really the best possible product we feel we can build right now. So maybe just another question on this. So you said like you see the cosmos chain scaling to, you know, like, you know, tens of transactions in terms of like, you know, on-chain transactions and then, you know, supporting kind of thousands or thousand transactions. How far do you think this can scale with this sovereign cosmos chain? And what do you think are the most important, I don't know, changes or directions that? that need to happen to reach a higher scale of it. Yeah.

So in terms of order of magnitude, I think there is definitely like a 10x improvement that we can get, probably both on-chain and off-chain at least, from just like tuning things basically. So you have to, in computer science in general, if you think about scalability, you think about bottlenecks. Okay, like what is the bottleneck that's causing, you know, my maximum off-chain order sent to the MMPO to be like, you know, a thousand or two thousand or whatever it is.

And I think one of the biggest things that we found so far that's been a bottleneck is the message propagation, which sounds like a fancy word, but it's basically like, it's okay, you have like a network of a bunch of validators, right? And then when you send a transaction to the network, it first gets sent to one of them,

but then it has to like propagate to all the other ones in the network. So it's like validators sending messages to each other, basically. And that right now on Cosmos uses a pretty, standard technology. This has actually been one of the big things we've noticed Solana has been focusing on to achieve improvements in scalability is their message propagation. So there's some learnings that we can take from there and some learnings we can take from elsewhere. And again, it's cool because we can literally build like whatever we want, right? Because we control the, like what we're building is like the software that the validators run.

So we can change things like the message propagation. And I think that can, it's just one example of something that can lead to, I don't know exactly what it's going to be, but probably like potentially a 5 to 10x improvement in scalability. Things like just improving the CPU usage of the validators themselves can improve a lot, the on-chain scalability. So there's a lot of different levers that we have, and a lot of them are pretty technical, honestly, pretty low-level programming.

But I think it's cool. And again, similar to my answer on MEV, you just control a lot more of the stack. So you have a lot more things to play with in terms of improving throughput over time. Awesome. Yeah, I think obviously also building this cosmos chain means you become kind of part of this cosmos IBC ecosystem. I guess first of all, maybe one question also, like in terms of your building like kind of modifying the stack there a bit, is there also something you're like that other cosmos chains could use like from what you're building?

Yeah, so in terms of IBC, for those who aren't familiar, it's basically a bridge more or less that's kind of built into Cosmos natively. But it's fully decentralized. It's much more secure than other types of bridges just because it's kind of, again, built more into the blockchains themselves than kind of like on top of it. And we are integrating with IBC in a big way. One of the exciting things that we help to push for is a launch of native USDC on. on Cosmos and the way that they're doing this is kind of interesting. They're basically building a sovereign their own blockchain that will run USC and potentially other things as well. But we can still on the DYDX chain, use this native USDC because you just IBC it from the native,

what's called the noble chain that they're building over to the DYDX chain. It's also cool because one other example that we're literally building right now and probably going to be using is for onboarding. So you might ask or try to solve the product question, like, how do you get USDC from Ethereum to the DYDX chain? And we're going to integrate with things like Axelar to be able to do this. But it's cool because you can use Axelar as a bridge. But at the end of the whole thing, you can end up with native USDC, which is much more secure. Like you don't have to worry about bridge hacks or anything like that.

And the way that we're doing this is you effectively use like Axelar to bridge Ethereum, USDC to actually osmosis, USDC, osmosis being just a separate cosmos chain that operates a Dex. Then you do a swap on osmosis's new stable swap feature for Axelar USDC to native USC, which will also exist on osmosis. Now like the user has, native USC, then you use IBC to move that native USC from osmosis over to the noble chain over to DYDX. And although this sounds complicated, but all the user is really going to have to see or know for this is like, you know, click deposit like 10 UCC and then like all this magic happens in the background, basically. But it's, again, cool technology and I think shows the power of composability

with things like sovereign blockchains. And I think this is a really good early use case of it. Is there also a use case that you would like tokenize some of the positions that people have on D-Y-D-X so that they can then maybe use IBC to like, let's say if there's some sort of lending chain or some other kind of chain and they want to like move it over and use as collateral or does that not work because you have to be able to liquidate on the D-YD-X chain itself? I mean, it's possible. Like, yeah, we've considered that before.

We probably won't build that initially into the chain. But I think we've kind of talked about like D-Y-D-X using other things that's made possible by composability. and I think at least so far we haven't been quite as focused on other things using DYDX as like building blocks for composability. But that is something I'm excited about for the longer term. It just hasn't really been a major focus right now. And that's still intentional because again, like if you just like listen to what I'm saying

and like the things I talk about is like mostly like let's just build like great product experience and like try to make it great for traders on DYDX. So that's been our main focus. I think one of the underappreciated parts, of composability with Defi, though. Like people talk about like financial building blocks on chain, and I think that is really cool. Again, that may be a

focus longer term. But one of the things people don't talk about as much is kind of the composability of just applications that are built like front ends, basically, that are built on top of the apps, the decentralized apps themselves. So things like

the new DAP browser inside Coinbase, I think is a great example of this, where now you can use uniswap. Hopefully, at some point, you could use like DYDX from directly inside of Coinbase. And this is really something that's uniquely made possible because the underlying building blocks are open and they're just technology. And now Coinbase has a uniswap integration. And this is unique, right? This is like really fundamentally different than what came before. Like, Coinbase is not going to have like FTX or Binance integration. It's like,

doesn't really make sense. They would just build the thing themselves. But from like an access perspective, it's really cool that a lot of these decentralized apps are just codes they can be plugged into from like a interface perspective into a lot of these different types of products. And I think that's personally, I think that's going to be like a big narrative over the next couple years as well and something I'm excited about. Yeah, that's super interesting. I guess because right now also the main front end is kind of D-YDX trading. I guess, yeah, people can interact with the cosmos chain or like the contracts there and

and integrate that or how else do you, like, do you imagine there will be like other front ends built by other parties and how do you kind of help people build that? I think there could be other front ends built by other parties for sure. And I think that will happen. Again, like with the YDXV4, all of the code that we're writing will be open source and, you know, licensed to be available by effectively like whoever wants to use it. So that means like even just for the front end that we or other parties develop, that will be open source, and that could easily be plugged into or white labeled into some third-party application or, like, independent third-party could host a version of the front-end.

There could be multiple versions running, and I think that improves, like, access and censorship resistance. So I think all these things definitely will happen. And I think they're, again, uniquely enabled by this concept that it's just technology and it's just like open-source code and anybody can run it. It's not like you need permission from us or anybody else. I wanted to ask one more question on this whole, like, cosmos chain L2 thing. I mean, I think you were very, like, clear, articulate,

like sort of explaining the, you know, your rationale and around the D-YDX and how you guys made that decision. But I'm curious, and of course, right now this is like big debate, and I think I have a sense at the moment there's, like, some narrative again gaining momentum that, you know, EF is the settlement layer and it's the ultimate thing and the L2s and everything will be there. And of course, you do have also this more thesis around sovereign chains. So I'm curious, like, how do you see this play out?

And are there particular types of applications or use cases where you think there will be more fitting in one world or the other? Or do you think that it's more going to actually go into one direction? Yeah, so I think maybe just zooming out a little bit, like one thing that muddles a lot of these discussions, and I think causes people to just yell at each other and say different things, basically, is time horizons in terms, and what I mean by that is like, okay, you would ask that question, but like, what time horizon are you asking that question on? Are you asking like, okay, what should we literally use and literally build if we want to build the best possible product in a year or two from now? or are you asking what is crypto going to look like 10 to 20 years from now?

And those are very different questions, I think. Again, like the thing that we really care about in terms of what we build right now is the first one, in terms of what can we literally use to build the best possible product right now. And it just would have been, it's like literally impossible for us to like build a product that we wanted to build on top of like an L2 for some of the reasons that I talked about before. And I'm not sitting here saying that like everybody, should build their own sovereign blockchain. Like, that's not at all the case.

Like, if I were building, like, a uniswap or, like, an NFT exchange or something like that, I would definitely just build on, like, an EF based L2, almost definitely. And there's different use cases for different things. And I think roll-ups do show a lot of promise long-term. I'm really excited about, like, Stark-based roll-ups long-term, like ZK proofs, basically. We're literally, like, the biggest user of them right now in the world. Like, where are the biggest use case of like ZK proofs in the world by like transaction volume, probably like anywhere.

So I'm excited about them long term. And they literally work. But there's a lot of limitations. And I think one of the big narratives of crypto, and I've been in crypto for like over seven years at this point, is that things take longer than people think they're going to take. And that causes people to kind of get into this like hype and bust. mentality where it's like, okay, let's like take L2s as an example. Yeah, they work right now. Like, they're great. Like, we're using them. Are they everything? Like, do they do everything? Like,

no, absolutely not. Like, and are there like answers to how we can make them better long term? Like, yes, absolutely. And great people are working them like StarCware and optimism and other teams. But technology is really hard. And like, especially emerging technology is really hard. And those problems are not going to be solved in the next year or two. They're probably going to be solved in like five years or 10 years or maybe even 20 years from now for some of them. So I think if you look at technology on different scales, you'll come up with like different answers of like what's most likely to be the thing that dominates long term. I think we've kind of seen, and it's actually been surprising to me too, this kind of narrative

of, okay, everybody like started on Ethereum basically, at least for smart contracts. And now there's been kind of this explosion to like a bunch of other chains. like there's a polygon, there's Solana, Avalanche, I don't know, there's like at least 10 of them that people use these days. And I think that'll continue for probably the next year or two, maybe three at least. And then I think we'll probably see like a consolidation more towards like the ones that are really working long term. What are those winners going to be? Like I don't know, like maybe I have some opinions. Like I'm more excited about like ZK based rollups and potentially some optimistic roll-ups than other things. But there are a lot of open questions there, right? Let's just take

an obvious one. Actually, roll-ups don't do very well on censorship resistance right now because they use a centralized sequencer. The sequencer sort of being like this service, basically, that runs the L2 roll-up and publishes the roll-up chain back to the base chain like Ethereum. But for all of them, basically, there's only one right now for like optimism and Starknet. And that's a problem and it's like people are working on solving that. I've talked to Starfer about this, talk to optimism about it. It's a hard problem. So it's probably going to take like some amount of years. But if you think about well, okay, from like DYDX's perspective, we really care about open access and censorship resistance. So, you know, we actually think in a lot of ways like what

we're building on DYDXV4 is going to be more decentralized from like an access perspective than if we we're building on like an L2 with a centralized sequencer. Again, is that going to be the case forever? Like, no, probably not. But like, that's functionally the reality of what we're building on right now. So I think it's important to just kind of take a step back and like think about like different timescales when you're asking these questions as well. And I feel like most people don't do that.

Yeah, I think it's super interesting that you guys have been like remaining so adaptable and kind of switch through the versions. Like you can even see it right. It's now V4. And it sounds like you're not kind of saying that there couldn't be like V5, obviously, and something else. And obviously, there's always like some sort of migration involved from the previous product. Can you maybe expand a bit how it will be in V3? I guess it will keep running and before we'll come up.

And maybe also like how you plan to do this migration, but also maybe some sort of tips of how to like do such a migration for other people building maybe on the wrong stack right now that want to do a similar thing, but you have learned doing it for so many times already. Yeah, I mean, I think there's different approaches to it. I'm more a fan of not so much like forced migrations from one product or protocol to another protocol, but more just kind of like launching the other protocol and then people can migrate to the new one if they choose to do so. And I think this is similar and kind of akin to how like a uniswoswad v3 versus like a uniswad v2 was launched. And that's going to be a similar approach to what we take with. D-YDXV-4 as well. So D-YD-D-X-V-3 will continue to operate for a while at least, and then D-Y-D-X-V-4

will at some point be launched, and then these things will probably run in parallel for some amount of time. There will be some process by which you could move your funds from D-Y-D-X-V-3 to D-Y-D-X-V-4. I kind of touched on that a little bit with the deposit and, like, the bridging process I was talking about earlier. But I think that's kind of been the best practice, is just to allow users to use both. Don't so much worry about, like, migrating, like, a live system because, number one, that's actually really, really hard to do from a technical perspective. Like, if we take an example of, like, the Ethereum merge, actually, like, they didn't

run both in parallel. They just, like, literally upgraded all of Ethereum at once to using proof of stake over proof of work as a separate example. But I think a lot of the technical complexity and the reason it took so long was because they tried to upgrade an in-place system. Not saying that was necessarily the wrong choice for them. Like it would have been really complicated, right, if there are two versions of Ethereum running,

but it does make it harder. So, yeah, it's going to be two versions running in parallel with hopefully some good product experience for moving funds over. Yeah, totally cool. And then I guess, yeah, what's interesting, right, there is the YDX token, obviously. And that will probably, I mean, it already has like certain utility.

And I guess that utility will extend when you transition to V4. Can you talk a little bit about what's the token currently used for how will it play with V4? Or like, yeah, how will it interact there? Yeah. So the main thing the token is currently used for is as a governance token on the DYDXV3 protocol. And similar to a lot of other DFI protocols, right, where anybody can put forward a governance proposal, people can vote on it and that's been going on in a big way

in D-YD-XV-3 right now. Actually, there's a little bit of an aside. One of the cool things that's been happening on on D-YDXV-3 with the token is pretty active governance, actually. And it's been cool because a lot of these trading firm type users that I was talking about before

have been playing a huge role in governance as well. And I think that's something that's been exciting and I think we'll continue to operate farther into the future. And this is a cool concept, right? where like if this exchange or this exchange protocol can be controlled by the users of the platform, rather than like an independent third party, like some random company,

that's like a fundamentally new concept. And I think we'll just align incentives a lot better between the users and the exchange itself. And I think that's something that's sorely needed, especially after FTX. So aside over, but kind of talking about like what's going to happen with DYDXV4. So obviously it will be an independent sovereign blockchain, and that requires layer one token for the chain itself. It's kind of undecided which token is actually going to be used as the layer one token for that chain. Presumably it would be the DYDX token,

but that's something that still needs to go through a governance vote. It's not really up to me. But yeah, that layer one token would be used for things like staking to validators, pretty similar to like an osmosis or some other chain that uses proof of stake, it would have things like voting on the new chain, like receiving staking income potentially from doing staking, things like that. So it's basically the same thing except with the addition of staking and choosing validators, which I think is exciting and is really important.

It is a core use case for the utility of the token as well. We also wanted to talk a little bit, like zoom out a little bit and talk a bit more about sort of D-Fi and C-Fi. I heard you say in another podcast that you think it might take around, you know, five to ten years for D-Fi to be able to, like, really compete with C-Fi. What do you think are the hardest challenges that have to be solved for that to happen? Yeah. I think the biggest one is probably product experience still. lot of people talk about it, right? It's just hard to build a product experience that's on par with centralized products. Just by the definition of the technology, Defy is always going

to be higher latency than C-Fi. I think we can get to a point where it's like approaching the latency you might experience on C-Fi and things like Salana and D-Y-D-X. I've done a good job with that, but it's challenging, right? And it's just harder to build basically anything in Defi than it is to build in C-Fi. I think the other thing that that Defi still really needs to solve is great product use cases. And I think we've seen some of them in things like MMs with Uniswap and others, like lending markets with compounds, things like NFTs, potentially things like derivatives, decentralized exchanges. But I think it takes time for a lot of these product narratives to play out as well. Kind of look at something like NFTs, for

example, like NFTs existed in like almost exactly the same form for like three or four years. Like I know the founders of OpenC and they were building OpenC for like the longest time and like nobody used it basically and they had no volume. And then it sort of blew up all of a sudden like in the past year and with basically the same product. And why did that happen? I think it's mostly just like kind of the maturation of that product took a little while. And then it also takes time for users to kind of wrap their heads around. What is this new concept? And then you know, these things kind of go viral. So I think it takes time for that to happen. Like, there's market cycles obviously. And I think we see the biggest influx of new users in the bull markets

all the time. And I don't have a crystal ball. Like, I don't know when the next bull market is going to be, like if ever. But probably it's going to be at least like a year or two from now. So, you know, that's going to take some time. There's been just a lot of pain in the past year for all crypto companies and defy included in that with the market down turn right. And, the FTX collapse. I think we've been fortunate to have done a good job with, like, our balance sheets and just our long-term focus at DUIDX, but some others haven't been quite so fortunate. There's negative regulatory headwinds, and there's a ton for us to do on the education side to policy makers and regulators to get them to understand, at least first of all, like what's going on. And then

for them to realize that there are a lot of positives, even from their perspective, for what they're trying to accomplish that defy can offer. I mean, everybody's like talking about this on crypto, Twitter after FTX, right? It's like, this is literally impossible. It could not have happened on a decentralized exchange. So I think kind of making that story and getting regular people more or less to understand and internalize that will take time as well. I think there's just like one of the other things that I think is really positive is there's

just a continuous influx of talent into defy and crypto. And I think that takes a lot of time as well. And yeah, I don't know, I think just all these things put together, like finding product market fits, like regulatory education, like more talents, the users understand more. This takes time and it takes longer than people think. Like, do you think the, like looking back in a few years, the FDX Collapse will kind of have helped the Defi ecosystem or like how do you see it right now? So what I told our team is I think it will certainly help Defi in the long term, again, like the five to ten year future that we were talking about. But I think maybe it helps Defi like a little bit like right now relative to CFi. But I think the damage it did to the overall space is so negative that probably, you know, plus and minus for DFI companies, it's negative in the next year or two.

just with things like obviously the prices got hammered, the confidence that traders and investors have in the space was super shaken. The negative regulatory headwinds that I talked about. So I do think it's going to be negative for the next year or two and you have to prepare ourselves for that. But I even felt this personally too. And kind of the day the FTX collapse happened, which for me at least was the most surprising day that I've had in crypto so far,

it really gave me a ton of resolve and confidence that what we're building like has to exist at some point like I don't know maybe I'm crazy but like once you kind of wrap your heads around like what a decentralized exchange is or not even what it is but like what it could be you're like this is like better right like this just has to exist like it's some form in the future long term it solves a lot of these problems it creates this open like equal be accessible financial systems for the world. And you talk about, you like look at like a lot of the mission statements for some of these companies. Like if you look at the mission statement for like a Robin Hood, for example, it's like they want to democratize access to like, I don't know, finance or whatever. And

Coinbase is at least for the longest time was they want to build an open financial system for the world. It's like, how are you going to do that? Like just look at like the technology that you have available to you. It's like, yes, that's exactly what we should be doing. But this is the technology that we should be building on to be able to make that future possible. And again, I think it'll take longer than people think and people will be disillusioned, right? And I think people have been disillusioned, especially with defy in the past year or two, because, and this happens with all these technologies, it's always the same thing. It's always like people are right, basically, when you start, it's like, we have the right idea. We don't know exactly what it's

going to be, but it's like, okay, yeah, we kind of get it, at least for the people that understand that we can build these open financial platforms, and that's fundamentally new thing. Like, this is amazing. Like, everybody should use this. But then it's like, oh, wait, like, this sucks at least for the first couple years. Like, this is, like, you told me this is going to be a new financial system. And I can, like, barely even connect my metamask wallet to, like, this website. Like, what the hell is going on here?

But then, you know, after years of building, and that causes disillusionment and in crypto, this is even more negatively reflexive, right? because there's prices that are attached to all of these things, and then the prices start going down, and everybody's like, oh, this is like an even bigger piece of shit, and it just causes these things to get higher and get lower than they probably otherwise. What if everything weren't priced? But I think it takes a while, and there are hype cycles,

but at least to me, like the FTX collapse, really gave me a lot of confidence that this has to exist long term, and it's something I'm excited to be at the forefront of helping to build. Awesome. Yeah, that's really cool. We do have like a question actually in the stream that is somewhat related, I guess, to this whole discussion. So I think we kind of mentioned, yeah, you're building obviously the product on this decentralized platform to enable this future, right, that Defi tries to promise. Now, is there also like something you're doing? I think you mentioned it briefly, like kind of educating the regulators or like interacting with regulators. Is this something the IDX foundation or someone is doing? Or are you? you like more focused on the product side? And if you're doing like what, can you expand a bit and explain maybe what you're, how you're approaching that? Yeah, absolutely.

So just to be clear, like the bulk like almost entirely, uh, the company and the project more broadly is focused on product and just building do IDX V4. It's important not to get too distracted by like any of these things in part. In particular, I think you kind of saw that happen in a pretty negative way with FTAX. like one of the biggest things like SBF was doing for the year prior to the collapses, like talking to regulators and stuff. And that wasn't necessarily bad, at least from their perspective. But I think you saw some of the product progress on FTX stagnates for a while because

of that. So we're really focused on the product itself. That being said, as I mentioned before, I do think this is a really important thing that needs to be done going forwards. It's just critical that developers can build. open source software in a legal way going forward. I think there's sort of this meme, at least on crypto Twitter that I see a lot, that's like, oh, just like everybody should be anonymous and like, we'll all be fine.

And like, you know, who cares about the law? Like, we're all just like move to someplace they can never catch us or whatever. And like maybe some people will do that. And like some people have done that empirically. I think it probably won't work for the bulk of people once like regulators actually start caring about this stuff. but even if it works for a very few amount of developers the thing I said that we need before is more talents into crypto right

and it's like how we're going to get like enough developer talents into crypto if like the answer is like oh yeah everybody's just like anonymous and like as perfect opsack and moves to like Malta or whatever like it's not going to happen like we're not going to build like this future that we all want to exist and that we feel like is better financial system for the world unless developers can build these things in a legal way. So that's kind of just motivating the problem. I feel like it's important.

It's like from my perspective, it's like literally impossible for us to build this future without doing this. So I think it's important. And we are starting to do this at DYDX. We're certainly not like the biggest player, like the biggest voice for DFI. There's a lot of other great people like the blockchain association, like the DFI education fund, other projects like I know UNISwap is doing a good job. in this in terms of regulatory outreach.

I think it's hard, right, because there's tons of different regulators and tons of different jurisdictions for us. Our team is based in the U.S., so it's primarily, like, U.S. regulation focused. So, like, we do go out and talk to regulators, but the thing we've really started trying to do more of is going out and talking to policymakers, which is like basically like senators or like their staff or like people that are actually in government. because the kind of realization that we've come to and a lot of people have come to over time is that a lot of the laws that exist right now

just literally don't make sense for crypto. It's like, let's maybe just take an example, like an obvious example. Like a lot of the derivatives regulation in the U.S. is based around transparency, which makes a lot of sense, right? Like you want to make sure as per like FTX that when you're trading on a derivatives exchange,

you know where the customer funds are. That's like pretty basic. Like we can all agree. that that should be the case. So a lot of the regulation focuses around that and there's like all these requirements and you got to have like people and submit audit reports and all the stuff, like whatever. But in Defi, like it solves it. It's like, go look on EtherScan. It's like 100 times better than like any auditing or like licensing we could ever do. So like why the heck like do you have to submit like an audit report for where the customer funds are with Defi? Like just go

look at like ether scan right now. Like it doesn't make sense. And there's like even worse than that like a lot. there's like just not a lot of what, there's not any ways to comply with like certain regulations, uh, for a smart contract because the regulations just weren't drawn up with that in mind. So like, you know, realization too,

like we feel like it's obvious that there needs to be new policy, um, not just like new regulation for some of these things long term. And policy also is just like much more open ended, right? Um, I think regulation a lot of times comes down to what literally are the laws. It's not really so much regulation.

job to like change the laws, right? More of their jobs to interpret the laws and enforce the laws. But if you're sitting here saying, well, like some of these laws just like don't really make sense, then you have to go to the root of the problem, which is policy. And I think policy, unfortunately, takes a really long time, right? But we still feel like it's worth doing it because of the time horizon that we're building for. And, you know, we want to ensure that developers can continue to build a lot of these things

in a legal way, like specifically. I've done some meetings with like staffs of various senators, things like that. We have just recently hired a really great head of policy that we're excited about. And I think it's great that there are so many third parties that are out there that are kind of making the case for defy to policymakers. And that's absolutely important and should consider. But I think one of the things that hasn't happened so much, but I think is really important going forwards is literally the people that are building

defy to go and make a name for themselves and like go and be the ones that are out there like educating people on what is defy like what like how could it possibly have solved the like ftx collapse what are the risks like how should it be regulated and i think just like starting that conversation is really important and something that we have really just started thinking about a dydx but i think we'll become more importance over time oh amazing well final question and there's maybe a little bit like different topic so aside from dydx and critical What are you most interested in at the moment? Yeah.

I mean, maybe in crypto. I don't know. I've just been working in crypto my whole career, so I barely even think about anything else or don't know that much at least. But I'm excited about NFTs. I think NFTs are something that's newer, right? And I think they probably will experience this trough of disillusionment. Every new technology goes through.

But, yeah, excited about things like OpenCee and Magic Eden. And excited for a lot of the infra work to continue on crypto. When we talked about some of it, right, just like the L2 is continuing to improve different L1s. And yeah, I mean, obviously I'm excited about Defi. I don't have any like super hot takes for you because I feel like one of the things in crypto is people just get excited about like random stuff that doesn't make sense to. And I think there's only like three or four narratives that actually do make sense to me at least in crypto, which is like, like defy NFTs, like infrastructure and like data basically. Like I don't find anything else super interesting.

That's not to say like nobody should build anything besides that. Like I would not have included NFTs in that if you had asked me a year ago and I would have been wrong. But at least for the things that are proven, those are probably the most interesting to me. And I think we have a long way to go on them. Cool. Amazing. Well, thanks so much, Antonio, for joining us.

I really appreciated how like succinct to the point like, you know, you kind of answer the questions and like went through kind of the YDX's history and what are you guys doing? Super excited. Exciting and you know I'm really excited to see before what are you guys going to do

in terms of building that chain and also in terms of just you know pushing forward the app chain thesis, the cosmos SDK you know, the tendament, like so many different things that I think are going to be like super excited. So yeah

excited to see what you guys are going to build and how it's going to develop. And hopefully we can resume that conversation at some point in the future, maybe when there's like V5 or something like that coming up. Sounds good. Well, thanks so much for having me on and great conversation. And thanks so much for the listeners for tuning in. And we look forward to being back next week. Thank you for joining us on this week's episode.

We release new episodes every week. You can find and subscribe to the show on iTunes, Spotify, YouTube, SoundCloud, or wherever you listen to podcasts. And if you have a Google home or Alexa device, you can tell it to listen to the latest episode of the Epicenter podcast. Go to epicenter.tv slash subscribe for a full list of places

where you can watch and listen. And while you're there, be sure to sign up for the newsletter, so you get new episodes in your inbox as they're released. If you want to interact with us, guests or other podcast listeners, you can follow us on Twitter.

And please leave us a review on iTunes. It helps people find the show and we're always happy to read them. So thanks so much, and we look forward to being back next week. No.