2953: AI-Powered DeFi: Insights from VirtuSwap's Founder

[00:00:01] Have you ever wondered how technology can transform your relationship with finance? Well in today's episode, the founder of VirtuSwap, which is a decentralised exchange platform, is going to share how it is on a mission to not only reshape the DeFi landscape,

[00:00:17] but also leverage AI to optimise trading and liquidity allocation. And with a strong presence in the DeFi ecosystem and partnerships with notable platforms such as Unizen and OpenOcean, VirtuSwap promises significant gains for traders and liquidity providers alike.

[00:00:37] And today's guest has also got a wealth of expertise, having spent more than 20 years as a professor at prestigious universities worldwide, currently leading blockchain research at Tel Aviv University. And today he's going to be sharing his insights on the innovative technology behind VirtuSwap.

[00:00:55] And we'll also discuss the pivotal role of AI in decentralised finance, and also the exciting future of DeFi in an AI-driven world. So how can these advancements change the way we interact with finance and investment? Now obviously hosting a daily tech podcast comes with its challenges,

[00:01:13] and so I'm incredibly grateful to our sponsor for their essential support. Defence contractors face immense pressure to comply with CMMC 2.0 security standards, something I'm hearing more and more about. Now finding a secure, easy-to-use file sharing solution that meets those CMMC 2.0 guidelines is quite a challenge.

[00:01:35] And the federal government and federal system integrators supporting the Departments of Defence have similar compliance requirements for improving cybersecurity and data protection. So why not get on the faster path to CMMC 2.0 compliance with a company called Kiteworks,

[00:01:52] where you can leverage the same zero-trust framework used for federal requirements and a platform that offers secure file sharing tailored for the defence industry's unique needs. With granular access controls, encryption and DLP integration, Kiteworks exceeds legacy tools in security capabilities.

[00:02:13] Kiteworks is a FedRAMP moderate authorized to give you that peace of mind. So why not accelerate your CMMC 2.0 compliance and address federal zero-trust requirements with Kiteworks' universal secure file sharing platform that is made for defence contractors? Simply visit kiteworks.com to get started.

[00:02:33] Buckle up and hold on tight as I beam your ears all the way to Tel Aviv, where today's guest is waiting to join them. So a massive warm welcome to the show. Can you tell everyone listening a little about who you are and what you do?

[00:02:48] So my name is Evgeny Landris. First of all, thank you very much for bringing me to the show. It's a pleasure to be here. So I sort of wear two hats. My first hat that I've been wearing for close to 20 years now is that I'm a finance professor.

[00:03:02] So I spent most of my career at Boston University in the US and about four years ago, moved back home to Israel. And I'm a professor at Tel Aviv University and also head of the Blockchain Research Institute. I guess we'll talk a little bit about that later.

[00:03:16] And also for the last, I'd say two and a half years or so, I'm as founder of a startup in a blockchain or more specifically DeFi scene, which develops and actually operates at this point a decentralized exchange of crypto assets,

[00:03:33] which is quite different from the other exchanges in that space. I mean, this space is full of competition, but I think we have something unique that we bring to the table and I'm very happy to talk about that in the next half an hour or so.

[00:03:47] And when I was doing a little research on you, when I was looking up and trying to learn more about Virtus.Swap, that thing that sets you apart really in a world of places like Coinbase,

[00:03:59] which is almost like the AOL of crypto that it seems to have turned into, Virtus.Swap is a decentralized exchange platform specifically for those smaller cap assets. So I've got to ask, how did the idea of Virtus.Swap come about? What inspired you to create it?

[00:04:16] So like many in this space, I've become interested in decentralized finance in general and decentralized exchanges more in particular in the summer of 2020. So that's what's called the summer of DeFi. So that's when basically DeFi became a thing. So DeFi was there before as far as 2017 and 2018

[00:04:34] with the protocols like Bancor and Uniswap being the trailblazers in this space. But really people started using DeFi and DEXs in the summer of 2020. So I really started to dive into this topic. Initially from the academic kind of standpoint, I've written a few papers previously about ICOs,

[00:04:53] which were a big thing in 2017 and 2018. But I was getting more and more interested in blockchain applications and economics and finance. And DeFi seemed like probably the most useful application on blockchain so far. And I think it's still the case for the most part.

[00:05:10] And one thing that I, as I was trying to understand more how DeFi works, there is lots of advantages to trading technologies that DeX has employed. We can talk about those. But one of the things that I noticed is that a big impediment

[00:05:24] to DeFi taking over the world, which is still there, is that the liquidity in DeFi is first of all small and secondly very fragmented. And basically that means that trading can be inefficient, which means basically expensive relative to alternatives such as Coinbase,

[00:05:42] that you mentioned, such as centralized exchanges, where liquidity is more concentrated. And this problem is especially acute in smaller assets, which we call the 99% assets. Pretty much anything outside of the top five of any given blockchain are going to have very small and very fragmented liquidity.

[00:06:01] And that basically means that for every trade that you're going to try to do in those smaller assets, you're basically going to do a couple of trades, what we call triangular or indirect trading. You're going to exchange asset A to asset B,

[00:06:17] and asset B to asset C, and that's going to be your trade. Now in the process, you're going to pay the cost of trading twice or maybe three times. And the cost of trading includes things like the pull fees that you pay for liquidity providers,

[00:06:31] the price impact of the trade, that's something that's common to any type of trading, gas costs, something that you pay on a blockchain for executing trades. And so if you pay this twice or three times, this becomes really expensive. The trading becomes really inefficient.

[00:06:46] So that's basically the problem that I and my co-founder in this American Virtus.book, Ronnie McKayla, did identify as one of the more important ones in this space. And that's something that we set out to try to solve. I love it.

[00:07:02] And for any techies that might be listening that want to take a little look under the hood, can you tell me a little bit more about your unique technology that Virtus.swap employs and how it's helping to transform the way that people are engaging with finance?

[00:07:15] I'd think you can share around that. So the problem is that to perform many trades on existing blockchains, our estimates, and we've done quite a lot of data work, we'll talk about that as well, is about 25% to 30% of trades belong to the category that I described, right?

[00:07:31] So trades involving smaller assets that are indirect. And so basically I thought about what would be a way to make those trades direct, right? So make sure that if you want to exchange even smaller assets,

[00:07:43] you're going to be able to do it without jumping through multiple hops, right? And the idea that we came up with was basically trading through what we call reserve pools, right? Or reserve routes, right? So let me explain in an example how that works, right?

[00:08:00] So let's say that you want to exchange asset A for asset B. And let me give an example of particular assets. Let's say that you want to buy, for example, Ether and they buy Link.

[00:08:15] That's the token of Chainlink, one of the big protocols in the DeFi space, right? And let's say that a pool like this does not exist, right? So you cannot go to the Ether Link pool and do the trade that you want.

[00:08:27] So you'd have to go this triangle route, so Ether to, for example, USDC, and then USDC to Link. Now, in our case, what we do is we constructed this new kind of architecture where pools can accept what we call reserve assets, right?

[00:08:43] So what is going to happen, right, is that you're going to go to the pool that has the asset that you want, for example, Ether, right? So you're going to go to Ether Bitcoin pool, okay?

[00:08:55] And you're going to take the asset that you want out of that pool and you're going to pay using the asset that you have, in this case, Chainlink, okay? Now, this Link is not part of the two assets of the pool, right?

[00:09:08] So a normal pool would not be able to accept this asset, right? Our pools are different because they allow to accept reserve, right? So reserves, you can think of those as basically safe deposit boxes attached to the pool, right?

[00:09:21] That are initially empty, but they can accept stuff into them, right? And part of this, or an example of this stuff is this Link token, right? Of course, we need to kind of think about what is the price, right?

[00:09:31] At which rates those assets are going to come to the reserves and we do this using triangulation of prices within our decks. It's very important we don't rely on external oracles, right? So pieces of information coming from outside of the protocol, right? So that significantly reduces the risk.

[00:09:49] What's going to happen is that they're going to basically display to the user what we call a virtual trading route, right? So the user would think that he'd actually be able to trade directly Ether for Link, right?

[00:10:03] What's going to happen in reality is that Link is going to go into the reserve of the Ether or Ether Bitcoin pool, you know, in this example, right? So that makes it the first part of the trade. And for a trader, it's a fantastic idea, right?

[00:10:17] Because basically for a trader, pretty much every trade is a single trade, a single hop trade, right? Without the necessity to pay duplicate fees and duplicate price input. Now what's going to happen under the hood is that those pools are going to accumulate reserves, right?

[00:10:33] And those reserves are risky, right? So the liquidity providers are not that keen on keeping those reserves. They're not keen on having larger reserves and also for those reserves sitting there for a long time. So the pools are going to basically try to exchange

[00:10:48] those reserves between themselves, right? For example, this pool in the example that I provided just accepted the Link token, right? So we're going to search if there is some pool where Link is one of the two main assets, for example, in Link USDT pool

[00:11:04] that has, for example, Ether or Bitcoin as a reserve. If that happens, if they're able to find a pool like this, right? Then basically those pools are going to be able to exchange those reserve assets between themselves, right? And each of those two pools,

[00:11:19] the asset is going to go from a reserve of a different pool to one of the main assets of the pool, right? And so we're going to search for this constantly, right? And as long as the trades are relatively balanced, right?

[00:11:30] Then basically different pools are going to have different reserves that could be exchanged among each other. And what's nice about it is that this type of activity does not involve pretty much any cost, right? There is no pool fees, there is no price impact,

[00:11:44] a little bit of gas, right? But that's about it. You can think about this basically as pools completing the trades in the background, right? So the user does the first part of the trade, right? And then the trade is later on at some point completed

[00:11:59] when the pools exchange the reserves among themselves. And we've done quite a lot of simulations basically to show that as a result of this exchange, that it's automatic, it runs in the background all the time, right? So there are some small incentives

[00:12:11] for whoever basically is initiating the exchange of reserves. But those are sufficient to ensure that this exchange happens constantly. And basically the conclusion is that the average reserve that pools have at any given point in time are quite small. They're on average something like half a percent

[00:12:30] of the value of the pool, right? So the added risk to the liquidity providers is there, but it's not large, right? And so in exchange, liquidity providers get much more returns for their liquidity because as a result of this indirect trading, those pools are exposed

[00:12:45] to a lot of different types of trades, right? So for example, in a normal debt, if you provide liquidity into Ether Bitcoin pool, you're able to serve Ether Bitcoin trade, right? In our case, you're able to serve, this pool is able to serve any trades involving Ether,

[00:13:00] for example, Etherlink, or any trades involving Bitcoin, so Bitcoin link, right? And that obviously increases the overall amount of the volume of reception both through the pools and increases the returns to liquidity provision as a result. So it's sort of a win-win for both traders and NLP, right?

[00:13:18] So less liquidity providers are exposed to more trades, and so they get higher returns to what they do. Love that. I think we should also mention this, so much hype around AI at the moment. So I've got to ask the question,

[00:13:30] in what ways do you see AI in enhancing the capabilities and indeed efficiencies in decentralized finance? And is there anything you're doing at VirtuSwap to leverage AI in your operations? I appreciate probably not too much you can share,

[00:13:46] but where do you stand on everything that you're saying here, and what are you doing with it? Yeah, I'm happy to share as much as I can. So basically so far, what we've been talking about the last 10 minutes is kind of the financial engineering part of things, right?

[00:13:58] And this is one of the two legs on which VirtuSwap stands. The second leg is, as you mentioned, AI. So as far as over two years ago, we started developing what people now call AI agents, right? And before the term existed, it basically meant optimization.

[00:14:15] But in our case, optimization of liquidity. So we noticed early on while doing our simulations that not only liquidity is fragmented in DeFi, but also it tends to be there not in the right places or not always in the right place, right?

[00:14:31] So sometimes you're going to see a lot of liquidity in pools of assets that don't really have a lot of demand for trading, right? And sometimes you have insufficient liquidity or most often you're going to have insufficient liquidity

[00:14:44] in places where there is a lot of demand for trading. Now, liquidity providers, for the most part, are not sophisticated enough to know exactly where to put their money and to do these reallocations automatically. And there are nowadays some protocols that help them do this, right?

[00:14:59] But for the most part, right, liquidity allocation is inefficient, right? So we basically started by saying the following, right? Let's try to help people put their money where there is going to be the biggest kind of bang for the buck. It's where the returns to the liquidity provision

[00:15:14] is going to be the largest. Of course, we cannot control this process, right? We cannot take people's money and put it into where we think it should be because we never have custody of funds, right? We can only try to incentivize the process, right?

[00:15:27] And so we developed basically a system of rewards where we reward different pools with our own proprietary VRSW token. So we reward different pools in such a way that we think will move liquidity towards places where we think the liquidity is needed most, right?

[00:15:47] So we think that a particular two assets are going to be traded heavily, right? So this is the pool that we are going to incentivize with our token, right? So our AI kind of agent, the first one that we built, which we called Minerva, basically has two elements.

[00:16:02] The first one is the predictive part, right? We're going to try to predict the trades that are going to happen, for example, over the next week, right? So we'll take a lot of information from past trades, from current market conditions, right?

[00:16:14] And we perform different kind of predictive models in order to try to see the whole distribution of trades over the next year, the next week, right? And then given this prediction of ours, we say, well, how should we allocate a given amount of rewards?

[00:16:28] Let's say we have $10,000 of rewards to allocate to liquidity providers over the next week. How do we pick the big bits and pieces of this $10,000 and put them into places, into pools that we want the liquidity providers to provide liquidity in the most efficient way?

[00:16:44] So it's a very multi-dimensional problem, right? It's not a very easy problem to solve. So we've spent quite a lot of time developing different types of heuristic algorithms to solve it. I think we so far kind of succeeded in doing something that seems sensible, right?

[00:17:00] So that's kind of what we have right now. And I'm not sure how much time we have, but we're working right now on multiple, well, several additional AI agents that are going to make this process even more efficient and also potentially decentralized, right?

[00:17:15] So it's very important to notice that right now everything I described in terms of the AI part is sort of centralized, right, and a black box from the user's perspective, right? So we have those algorithms. We kind of designed the reward distribution, right?

[00:17:30] And other kind of respond to those incentives. But if you really want to follow the blockchain or DeFi ethos, you want things to be decentralized, right? So we're working right now on a system that we call decentralized Minerva, which is basically an ecosystem of different components, right,

[00:17:45] of basically what we call solvers, right? So people producing or proposing different solutions to the problem that we pose, which is distribution of rewards between pools. And we think we have a decent solution, but I'm sure it's not the best.

[00:17:58] I'm sure there are people who can't come up with better solutions. So basically, people are going to submit their possible algorithms, right, for optimal liquidity distribution. There are going to be solvers that are going to basically run those algorithms, right, and supply the results, right?

[00:18:13] And of course, we need to make sure that those results are truthful, right? The whole kind of system is both incentives for truth-telling, right? And then basically we're going to be choosing, the system is going to be choosing the best algorithm, right, the best distribution of rewards, right?

[00:18:26] It might be ours, it might be someone else's. We don't care, right, as long as this is done in the best possible way, right? So that's one thing that we're working on right now, but we're working on several other kind of AI agents,

[00:18:38] most of them related to optimization of liquidity, but not just. A few years, well, maybe they've been a part of a Discord, Telegram, or WhatsApp, crypto group chat. And sooner or later, somebody's going to say gas fees, and there'll be a whole debate that sparks from that.

[00:18:55] And one of the things that put you on my radar was VirtuSwap's impressive achievements in reducing trading costs and also increasing returns for liquidity providers. So what are the key innovations that enable some of these benefits? Yeah, so as to the gas fees, right,

[00:19:12] the reason ours are smaller is just that because we have fewer kind of parts of a fixed transaction, right? So since we have less ops, we have one transaction instead of two or three that would have happened on typical debts.

[00:19:25] I mean, that reduces all the costs, including gas costs, right? And it has to be mentioned that despite the fact that we have one transaction instead of two or three, right, the side that we use, so those reserved type trading increases the gas costs a little bit, right,

[00:19:39] relative to very simple transactions. So that's something that we have to acknowledge. However, the savings in the other parts of the cost, which is the price impact and then the pool fees, right, which are actually more significant in layer two blockchains on which we operate.

[00:19:53] So right now we're operating on Polygon and on Arbitrum. So those savings basically trump the slight increase in gas costs that are there because of our slightly more kind of complex transactions, right? So I think that's kind of the main driver of the reduced trading costs for traders.

[00:20:11] As to the LP, right, there are two elements, right? And I think we briefly discussed both. The first one is that pools are exposed to more trades, right? So pools can serve more trades. Every trade comes with a fee. The fee goes to liquidity providers,

[00:20:25] so clearly LP is benefit from that. The second part is AI-based optimization, right? So if we kind of induce liquidity providers to put their money into pools that are likely to receive higher trading volume, right, this is also going to increase returns.

[00:20:43] And again, we have quite a substantial body of real market kind of simulations of how things work. And our conclusion is that liquidity providers' return can increase by a factor of almost five, right? So 400% higher than the similar liquidity deployed on traditional DEX. And that clearly is important,

[00:21:02] especially given the fact that there are many kind of research studies, academic and otherwise, that show that on average, liquidity provision is not a profitable business, right? So liquidity provision is subject to a significant risk that's called the impermanent loss risk that basically is the loss of liquidity

[00:21:21] to arbitrageurs that constantly operate in the market. You might have heard of the term MBV bots, but basically bots that scan the ecosystem and try to capitalize on changing exchange rates between assets, right? So if you consider the impermanent loss risk,

[00:21:36] the returns to LP is typically are not positive. Now, if you increase those returns by a factor of two, three, four, five, right, then it will make liquidity provision a profitable endeavor, right? And in the long run, can increase the overall participation of people in this ecosystem, right?

[00:21:54] So basically increasing liquidity overall. Love that. And I'm curious if we look to the future, the speed of change is moving so fast at the moment. How do you envision the future of DeFi, especially in the context of rapid advancements in AI,

[00:22:09] how these emerging technologies begin to converge and what role do you see Virtus.Wap playing in this future as well? Yeah, yeah, yeah. That's a great question. So if you ask me this question in the fall of 2020, right, or even 2021, right,

[00:22:26] I would basically say that this space is going to explode, right? You know, from a hundred billion dollar TVL and going to reach a hundred trillion, right, like the global stock markets, right? You know, the sky's the limit.

[00:22:37] Now, you know, it's already three and a half years since then and it hasn't happened yet, right? So the TVL, the kind of the total liquidity in the DeFi market is less than it used to be in the summer or the fall of 2020, right?

[00:22:51] And there are reasons for this. There are many of them, I'll mention two. The first one is that kind of operating in DeFi is still very much not user-friendly, right? So for anyone who experienced kind of doing stuff in DeFi, you've seen things like transactions fading, right?

[00:23:10] The crypto wallets being kind of not operational part of the time. The UI in most protocols is just terrible, right? And there's also a lot of kind of financial kind of risks, right? So you can send transactions to non-existent wallets, right?

[00:23:25] Then basically that's a sure way to lose money. But it also risks security breaches of different protocols. You hear about them pretty much every week, right? Some protocol has been breached and some funds have been stolen, right?

[00:23:36] So there are a lot of kind of things that a typical Web2 user is not used to when operating in Web3, right? So the people who are using DeFi are still sort of mostly kind of the enthusiasts of the space, right?

[00:23:49] It's not the general population that operates in DeFi, right? The second thing that I think is crucial, crucial impediment to this kind of ecosystem exploding is regulation, right? So regulation in most countries, in pretty much all the vast majority of countries,

[00:24:06] is either not very coherent or inexistent, right? So there is no good regulation and there is a large kind of regulation uncertainty, what's going to be the future regulation. And in particular, regulation right now precludes things like tokenization, right? So representation of real-world assets on the blockchain, right?

[00:24:24] For example, I cannot really trade Apple stock right now on the blockchain. I'm forced to do this on NASDAQ, right? Or wherever it's traded. If we could find, if regulators would find a way to basically ensure fair representation of real-world assets such as stocks, bonds, real estate, art,

[00:24:42] any type of asset on the blockchain, right? That would basically increase the potential space of trading from just the crypto assets, which is the case right now, to all the other markets, which are two to three orders of magnitude larger than the crypto market, right?

[00:24:57] So if and when this becomes possible, I think DeFi is going to really see a huge growth, right? Because DeFi actually provides a lot of advantages relative to traditional kind of trading mechanism, right? Or trading venues, right? It's decentralized. There is no risk of custody in DeFi.

[00:25:16] There is huge interconnectivity between different protocols, right? So it's very easy to do things like arbitrage, right? There are unique financial solutions such as automated market makers on DEXs or flash loans, that's something that's inexistent in traditional financial markets.

[00:25:32] So if only we can enable kind of all the financial innovation in DeFi and it's happening with a tremendous space, as you noticed, to be able to apply it to a larger population of assets, larger than just the crypto assets, which is generally the case right now,

[00:25:48] this will be the moment that DeFi is going to win basically. Now, how do you virtual swap as part of this revolution? So, of course, virtual swap is just one of the many DEXs or many DeFi solutions that exist right now.

[00:26:02] We do provide large advantages to a particular type of trading and I think this advantage is going to become more and more substantial as more types of assets or more assets are added to the DeFi ecosystem. So that's one.

[00:26:17] And the second, AI has become kind of really a cornerstone of our lives, right? All types of AI, right? So you can start it from large language models, but not just. And so doing things right, doing things in the most efficient way,

[00:26:31] in particular, putting liquidity in the ball, assigning liquidity to places in the most efficient way, I think is going to be a big deal, especially as the amount of trading in DeFi is going to start rising, right? So I think we are at the forefront, I think,

[00:26:47] in this space of optimizations or AI type innovations. I think there's going to be a significant place for virtual swap as part of this DeFi revolution, and I really hope that it happens sooner rather than later. I think there's no doubt it's going to happen someday.

[00:27:02] I just hope that it's going to happen within a year or two and not 10 or 20 years. And of course, when we're building that future, partnerships are crucial in the DeFi space, especially around increasing adoption, bringing more big business into the space. And obviously your community want regular updates,

[00:27:20] often you get these partnerships, but you're locked down to NDAs. It's an incredibly tough balancing act. I don't know how you survive in that part of the world, but are you able to discuss the significance of some of your recent collaborations with entities like Unizin

[00:27:34] and OpenOcean and so many others and how they ultimately contribute to your success at virtual swap? Sure, sure. So I think one word that I mentioned a couple of minutes ago that I think is crucial to DeFi is interoperability, right? So different protocols can work together, right?

[00:27:51] Pretty much seamlessly from a user's perspective, right? And what it means from our side is that when you want to do a swap, right? When you want to kind of swap two tokens, typically you're not going to go to virtual swap UI UX, right?

[00:28:07] So I think we have a nice and kind of functioning UI UX, right? Where people can go and do things on it. But most often you're not going to go there. You're going to go to what's called an aggregator, right? So an aggregator is basically a platform

[00:28:22] that scans all the different kind of DeFi protocols, Dex's, Decentralized Exchanges in particular, and basically tries to ask a question, right? So let's say a user wants to swap 1,000 USDC for Ether. Okay, where should the user do it, right?

[00:28:38] What would be the way to maximize what we call the amount out, right? So the amount of asset that he wants to buy, right? That he's going to receive as the result of this trade. And most often, right, the best thing to do

[00:28:52] is to break the trades into small pieces. Some pieces are going to go to Uniswap, right? Some pieces are going to go to Virtus.org. Some pieces are going to go to another exchange, right? Again, with the goal of maximizing the output,

[00:29:04] the reason it makes sense to break trades into pieces, right? Is that enables to basically reduce the price impact of a trade, right? You can think of basically taking all the fragmented liquidity that sits in different pools and different protocols and bringing them together to one trade, okay?

[00:29:19] And when that happens, obviously, that creates the best outcome for the trader, for the user, okay? So it's extremely important for Virtus.org, as for any other kind of Dex or DeFi protocol, to be part of this integrated system, right?

[00:29:34] And that basically means being integrated into the aggregators, right? Because most people, again, are going to go to an aggregator, right? And try to search for the best route, right? Or ask that reader to search for the best route.

[00:29:47] So integrations that with the entities such as OpenOcean and Odus and Uniswap, right, and many others that we're working on basically is our way to become part of this ecosystem, right? And have trade come from the aggregators to Virtus.org.

[00:30:02] And we can see right now that larger and larger fraction of trades that are settled on Virtus.org, right, basically executed using pools at Virtus.org, are directed from the aggregators, right? And this share is going to keep growing as we continue

[00:30:17] kind of collaborations or integrations into those aggregators, both on the SAIT chain and what we call cross-chain, right? So we also have integrations cross-chain aggregators that basically allow swapping from one asset on, say, Polygon to another asset on Arbitra, right?

[00:30:33] So both types of integrations are super useful for us and we're working hard on ensuring that we have as many as we can as soon as possible. It's a very competitive space, right? So basically when we talk to aggregators, right,

[00:30:47] we really need to convince them that our solution is unique and somewhat different from existing ones, right? Because otherwise it would not make much sense to integrate us. So that's kind of the battle we're fighting right now.

[00:30:59] And on a more personal level, I was doing a little research on yourself before we came on the podcast. And given your extensive academic background and indeed your current roles as the head of Blockchain Research Institute at the Tel Aviv University, I'm curious,

[00:31:16] how do you balance academic research with your entrepreneurial ventures and how do they inform each other? They seem to be a lot of synergies there, but I'm curious what you say. Yeah, so you're exactly right that there are a lot of synergies there.

[00:31:30] I'll give you one example, right? So as part of Verges' work on AI agents, we kind of analyze lots of data, basically. Lots and lots of terabytes of data trying to optimize things, right? Now it happens so that for academic questions, right,

[00:31:49] you also can answer many of them by analyzing a lot of data, right? And in my case, it is basically the same data, right? So the questions that I ask in my academic research and the questions that are asked at Verges' work might be different,

[00:32:03] but the data and the tools and much of the analysis are actually quite similar. I'll give you one example, right, of a very kind of broad question in economics and finance, right, that can be answered with the type of data that we analyze at Verges' work.

[00:32:19] So there is a pretty well-known phenomenon, right, that financial market development is good for economic outcomes, right? So financial markets typically lead to better economic growth and good things for countries in which markets are developed, right? Now it's not very clear what is the reason for it, right?

[00:32:38] So people have basically proposed many, but one seemingly important one is better capital allocation efficiency. So developed financial markets allow to allocate capital to different activities more efficiently, right? And that clearly promotes kind of good economic outcomes, right?

[00:32:54] However, it's not very easy to try to find empirical support for this conjecture, right? So traditional data and traditional markets typically are not enough to be able to answer this question. Now what we do is we use DeFi data, in particular data on what's called concentrated liquidity pool,

[00:33:12] to develop DeFi-specific measures of both efficiency of markets, right? And we can do this at a very granular kind of level, looking at each liquidity pool as a market, right? So we can build a measure of market efficiency

[00:33:27] and we can also build a measure of capital allocation efficiency, right? Again, at a very granular level, right, both cross-sectional and in the time series. And that allows us to answer a question, right, whether market efficiency, which is one of the more important characteristics

[00:33:46] of how we develop financial markets, right, whether it contributes to capital allocation efficiency, right? So this is a question that we can answer, try to answer using DeFi data, a question that's very difficult to answer using data from other kind of more traditional markets.

[00:34:01] And we actually argue that there is some external validity to the find, right? So there is nothing particularly crazy about this element of DeFi markets that would not allow transporting our evidence from that market to other kind of services, to other markets, right?

[00:34:16] So this is basically one example of an academic research, right? So trying to answer kind of a very general kind of economic or finance question, right, that uses the same data, the same tools as what we use for a very particular kind of practical problem

[00:34:32] of liquidity optimization in Virtus.org, right? And there are many examples like this. So I think there is, like you said, there are many significant synergies between what we do at Virtus.org and what I do on the academic front.

[00:34:45] And yeah, I think there has been benefits both ways. So yes, that's the reason that I think I am able to balance, to manage kind of the workload in both spaces. Well, based on that, I'd love to find out more

[00:34:58] about the kind of information you digest and read because we have an Amazon wishlist. And I always ask my guests to maybe leave a book, something that they've enjoyed reading recently or just would recommend other people check out. Thank you for sharing your insights today.

[00:35:13] But would you mind leaving everyone with one final gift and a book recommendation we can add to our list? Sure, sure. And I'll go outside of finance for this. I think throughout this chat, right,

[00:35:25] I think the thing that I try to emphasize the most is the topic of efficiency, right? So we're trying to do things in the most kind of efficient manner possible, right? And it turned out that I was thinking about it

[00:35:38] is that efficiency is something that I kind of have been interested in for the longest time, right? Not just in the context of finance, right? So in particular, I very much like efficient writing, right? So writing that doesn't have kind of superficial, unnecessary elements to it, right?

[00:35:58] And I think the author that I like the most is Jorge Borges. It's an Argentinian writer from the previous century. I think I've read all of it. I mean, I generally like South American literature, right? He's kind of one of the most important figures there.

[00:36:13] And the book that I like the most, I think my favorite book of all, is a collection of short stories that is called The Alice, I think The Alice and Other Stories, right? Which basically a collection of roughly 15 fictionary stories

[00:36:28] that are very short, probably maybe 10 to 15 pages each, right? But tell a complete story from the beginning to the end with lots of details, right? But in a very efficient fashion, right? So I don't think you can take any single sentence out of it

[00:36:44] without losing a meaning, right? And so I think this is probably the most beautiful kind of example of writing that I found, and that's my recommendation. And actually, this particular book, I'm rereading every few years just because it's just beautifully written. So that's kind of my recommendation.

[00:37:02] Oh, wow. I'm going to be checking that out. I will add it straight to Amazon Wishlist. And for anybody listening who just wants to find out more information about VirtuSwap, maybe they want to find out more information about the work that you're doing or even join your community,

[00:37:16] where's the best starting point for everything? Fantastic, yeah. So we have a website that's kind of a landing page to our protocol. It's virtuSwap.io, so VirtuSwap.io. We also have a Twitter account, an ex-account. It's called VirtuSwap, where we post pretty much every day

[00:37:37] kind of updates about what we do, engage with our community. We also have VirtuSwap Discord, which is quite active. We have a Telegram group channel where people kind of engage and ask questions and get responses. And we also publish kind of more detailed ideas

[00:37:56] and developments in the form of articles on Medium, right? So there is VirtuSwap Medium, where all of our ideas and developments from the very beginning until now are on display. So whoever wants to spare more than one minute that's typically required on Twitter, right?

[00:38:13] Whoever wants to spare 10 or 15 minutes, that's the place to go. Awesome. Well, I'll make sure there's links to everything so people can find that nice and easily. And we've learned a lot today from all about VirtuSwap, the story behind it, and also looking into

[00:38:27] the future AI's role in DeFi, the potential it has, how best to utilize it, and also ultimately what it means for the exciting future of DeFi, especially in advancing that entire landscape and increasing adoption. So just a big thank you for taking the time

[00:38:43] to shine a light on this space and the great work you're doing. Thanks for joining me. Thank you very much, Dale. It was great to be here and very much appreciate the opportunity. So what does the future hold for decentralized finance? Especially with the integration of AI?

[00:38:59] I think one of the things I learned today from my conversation is the advancements of VirtuSwap's technology. And with AI optimizing liquidity and trading efficiency, the potential for DeFi is vast and full of opportunities. And my guest's insights today provided a glimpse into how these technologies can reshape

[00:39:21] our financial landscape. And I'd love to hear your thoughts on anything we talked about, because as this world of finance continues to evolve, let me put the microphone in front of you for a moment. What are your thoughts on the role of AI and DeFi

[00:39:35] and these emerging technologies converging? How do you envision the future of decentralized finance? I'd love to hear from you. Share your thoughts. Email me techblogwriteroutlook.comx, Instagram, LinkedIn, just at Neil C. Hughes. I'd love to hear from you. But that's it for me today.

[00:39:55] So until next time, keep questioning and stay curious. And hopefully join me again tomorrow. But that's it for today's episode. So speak to you nice and early again tomorrow morning.