From Market Making to AI Rails: Alexis Sirkia of Yellow

Yeah, agents is going to be like a 20 trillion market according to Gartney and McKinsey. Five trillion for retail, 15 trillion market. And this is within the next three to five years. And imagine that the the US dollar cash economy is a 10 trillion market. So within the next three to five years we're going to have double the US dollar cash economy. Just between agents buying things and trading things with each other, you know. Hey everyone, welcome to the SM podcast where we ask your favorite builders the questions you wish you could ask yourself. This is a production of neonestoday.com and I'm your host, Dylan Grabowski. In today's episode, we're joined by Alexis Cirkia, the captain of Yellow, a decentralized clearing and settlement infrastructure built for trading AI agents and gaming. In today's conversation, we discuss the lessons learned from market making at gsr, the builder over speculator mentality, how AI agent will impact on chain activity, providing counterparty risk solutions, how Yellow is building out its ecosystem, and so much more. Just a reminder, nothing said on this podcast is a solicitation to buy or sell any tokens and that the guest or host might hold tokens discussed in any given episode. You can check out a disclaimer@www.neonestoday.com and if you enjoyed this conversation, please consider rating and reviewing it on your favorite podcasting platform and sharing with your friends and family. Every rate, review, like and comment. Really helps us reach more people, bring you more of your favorite guests. With all that said, I really enjoyed this conversation with Alexis and I hope you do too. Hey everyone. Welcome to the Smart Economy Podcast. Today we're joined by Alexis Cirkia, the captain of Yellow, a decentralized clearing and settlement infrastructure that operates as an L3 overlay on top of existing blockchains. Alexis, thank you so much for joining today. You're kind of an OG in the space, so it's really nice to be able to share an hour of your time. How are you today? Great, great, great. Happy to be on your show. Thank you for, for inviting me. Yeah, and thank you for dealing with the back and forth. Ben over at Waxman has been fantastic. We've probably tried to align on two different episode interviews, but you were traveling all over the world and the timing just didn't line up. So the first thing I want to ask is what have you been doing on this world tour over the past few months? Before we hit record, you mentioned that you've been to like 12 different countries in the past three months. What are the events you're traveling to. Is this all for conferences? Is this all for bd? What's the purpose of kind of like the Yellow World Tour right now? So there's been a lot of conferences. We, I started, I left from Andorra, from my hometown and then we're doing a lot of business. I have a few people in Andorra and a few people came to Andorra to have all these executive meetings, like deciding what we're going to do for the future of Yellow. And then in Hong Kong there was a conference, flew all over there, another conference in Sydney, that was another TAF ride. Then there was another conference in Bangkok that was. We went first to Chiang Mai for the launch of Yellow. We have a team there as well. And a lot of people joined over from different countries, from China, from Denmark to. They joined at our offices in Chiang Mai. And so we launched the 8th of March. It was quite a success. And after that I went to Money20 20 in Bangkok. That was also very cool because it's a very different conference. It's not so crypto, it's more institutions, banking and fintech, but fintech in general not. And they don't 100% understand neither crypto nor AI, you know, even though they're trying to, to start to adopt all these traditional trade institutions and then all the way here to Miami and waiting for consensus starting very, very soon. So I guess the big question I have is when you go on like crypto Twitter or basically crypto Twitter, which is where most of us get our information, it's really easy to see a lot of negative sentiment. But then when you're on the ground at these events, and it's always like this during bear markets, the people at the conferences are still excited about the future of the industry, they're still excited to build and they're still excited for what's next. So why is there always this disconnect during bear markets where we have the online community overwhelmingly negative. But then when you go to all of these in person events, you're just met with super excited individuals who are willing to build. What's the reason for this disconnect? Look in these conferences, just to name a few, like for example, I met Joseph Lubin, co founder of Ethereum and Consensys and Metamask and now Linea, discussing about the launch of Linea that he launched just before the bear market and it didn't go too well and, and everything he could have done to improve. But he has a solid project, you know, and he's very excited about the future. And I've built with, I've spoken directly firsthand with some of the institutions that are adopting Linea. And when this is news, you know, so the people are building now and that's what I wanted to relay. I spoke with Yatsui from Animoca Brands and he, that, that was in, in Tokyo I think. Yeah, in Tokyo. I spoke with Yatsui and he's just like us, he's focusing a lot on AI agents because we think that, that crypto and blockchain and this proof is like a really good, this trustlessness between AI agents is really the use case of crypto. You know, crypto was set up because it enables us to not trust a third party. And now we have AI agents suddenly interacting with each other. And of course there's going to be so much trustlessness between these. So crypto and state channels, that is the yellow specialization, really bring a use case for these AI agents. So there's a very, very strong narrative on AI agents. Everybody's talking about AI agents and money 2020 which was the institutional one. I was surprised as well. More AI agents. But then the directly AI agents is going to be like a 20 trillion market according to Gartney and McKinsey. Five trillion for retail, 15 trillion market. And this is within the next three to five years. And imagine that the, the US dollar cash economy is a 10 trillion market. So within the next three to five years we're going to have double the US dollar cash economy. Just between AI agents buying things and trading things with each other. You know, it's, it's going to grow pretty fast. And of course AI agents, you know, they're going to try to, I mean there was just one agent that just deleted the whole database of, of, of his, of this company that is doing rental for cars. You know, so really like the power we're giving to AI agents. And then there was all these other companies, you know, that are relying on these software that cannot book a car and they lost like three months of bookings because the AI agent thought it was also a good idea to delete the backups. But somehow they managed to get some backups, some physical backups that the AI agent didn't have access to and they lost just a few days. But this is the power we're giving to AI agents. So really like they have so much power that they can put your business down. So really we have to be careful how they interact. And crypto is such a good use case. So I think there's a lot of talk with, by builders and speculators. They're still on the last wave. You know, they're not seeing the future. And all these online forums and people talking, you know, they're talking about, about the past, about maybe the meme coins, that it's not such a thing anymore. And they all lost money with meme coins. And there's a bad sentiment. The Bitcoin, of course, took a hit probably with Iran. And the sentiment of like, you know, you know, when there's a war or a conflict, it's never good for crypto. Like when the economy is bad, gold goes up and crypto goes up. But when there's like a conflict or a war, crypto is not seen as a refuge currency as the gold is. So gold usually goes up in something like Iran, but crypto is not seen a solid currency against wars as gold is. So crypto usually goes down and so do stocks and all the other techs. So crypto has these very funny correlations. In some senses it correlates with gold. In other senses it correlates with fintech, in others with AI. And so today bitcoin is down probably because of Iran, because the, they are doing good things in the US for crypto and this administration is doing good things. But on the other side, it's so volatile that it creates a lot of like FUD and people not sure of whether to go in or out. So all these negative sentiments, I guess, and the previous bad experiences and the previous cycles that I think are totally like, you know, the NFT cycle is over. The meme cycle is, is almost over, you know, and so really the solid coins have left. And now it's, it's very interesting to see that coins like XRP that have always like, pushed for institutions, you know, seem to be the ones that are, are getting the hype in, in conferences and people are copying. You see even Solana going to institutions when Solana was always like this meme coin is the foundation of Solana. Is all these meme coins built on Solana. But now you see even Solana going to institutions and being adopted by institutions and Canton network. Canton is. Everybody's talking about Canton, so everybody's talking. So the narrative now seems to be institutions and AI agents. It's the two big narratives and AI agents. We still don't have a blockchain that has. We don't have a protocol. Hopefully Yellow is going to have some leadership there because we really think it's the best technical solution for AI agents that don't trust each other. But yeah, everybody's focusing on institutions. And as I told you, for AI agents, institutions is 3/4 of the pieces. And the retail pie is really being pushed hard by Stripe with Link and Visa and MasterCard. So the crypto retail, you know, has a lot of competition from the trade five that is already established in retail, like Stripe and PayPal and Visa and MasterCard and American Express. So they're all. The retail is going to be a tough one. But institutional rails for AI agents there, there is a big door open for crypto because institutions don't really have digital rails they can rely upon. And banks are not connected to AI agents just yet. And I think this is the opening door that crypto has been waiting for like the past 15 years to really establish itself as a powerhouse. And people will use crypto without realizing and they think it's coming. Crypto or yellows, which is a state channel, which is a different type of trustless protocol. I really think it's what's coming. So we're excited because we see this coming and we have all these reaffirmation from Gardner McKinsey, all these people that are doing this research and they say, yeah, it's going to be a huge market and our solution is the perfect solution. There is no really the competition landscape on institutional is very like desertic. There's nothing in retail. There's a lot in retail. Everybody's talking about retail AI agents, but institutional AI agents, there's really, really little. There's not even a good AI agent that is institutional. You have open claw. But you know, I can't have my open club for. With all my skills that are approved by the company and everything. So like really, I think we're very, very happy to be growing with an industry that seems to be staying is not like meme coins or NFTs or something that, you know, we, they, we build some fundamental use cases or ICOs, you know, all these previous cycles we have had, they seem like a proof of technology. But the difference with AI agents is that it's like a technology that is here to stay. And that's like everybody understand that it's not even, there's not even a discussion about it, you know, that it's not like meme coins that there was so many non believers. But like with AI agents, everybody knows it's here to stay. And the fact that crypto has the perfect solution for AI agents to pay each other and to transact with each other, to trade with each other, is that's why people are so excited in conferences and I hear this over and over again. But of course the communities, these solutions are only being built now. Our solution was only ready like a month ago, you know, so it's like so novel that you know, it will take some time but I think it's going to grow so fast. I don't think we will have an industry that is going to be growing faster than the agentic and the agentic payment industry. It's going to because it opens so many doors. Yeah, I mean I just started down my open claw journey two months ago now and I've already had my entire home directory deleted once because the agent just, I was just given bad code and so I can only imagine that institutions and enterprise are terrified of something similar like this happening. Like we saw with the example that you brought up earlier. I cannot wait to dig into your philosophy on state channels, how they can be used for agentic payments. But I feel like there's. The listeners would be remiss if we didn't just take a quick little step back and just walk down the history and the experience that you're bringing to this conversation right now. You were the co founder of GSR and GSR was one of the first market makers in blockchain and crypto. You helped create markets for Ethereum, you helped create markets for Ripple. So there comes with this kind of outlook for the future and this excitement for what AI can bring, both retail and institutions. You also have this plethora of experience not only just in the blockchain space, but you've been working here for almost as long as blockchain and crypto has been a thing. So maybe really quickly we can just take the, the professional journey that you took when you were working in aviation and space and then your pivot to E commerce. And then what happened in 2013 when you saw crypto and you said this is a place I need to be building in right now. It was very, it was very like by chance, you know, I was there in the right place at the right time and I had the right connections. My, my school friend that had been at Goldman Sachs before, one of the guys from Goldman Sachs went to work for this startup in Silicon valley with only 12 people and they were printing money and but this money was xrp, you know and we, I, we just thought it was so amazing. And they asked for a market maker and he was a market maker before and I was technologies, the software engineer. So between the two we build a great, between his financial knowledges and my technology knowledge. We built an amazing fintech company that became very, very successful. And you know, like, we helped companies like Ripple a lot to really like build markets as solid as traditional finance markets. But this was only 2013, 14, 15, you know, and at the time in crypto there was nobody from the traditional finance industry like, very interested. So we really had an edge. It was like putting professional poker players with rocket engineers software in an amateur tournament, in a poker amateur tournament. Like we had the best software and the best skill set from the traditional finance. And at the time there was no other like match like this. There was some people with traditional finance getting interested, but it was mostly manual trading and there were some people with tech background trying to do bots, but very, very, very, very lousy technology. So really, so really this is what, what made the great market making company that we built. And around 20, 18, 19, you know, I saw that there was this big problem in crypto. The thing I wanted to solve, that we have solved now is that I saw that all these people were trading in exchanges and they had the risk of the exchange. And the thing about they had to trust the exchange to trade. And today is like 95% still of trading that happens on centralized venues, even though it's moving to more decentralized venues, still over 95%, you know, and, and things like Hyper Liquid offer some sort of, some sort of less risk, but still you saw how they closed the market when they wanted. So it's like your funds are still at risk. They're at the same risk or more risk than with Binance, you know, on something like Hyper Liquid because really in the end they control many things and even though you have custody, they can close your market and they can screw you worse than someone like Binance that has a big, that has a big safu, like this SAFU fund, you know, that if things go wrong, at least they have so much cash on hand that they can pay for whatever mistake. Hyperliquid doesn't have those sorts of reserves. So I would prefer Binance and Hyper Liquid for example, if I had to choose one risk over the other. But what we're really trying to solve with Yellow is, is being able to trade with, with a centralized exchange without the risk of decentralized exchange. And that is the, the essence of crypto. Two parties interacting with each other without trusting a third party. For example, with Bitcoin, I send you money. I don't need to trust the government to issue that money and I don't need to trust A bank to transfer that money. And with Ethereum I, we sign a contract and, and I don't need to trust a person to execute that contract and I don't need to trust a judge to arbitrate if things go wrong, you know, so it's all about this automation and this trustlessness. But when we're trading, you know, that's what I, that's what sort of hurt me, you know, when we're trading so much in GSR and then where there was this, this Norwegian exchange, it was called just Coin. It was a small exchange, you know, but it said the logo was the safest exchange in like the world. It has never been hacked, you know, and from one day to the other they closed and they got hacked and every, they lost everything from one day to the other. And their, and their slogan was the safest exchange in the. In, in the world, never been hacked, whatever. So this was just Coin and before there was Mount Gogs and of course we've had Bybit recently and Kucoin and so many other hacks in so many exchanges. So, so how can you expect a bank like let's say HSBC to want to trade the crypto when the counterparty is somewhere like Binance, which is probably my favorite exchange. But how can you, how can you trust like 10,000 people working there that they have all the, some, some people eventually must have some keys, you know. And then there is all the outside hacks that can happen from even things like smart contracts like the Parity contract got hacked. So it's just too, the surface of attack in crypto is always so huge and we're not talking like a bank heist which is like, you see this Ocean's Eleven and, or Ocean Something and it was like $8 million. We're talking like billions and billions and hundreds of billions of dollars. Imagine if the wrong people with like, with weapons and everything, they really want to get it. You know, it's, the pot is so huge that, and it's so juicy that bad people, hackers, everybody's like, you know, it makes sense to, it makes sense for in a region like North Korea to, to, to have all these people just trying to hack and steal money because it's, it's such a blue collar crime. But you know, it's like really people suffering with their life savings on the other side, you know, so this is the big problem that I wanted to solve and that we solved. We solved it with the technology and for very, very long, until 2021. So we tried initially with blockchain. And until 2021 we were very focused on using blockchain. But the problem with all blockchains is that including the better ones or the worst ones, let's say you have 30,000 validators or even 20 validators that are going to sign every transaction. And when you're doing high frequency trading, like we were doing, or we, we still do it yellow, you know, like we're doing so much volume that is just from the efficiency perspective, you can't have 30,000 computers supervising every transaction that happens. Or even with like, or even a set of 20 computers, or even like delegation is going to. We're talking about microseconds and in the future, nanosecond speeds when we trade, it just doesn't work. So what's really co is that we have a very simple solution where we use blockchain only when things go wrong, to arbitrate and to settle. And it's so simple that I can explain it like almost everybody understands it. I buy like a bitcoin for you from you, let's say 70,000. Let's say $70,000 today, and we both cryptographically sign, okay? And we keep a proof, we keep this ticket, this invoice that we have both agreed to this transaction, but everything stays private between you and me. And then I buy an Ethereum from you, let's say $3,000, and we both cryptographically sign, okay? So we're trading with each other crypto. And let's say I do a thousand trades or a million trades, like a normal trading session, okay? But let's just keep it simple with these two trades. And then I owe you 73,000 and you owe me one Bitcoin and one Ethereum, okay? And then I ask you to settle, and then you send me the. So then is when I ask you to settle, you send me the bitcoin and Ethereum and I send you in one transaction. I send you the 73,000 USDT, let's say. And if you don't want to settle previously, we have put a security deposit to both of us, okay? So it's this escrow mechanism, okay? So you will lose your escrow and the, the security deposit is always at least the collateral you're willing to lose. So when you're trading Maybe to buy one bitcoin, you put like $20,000 in collateral and you're trading like with forex leverage, okay, Today, So maybe I would have put $20,000 of collateral each. Okay, so then this allows us to trade and I know you're going to settle. And if you don't want to settle the smart contract, we will use the lock collateral to pay me my, my profits and losses, my P L. So whatever I, whatever I gained, the smart contract will ensure, and it will do so by me showing the smart contract my, my cryptographic proofs of everything we agreed. So you see, in this very, very simple mechanism, we agree to everything and we sign cryptographically every time. And if you settled, I wouldn't even have asked the smart contract. We can cycle privately and I don't even have to ask. But if you don't want to settle or if you close the market like hyper liquid, my client will just go to the smart contract and will, will ask to execute. So you can see that with this method we can trade at ultra high frequency because we're on, we're only cryptographically signing which is like a very, very basic computing operation that can be done locally and we don't need to. And we only use the blockchain for execution and arbitration which was the intended use of blockchain, not for private transactions. In history and business in the world, the business starts always privately. Okay. So when you ask like somebody to, to do like you for your wedding, you say okay, I want this, this, this, this, this, this. And then you're going to, to have everything that you asked for and then everything went okay and you're going to settle for your wedding in one go or maybe like in, in, in many goes maybe 50% but anyway. But one thing is business and the other thing is payments and settling. And with Web3 until today everything was settled when there is no real need to. You really only need to settle when you have to tell a third party about it to tax authority or whatever. Otherwise you just talk between. We do our business privately with each other, you know, and we don't tell a third party about our private business. So with this way bank like HSBC can trade against Binance without trusting Binance. Okay. Through these collateral mechanism and with my background in market making, we bring the liquidity of binance and we make it available on the yellow network and people trade against binance liquidity on the yellow network without the risk of binance. Okay? So it's more volume for Binance. It's not an aggregator like one inch, you know, we bring the liquidity from different venues. Okay, so now for example on Yellow Dot Pro you can find the liquidity from Binance as our first use case. And you can trade Binance liquidity without the risk of Binance, this is at the same price as Binance and this is our first test. We have only Bitcoin, Ethereum and yellow markets open and it's not custodial. And it's like you can trade at the speed of light. And then the extension of this, this is one use case trading, which is the one I wanted to solve. But when we launched, when we were, when we built this, this was around 20, 23, we told each other, we said like, you know, this underlying technology that we use for trading, we think it's huge. We think it can be bigger than the trading, you know, than this broker mechanism. It can be used for gaming, for AI agents, for whatever. So we opened up the SDK, you know, a bit like when Amazon.com opened up AWS for everybody to use AWS and today three quarters of the profit from Amazon come from AWS. So we thought, hey, this SDK that we're using internally, we think it can be bigger than the trading yellow network. So let's open up the SDK so everybody can use it for their own apps. And we got about 500 projects already that have developed apps with the SDK. And there is a lot of trading venues like Yellow Pro, but there's like others, there's one very cool one that you swipe right to go long, you swipe left to go short. That's a very cool one that I love of and I think it's called surfing. And then I don't know, there's like gaming, there's trading, there is. And the one that we saw in the last three since January, I think probably because of Open Claw, is AI agents. The we saw the growth in usage of our SDK for AI agents. Like really it started to become the first, the first use case of this technology because the same as the bank doesn't trust Binance, an agent isn't going to trust another agent. And the thing about using blockchain for this is that we expect there's going to be billions of AI agents. And if there's billions of AI agents, it's very difficult to use a solution like blockchain, like full blockchain. I would say ours is like blockchain only the start to set the rules and in the end to settle or arbitrate. Now when we're talking like that state channels, which is like what Lightning Network was and the thing I was discussing with Joseph Lubin, because Joseph Lubin was the guy that first invested with consensus on state channels, generalized state channels for Ethereum and he told me he spent like $8 million on it before sort of abandoned it because he didn't have the clear thing. And then I explained our solution and he loved it. You. And he loved it because we, we solved the capital efficiency problem that was the big, the big problem of state channels. So anyway, so agents, agentic payments and agentic trading is the one that we see as the first use case now. And it's been our focus. We have like a demo app, which is very, very cool, that you buy an airline ticket, okay? One agent, you tell him to buy an airline ticket, okay, to your, to your own agent. And today when you do this, it will go to all these airlines and it will tell you, okay, this is the travel agents. It'll tell you, okay, here, travelocity or kiwi.com or booking.com is the cheapest, or Kayak is the cheapest. But sometimes the cheapest will be someone like cheap tickets.com and it's not a very trustable travel agent. So this agent will go and look all the agents from the travel agents and will try to find the cheapest ticket. And it will put money on it will put the security deposit, and then the travel agent will send him a booking code to your agent. And your agent will verify with the airline that the booking code is valid. And if it's not valid, it will raise this to the, to the smart contract for arbitration so that the travel agent will lose their security deposit. And if it's valid, which is the normal use case, your agent will tell the smart contract to settle and the travel agent will be paid. So this is a use case of me buying the cheapest ticket from cheap tickets.com or whatever. Very scammy name, scammy. But, you know, as long as you can verify that it's a valid booking code from the airline directly, you know, and it's good. Why not buy it from the cheapest, which is not what we do today. You know, today we go to the most trusted. But in the future, your agent, with this technology that we bring will be able to save you money and get you the best deal or so. So we have this demo app. It's the, the one we showed in the, in the, in the conference. And it's, it's a really good use case of blockchain. And people, they don't really realize that it's. Everything is cryptographically secured in the back and their smart contracts are going to arbitrate and execute and everything really. It's the dream of Web three Yeah, I, I love this use case. I'm personally buying some tickets right now for some flights and because of the, the conflict in the Middle east, just ga gas is going up. The price of flights is going up. Just in a week's difference, a ticket from here to Spain has gone from $1200 one way to $1700 one way. So this is an amazing use case that you're talking about with insert scammy cheap tickets.com here for anyone who's traveling right now. That, that's going to ring true. And, and obviously you're starting to see all these different uses for, for Yellow, particularly with. And I want to double, I want to, I want to zoom into that in one moment, but I want to take one quick step back because right now when we talk about why yellow was built, it was for counterparty risk. And you mentioned that a lot of your use cases are going to involve HFT high frequency trading. My question is, do you also have hardware infrastructure providers that you're partnering with? Because when we talk about something like hyper liquid, a lot of those nodes are located in AWS's Tokyo region. And if you're trading here in the US there's a 300 millisecond differential and so there's an edge to a lot of the traders that are in Tokyo. So two things. When you're working with HFTs and folks who want to do high frequency trading, are you also incorporating infrastructure partners to help help reduce this edge from a global perspective? And secondly, and I'm sure this is going to open a can of worms, why state channels? The last time I heard about state channels other than lightning was like 2018 and that was the scaling solution. So why do state channels make sense today in 2026 for a technology that seems like it was left in the past six years ago? State channels are the only ones that allow you to trade between two parties without the third party supervision. Like Zksync and all the other crypto technologies. They only allow you, you're always like, everything is being supervised so private between two parties. That's always been state channels. But it's a very simple technology. It's not really, as I explained before, like anybody can understand it, it's way simpler than blockchain because it's just a hash of every transaction. But how can you. So this is, I think this is probably the first time that I explained publicly what is in our roadmap and what we have been also building in parallel. We try to not explain a lot because we don't want to hype anything about the future to not be considered as security. Like we, when we release something, it's already built. So one of the things that can be done. Once you have this technology and you bring the liquidity to a yellow node, for example, like now in Yellow Pro, imagine a future where a trading firm like my previous trading firm GSR Winter Mutual, instead of connecting locally from one AWS to AWS Tokyo, imagine it compiled their trading strategy locally. You build your trading strategy with a trading strategy builder. You backtest it with the history of Binance and then you send it to a matching engine, which is basically a yellow node is a matching engine to be executed locally. And then you're not talking melee or even like hundreds of microseconds, you're talking nanoseconds and you're talking that every trader, no matter where they are in the world will have the same advantage because they will be running their trading strategy locally in the same computer. And if you go today to Body Trading, that is one of the subsidiaries of Yellow that is built in, in partnership with. And so if you go to buddytrading.com there you have the strategy builder, the backtesting and like very, very soon you, you can already you'll be able to execute it on, on the yellow network. So today you can execute it on Binance or something. But very, very soon this body trading, which is like this strategy builder, will be able to, will allow quant, quantitative trader quants to build their trading strategy and execute them locally on trading venues. And this is really going to revolutionize the, the, the trading industry because we're going to. It's like cloud computing. You know, it hasn't happened for, hasn't really happened for trading because we're always like interacting from one computer to another. Even if they're both on aws, we're still talking between two computers. We're not like executing our strategy on the matching engine. And this is something that we have developed for the last, I think about four years already. It's been a parallel project led by Yellow that has been four years in development that you can already test the strategy builder. Anybody can go to budgettrading.com and trade the, and test the strategy builder and you can create your own models. And actually if you're very successful, you can allow other people to piggyback on your models and you can sort of rent your trading models and it gives a new venue for a new revenue stream for exchanges. Because today exchanges like Binance, they have a trading fee, which is usually quite small. And maybe you trade a million dollars and you make $100,000 profit, or maybe you trade a million dollars and you make $100 profit. It doesn't matter. You're going to pay the same small, let's say $1,000 fee. Okay, so even if you're making $100,000, Binance is still earning only $1,000. And if you're making, it doesn't matter. So, and if you look at the traditional finance industry, you check in an exchange like Nasdaq, this is a number from some years ago, but their, their profit was like around like $6 billion only. But the trading industry is between 20 and $30 trillion profit in the traditional finance. So the, the, the, the amount that NASDAQ is making on trading fees is very, very small. The big part of the piece of, is assets under management, trading, like brokerage fees, everything, you know. So the exchanges with this model will allow trading firms that are making more money to pay them more to rent a slot to have, because those slots are going to be limited a bit like AdWords, you know that they have a ranking and the one that pays the most goes on the top. So in the future, the trading firms that paid the most will be able to have a slot in the trading venue and it will be a fair game. Just like AdWords where you pay them more because you're, and you're making more money, so you are willing to pay more to the exchange. So it means that some of that revenue is going to translate to exchanges. And exchanges will make more money compared to the whole industry, which is today in traditional finance. Exchanges make very little money compared to what the industry is making on top of them. You know, like NASDAQ is not making so much money. Actually all the money is the brokers, the prime brokers. Those are where the big money is made in traditional finance. And I'm talking like a dozen billion on one side exchanges and trillions on the other side in profits. So yeah, so we think that with this technology, some of the profit will be translated from trading firms to exchanges. So we think it's going to be very, very welcome by exchanges once we deploy it and you can send your trading model to a matching engine that is executing Binance. Liquidity is going to be amazing in the future and it's going to be faster, so you'll be able to trade faster. So it means more trading fees, you know, so you trade faster, it's more trading fees, your rental slot is more money. And then you can trade larger as well, because you can within the yellow network, you can do what they call like clearing. So you can have like 100 bitcoin long on Binance and 100 bitcoin short on Kraken. Today this requires like 100 bitcoin exposure on Binance and 100 bitcoin exposure on Kraken. You know, to do this operation. But in traditional finance usually you have a clearinghouse and the clearinghouse is going to, they just care about your total exposure. And the net exposure on this operation is about almost zero because the risk of finance and Kraken is very similar. So it's almost zero. So the clearinghouse, usually for very, very little money will allow you to do this operation. And then you make money based on this location. The difference is that when yet when you do this on the yellow network because you have Binance liquidity on one side on the yellow network and Kraken liquidity on the other side on the yellow network. A third party clearing agent can give you a credit line for this operation for very, very cheap, which means that you can trade larger. It's more capital efficient to trade on the yellow network than to trade directly on the different venues. Because you can use like in traditional finance you can use the concept of like clearinghouse that just cares about your total exposure. Like if you have Amazon and Tesla and Microsoft, whatever, they just care about your P, L, you know, like your total P and L. They don't care about the individual positions. So I'm curious then we're mentioning big, large name brand exchanges, Kraken, Binance. And your name carries weight when you go into these rooms to have conversations about the yellow network. So are these centralized exchanges also acting as node operators on the yellow network or are they just providing liquidity? And if they're not operating as node operators, who are the node operators on the yellow network? Currently there's very few node operators. It's one of our missions in 2026 to, to increase the number of node operators. And there is some, how to say, some yellow rewards to, for people to some rewards from the protocol for node operators in the future. So the idea is to have each node operator is bringing liquidity from a certain venue. The exchanges can operate them directly, but also market makers can operate them because many market makers are receiving like, how do you call this, a rebate from exchanges. So for the volume they are bringing to the exchange, the market maker is actually getting paid maybe one basis point, you know, and it means that if you bring the, if you you can put it on the yellow network and actually with that one basis point is just like more volume for the exchange. So you can, you can have it at the same price as, as the centralized exchange outside of the centralized exchange because exchanges are paying you to bring real volum to exchanges because there's a counterparty always on the other side. So as long as there's counterparty. But I think in the future exchanges, it will be just a new way for exchanges to get more their liquidity. So today they have their web apps, they have the API, their desktop apps, so their desktop website. So they have like three, three things which is like the mobile app app, the web interface and the API. So today you can access mainly through these three. So yellow will just be a fourth way to access the same liquidity from exchanges. And the other reason is that we think that exchanges will integrate is that we are doing a considerable effort to integrate our yellow network into legacy traditional finance software like core banking and all these software. So once we have the big buyers, of course exchanges will want to have the exposure from these big buyers. When I was talking about like HSBC or something, so we are not working directly with banks, we're working with the core banking software that all these big banks are using and we're integrating yellow there. So that's a big part of our, of our effort now. Like it's starting now that we released a protocol now our effort is to integrate into traditional finance finance, core banking and core brokerage software that banks and brokers use. So once we have this, of course the exchanges will want to have direct access to those purchases. And forgive my naivety, I didn't get to dig too deep into the docs. What does it take to become a node operator? Is it an allocation of yellow tokens? Because as we talk about, about HSBC or other large international tradfi or banks getting involved, getting skin in the game. As me a crypto native, it's super exciting to hear that they need to allocate a portion of tokens so that they can become a node operator in addition to like the liquidity that we're bringing to the table. So am I correct in understanding that there will be a certain amount of yellow tokens that will need to be staked for being a node operator? Where, where might my assumptions be wrong? Yeah, initially it was a number that was not too big. I think it's between 200 or 250,000 yellow tokens that if you divide it by 20 or so today. So that's like 200. It's about 10 to $12,000 that you have to stake to become a node operator. So the number is not too big today. But we are giving grants of this value to companies. Well, we're not layer three, the foundation that owns the yellow tokens is giving grants to people to become a node operator. If it's like for example, an exchange or something and they are bringing real users that are going to be. Because the main purpose of the yellow token is this security deposit for the escrow mechanism. That's the primary utility, not the, not the, not the deposit of yellow tokens for becoming a node operator. That's not the primary utility is this security deposit that you use when you do escrow. And the protocol will reward the people that do escrow. Okay, so if you have put, if you have locked yellow tokens for the escrow, you will get a reward from the protocol for having taken this risk. And if you try to, to defraud your counterparty like in this trade, in this travel agency and you lose your, then you will lose your security, your yellow tokens. And that's a very similar use case to Bitcoin or Ethereum. So you're a very good actor. You, and you sign the transaction faithfully, you get rewarded, Bitcoin or Ethereum. And if you're a bad actor in Bitcoin you lose the, the Bitcoin you were rewarded and in Ethereum you lose the, the Ethereum you staked. So it's this rewarding good actor, penalizing bad behavior. Crypto. Crypto, very crypto use case that we have copied in yellow. So yellow is, that's the primary use case, rewarding good behavior, penalizing bad actors. And the staking for the node is something that we had considered. I think it was 200 to 200 or 250, I don't know in the end how much, but I'm not sure it's, I'm not sure it's important. I think the more node operators we get, the better we want real businesses running node operators to make it sustainable, like business sustainable. You want to have profitable businesses just like mining operations or something. So we want profitable businesses behind the nodes that are making a profit from the yellow economy. So, so initially there is a grants program and they might receive it whatever yellow they had to stake. They will probably receive it as a grant if they are bringing users, but they need to bring users. That's what we want, users adopting the yellow protocol, indirectly, of course. Quick question. Is the grants program a part of yellow's acceleration program. Are they separate? Separate. So the. Okay, so this is a bit complicated. So think about like Solana Labs, I think and Solana foundation or like Ripple and xrpl like XRPL or like so, so one is the company and the other one is the foundation. So layer3foundation, the company behind yellow.org that is the one that is giving grants to the ecosystem and one of the biggest ecosystem players in today we have two accelerators. One is going to be under COIN Cilium, which is one of the. And the other one is yellow.com which is like a private venture that is running an accelerator and giving grants. But those grants initially come from layer three, which is the foundation. But Yellow.com is a private venture. Future is called Yellow.com Inc. And that is running an accelerator with the purpose of building the yellow ecosystem. And hopefully it will become a Treasury company in the future and hopefully publicly traded. That's our goal, to become privately publicly traded company with Yellow treasury but reinvesting in the yellow ecosystem and hopefully getting a profit. So that's the acceleration. And coium has similar objectives in the uk. Okay. And they, they plan to do the same thing in the, in the UK and so far we have these two accelerators that have applied to become like ecosystem partners and, but also people can apply directly to, to Layer three if they have like, like an idea. So it's, it's, it's a hybrid of, of Layer three that actually has the tokens and, and either coium or yellow.com that has the resources, the people to run the accelerator programs. And, and they are Both Consilium and Yellow.com are for profit companies. Okay. So it's a for profit accelerator that is investing in these companies and expects these companies to make a profit. And Layer three is a nonprofit that is giving grants for the ecosystem system. Yeah. As we kind of like zoom out and wrap up, one of the key themes that's, that's sticking with me from this conversation is that Yellow started with an idea for removing counterparty risk and providing opportunities for access to liquidity. And you've just kind of grown and expanded to address where the market is going. You're addressing AI, you're addressing counterparty risk, you're addressing gaming. And as is such, there's a lot of different ways for developers, users and contributors to gain access to the ecosystem. There's the SDK that was released, there's the different types of accelerator programs you were mentioning. There's the grants program, there's becoming a node operator in Addition to all of this, there's also yellow media. So I'm curious to hear what your perspective is for also integrating kind of like a media arm for the yellow ecosystem. What's, what's the significance, the importance of having media outreach? Yeah. So there's two things like well some, somebody said, I can't remember who this quote is from, that every company is a media company. Okay. And what, what we can give so to the Idea purpose of yellow.com is to have this accelerator program. But one of the best things they can give to these entrepreneurs applying is exposure. Okay. So I have some statistics. I read that it's like 4 million users, 4 million unique users and there's like 500,000 from US and Korea and whatever is really high quality destinations that are reading this media. The media is now in Google News. So whenever something is published in yellow.com in the media section, Google News republishes it. So it's, it's pretty and then coin, market cap and others as well. So but it's, it's getting some, so it's real journalists also behind the yellow media and they're doing a great work. Like there was one in particular that I loved, one piece that was from an SEC commissioner. So the, the journalist interviewed the SEC commissioner and she said people will not be able to hide behind their AI agent or behind their AI. Yeah, I think this was her, these were, there were her words, you know, and this was requoted by a lot of media as well. And so little by little we're growing today there is like we're still, we're still like not, not so popular but we're like already like number four or five in the crypto industry. You know, we have grown like significantly in views and these are like high quality views, you know, like from the, you see the countries, it's like high, high paying countries. High, like high paying countries that are important for entrepreneurs. And we thought that you know it's, it's good for us to have this media arm because it allows us to, to broadcast the message of yellow. But it's going to be great for entrepreneurs to as part of the acceleration program to get all these free exposure is not free because it's part of the package that yellow.com provides. So it's this exposure that they can get as part of the package of adopting yellow or whatever. We will give them all this exposure. So we've grown it for the purpose of growing the ecosystem. Once it becomes like a good media, the all the entrepreneurs building in Our accelerators, they will get exposure there. I don't know. What do you think about this idea? I love it. I came into crypto in 2018 to be a content creator and an educator. I've written over 1800 articles, I've hosted over 200 podcast episodes. I'm a big fan of using media to educate people. And a lot of developers and people who work, work want to be able to share their story. And there's not always a, a venue for them to be able to talk with a journalist or reporter, somebody to cover what they're working on. And so anybody who's building this is a positive externality for them to have somebody interview them, talk about their work and then share it. I think it's fantastic. So I'm, I'm def. It's, it's on my bookmarks now. The, the yellow media and we have three co founders. So we have one co founder, he's focused on Camille, on media and ecosystem. The other co founder built much of the SDK, but like he was focused on the SDK before, but now he's focused on Yellow Pro. And myself, I'm focused on layer three on the foundation and letting people know what the yellow SDK is, what you can do with the yellow token, everything you can build on top of it and bringing people into the ecosystem and yeah, so it's also a good separation of the co founders tasks. We are three co founders, but actually we're all almost running a whole venue by ourselves, like a whole entity by ourselves. But there's so many synergies between all the entities. So we work together all the time. Yeah, I love it. We've hit time. So I'm going to ask you what I hope is a simple question, but maybe it's a little spicy. When I first got into blockchain in 2017, we were talking about onboarding the first billion users onto blockchain. And with the rampant growth of AI, I think the first billion wallets will belong to agents, not humans. So when do you think we're going to have a flippening of AI agentic wallets, that there's more of them than human wallets? I think it's going to happen like within a year or two, you know, like within, within 12 months probably. Like it's because like we see it with what we're building because it's so easy create for an agent to create a wallet. And creating a wallet is free. You have to, you have to fund it. That costs money. But with yellow it doesn't cost money, actually, because it's like only cryptographic signatures that are free. So I think probably, probably within a year or two, you know, and imagine how funny it would be that while you're creating this wallet because there's not so much randomness and we're creating so many billions with agents, one of these agents suddenly creates a wallet that, because it's like random, so it can happen, creates a wallet that belonged to Satoshi, you know, and then he gets all these thousands of bitcoins, you know, whatever, whatever that wallet has, because there's some wallets with many thousands of bitcoins, you know. And now he sudden said, wow, now maybe with AI agents working and all creating wallets, maybe it's going to happen, you know. Oh my gosh. Now I have, I was just going to say now I have to worry about quantum hackers and an AI agent magically creating, recreating one of satoshi's wallets. Well, because Satoshi's wallets, they don't have so many. So many bits. So those are quite easy to create, you know, like. Yeah. Awesome. Well, Alexis, we've hit time, so thank you so much for your time today. What's the best way that somebody who listens to this can reach out to you or get in contact with you? Yellow. So I think crypto Twitter, like you mentioned before, like our handle is just yellowello and this is as easy as it gets. I think we are quite responsive on Twitter. We also have a telegram. Just search yellow in our telegram and you want to contact me directly also on telegram. I'm always active on the chat. Awesome. Twitter telegram. You heard it here, folks. Alexis, thank you so much for your time today. This was a pleasure to chat with you. You, you've earned your stripes in the blockchain and crypto space. So it was nice to to have an OG and pick his brain for an hour. So thank you so much. Thank you. Bye bye. Thank you so much for tuning in to the Smart Economy podcast to stay in the loop with all of our latest guests and insights. Head over to www.smarteconomypodcast.com and don't forget to subscribe to our YouTube channel for all of our videos. Content. If you enjoyed today's discussion, please consider showing your support by liking, commenting and reviewing this episode on your favorite podcasting platform. Your feedback helps us reach more listeners and bring even better content your way. We'd also love to hear from you, so please drop a comment in Spotify or YouTube to share what topics you'd like us to cover next, or who you'd like to hear as our next guest. And if you or someone you know would make for a great guest, please reach out. We're always on the hunt for fresh voices and new perspectives. And for our Neo Token holders out there, please consider voting for Neones today as your Council representative. We've proudly been serving the Neo ecosystem since 2017. We'll continue to do so by putting portions of our Council rewards directly back into ecosystem growth initiatives. Once again, thank you for listening, Smart Economy Podcast. We will catch you next time.