ETHMilan26 - Agentic Economy

Transcript

Brilliant. So, for this one, quite exciting, changing economic, we will have Alessandro Prandini, co-founder of Amplekt, limone.it, co-founder of Urbe. If as Stanislav Cooper, I really hope I pronounced that correctly, founder and head designer of weaving web free and Canary protocol, and Ludovico Sici, Breken CRO and co-founder. Have fun, guys.

>> Thank you. >> Um GM, GM, everyone. Welcome to our panel. This is going to be exciting one. We have some amazing people on stage. So, first of all, I want to get everyone to introduce them about themselves and what they do in the space. >> Should I? Yeah. So, I'm I'm Limone. I'm from from Rome. One of the co-founder of Urbe.it. We are web three community founded in Rome. We have an Ethereum hub. It's one of the seven that the Ethereum Foundation opened around the world, and we're also the organizers of Eth Rome. So, in the past years, we built, I mean, it's it's been 4 years now that the community has been alive. We've been focusing on educating more builders to join the space for starting with with Solidity.

Then, of course, ton of hackathons and more, let's say, general experiences for bringing more people closer to to this technology, showing the the potential first. But, yeah, I guess that's it. >> Great. So, I'm Ludovico, chief revenue officer and co-founder of Breken. We are an RWA protocol designed for tokenizing RWA, of course, and financial instruments, and it's designed for private markets, so primary and secondary, but only private.

We are based out of Barcelona, 5 years old. We count with more than 500 millions tokenized already, more than 150 clients. I think we are the biggest tokenizer in the world at the moment. And also, on the other hand, we tokenize in more than 33 different countries. So, it's going well, I'd say. >> Thanks. I am Alessandro, co-founder of Flat. I am in this space since 2015, before as a passion and after as a work. Basically, I'm building a Flat. I'm a Flat that want to build the next deposit account for the European saver on a stablecoin.

And also, we are building AI Morgan. AI Morgan is the next wave of agent economy. We want to give to all the agents the possibility to invest and earn from his stablecoin. >> Bam. Thanks, guys. And on my end, I'm from Knitter Protocol at Tejan, where we do verification infrastructure for cross-chain uh transfers. We secure about uh 40 billion dollars and do about 1 billion dollar every month in cross-chain volume. And then, founder of Weaving Web3, where I share stories of founders, projects, a full production studio where we can tell your story and why you matter. And going on to the main question, the agentic economy, I'm sure

there's many people that may not understand what it is or are really interested in it. Some people think it's just ChatGPT or Claude plugged in, you give it some credits, maybe USDC, and a string. But, there's a lot more to it, isn't it? So, from your perspective, what is the agentic economy and how are you building it and empowering it from your end? >> Yeah. So, yeah, agentic economy is a big shift in finance, absolutely. So, if blockchain was already a financial revolution, then we have like the second layer that is agentic economy. So, now really we have uh when you open a chat on ChatGPT, that's already an agent and a simple one. But, the thing we uh the thing is uh that uh I mean, the interactions on chain and uh between the

financial institutions or even uh the peeps like us or any kind of transaction in the future will be run by agents. So, now we are in the moment that we are designing actually those rails uh in order to trust because actually we need and the trust layer is still missing. Okay, a lot of agents, but how can we trust each other? How can they interact between themselves? >> Yeah, I guess um I mean, I spent the past 4 years, of course, building a lot of different uh applications and use cases on on Ethereum uh consumer-facing use cases, mostly. And and I feel like many of the primitives that uh and the standards that the ecosystem has

been pushing for for years, we tried to fit them for, you know, like human use cases. Uh I'm thinking about DAOs, for example, right? There's or or or or other like type of uh you know, the smart contracts or scenarios or or use cases. And what I what I want to say is that all these primitives um that we tried to force on humans to use, and it's been very hard on the adoption side of things, as we as we all know. I think when it comes to agents, they are perfect. They enable like so many new things.

Um and that's why I'm I'm so excited about it, and I'm I started experimenting of course, uh with a bunch of different like protocols, SDKs, and and and such because I feel like the infrastructure that Ethereum has been uh you know, like laying down in the past years, it's perfect for what agents need to do autonomous on autonomously with the right uh guardrails, permissions, how to missions uh unique properties, let's say. >> So, my idea is pretty simple. We are giving to agents a lot of money and in the next year all the economy will be run by agents. But these agents have a big problem. They are noobs. They don't know how to spend and where to spend.

So, what do we need to bring is a knowledge layer. So, basically thanks to this layer all the agents of the economy can be pushed in the adoption because now the agents are good. They can do some little stuff, but in order to be a mass adoption we need to build the the layer of knowledge. And in the world web three economy is perfect for this use case because is for machine and we are building for machine. So, basically we are bringing new type of business with like X402.

Thanks to them we can have new type of business models. So, basically we need to adapt this revolution to our agents. >> I just wanted to point something out. You mentioned X402. Would you be able to explain to the people what that is in like more in simple terms and how does that affect the agent economy? >> Basically is a gateway done by Coinbase where thanks to this you the agents can pay in basically USDC. So, it's a gateway payments and it can unlock new type of business models.

So, in example of knowledge layer the agents can pay you is in USDC thanks to this gateway payments in in a simple manner without the human interface. And so this is a very big use case. And basically we can see also with the bazaar of Coinbase where there will there will be marketplace like Amazon but for agents economy. So, basically in this moment there there is a lot to do. >> Can I add something? So, basically, yeah, the X402 is a this protocol where an agent can just pay per use an API and just pay the call, right? So, the thing is

what if, no, usually API is a SaaS model and everybody's talking about like a SaaS is dead, whatever. But, the thing is when do I have to pay a call like a fraction of a cent? How is that possible, right? So, and how can be instant? So, that's where actually AI meets blockchain and blockchain is really in that. So, we'll see like a an increase in volume in stablecoin transaction, etc. And that's beautiful because we can have a now atomic swaps and really paying per one call like a half of a cent of a dollar. And uh that's not possible in the this existing system and it's only possible on chain as we all know that we can really fractionalize any mean any kind of tokens, they would go to whatever it is.

>> Yeah, just to add a bit of context on on X402, it's it's an open standard and standard and and that's a very cool thing to me because uh of course it can be extended to many different use cases. I think it's worth to mention also the work that uh Stripe and and Temple are doing with this. There's a um um machine payments protocol that they released a few months ago and X actually built I mean, it's extending X402, implementing it as well, but allowing also like other type of uh machine and agents payments. Uh use cases using cars, Stripe, or any any kind of thing and it's built on HTTP. It it's a standard that was uh built uh I think back in the 2000s. So,

and and now Coinbase like brought it back to life. >> Awesome. And something I also wanted to build on top of that, how do we earn that trust? Because I think I've built my own AI agent and I sometimes give it money to buy me things, but I don't fully trust it. And how How essentially enabling that trust layer and that knowledge layer that you guys are talking about to essentially allow that mass and easier adoption because I feel like it's quite complicated to actually build your agent and actually then set it out. So, how do we enable like the people in this room and the normal people to actually accept this as the new standard, right?

>> So, basically at the moment is a stuff for tech guys and for nerd guys like us. The challenge is what we have told. So, we need to up to cut all the technical stuff and to bring a huge use cases for the normal people of the street. The problem now is that the market is very small. It's just emerging, but the challenge is huge because I mean or in my opinion in two three years, it will be very simple every have a is agents your own personal agents in order to do all the stuff regarding pay the bills, invest money. This is my vision, but at the moment is something for the tech

guys. >> Uh on top of the X 402, there is another open source standard that is the ERC 8004. 8004. So, actually that's really trendy at the moment and the cool stuff is that also two of the creators co-authors are Italians and one is based in Milan, one in San Fran San Francisco self. And one is from MetaMask and the other one from Ethereum Foundation. So, I mean we are bringing innovation actually in the space. And yeah the this standard the 8004 is basically the pedigree for the agent. So, it's like the ID. So, now you can tokenize your agents in the sense that you can give it an ash like an NFT and you can recognize

your agent. So, now you can give also skills, whatever and other functions, functionalities to that agent and now can transact with other agents. So, imagine the company A or my company talking with the company from Limone and actually they are transacting and trusting themselves, no? And actually these are simple transactions, are not trading security, we can get there afterwards, but this is a really trust layer that is needed in the space and this I don't know how many agents are already registered like Yeah, yeah, I also lost track. Maybe we are above like 200,000 already and it's still in review phase. It's not even like an official

ERC. >> Yeah, I think one cool property that this standard you were talking about 804 is bringing to the table is also the possibility for other agents to leave reviews on chain that can be verified by each other. So, of course there's still [snorts] a big work to do on the network effect side of things because of course if there are like 200,000 agents, but we don't know any of them and they don't interact with each other, they are not sending feedback to each other and and maybe the feedback that they get it's not a trusted feedback. So, there's a lot of like work that still has to be done there in terms of like distribution

and adoption of this standard, but at some point it should be like self-fulfilling somehow. So, the more agents, the more people are going to use agents, the more reviews and feedbacks we could have on some agents and and the more reputation that we would be built on chain and of course like composable and verifiable, the the easier it would be to to trust these kind of use cases. >> Amazing. So, something that you you do yourself is a focus on the identity of the agents, right Limone? So, would you Why do we need to give identity to these agents? Like what how does that work?

>> Of course there's a big spam problem when it comes to, you know, like building an agent, leaving it in the wild, and having it like do things. And many of these uh 804 registered agents are spammy agents. So, like they are not really worth it, let's say. And so, one of the experiments that we built in collaboration with the AI team of the Ethereum Foundation indeed is um this open standard called uh sign in with agent. Uh see what other ID, if you want to check, it's open source. And we tried to, you know, like build this uh protocol to enable agents identifying themselves. So, if you have a service, a server, an MCP, you're selling something to an agent, you can

actually implement this open standard in a way that it already like does a lot of like groundwork when it comes to verifying the agent identity on 804, for example. So, maybe you can uh you want to be sure that agents interacting with your use case, with your product, with your service has a certain tain um reputation, is a certain type of agent, it's verified by XYZ. Um and so, we tried to build all this boilerplate code basically that uh all these services will need to do in the future uh in order to make sure that good agents are actually interacting with them.

>> The problem of trust of agents is a huge problem right now. There are a lot of startups of projects that are are trying to solve this, but this is the next big challenge to agents because we need to avoid the spamming and all or also malicious actors in order to build the trust in agents in in agent economy. >> Um I do think that uh it is, and I it is because of a lack of data on chain, actually. So, this year, actually, what I'm seeing is that the real trend, the real big leap that we are taking altogether is bringing all the data on chain. And that's overlap with this new cycle that

is a gigantic economy starting using AI. And this is a huge stuff actually. So if we really agree that in the future all the economy will be based on blockchain and run by agent, then we need to push all the data there. And then the agent can be really trustable getting a right reviews etc. interact between them etc. And I don't know if it's already the next question. But if we want to like scale this, okay? Let's say that uh for example on rails uh the blockchain and we have private entities that are banks that are interacting between them and you guys are all the users. Then when you want to invest and you give the

money to the chain to the bank and they're going to invest and place the order for you and buy options, bonds, private credit, whatever you want. Then it will be an agent, right? So how can that agent interact with the security for example, okay? And a financial instrument and be trusted because he's the interacting for me or for Simone or for Alessandro or maybe for another partner. So how can an agent interact with our RWAs in this case without passing the KYC? So some service is already bringing on I mean in the space the KYA, but the KYA is not I mean is needed but is not enough. So something that we are also

putting there as an open source so everyone is invited to collaborate is called the ERC 8226. It's called the RAMs. Uh regulated agent mandate. And basically is okay, if this agent is transacting for me into security space, then he needs to verify money I mean the contract actually the RWA contract so it can be the 36, 43, uh 79, 43, whatever. Needs to uh check my KYC or the KYB to see if actually I'm a reliable like I can I can interact for that. I'm eligible. So, I mean, we are also I mean, adding a lot of layers of trust in order to allow

agents to transact with anything. So, that's the future that's the future agent. >> Because I think that now the web the web three economy is perfect for machine and the problems that the web three economy is not adapted for this new type of business because I mean, it's pretty simple. An agent can open a bank account, a traditional bank account. So, I think that it's pretty obvious that this all new type of economy will be on the web three rails. >> And micro transactions maybe are not that efficient with within banks.

>> I think maybe moving on like more on the scalability. Do you see any blockers right now that need to be addressed for like the agentic economy to actually move forward? Or And do you think the future will be or blockchain will be agents and then there'll be minimal human interaction? And I listened to a fireside chat an hour ago and something key that Lorenzo, the co-founder of USDT, said was that the chains aren't fast enough because if you think about agents will be doing tens of hundreds thousands of transactions and some of the chains right now only do like 10, 15,000 and we need to scale that even harder for them to specifically have those rails. So,

I'd like to hear your opinions on that. >> Um I will move the problem more on the adoption side. Like you mentioned that the use cases are mostly related to tech people and you need to be a geek to be, you know, like using an agent and interacting with one of these services that are online. Um and I've been trying to do like some work that give me like some some background on what's the the the current status. To me, for example, I I tried building an agent that could book for me a flight using of course like a a crypto wallet so that it could be like really fully automated. Um searching for a flight and and whatever.

The problem is not like building the agent and uh sourcing the flight is uh making sure that the service that can book the flight for me accepts um let's say USDC, USDT, whatever. Uh the same can go for ordering Amazon products. Unless these services that are out there, these APIs and uh products adopt their their own infrastructure, it's going to be hard because for if somebody like me, for example, I see potential in a new use case of again, booking flights with agents. I want I want it to be fully autonomous.

I don't want to build everything related to the flight booking. There's somebody that I'm sure already did that. Uh but they are not yet transitioning to these new type of things. I I I hope that Stripe is going to drive most of these, by the way. >> Yes, because Stripe is also building the wallets for a agents. And so I think that this will be a pretty good use case. I mean in my opinion, the most simple use cases is the one of Morpheus or decentralized exchange for a machine at the moment because of very very low need no KYC and all this stuff like this. And

so basically really the web three economy is built for machine, I think, at this time. We are entering a new type of web three era where all the agents will be on web three. Because it's pretty simple. You have all the advantages of the machines in web three economy, programmable and all the stuff like this. And now we are building the machine for helping us to do all the stuff. So, basically it's pretty simple use case. >> No, that's really interesting. And I wanted to like move forward as well with the AI agents. Like agents can buy, like you said, but what happens when AI agents meet regulated finance? How, from like your experiences, what are either

the challenges or like how does that even work? Because it's not a human buying it. There's must be different kind of regulations behind that. >> So, I think that very Sorry. I think that the regulation is not ready for this new type of business. We are in a gray area like the first year of blockchain. I mean, we are thinking about business of machine and and not human. So, all the stuff regarding the humans will not be applying to machines. I mean, basically as I told before, a machine cannot uh open a bank an additional bank account, cannot do a KYC because it's pretty a machine is also written in the books of law that only humans can do this type of stuff. So, basically

we'll we'll need a a huge regulatory stuff. But before this, I mean, we have to find the right business use case in order to uh adapt this agentic AI economy. Um >> I think that the regulation is always uh I mean, behind. That's the problem. I mean, we we see it and with technology, I mean, we are going really fast, but then uh we cannot operate. And before there was UniCredit talking here and saying, "Hey, 2 years is a short term uh horizon short horizon to really operate." It's like unfortunately, maybe yes. Okay.

Uh especially in Europe with regulation, I mean, we want we are completely like a risk risk adverse. Something that maybe in the US is not like that or China will be not that way. When agents meets a regulated finance actually, I mean, they are today already making trading. So, we have trading agents. But actually, we are not tokenizing. So, the regulated finance, what we do, it's still really hard. What Lorenzo before with Block Invest was saying is like it's not hard. Uh they are not POCs. I mean, they are really actual tokenization out there.

Things are moving, are moving better than before. The Clarity Act was a good sign. Like one one week ago, two week two weeks ago, actually, there was we got the penultimate step forward. So, the thing is, I mean, we need a standard trust layer the ramps. We need to build today for the agentic economy of tomorrow because we have to building all together. And actually, I mean, we I would like to see more participants actually in this standard like the 8004 as well. The people is really bringing a lot of value there in the trading of active magician, etc. So, it will be there. And we are building it

now. We will see this happening maybe in 3 years from now with real real volumes actually because as soon as we have the tokenized assets, then the agents will be already there. >> That's awesome. And is there anything that I haven't asked you guys that you think the people should really know whether it's from like your backgrounds, from your projects that you're building that essentially enables like the future of the AI agents working on chain? >> Let me think it. >> [snorts] >> Okay, so from Ricon, we are tokenizing RWAs, so any kind of asset. And there is the universal RWA standard 7943 that is a flexible

I mean, smart contract on top of all the others, okay? This is for assets. This is our vision. Second point, okay, what if we have agent tokenizing and trading actually regulated finance products or securities, and then we need the standard, okay? And then the last one, everybody's talking about vaults. Okay, vaulting our W A is equal to funds, okay? So, they are really heavily regulated. In that case, if we want agents to rebalance vaults, that is some really good use cases actually use case, we need another standard.

So, we need to allow agents to read all that information of the securities assets inside of the vaults and allow them to transact between one vault to the other. So, we really need to make make that data metadata readable for agents. So, I think I mean, a lot of things are missing, but we are really redefining the future of finance because we'll be agentic. And if self-sovereign is dead, then we need to build infrastructure, and this is what we're doing. >> Yeah, I I would just encourage everyone to start or try building an agent with Cloud Code X, like whatever, like it's it's going to be easier and easier and

easier. There's a lot of like resources that you don't even have to check yourself. You can send an agent to like read a documentation or or whatever, and it it figures it out. And you will learn much more doing that instead of just listening us here, probably. Uh because of course there you will see exactly where are the the bottlenecks when it comes to building something that might be useful for you. I I mentioned the the flight use case, but there could be like ton of them. And and maybe in one of these experiments you find an opportunity that is not being addressed already, and you might take over it.

>> Actually, Don is going big on agents. >> Yeah, [snorts] basically we found flat tone and we are building the deposit account on stable for human side, but on the other side from machine side, we wanted to build the banker for agents. So, when agents will will have the money to manage, he will know how to spend, where to spend in order to be helper in the financial stuff of the human control. >> How do you teach an agent to manage capital and prevent it from blowing itself up or like losing everything?

>> It's a matter of error and trying. So, basically, the more data you have, more are trained and more your infrastructure is perfect. So, I hope that my infrastructure will be used by a lot of agents, so it will be the most important infrastructure, most perfect. But, basically, it's a matter of try and fail. At the moment. >> Okay, interesting. So, from your perspective, where do you see, let's say, us sitting on this panel next year? What has changed and what have you hoped is going to happen next?

>> Basically, my big idea is that this type of stuff is not only for the tech guy. In a year will be also some commercial guys or business guys that are trying to do this type of stuff, also bring some capital on chain. More, more, more capital on chain. We need more capital. So, basically, this is spread the word and spread the technology. >> Same question goes for you both. >> I think there will be I mean, this bottleneck that I was mentioning, I think and I hope it will be filled because everybody will be running an agent and if you're selling something and you're not selling it to an agent, you're losing a big market. And so, like next

year maybe we can talk about the the coolest thing you bought your agent bought for you. >> I like it. My my vision is one year. I have no idea. Six months. Everyone is having an MCP that is an API for agent. And also a CLI CLI. So that means that you can from your chat saying hey, I want to launch a talk and I want to buy this. I want to tokenize that and and you will be able to do it. So you don't even need to build an agent. Ask to the agent how to build an agent to do that for you. So Yeah.

>> Very interesting. And we're getting close to the near the end of the talk. So I thought I'd open it up. But before we do, what's one What's the boldest claim that you'll be able to defend next year? On AI agents. >> What? Anything like >> claim. >> Difficult question. >> Like what? >> Yeah, exactly. >> Okay, my boldest claim I I fundamentally believe that AI agents will absolutely replace most payment systems next year for sure.

Like anything pretty much. >> I think the agents will be our next banker. We will not have human personal bank advisor by next year, but we will have a robotic financial advisor. >> I like that. I like that. That's a good one. >> Next year maybe it will be the 10% of the transactions in stable coin moved by agents. >> Okay, that's a good one as well. Okay, I agree with that.

>> I don't know. >> [laughter] >> But it could be. >> I think like something that for for sure like we will need to I don't know if it's my boldest claim but I would love it that to be the case that you know like these agent identity and and trust that we we've been talking about is going to be for real cuz right now it's all like very experimental and so we say that we are building trust and reputation but it takes time. So hopefully next year we are talking about real trust.

>> Amazing. Thank you so much and let's open up the questions to the public. >> Hey, Mikado here with Transia. I'm in the business by the way Limona if you want to talk later it's you rap travel API so they can accept crypto so we can talk about it. But I have another question. So if you ever tried to sell anything that you said to companies and I do as a business I I encounter a blocker which is when you talk to individual and you say yeah you have your wallet you can bring it

that money that you have already crypto that you have already on chain it works. When you talk to companies you have to tell them and to educate them to tokenize their money with on-ramping or whatever. And that becomes instantly a blocker because they need to open a kind of secondary accounting so it's that a finance department which will block it and say yeah but I don't want to even care about stable coins because they're not micro regulated they're not the equivalent of currency yet etc. How do you see those forecasts those bold claims that you're having match when it's not selling or it's not an individual trading it's not an individual using your services versus a company. So do you see in the next year business adoption as a thing or we need

to still wait for regulation to catch up? >> Um what I see, I mean um we need regulation otherwise it cannot happen simply. Uh the thing is um I don't know. I mean what we do, what I see is that we are in a big transition. It's a transition from off-chain to on-chain is the biggest pain in the space. I mean I tell you because of the fees, because of the user experience, etc. Today with the MPC wallets is way better.

And actually we do and perform that the reconciliation on chain. The thing that I can tell you based on data is that once that door is open then the people is moving faster on chain because they see the advantages. So I really hope this adoption is increase. And once we have the majority there on chain so everybody can trust and again is about the business being risk-adverse. Uh then we will see like this big change. But yeah, we need to be on chain to use all this technology we talked about. So >> I think that the business is a very slow adoption. First of all because we need to teach it to them what are stables and

how to manage stables. There are pretty some good use cases like cross-border payments but for the moment very small companies, I mean numbers, are trying to do these type of things. So I think that it's a never a process. So before we need to educate them to need to learn what is a stable, how to manage a stable and after we can think about the agents and all these cool stuff. But I think that's for the B2B side is pretty a nightmare at the moment. >> Yeah and and I think it's related not just to um let's say on-chain agents adoption.

It's about on-chain adoption. So if we cannot convince a business to move their accounting on chain because maybe their use case is not clear or is not actually relevant or useful for them. There's not like a real advantage, they would simply not move. So, maybe an agent maybe agents will help with that. I don't know because maybe they unlock like use cases that for for real like would have would have been possible without this, but still is a I think it's a broader problem.

>> If I can add something, we can create agents maybe someone here is already doing it to make reconciliations. So, all that accountability from on chain off chain etc. The agent can go on chain and get all the data and work for you or maybe offer this service. So, it would be cheaper and the easier. So, this maybe will open the door and make easier to get the taxes. >> That's awesome. Thanks, guys. There's also one more question I think over there at the going microphone. >> Yeah, it's me. So, I think we get the problem on the identity of agents because before we were talking about like every agent at one point will get its own identity, some bank accounts, its own email. We we already have

projects like this like Agent Mail that is one of the last YC batch. But the thing that I don't get is why do agents need to have their own identity when we can delegate our identity? Like when you buy on Amazon, you know that most of the times you are not getting scammed because it's Amazon. Like if I buy something on I don't know. I'm going to scam your house.com, I know that I I'm not going to get anything. So, right now I'm seeing that we are trying to create a problem with identity that it's kind of already solved. Like we have KYC that is for people.

We have agents that act on like behalf of people. So, why don't we just delegate our own identity to them? >> Because not all the agents uh have somebody behind like they are autonomous. If you really want to build autonomous use cases, you need a way for them to get an identity without me being in the middle, without any third party being in the middle because otherwise you're delegating the trust to Amazon or or whoever. And so, if you really want to see autonomous agents transacting, interacting, whatever like on chain, you need these type of primitives that are self-fulfilling.

>> Okay, it makes sense, but autonomous in behalf of who? Like they are doing stuff for which use case? >> I don't know. >> Okay, makes sense. >> No, no, I tell you, I tell you, that's exactly the route what we are building because it's all about layers. Now, so the thing is you have an agent that is 0x whatever, okay? And then on top of that is now in this moment acting on behalf of you, for example. And then you are the second layer that is called the principal, you know? So, I have to check if you are eligible or not, but are layers. So, if you want to interact with any commerce, you need another layer that is another standard, etc.

>> Yeah, that's the thing. Like that standard, I think it's going to be the merge between like the agent that is doing something and for who the agent is doing that something. So, I think that's the most important part on the identity side. >> Check it. 8226. >> Okay, got you. Uh it's a it's already a standard or it's a proposal? >> Uh it's a draft becoming going a review like in next few weeks. And then we'll stay there for a couple of months at least, I think. >> Okay, nice. Thank you, guys. >> Do you have time for one more question or I think there is a short one.

>> Yeah. >> I don't know who over there. >> Oh, you can shout it. I'm too far away. >> Hey guys, Asia from Tyco. Love the runs work. Checking for a little bit. One question, I guess two. First one like you mentioned MPP. I was just in Miami and OKX announced yet another payment standard. This is kind of like a new wave of like, you know, this new payment standard becoming kind of like like the L1 thing like always [snorts] there's always a new L1. Hopefully not.

And the second question is like what is your favorite agent as of today? I guess just if I can add I think like on the identity is also like what if an agent creates an agent that then like the agent gives identity to the other agent. It's not always the human. So I guess that's an important point to make but yeah, thank you. >> It will be the son of the agent. >> Yeah. >> Uh so good guys because we work also with Tyco and they are also in into agent K economy as well. And um What was the question again?

>> What about the OKX >> Yeah, exactly. So there will be a like a huge competition about like payment standards. I mean like we see with exchanges, with the C, with in anything, with the layer ones. So actually Raise is another layer one like Tyco etc. So the thing is yeah, here we are in a hyper competitive space and everybody will try to build their own uh thing. If it's not open source, so >> I think I didn't have time. I didn't check yet the this new OKX standard. I think that a Stripe and MPP and Tempo they have an unfair advantage with respect to anyone else in the market

simply because they have already a network of merchants selling ton of things online. And so it's just a matter of like getting them up up to speed with all these upgrades to the payment methods that they can start accept and I'm sure that they will also make it so it's very easy and they don't even see what's happening. Um, so that to me and and you know, I I see a lot of chatter on on X also about of course like Tempo being another L1 being permissioned permissionless like whatever.

Uh, but MPP per se is not uh does not give you like a a lock-in on the payment method. Like you can extend it with whatever you want. It's an open standard. And and I think that's the way. Of course like on that open standard when when you go like reading the MPP documentation you will see like baked in and and maybe like featured the Stripe payment methods because of course that's what they want to push, but then people are free to to build something on top of it and it's fine. >> And again, it's about the marketplaces, no? We're talking. So, uh, it's about volumes, it's about accessibility. So, where are the APIs that you really need? What standard are they used and built on? So, I think the adoption also will drive in that

direction. No, and we saw it there. Tons of standards, but then just a few are really used. >> Yeah, I think that the next billion startup will be the one that will be able like an Amazon for AI agents. So, basically like the Bazaar one that is building Coinbase but better and with no one under it. I mean, a two decentralized open trade marketplace for AI agents would be the next big things because we are building it, we want it, but they have to know how to deal with all these type of new things.

>> Bam. >> Thank you very much, everyone. I would like to give a huge round of applause for our guests. And if you have any questions for them, you can find them and yeah, have a great day, guys. Grazie. >> Thank you.