Interoperability and Scalability in the Age of AI | EBC11

Okay, nice. Thank you everyone for coming here today to this panel scalability and interpability in the age of AI. Um, Ethereum just turned 10 years this this year and it's amazing how we have been talking or at least I have the impression that we have been talking for scalability and interpretability since forever and it's amazing because it's still a relevant topic. A lot of things has been implemented, deployed and delivered and it's still something that we are we are seeing and we have the opportunity to to keep talking uh today and right now we are living in a multi-chain scenario where we have a lot of different layer ones lot of different layer twos that are working together and to to try to get all these moving parts uh we have seen how the AI has become a hot topic right And the the main question for this panel will be if how the AI is going to be able to help this scalability and all all the problems that we are and the challenges that we are facing. And before diving in in this topic, I would like to give uh some some quick time for an introduction of this amazing star lineup of panelist.

Uh Wes, do you want to start?

Yes. Uh Wes Crook, uh CEO of FP block. Uh we've been building in the blockchain space now since 2017. Uh just a little background history, we're the original auditors of Cardano. We have been a public the auditor of the swirls uh public blockchain for Hideera for the past 7 years.

We have about 110 projects under our belt and we just uh launched our come framework in August that help bring interoperability uh to web3 space. Hello everybody. My name is Alex. Uh u in blockchain for 10 years already building infrastructure. Um the most notable things about myself uh is that they've launched Rainbow Bridge, the first decentralized uh bridge in between Ethereum and other network.

Uh and the latest things that I'm working on is Aurora. This is the EVM compatible solution on top of Near as well as Near intents. uh probably one of the biggest uh crosschain trading uh platform at the moment.

Hi, I'm Mari. I guess we're working in blockchain as well for too long. Um for Rakitin, Rakton has a big interest in blockchain particularly in Japan uh where they're based. We run a crypto exchange in Japan. We did sort of play around with an NFT marketplace for a while, but collectibles kind of, you know, ran ran their time really without utility.

Uh we then worked on we're doing a web3 wallet so an MCP based consumer wallet uh really now trying to focus on opening up um partnerships and web 3 and DAPs to to rack and customers. Okay.

Hi everyone. I'm Norbert from ZK Cloud Zenith. Um in the like in the past 2 3 years we've been building ZK proving infrastructure with ZK cloud and in the past couple of months we've been focusing on adding uh an execution layer on top. Um that's what we call Zenith allowing enterprises to spin up natively verifiable multivvm blockchain environments in a in a very customizable and modular way. Um happy to be here.

Thank you.

Amazing. We have a so nice speakers with different backgrounds and for all of you to try to give some context. Maybe some of you you are not so technical and could have the the same feeling that I was mentioning at the beginning that um these tech guys these persons are always talking about the same topics. They are always talking about the scalability and how they are going to get the fully crosschain compatibility and all of these stuff. So to give some context to for the people who is here, why do you think that we are still 10 years after Ethereum came uh why is still so relevant?

Why why do you think that we are still talking about uh these topics and what are the next things that we could start building?

I think uh from our opinion our p perspective there's just too much fragmentation. The industry in our opinion is upside down. We have ecosystems that are funding uh application builders that are then bribing the users to come to the space. And the reality of it is we have no users in web 3. Uh very very small, you know, class of people are using the applications.

They're all built to drive tokconomics. They're all built from a ecosystems perspective. So the whole idea of interoperability is becoming prevalent because that's where the capital markets are driving us because they're tired of seeing our industry not look at the user experience and building applications that users actually want to use and then the enterprises don't want to come into this space because why do I want to build an application that I'll have to support on three or four different chains that's just not feasible for an enterprise you know with limited you know with even though they have large IT budgets, they have limited R&D money uh to go build on. So what we see is until we as an industry have a common set of rails, a common set of standards, and a common infrastructure, we're not going to solve the interoperability problem. To us, interoperability will come once we have a common platform.

Um do you agree, Alex? Sure.

Uh in general, yes. So the thesis that a crypto doesn't have users I tend to agree with this because the majority of people who are in crypto are not users are they are investors and that's a different type of a person um in terms of the uh interperability standards and the way how how we are able to make the user experience much better than we have at the moment. I do believe this is just like the age problem. Um, crypto for me has always been reminiscent of the previous big technology that I was working in, which is the general purpose computation on GPUs, which now is known as like powering AI. Um, and now there is a very specific and straightforward point of application.

But when I was starting to to to go in there like we needed to do very low-level things and in case you are going to change the nested loops um then you're going to have instead of the acceleration of your compute you're going to have a deceleration 10 times. So yeah, I think the technology is a little bit too early and because of the like very big focus on figuring out the base layer of the technology, we are forcing users to think about things like stuck transactions, nonsense, you know, gas payments, all of these things is not something that is happening with your Apple Pay, right? And uh um we're I'm pretty sure that we are going to resolve this. Um, and there are lots of very promising developments in the interoperability space, especially specifically with the focus on AI.

Marty, okay, it's probably slightly different view in this. Um, when we first started out, we built our first app on top of UTXO and at time it was our background. We came from Bitcoin and so we kind of we kind of stuck with what we knew. Um, but we came to the realization we're building a custom scheme that really was never going to be integrated to by anyone else but ourselves. and looking at what Ethereum were doing.

uh for us really the fact that there was you know published interfaces that ERC's and ERPs were there that we could build our own sort of collateral against these interfaces and consume other people's tokens on the same interface really made a lot of sense for us for interoperability because now really we don't have to provide so much of that the integration you know that's just chain level integration if you know for us if you give me an API and an RPC endpoint I can probably talk to what's behind that So that for us really made a lot of sense and really for Ethereum has grown the ecosystem just by having interfaces that people comply with. So I think that's been really valuable. I think that the problem we're starting to see is yes there's fragmentation or or there's diversity. There's now a lot of chains and a lot of tokens and certainly from an enterprise perspective scaling to support all of that is difficult. you know, we we kind of went from certainly the early days of running infrastructure, you know, is to now almost looking for a provider.

So, we try not to run infrastructure just for RPC connectivity because there's too much of it, but but certainly those interfaces make it easier for us to connect different tokens on different chains um pretty easily to be honest. I mean there is there is friction. And I know we're getting into but from an interoperability that has really helped um how we build our stacks and even sideways integration things like wallet connect how we integrate with DAPs know having those interfaces means we're not building a custom integration to to anything where we try not to stop there.

Yeah. So from my side I agree with uh what the others have said. Um I do think that UX has to improve a whole lot. Um because like blockchain itself should actually be unnoticeable for the end users and it should actually be like kind of limitless or like with have unbounded performance. And um with ZK cloud um we started focusing on uh on ZK proving infrastructure because we do believe that zero knowledge as a technology can actually help both in interoperability and scalability which could make the UX more smooth and seamless and un like blockchains becoming more unnoticeable behind uh the front ends and also more performant.

Uh and what I mean is um just I don't know how familiar you guys are with ZK technology but on a high level um with zero knowledge proofs you can prove the correctness of any data any computation any statement um without revealing the underlying data per se if you need to hide that. Um, and what it means is you take a ZKVM like it's a VM executing your your code, your program logic, but it doesn't just execute it, it also creates a proof that it has been executed correctly. And with this, when we talk about interoperability, um, the performance of these ZKVMs have been going through a tremendous improvement and we are seeing real time transaction proving happening even now. And with that you can actually you know prove the execution of an event on one chain to another chain where it can trigger another event. You can prove uh that finality has been reached here and you can reach finality there too.

uh the the mechanisms around that need to be figured out I think because the solutions that we have now partially already using ZK are good for the current rate of adoption but for like billions of people using uh blockchains I guess there is still quite a bit of improvement needed on the other hand for the scalability aspect with ZK you can move away from the traditional functioning of blockchains as in like re-executing the in block to verify that the block proposer has done everything correctly. You can just do proof verification and off of that um uh a test to the block and like proof verification has very minimal requirements. even your smart fridge actually is soon being able to uh verify a proof and then the other thing is with ZK you can do a lot more computation offchain increasing like you don't have the limits of the EVM for example when we talk about computation capacity but um you pro you you send back the proof or or verify the proof back on chain and with that you actually can scale the the the size of the computation that's possible through through blockchains as well. So I think um ZK is going to be core in both.

Okay. And talking and yeah centering the the conversation around uh interpretability. You have mentioned things related like uh UX which is is a a key point in my opinion it's one of the key points that will um make things easier for the final users. uh do you think that the right now the challenges that we are facing to achieve a common or to try to achieve this utopical world where we have true crosschain compatibility that is seamless for the end user? Do you think that the current challenges that we have right now are more of um technical thing that we have technical things that we still need to to solve or do you think that is a coordination or an offchain thing that is not related with the limitations of the current technology?

If I understood the question correctly, you know, I think the user experience, the user community doesn't care about all this. All they want to know is when I hit the button on my phone, does my application work? And you know the fragmentation within the web 3 space with all the different chains and we're talking about interoperability. You know I believe personally and I you know I'll say it 85% of the names that we talk about today won't exist in 5 years. They'll be like AOL they'll be like you know um you know AOL for example.

I think what we're going to see with the technologies because there's a lot of great technologies. We don't have a winner yet. We don't have a clear who's going to provide that infrastructure standard. We're going to have a Netscape moment. We're going to have a DOS moment.

Somewhere in our technology cycle that will set that standard for us. Um, you know, if you're familiar with the internet, up until Netscape, it was all about scripting and it was very difficult to get to it. Then www came to life and here we go. We'll have that here in this space because users again don't care about the infrastructure other than they want to know that when they hit the button it works, their money is secure, their transaction happens and it's not difficult to do. And in web 3, we have a lot of challenges with wallets and wallet connects and interoperability challenges that we all talk about, but we're not really solving those problems other than from the perspective of each chain.

Each chain saying, "Hey, I have an interoperability solution, but it's not going across everybody. It's not a common standard. Everybody's doing it a little bit differently." So, it makes it very hard for enterprise. is it makes it very hard for people building applications for users uh not investors um to really figure out how to go build and build it correctly.

So users want to use the applications and want to drive the revenues into web 3 that we have today. So, you know, that's one of the reasons we did Colme was to help, you know, we're not saying that's the end all. We're just that's our vision of how we can get infrastructure commonality and drive some of the technical challenges and and some of the roadblocks with gas fees and congestion and things of that nature out of it for the builders so they can just build applications wherever they're coming from web 2 or web 3 native and move forward. So from a user perspective, you know, all the conversations around interoperability, all the technologies, they don't care. They just want it to work.

And, you know, it's our job as an industry to figure that out. And I don't think we're there yet. I think there's too much infighting, too much uh fragmentation, and then over the next couple of years, you're going to start seeing the real players come out. Don't know who those are yet, but somebody will come out.

Okay. uh Alex do you think that we have that we went over that barrier and that we don't have this technical challenges as it's more of a coordination problem

so I think that obviously different ecosystems have their own uh uh standards and their own approaches uh and uh the thing that reinforces their focus on these approaches is the underlying tokens and the value of this token So there is this concept of these thick protocols that are acrewing value through the token and because of this they are trying to you know enforce their standards everywhere and uh however this is just like is going to be resolved on a higher level right so um for example with near intense uh at the moment every day you know more than $50 million are traveling crosschains and uh at extremely low rates uh and very very quick. So the system is built in a way that people do not need to think um about how to fund transactions on Cardano versus Salana versus Tron versus Dogecoin or Zcash. So all of these is just working for them. It's very straightforward and simple and uh you can't build systems that are going to abstract away things from the users and uh and like that's that's a pretty it just takes time. However, also another question here is that what is the world is going to be in couple of years from now or in three years from now.

Do we need actually to build systems for humans to enable a good user experience for them? Is it okay to have you know semi-done systems that are going to be operated by user AI agents which are going to find the most efficient way how to transfer value uh even without a good user interface and um there can be different points of view on this but clearly what we see from the development in the AI vertical over the last years is that AI becomes more and more powerful and uh more and more companies around the world are experimenting around the agentic economy which is going to fully abstract away all of these complexities of the low-level technologies. So I think that we are at the tipping point where everyone need to honestly question themselves whom they are building their products for whether it is humans or it is going to be AI and depending on the answer to this question the design principles for the products are going to be very very different. Um, Marty. Okay.

Um, that's that's an interesting segue because we were we were thinking about the same thing. I know Alex uh we were chatting last week and the point you made to me that I thought it's quite quite interesting is that a really rubbish user experience for a human probably doesn't care to be an AI. Um, which which is quite good. Um for us on the UX side, I think there's starting to see enough of standards of capabilities on chains to let us abstract away some of the friction. So you know native tokens is usually one of the pain points cuz you know people generally want to use a some sort of token whether it's for a game or for whatever for utility for marketing.

We are we're being experimented with sort of different tokens but it's the token that has the value. It's the token that people are interested in. The fact that we also have to deal with native tokens on chains is friction and if the user has to deal with that that's more friction. solely. But now with you know gas abstraction and account abstraction, we can start taking some of that that pain away and get a much better UX cuz I think everyone's right until web 3 gets as convenient and as usable as web two never really going to get adoption and and so we get to take these friction points away.

interesting on on the AI side of things where we've started experimenting with Agentic payments um taking AP2 and X42 and just they're they're a framework that kind of lets you reason about how that payment flow could work from an AI agent down to a chain and there again when you you peel back some of the layers behind that it's rooted in two more ERC's that that that help you sponsor transactions particularly on stable coins. Um the friction now that we're kind of trying to work our way around now is, you know, generally to date we've had, you know, token A on chain A, token B and chain B. Now we're dealing with, you know, a USDC coin that's on God's amount of chains. No, it's everywhere. So how do we handle that now?

Do we restrict ourselves and say we're only going to accept stay on base, but yet market cap for USDC, I think the second biggest is Salana. So, we're going to say, well, we're not interested in in customers who use Salana or now we're back to how are we going to bridge how are we going to, you know, have an experience that just says I just want USDC. I generally don't care what chain it's on. As a payment rail, I want to accept it as a consumer. I have it.

So, therefore, we kind of need to work through the integrations of to make that work in a seamless way. And I know again I know near does it sort of at a protocol level and then there's obviously the RC's around intents uh sort of crosschain level. So these things are here we I just not there yet with now what's the best way to jump because we have multiple ways to get around that and for app developers in enterprise it's quite a big cognitive load to get over um around scaling that across multiple chains. Um but that that's the challenge. Yeah, not worth.

Yeah, I uh I do agree with you guys have said. I I would mention maybe one quick thing from my side as someone um dealing more on with with zero knowledge, how it can be used in blockchain and outside blockchain as well. I think one key technical blocker has mostly been solved. It's the performance of of general purpose ZKVMs and that's that's really uh unlocking um a lot of things that have not been possible even a year ago. Uh but I want to mention one thing which is kind of technical but also conceptual a little bit related to maybe a future standard.

Uh and then last week when we chatted, Alex mentioned asynchronous models like current interoperability solutions like I do agree with all the all the agentic uh economy coming up but they will rely on existing or or like the infrastructure that is that is underneath. So most of the interop um solutions are currently built or a lot of them with synchronous uh assumptions and like that is a blocker for scaling and I think I I don't want to go too deep and I think Alex has a has a much deeper view on this but just the fact that you could start moving to a standard where you don't need sequential execution and one ch one event waiting for the other event to be finalized and then taking a lot of time to to actually execute the entire process flow. You could actually have asynchronous composibility. You could have the different events proven and then uh basically eventually get finality for all of them. I think it's it's kind of like a rabbit hole.

I don't want to open that. Maybe Alex has uh some view to add on on that. But I think this is still something that would be good to explore and maybe to become a standard in the future. Uh I don't know.

Yeah. Uh you have opened the Pandora's box. I I think it's a super interesting topic to talk to talk about or to start thinking in what in the future we could face around yeah synchronous operations in EVM or repeatable machine. Uh but yeah it could be a rabbit hole. So uh you have mentioned you have brought Alex brought two really interesting topics.

The first one he start talking about intense and also about AI. So uh there's some people right now that we have seen that everybody's talking about AI. Everybody knows what it is. Everybody have a barely understanding of of it. So uh but there's also some people that thinks that the we have super hype technology or that we may over that it could be a phase that after the few or in the coming years we are trying to get some or some people could have the feeling that we are trying to put AI in everything and yesterday I was listening one of these panels and there was uh a panelist who mentioned that AI will be the killer application for blockchain.

Do you think that I think you kind of agree that AI has a a place in blockchain, but do you think that we have a lot of hype or do you really think that AI is going to help improve this interpolability and scalability issues that we are facing and also super interesting that we may not have this user experience in mind and we will be open the door to this agentic experience. Well, I think you know AI is here to stay. Um, we are doing a lot of work in AI. We've actually built our first application strictly using AI and it's been out in the market performing very well um for sports analytics and you know we we have incorporated AI into our business operations. So we do our proposals, we do our test suites, we do our um documentation leveraging AI.

But the one thing that we have experimented with and we see and this goes back to the agents that you're talking about, we need to make sure AI has integrity because if you look at a lot of the of the tools that are coming out right now, they work great the first month. And as you kind of go through, that's why you're seeing so many versions come out is they're having to clean up their language models. So what's happening is like we were working with chat GPT5 the other day with something and it just went awful, you know. So there's there's a integrity issue with the language models and making sure when you build it that you're not ingesting garbage that will degrade the use of the the model. So there's a lot of work that still has to happen in AI.

It is a huge bubble right now. I think uh somebody yesterday was talking about how 26 is going to be the year of the AI bubble b you know blow up but um you know it is here to stay. It has its advantages. We will use it from a developer standpoint to iterate faster. Um it's going to cause a you know we're not people are worried about losing their jobs with AI.

They're not. We're going to need more engineers that have to have the knowledge on how to manage the AI outputs. But what it's going to create is a gap. We're going to lose the ability of junior engineers, people coming out of school wanting to be in this space. You're seeing it in the US now when you hear Zuckerberg spending $100 million hiring engineers for AI building.

It's there there's going to be a level of these senior engineers making huge amounts of money. And over time, we're not going to have this this hierarchy that we've had for years and years and years where you come out of school, you start off, you're writing scripts, you're doing testing, you're doing all the things and learning how to be a good engineer in writing software. We're going to have a talent gap there and there's going to be a huge gap here that we got to figure out how to fill and AI is is going is causing that. So we got to figure out how do we train the next generation of user of engineers in how to use AI. You know, we've we're losing critical thinking skills as a as a society.

Um I see it every day, you know. So things that I had, you know, engineers do before at a junior level. We're just letting the AI agents do it for us. So, you know, what do I do with this guy? I can't pay this guy, you know, $75,000 a year to sit around and look at the walls.

and I and I can't wait for him to learn the skill sets to manage an AI output. So, we really have a that's a big problem to me with the AI um in the integrity issue, you know, so we got to solve those problems uh or we're just going to, you know, just kind of have this very select group of people controlling what's going on out there. So, it's it's an interesting dynamic for us today. So from my point uh the interoperation between AI and blockchain is uh almost inevitable. Uh autonomous AI agents need to transact value.

This is a core thing that everybody do. All of the businesses, all of the humans, if I want to get something, I usually need to pay for it. So AI agents would need to be able to transfer value. And we have two options right now. for them either to apply for ordinary and use the ordinary financial rails for the value transfer or use the blockchain.

Financial rails the usage of it was for hundreds of years perfected to focus on humans and everything around humans. You cannot open a bank account without a passport, right? That's a very straightforward thing. And AI agents, they don't have passports. So in order for AI agents to be uh usefully using financial system there needs to be tons of changes in the regulation and governments are not the quickest moving entities.

So where is my bet on what is going to be the natural value transfer for AI economy? Is it going to be the old very uh stable but also slowm moving traditional financial rails or it is going to be the magic internet money? Uh in order to use them you need only to have a private key that any AI agent already has. Right? So my bet is clearly on the second half and I do believe that uh many AI companies are actually understanding this and that's the reason why Sam Alman is partnering with stripe that's the reason why stripe is building their own blockchain that's the reason why circle tether are building their own blockchains everybody understand it blockchain is the enabler of the value transfer use case for the future AI economy And uh like from my point of view, blockchain vertical should not hack it up and we need to keep up with all of the increasing demand from the AI because they require scalability.

They require interoperability in between the blockchains. Uh and we need to meet this demand. This is one of the best opportunities for the blockchain to get a proper product market fit besides only uh gambling uh on the uh on the on the tokens right so um I encourage everybody to build towards this goal

market one non well sorry non-blockchain related I suppose part of that for Wesley I mean I I read a really good quote that AI makes a good engineering team better than a bad engineering team worse because we get lazy you know uh relying on it. Um for us the AI use case u we're starting simple cuz say we were experimenting with the J pay genantic payments and really how we could create MCP services that we could monetize through microtransactions um using stable coins that's low risk you know low value but for us a way to learn how that technology all hangs together and and for us it takes out some of the non-deterministic piece of it you know for the AI still nervousness about it nondeterministic properties you know it just can go slightly rogue on you every now and again so and and and again is that a laziness thing you know are we going to let AI patch the things that we find too hard to engineer properly um don't know it's just just a thought in my head at the minute um but I am conscious that we see AI as you know it's the new hammer and everything turns into a nail and we apply it to everything. Uh sometimes, you know, some of it is just good engineering. Um the other thing I was going to say on all of this was funny on AP2 what they again sort of that AI intersect with blockchain. I got asked this about 3 years ago and I I honestly said I just don't know.

I really couldn't figure out what the intersect looks like. And you know now I'm a bit more clear around know data verification, authentication and payments. And on AP2, they see the AP2 layer as the authentication, the trust layer for payments. And interestingly, they're using verifiable credentials to implement it. So there's another intersect with blockchain where we can use VCs to control and verify what an agent does that it's doing what was intended by the user.

So, so that's where I am with it all.

Just a maybe a last thought on the comment. I think the biggest impact of AI merged with with blockchains through the agentic economy will be for the users. And I think this is also a type of like scaling and and maybe improving interoperability or or more making blockchains to become unnoticeable and seamless. But just the fact that on existing infrastructure agents can find the most efficient routing paths for user intents and uh you know make user transactions faster, more scalable, more efficient, make it cheaper. That's going to be a that's going to have a huge impact.

So it's not directly impacting the infrastructure behind interoperability and scalability but rather using it in a way that it boils down to uh a lot of efficiency gain for the user.

Yeah. uh we are running out of time so I would like to introduce one last topic and it's related with um enterprise adoption and I would like to ask you quite briefly um if you think that we have in this conference we have a lot of enterprises lot of institutions and for the last years we have been hearing that um the technology is not material yet that some companies that will want to go from some web two companies that wants to jump into the web3 they will they could experience some bottleneck. Do you think that we are still in that situation or we are on on a perfect time where some web two companies could fit in the web 3 space?

Again, I think it boils down to, you know, enterprises coming into this space. It's about choices and there's too many choices. We need to provide a common infrastructure that people can build on. um you know an enterprise uh you know you take a large enterprise financial institutions I've talked with a lot of folks at Bank of America JP Morgan other state street bank they have their own blockchain groups going but the counterparty risk the compliance issues the regulatory issues are keeping them out um and and again from a cost of operations you know you can only build an application once and maintain it you can't keep building and changing as the blockchain landscape shifts. So having one, you know, infrastructure and set of rules that these guys can build to, then we can drive to the regulatory levels and the compliance levels and the counterparty risk levels that will allow these bigger institutions to come in and play and be a part of it.

And then when we get that, then we're going to really take off as an industry because that's where the real money is. You know, we have a lot of people slloshing around and doing investments now, but you know, when you talk to, you know, folks at Fidelity and all the big financial houses out of the US that I I talked to, you know, they're just waiting. They're waiting for some kind of uh epiphany where we have one set of rules that we could go play by and then, you know, then they're going to try to make their own rules, but that's a different story. But that's why you're seeing regulation, you know, slow adoption. the European regulatory environment's gotten really bad.

Uh US is, you know, we've seen it with, you know, the axe with Trump, but you know, the problem in the US is I can go do something in in Wyoming, which is a very pro- Bitcoin uh blockchain state, and I can get sued in Oregon for the same thing. So, we still have a lot of regulatory mess to clean up. So, enterprises don't want to get married down in that. It's too much money to be spent. So, we still have a lot of work to do.

um you know getting a common you know interoperability whatever you want to call it but having a common set of rails a common set of infrastructure you know using agents whatever but we have to be able to have one you know it it this is what we do for infrastructure in web 3 and that's how we do it

super quick uh

so we are out of time but super quick from my point of view enterprises started to adopt blockchain and AI and the best example of this is the person from rakuten who is sitting just on the left hand side from me. This is the best uh example of that this thing is happening

really really quick. Um it it's hard to scale in enterprise and that's why Racketin put I guess a a a small team it's a small engineering team in Belfast um to help them work through this space.

Yeah. So from my side just very quick I think the complexity that enterprises are facing is not going to be like pure for example TPS related uh but more like how do you manage state consistency across a multi-chain transaction or in a multi-chain environment how do you manage finality how do you manage u uh all these types of things so I think like what we have seen is enterprises want a a more centralized solution most of the time or that's what we've been uh coming along a single entity and a single operator in a blockchain can scale much easier than in a decentralized setup. So that's why I think it's going to be more complex challenges in terms of like transaction composition etc etc uh that that these um enterp enterprises will face. Okay, nice. Uh, thank you very much for attending this this interesting topic.

Uh, please big applause to our panelist and I think I have to say that this is the last uh panel before going to to lunch and that we will have around 1 hour and everything will be back in at 15. Yeah, I'm your new host now. But um thank you guys so much.