Welcome to the show. We have on the show today, Alex, who is uh the head of Aurora Labs and also one of the OG developers of NEAR, and he's focused right now on the NEAR Intents product. Welcome, Alex. Hello Peter and thanks for having me here. I want to begin with a quote from AIXBT on Twitter that um it or he or she shared a couple of days ago um and use that as a jumping off point for our discussion. It said, near protocol just flipped to number one cross-chain infrastructure with 35 % market share processing 30 billion annualized volume. ThorSwap, Zcash, Stableflow all routing through near intents now. um Ocean Pal raising $120 million to buy 10 % of supply protocol doing chain link job at 0.35x the valuation. So the other stuff doesn't matter so much, but what's really amazing is that 35 % as of this writing, think it's higher now. Market share. That's amazing. Well, let's get into it. How was NEAR through NEAR Intents able to capture that much cross-chain? I guess, transactions in such a high market share. Maybe give us a sense of what near intents is and tell us about that. Yeah, absolutely. So probably everybody who are listening know a little bit about NEAR, but the whole idea is that this is a scalable L1 that was launched in 2018. The idea behind it was to build a scalable blockchain infrastructure. And the approach that NEAR has taken there is through the asynchronous runtime that allows for sharding or like parallel execution of the transactions. And the protocol has been innovating a lot during this time from 2000 for the past seven years. And one of the innovations were, one of the set of innovations were focused around the cross-chain interoperability. And you introduced a thing that is called chain signatures. This is a primitive that allows near-site smart contracts to control uh that you have private keys. And these private keys can be EOAs on other chains. It can be like admin keys for some smart contracts on other chains and stuff like that. And on top of this then, another wonder was built, uh which is called Omnibridge. And this is the solution that allows you to kind of lock assets on uh one chain. and then mint the corresponding wrapped assets or synthetic assets on the Neer chain. And what Neer Intents says at this moment in time, this is a cross-chain index that allows you to trade these assets and move assets across different blockchains. This is a core value proposition, again, as of now, of Neer Intents. uh And that's why... uh The numbers around Neural Intents are the numbers around the cross-chain, decentralized cross-chain swapping or the volume or like the inter-rope thing uh that is happening between the chains. Now, if we are going to go into the question how we are able to capture this volume, uh the kind of secret sauce for us is because of these chain signatures and because of this feature. that near smart contracts are able to have private keys that correspond to the accounts on other networks. uh For us, in order to connect different L1s and L2s and any other networks uh to the near-intense protocol, there is not a lot of work that needs to be done. So in comparison with other protocols that are mainly focusing on EVM compatible chains and they are developing their smart contracts for EVM compatible chains, which means that they are... extremely hard to convert into, for example, Cardano runtime or Stellar runtime, because these are custom runtimes. Since NeurIntense doesn't need to develop the smart contracts for these chains, we are able to uh connect these runtimes very, very quick, including the chains without runtimes, like Bitcoin or Zcash or XRP uh or LTC. And all of these chains are the chains that are connected to the Neuron 10s. And instead of going into like red sea of PVP in between the cross-chain protocols that are allowing for interoperability and the liquidity movement in between the EVM chains, we focus on all of the biggest markets and we enable them. Yeah, we've had a number of interoperability infrastructure projects on the show, uh Hyperlane, uh Axeler. Maybe we can back up a little bit and talk about like the interoperability kind of cross-chain market and kind of various approaches to uh cross-chain transactions and then maybe put in like how near-intense is different from them. I think that would be helpful for the audience. Is that something that you'd be willing to help us with? Okay. absolutely, absolutely. So I actually has been debating with myself on writing like uh public cry to the crypto Twitter. to say that our categorization is extremely broken within this kind of cross-chain and in general swap and thin because from the standpoint of the user, DEXs, bridges, cross-chain interrupt solutions, DEX aggregators, bridge aggregators, they are all doing the same thing. The user puts one token in and gets another token out. So from the standpoint of the user, this categorization doesn't make any sense. But also from the developer's standpoint, this categorization is also not making sense because there are very different approaches how to build like these kind of swapping solutions. You have on one hand, uh AMM style DEXs. where the liquidity that is used for trading is held by the smart contract. There are prop DEXs that are on the raise at the moment. This is a different type of a solution. There are like native slash wrapping bridges. These bridges are... locking assets on one side and then they're minting assets on the other side. So for these uh solutions it is extremely simple to move assets they don't need liquidity it's just like straight execution of like do first thing and then do second thing. And these solutions are the solutions like, for example, layer 0. And their product, use the t0, is exactly like that. So it's so simple to move use the t0 in between different chains. you literally can put. any volumes on the USDT zero just because it's just simple. You don't need to have a fee. Nobody's taking any risk of holding these assets and holding this liquidity. But the most complicated solution is the type of a solution where there are two assets that traded that are not wrapped. So if you're trading Bitcoin into Solana, You cannot wrap Bitcoin into Solana. These are completely different assets. If you want somebody to allow for this trading, you need to find a counterparty that is actually providing the output token and taking the input token, which means that this counterparty is having this uh kind of position that is going against their initial holding. and which means that they need to think somehow how to neutralize this position, what to do with that stuff like that. This is exactly what near intense is, and that's why this is something that uh is described inside of the intense uh words or intense buzzword for the cross-chain when the user is of locking one asset or providing one asset to the market maker on one end. and this market maker is providing the assets on the other end. Now, the decentralization part of these protocols are about that there can be many different market makers that are connected. The protocol is ensuring the safety and security of both market makers that are participating in the trade and the user, and then that it is an open system, so anybody are able to use this thing. Yeah, so this system is the most complicated. that is kind of like a trading, proper trading system. However, it is the most lucrative just because the biggest assets are not wrapped assets. The biggest assets are Bitcoin, Ethereum or ETH itself, which is available in multiple chains. And then stable coins, BNB, XRP, near Aurora, obviously. And these assets are not wrapped. So in order to allow users to convert one into the other, you need to have a trading body. You need to have a counterparty that is going to give you this liquidity. So, uh yeah, this is the kind of categorization in terms of the kind of functional or like the way how things are working. Unfortunately, everything that we have at the moment on Defile AMA, on token terminal. Gecko terminal. These kind of categories are not aligned with this stuff and um yeah, I hope that at some point in time we're going to clean up this mess and somehow put proper names for these things. However, maybe we actually don't need to do this. We just need to put their swap in provider and that's it. But I do think that there is a little bit more granularity into it than just a single category. because like single chain swapping is something different rather than the cross chain swap. Yeah, the from you know, various perspectives like from the user You know, it's always been weird right to have to kind of use a bridge to be able to you know use that asset on a different chain just having to do all of the There's just a lot of work. And you and I talked before the show started around about Infinex and how Infinex kind of, it encapsulates all of the complexity and hides it from the customer. And I love that approach and that, and it's able to do that through near intense, which is really amazing because as a user, like I don't, it almost doesn't matter to me, like how it's done. I just need it done. And I'd say that probably exp... That probably describes, I don't know, maybe 75%, 80 % of users. But there's going to be a set of users that actually want to know how it's done, make sure that there's security guarantees, et cetera. But I'd say that's probably a smaller percentage. And these are mostly folks that are really interested in that. But for most people, they just need something done, and they want to move on with their life. I'm with you, I'm completely with you here. this is probably the users who are very, very in-depth how the systems are working. In the majority of the cases, they are probably not the users of wallets. They hold their assets in a custodian solution and they are just somehow distances. These are managed. And they are requiring these custody solutions to... to be super, super secure. But also this bridging uh thing is extremely, extremely, these complexities, these low-level blockchain complexities are extremely far from what we experience in a day-to-day life. If I'm going on a uh conference uh in some Latin American country uh or in Argentina, then I'm taking my European card with me and I'm just using it to pay for coffee or something. And in the background, within these three seconds when I'm tapping my phone to the point of sale terminal, there are a bunch of things that are happening. Like there are these different complicated things like acquirers and then the banks and then... they somehow connect with each other. There is a conversion of my euros in my account into whatever currency that the merchant is accepting, which may be determined by the local legislation. then there is a settlement that is happening sometime later in between these different banks. And all of this complexity is not something that people are thinking about every day. It just works. and web2 as delivered here. finance has delivered here, extremely good user experience. And they are hiding things very, very good from the users because people just want to buy coffee. They don't want to think about which acquirers to use or through which banks to route the transfers or when it is going to be cleared and settled. This is not... something that is comprehensible by the uh ordinary person. And they want the solution to be delivered. They want what they want to be delivered to them uh right here, right now. And I do believe that this is just, for crypto, it is like a childhood problem. So we're just going to grow out from our little pants and we are going to build a better user experience. My whole 10-year career in crypto, I've been pro building good user experiences. And I think that's what the technology level that we have at the moment, we are able to get closer and closer to this very, very good user experience. That is at least... at the level of the Web2, but in many cases it is actually better just because the settlement time of the blockchains uh is usually seconds rather than days, which is for the traditional financial rates. Yeah, yeah, I agree with you we are as an industry we are maturing, you know, we're still you know, little children, but we're kind of growing up into adolescents and eventually we'll be as mature as, uh you know, the web too that all of us experience. The example you shared of, you know, buying coffee, like in the crypto example, like you'd have to sign every transaction, you know, like I want to use this route. I want to use this bank, this acquirer. Let me approve that. It's like, it is... it would be a terrible experience, but in web two, it just happens. And it's refreshing to see that your approach with near intents is to mimic that kind of user experience. Let's, maybe we can do like a side by side. I think most people are familiar with uh the bridge, kind of like the bridge approach to cross-chain uh messaging, cross-chain swapping. How are near intents different from that? Yeah. So the way how NeurIntents are working is that there are like three steps to every single swap that is happening uh in the majority of the cases. So all of the B2B partners of NeurIntents, so all of the distribution channels, big wallets, different cross-chain indexes, frontends, aggregators, they are taking the API of NeurIntents, which covers these three steps. The first step is the deposit of user funds inside of the Intents protocol, then the actual swap, and then the withdrawal of the funds from the protocol. The first and the last leg here, deposit and withdrawal, is handled by this deposit and withdrawal service, which is built on top of these bridges, these wrapping bridges. But the core thing, which is the actual swap that is happening, where... There is a matching algorithm for this particular trade that the user is willing to get. This is something that happens extremely quickly, uh within 1.2 seconds. And by the end of the year, it is going to happen faster than half a second. And this trade is settled on chain fully, de-centrally. And this chain is near blockchain. That's why it's near intense. But for the majority of people, it is completely invisible and you don't need to have a newer account to be able to swap Bitcoin to ETH. It just works. Now, if you would think of this flow, then it can be reminiscent to something that many people are using in crypto, which is centralized exchanges. So if you want to interact with centralized exchanges, you first need to deposit assets there, then you can trade potentially many times, and then you can withdraw assets from there. uh Near Intents is this decentralized version of a centralized exchange with the backend of a centralized exchange where it is written who owns what uh being a smart contract on the near blockchain. So it is open, all of the settlements are happening on the new blockchain. They are not happening because some kind of single authority is doing this, but it is actually a smart contract that checks that all of the participants of the trade have given their consent or authorized the action. They have enough money to spend. then within this transaction, there are no tokens burned or minted. So there are some invariants that are checked. It is a simple idea, but it is extremely powerful because on the scalable blockchain, like in your protocol, we are able to do this extremely quickly, in seconds and soon, in sub-second time, till the full finality, and at extremely low cost below one cent, which means that with assets deposited into the near-intense protocol that are held... non-custodially by the users, so I can deposit my soul, bitcoin, eth, zack into the protocol, I will still be holding these assets in my hand because smart contract would not allow anybody to perform any actions with these assets except for me signing with my key the particular uh intent of what I want to do, withdraw or trade or a bunch of other type of intents. So non-custodial, I'm holding all of these cross-chain assets non-custodially inside of the protocol and um I'm able to perform the swap of Bitcoin into ETH in 1.2 seconds. That's amazing. It's funny you actually use the uh centralized exchange uh as a model because I've actually been part of groups where rather than using a bridge, we've referred to a centralized exchange as a poor man's bridge because it functions the same way. It's cheaper and it's actually faster. And so and it also gives you additional privacy so that there are no linkage of distance. Yeah, which is really interesting, right? And I haven't heard other people talk about the, that's actually a huge benefit. You know, the privacy piece and it's faster and it's cheaper, yet it's like, but it has this like stigma of like a centralized thing that most crypto people don't like, but you know. just need to encounter some kind of situation when, because of some reason, there are some questions of the compliance department to the funds that are coming in. And then sometimes on some exchanges that start with and ends with C, the assets are not released to you for a long period of time. What is happening? You are not in control of the thin and this risk is real. This is something that people have seen with FTX when the assets were misappropriate. And uh this is like, I would quote Naval here. There is a famous tweet by him. Like, decentralization doesn't matter until it does. So, yeah, that's exactly that case. Yeah. Let's talk about distribution. mean, the 40 % market share is incredible. Yet, know, NEAR has been able to do it without like a lot of fanfare or promotion. It's just kind of happening quietly. were zero incentives that were deployed to, like they were marketing incentives. Yeah, so how was Near able to do that? Because I'm seeing, like in some cases, I know about Near Intents through kind of super apps I'm using like Infinex. But in many cases, people are using Near Intents and they don't even know it. um How is that happening? So that's the... This is, think, if I'm going to write a book at some point in time, know, all of these entrepreneurs writing business books. Well, first of all, it is going to be small. It is going to be 10 pages book because I really hate this lots of water inside of it, inside of many books. But NeurIntense progression up until this point is a, like... A book example of hyperscalen. For the first six months we were growing 100%, one month over months in volumes. Then it was going a little bit down, but it still was like 40%, which is also insane. And then we were just, we were getting kind of like a lucky ticket. uh with Zcash, but we have been working with Zcash for a very, very long time. We've seen lots of potential since one year ago with Zcash as a proper privacy solution. And already for multiple years, I was pondering about different privacy products to be shipped, and then maybe some of them are going to be shipped soon. um And they have just the proper tech. And uh for me, as a deeply technical person, uh I really like the design of Neer. think that this is a proper approach to scaling. And then I really like the design of Zcash as a proper approach for fully encrypted balances and fully encrypted cryptocurrency. So we have been working with them quite a lot and then it happened that we were one of the very few providers that were given the access to the asset. And then working together with Zashi team through NeonTense, we were able to allow people to really onboard to Zcash and also use it in day life. oh where with Zaxia Wallet you are able to pay for your burritos, which is the founder of Zcash was doing. And so they have like an off ramping solution that they're using, but within the solution, obviously nobody's off ramping Zcash, but they are through near intense, they are swapping it into a stable coin and then the stable coin gets off ramped and then it's pretty straightforward uh for them. oh And yeah, so you just go like very small steps here. is, you start with like, first you try to get to somebody and talk with somebody and then people just don't want to talk to you. Who are you? Like you have nothing. And then slowly you're building like, with every next month you're building a bigger case and then a bigger case and then the bigger case. And then you can talk to people with whom you wanted to talk in the beginning. then the people whom you wanted to talk at the beginning, they start to come to you by themselves. And this is exactly like, this is what changes over time when you're building up the case and you're growing, you're able to show some consistent results and consistent growth trajectory. ah Yeah, so. my advice here to all of the founders is that everybody wants to grow open. thing that you can show to uh the potential partners is the hockey stick pattern on all of the graphs that you provide. And you need to start step by step and always think, how can I double this thing in the next month? And then you will be able to get to the results with this type of mindset. It sounds like with the Zcash example, you had worked with both the Zcash team, Zucco team, and also with the Zashi wallet. And that seems to be kind of the distribution model, working with specific wallets. Is that correct? This is the case. So uh we are working with partners like Trust Wallet and many other hardware wallets. Some of them are in the integration phase, some of them are already integrated, but the press releases are not yet out. And I would say that by the end of this year, uh all of the majority distribution channels in crypto are going to get near and dense integrated. em Yeah, wallets is a natural thing. It's just because already for some period of time, like a year or maybe a year and a half, they started to integrate some kind of swapping capabilities inside of these wallets. And what they have seen is that users are pretty, I was going say, elastic or non-elastic, like users are pretty sticky to the wallets. Yeah. they don't quite care about the fees unless they are completely unreasonable. So there is a hardware wallet that takes 1.4 % on the swapping fees directly inside of them. And it is absolutely OK for the users. And I'm pretty sure that this hardware wallet is earning more from the swaps than from the actual sales of the hardware wallet. So the hardware wallet for them, the device is actually anti-economic. They are sending here to the people that they're going to get hacked. This is a proper thing. This is a proper concern. So now they are getting a hardware wallet. They put in there all of the... all of their net worth in the majority of the cases with crypto bros. And, and they are okay to take this 1.4 % hit. knowing that it comes from the company that is providing the security for all of the assets that they own. But in the back end, there are still some swapping providers that are working, and then these wallets are just putting additional fee on top of this stuff, and that's it. So this stickiness of the users to the wallets, the laziness of the people to check What will be the price on these different multiple things? And then doing additional steps, withdrawing funds from your wallet into something else or interacting with third party website, which can be a phishing website, which can be uh forcing you to blind sign some kind of random call to some kind of random transaction that can, that as we know, can result in 1.4 billion hack. scary. That's why with the neural tense, one of these approaches that we have used to design the system is that. there should be no smart contract calls. Users should not call smart contracts. The only thing that the user needs to do is to transfer assets to a particular address. And this is something that all of the major wallets, including hardware and software wallets, they are able to properly display to the user. So the user is able to verify the address and other things. There are some protocols that are used to create this one-time used wallet addresses. So if the user really wants to uh double-check that this is a correct address to deposit, so they can do this. But this really removes the risk of uh calling a random smart contract and getting all of your assets drained. Yeah, I'm still thinking about the hardware wallet example that you gave because in marketing, you know, we when we're thinking of like, what what is the company or project actually selling? And when you think about this hardware company, they're actually not selling a swap feature or even a hardware wallet. They're actually selling peace of mind. And that kind of peace of mind is it has a very high price. Like people are willing to pay a lot. to have peace of mind, which is actually a really great lesson for a lot of crypto projects because my observation is that a lot of crypto projects, overly focus on the complexity. They use really lots of jargon and confuse the user. But the truth is instead of focusing on the benefits, they focus on the technology. And I think as an industry, that's part of our issue. It's self-inflicted and we need to mature and grow up. Um, yeah. uh Since I joined in crypto 15, it was all technology driven. uh Ethereum is just starting. uh I remember how I met in the very, very end of 15 Vitalik on some kind of small meetup. And I didn't know a shit about it. And he was talking about different things and people. Somebody was asking, so when we're going to move to proof of stake? Because this was a new thing. And he said that, I don't know when exactly, but it is going to be soon. So soon was seven years later. But the... Like everybody were focused on the technology and do you remember everybody were reading through white papers and white paper is a necessity to do the ICO and stuff like that. exactly, nowadays people don't have... more like their attention span of people is like maximum three words in the tweet so you you need to condense like the message into something that is very short and very simple i do believe that this is one of the reasons why why zcash got picked up The technology didn't change. It's just the crypto Twitter picked up the privacy narrative. Yeah, there were some additional questions around privacy. There were about regulation, genius, act, stable coins, all of this stuff. uh Yeah, and then there were like some fud around privacy with Tornado Cash and Roman Storm, sending him lots of love and regards. uh But they were able to pack. the privacy thesis into a very simple and very easily consumable and comprehensible pitch, which is Zcash is an encrypted Bitcoin. And this is very, very simple and this is very suitable for people with a very short attention span. And it is like purely distilled core use case. uh If in this market, if you're as a crypto founder are not able to do this, you're not able to find this extremely short description of the value that you're providing, uh it would be extremely hard to find users. Yeah, at the point of wallet, let's just pick any random wallet, Metamask for example, I believe it's using a number of infrastructure providers on the backend for swaps. Will near intents be, I'm just making an assumption, but will near intents be another route that will be, okay, so as a user, I get to choose, okay, here are the routes I want. or I approve or it just happens. How's that going to work? eh I'm not quite sure what is happening in MetaMask. I moved out, I'm not using MetaMask. m I use that as an example, but really any wallet that I'm making an assumption that most wallets are going to have kind of uh bridge partners. They're also going to have near intense and potentially, you know, one or two other, I don't know, types of infrastructure. usual setup, how it works, is that in the majority of the cases, the wallets or whatever the front ends that the users are using, they are tending to uh show the list of the options of different providers what to use. and they're given an opportunity for a user to choose it. But they are also put in the by default provider, which usually is some sort of a combination of the best price, lowest time for execution. and maybe some internal parameters that are connected to the reliability of this partner. So for example, like a failure rate and the refund rate so that the partner is not able to perform a swap and things like that. I do not see that this thing is, at the moment, the wallet providers is sufficiently advanced. that they are doing the proper tracking. The reason for this is very simple. For wallets, they don't care which provider the user is going to choose because they're going to get their fee anyways. So the thing that they care about is that these providers are reliable enough so that they don't need to think of the support cases for the users. And this is very, very important. the state of the market for cross-chain providers is that many of them are very reliable. uh This part is like, this is where the cross-chain providers are providing very high quality of service. And also I wanted to say that in the majority of the cases what we have seen, it's like 95% of the users, are not changing the default swapping provider. They're just going with the default option and that's it. Yeah. want to think about it, which also uh gives an opportunity for the wallets to optimize on this behavior and remove one additional click that users need to do. But there might be some legal considerations around choosing for the user which swap provider to use. And there might be some additional thinking behind that. So here's something that I've seen and I'm curious to see what your opinion is. I've seen some business development agreements where, you know, make us the preferred route or choice and then we'll cut our rate by, you know, 100 bips or whatever, or 50 bips. Yeah, so how do you view those types of agreements? Because it's in effect making a decision for the user. and at the benefit of the wallet provider because they get more in fees. um But then the infrastructure partner gets paid less. Well, that's how to say this is like, this is a pool uh with sharks, right? So everybody are trying to grab their own land and figure out their way. ah That's the case. The thing is that once again, the wallet providers, they are not incentivized because m the stickiness of the users to the wallets is so, high. users are not going to switch their wallet because they're getting in this particular wallet 0.05 % more fee than in the other wallets. They're not going to change their whole lifestyle into the completely unknown interface that might take days to own properly on board. uh And if you don't believe me, just think of the times when you're uh trading on decentralized exchanges. Are you checking the prices on other centralized exchanges at the moment when you perform the trade? Do you know whether the markets are kind of... The prices are different, clearly. This is why the... uh like there are these arbitrage arbitrage that are making their money but you don't do this right so because of this stuff uh this is this is possible in the future in the era of the AI agents they are going to have much more time to think much more time to take care of the route and and you know optimizing for the fees this inefficiency of the market is going to be eliminated and It's only the most efficient swapping providers, which are the biggest ones because they have more liquidity, better market makers, market maker agreements, uh all of this stuff. They are going to survive. We are going to see the consolidation in the space, in the cross-chain space, definitely. Yeah. Tell us, uh I guess, what's ahead for near intense? And maybe we can get into Aurora also, since you're leading that um Aurora Labs. I guess, like, what's coming up that you'd like to talk about, any catalysts coming up that you'd like the audience to know? Yes. So cross-chain spot market is nice. And this is one of the use case. the majority of the, like almost everybody that we know in this space, all of our competitors, they are focused on this particular space. uh For us, our thinking behind the intents is much wider than just cross-chain spot trading. It is a good use case to start because it is very easy to design this smart contract that is going to handle the settlement and there is no disputes that are happening because user is approving a particular thing and you can check it on the level of the smart contract right at the moment of settlement. But we're thinking of intents in a wider setup. There are many different financial use cases in crypto. People are staking, people are putting money in vaults, but let's see what is going to happen with vaults after this week. um People are providing... one-sided liquidity into the lending protocols, people are providing two-sided liquidity into the AMM-style DEXs, and they're using all of these different financial products to earn yield. From the standpoint of intents, these are not very different from the spot positions. You're just buying with a particular asset an exposure to a particular financial instrument. So we are working towards the direction of widening the new intents to all of these products and making sure that with new intents people are able to get the exposure to any financial product that is available on crypto. including if you're going to do spot, then I'm guessing perps too. Yeah, obviously, obviously. But the financial products on crypto is not the only thing that people want to do. People are willing to do payments. And payments, it's not just transfers of the assets from one account to another. As I was saying to you in the beginning, during this payment, there is like a settlement period. There are these acquirers and stuff like that. And maybe the merchant wants to get fiat. Connection with Fiat is in our roadmap and we're working on this with our partners. And the additional primitives that would allow to implement e-commerce use cases where users are locking their funds with the platform and the platform is not able to take these funds. but rather say to the particular, this escrow smart contract, when these funds need to be released to the supplier at the moment when the service or the goods are delivered. Besides that, there are obvious natural language intents when people would like to get some service, but there is no particular uh product that they are buying. So... I want some food. This is a natural thing. This is how you can think of you're going to speak with your AI assistant that get me some food. Like, I'm hungry. I wipe code and get me some food, please. And then it is probably going to check with other AI agents or it is going to go and search for it probably would know my preferences in food. It is going to. uh try to remember what I have been eating before, so to get my meals more interesting and then it is going to get some food for me. So this natural language intent is also something that needs to be handled. It can be handled by my particular assistant or it can be handled by somebody else. You easily can imagine like an add-ons or like widgets to the website. where the user next to any photo that is there, user can just click a button and say, I want to buy this. uh And these solutions are already in the market. They're starting to get to the market. So any of the AI uh agentic commerce use cases, at this moment, they are being built. There is no. It is not a random thing that uh OpenAI is partnering with Stripe. It is not a random thing that Circle is building their own blockchain. It is not a random thing that Tether is doing the same thing too. Everybody see this new big market for the agentic economy. and are optimizing our processes for this to make sure that the AI agents are able to interact through the Neural Intents too. So our thinking of Intents is pretty wide and we just started with the cross-chain spot market as the most simple use case. Yeah. And I remember when NEAR first came out, it was actually very, very early in the AI space and um has continued to, know, part of its messaging has been around AI. And so now that AI has kind of caught up after a few years, know, NEAR is well ahead of the rest of the players. um And so the agentic space makes a lot of sense, plus NEAR Intents. Yeah, absolutely. is standing on these two legs. There is this uh cross-chain interoperability part or like financial part. This is something that I'm in charge of and I'm moving forward. And then the founder of NEO, one of the co-founders of the NEO protocol, Ilya Polosukin, who is one of the co-authors of Attention Is All You Need, the paper from Google that changed the world uh of AI. m uh He is leading the AI direction. the core thing that NIR is... So NIR is kind of standing on these two separate legs. But they are not very far from each other within this agentic uh commerce use case or agentic economy use case. And the thing that NIR is bringing within the AI space is something that is called DCML, Decentralized Confidential Machine Learning. Mm. And the pitch here is pretty simple and straightforward. There are quite a lot of GPUs out there. And these GPUs are in the data centers that are not controlled by the big corporations that are producing models, like OpenAI or Entropic or Mistral or any others, or DeepSec. So now... These companies who are developing these models, they have a big threat of moving these models to the external infrastructure for inference, because this is the place where they are earning money. But if somebody, if the infrastructure provider is just going to copy paste the weights of the model, then the whole AP is gone, right? So how to make sure that this can be done? done. So it's good that the hardware producers in this case, we then have to thank the producers for some period of time and they were introduced to this index called trusted execution environment. a special mode in which you're able to execute your code in GPUs and also on CPUs. There are also similar solutions. And you're actually able to put in an encrypted state these model weights inside of this solution. what allows companies to be able to offload their inference or the inference that they serve to other customers. And the reason why this thing needs to be decentralized is because if your provider is because of some reason cut off from the internet. Maybe it's hosting on Amazon's services in the US. uh you would like to have a redundant system and you would like to be like really easily to switch to something else. So that's why you need to have like a platform where multiple providers of GPUs are able to connect. So a decentralized system and then there needs to be like some kind of economy around there because you need to pay for this inference for this execution. So there needs to be something there and this is a moment where the Come together. connection with the particular AI use case. me, exactly. For many people, AI and blockchain is all about meme coins that many people are saying that like AI doesn't need blockchain and these are completely different things. But here is a very straightforward, very good example. where NEAR is closing partnerships with top AI companies already right now. And already these things are closed and already biggest company, AI, biggest producers of the AI models are using the DCML in the beta mode at the moment, but it is soon going to be released in public and available for everybody. That's exciting. Alex, thank you so much. This has been so fun. I think the audience, know about near intents. They're probably users of near intents. now they're aware of it, since so many cross-chain messaging and swapping is using near intents underneath. Thank you so much. And I appreciate you spending some time with us. Yeah, thank you, Peter.