From Protocol to Payments: Agentic Commerce and the Universal Ledger | DAS London 2025 | Day 3
Summary
Google Cloud's Web3 Initiative: Google Cloud has shifted its focus towards integrating Web3 technologies with institutional finance and capital markets, highlighting the development of a universal ledger and agent payment protocols.
Agent Payment Protocol: This protocol facilitates agentic payments and commerce, allowing AI agents to transact on behalf of consumers and businesses, potentially unlocking new business models and micro-payment economies.
Universal Ledger: Google Cloud's Universal Ledger is a layer 1 blockchain designed for high transaction throughput and interoperability, initially deployed as a private, permissioned ledger to meet regulatory requirements.
Micro-Payments and Blockchain: The discussion emphasized the potential of blockchain to support micro-payments and new business models, with blockchain's cost-efficiency being a key enabler for agent-based transactions.
Stablecoins vs. Tokenized Deposits: The conversation touched on the future of stablecoins versus tokenized bank deposits, with a focus on how blockchain can facilitate efficient and programmable financial transactions.
Regulatory Considerations: Google Cloud's approach to blockchain technology is shaped by regulatory requirements, aiming to provide a secure and compliant environment for financial institutions to transact.
Interoperability and Innovation: Google Cloud is committed to building interoperable systems that empower financial institutions without competing directly with public blockchains, focusing on enhancing existing financial infrastructures.
Transcript
Uh James, thank you very much for joining. Thanks for having me. >> Yeah. Uh we have uh 20 minutes here to to learn as much as we can in that short time about Google Cloud Web 3 digital business from you. Um maybe just by way of context start with a little uh introduction on your background and then how uh Google is now approaching the space. >> Yeah, sounds good. So hopefully you can all hear me uh just fine on the headphones. Um so yeah, my name is Dr. James Strowmans. I've been at Google about six years. Before that I worked at city for about six years and then prior to that I did a number of startups and my own academic background is in AI which is the topic dure at the moment. Um so I worked in AI for a long time from fundamental research during my PhD to about 2019 and always joke that when I stopped working on AI that's when chat GBT and uh the GPT models really took off. So I'll let you know when I'll get out of crypto because um it'll it'll go to the moon. So um yeah so we started the web3 business in Google cloud um about um well 2021 late 2021 uh and in that work we focused traditionally on the uh digital native uh space um indexing data running nodes helping people run validators more recently we run faucets for people to do engineering work to get access to test net tokens um but over the last year we've sort of slowly transitioned the business to focusing on the intersection of web3 technologies and institutional finance and capital markets And so, uh, the two topics that we'll probably talk about today are the universal ledger, which Tim kindly mentioned earlier, which you're right, I think it did, he he's right, it did go under the radar a little bit in terms of the significance of that work. Um, I think a lot of people didn't expect Google to build a layer 1 blockchain. So, I'm happy to share a little bit more about that. And then similarly, the work we've been doing around agent payment protocol. Um I think that's a really interesting um and technology and business agnostic um way to facilitate agentic payments and agentic commerce. So we'll talk about those two today. >> Yeah, let's let's uh start with agenda commerce a bit. Um I get a lot of pitches these days uh on AI agents at Blockworks. Um but uh it means a little bit different um has a different meaning to different people depending on the context. Um so let's just start by defining uh you know what is it? was agent commerce um in the con in the lens of payment the payment landscape um you know we're not just talking about AI agents assisting users but maybe even transacting themselves. >> Yeah. So I think firstly the work we've been doing in this space has been to uh empower in a safe and secure way agents to transact on behalf of consumers and potentially businesses as well. um such that the intent is clear and the agent uh doesn't necessarily hallucinate um uh a false uh capability or intent on behalf of of the the buyer and indeed the the merchant. Um the work we've done in the space agent payment protocol is an extension to our A2A work. Um so there's obviously lots of people are familiar with model context protocol. A2A is a complimentary piece of technology um that I think when combined with agent payment protocol and um blockchain rails are a really great fit to support the future of commerce. But it also supports credit card payments, you know, the heritage payment capabilities that the industry um has today. But I think if we're going to get into the world of micro payments, we're going to get into the world of maybe billions of agents transacting, we need a set of um payment rails that can support that. And in my view, um, if you want to pay for something that's maybe even sub cent, certainly sub dollar or pound, uh, you need a set of technologies that can do that cheaply and efficiently. And so agent payment protocol is is basically agnostic to that. We did some work with Coinbase on the X42 standard as well to link the two. And the idea is that when you want to have agents paying on behalf of a consumer um, with a merchant, you can follow these open- source standards. our partners can contribute to them and it can unlock um potentially new business models uh and certainly uh take us forward into the future of what you know agents might want to do for us. >> Are you seeing this first as primarily a B2B um advantage or a consumer focused or maybe both at the same time? >> I think the obvious like first order implication of this technology as it sort of presents to us as individuals is probably on the consumer side. Um, but I actually don't think that's where we're going to see uh it end. And I think actually some of the most interesting value unlock will observe will be in the B2B space. But of course, it's very obvious to us sat here today, even if you're not deep into AI or of course crypto that you might want to pay for your Nike shoes. Saw a great demo from the Algorand team recently where you would be be able to instruct your agent to go and do that for you. And that's something that you would probably do anyway. And it could be a holiday and it could be something else. and you can understand how you want to provide some guard rails for your agent there um through digital payment credentials and verifiable credentials to make sure that you're signing for something that you you actually intend to and that's you know a consumer concern. But what we haven't spoken about and what people maybe would consider the second order implications are well when businesses can pay for in the moment. So not net 60 you know in an invoice but you know very much in the moment in an untrusted way. So these folks don't have to trust one another but they can use blockchain capabilities to transact for something that is micro payment. You can imagine entire economies that don't exist today forming when you give the technology capability for these agents to orchestrate commerce between one another. And so you can imagine a whole bunch of agents paying each other for information through microtransactions on a blockchain. Um that just couldn't be possible today. And so if it's too expensive to do today and we didn't have agents, what's possible when payments are super cheap and you do have agents? What are the new business models that will come from that? And I think that's a little bit different to just paying for some shoes, right? I think that can be an entire section of autonomous agents paying uh one another for things that don't even exist today. And it's difficult for us to think through all the to get to the second order implications, certainly the third order knock- on effects of that. Um, but that's what we're building towards. You know, we want to make sure that the technologies there to enable that and to enable our customers and partners to build for that. >> Okay, this all sounds very sci-fi here, but uh maybe just back up a little bit if I want to look at a typical user story. Let's start let's go from the simple case first. where is in practice a user benefiting uh you know in the buying shoes example and then I want and then I want to delve a little bit deeper into the future economies that might evolve here as well but >> well I mean I I'm definitely not in the business of predicting the future but in terms of just how uh a consumer may use the agent to agent and AP2 protocol and they probably wouldn't know that they're they're using it right like the idea would be um people build user experiences on top of this technology of course they can run it on Google cloud orever they like um but the idea would be you You might have an agent. You might instruct that agent to say, "Hey, I'm really would like these shoes, but you know, they're too they're out of my price range right now. Um, could you perhaps watch those for me uh and alert me to when there's a price drop?" And we obviously assume that when a merchant or a vendor reduces the price, they're doing it because they want to shift stock or they want to sell uh more sort of simple supply and demand economics. So, in which case, your agent can track that for you. It's already been given permission to buy that for you. So, you actually might not even have the human in the loop. Um, the agent can then go and make that purchase. It's all pre-authorized and they can go ahead and it's exactly what you might expect. So I think that's a a nice sort of simple easy to understand use case that would resonate with with anybody here. >> Well, for the traders in the audience, you're basically setting a limit order on your Air Jordans. >> Yeah, that's exactly right. And uh I think when we think about payments, of course, payments for shoes, you could do payments for forex, you could do delivery versus payment, you could the technology could be scaled to other use cases in the institutional domain. Um I don't see why it couldn't go beyond just businesses buying things from businesses. you could absolutely take it through to agents transacting on markets if if someone wanted to do that. >> Okay. So then in in terms of micro payments um I mean the one that I typically think of is maybe uh compensating uh content creators. This is something that obviously could be applicable on YouTube. Um what what is uh what is this uh technology that is very close to being a practicable um unlock uh that we don't have today? >> Yes. Yeah. So, I think we're sort of maybe transitioning the conversation is naturally transitioning away from just the agent to agent piece. I think the the opportunity of blockchain stable coins is the the topic dure. Um, you know, I I have a well, I think stable coins are a great technology, but I actually think that they're perhaps um being uh conflated with solving problems that um we have already solved with commercial bank money. We can talk about that in a moment. But I think the idea that we can have cheaper rails to facilitate these types of things is only possible with blockchain. um blockchain uh is the is the technology that makes these uh capabilities um affordable so that an agent could do this on your behalf. And so I'm very excited about how um payments which I do think remains the killer app um for for blockchain and web 3. I remember maybe 2014 I bought something online using my Bitcoin wallet and the QR code flashed up. You know, I uh took it with my camera, the wallet autopop populated, clicked pay, and 250 milliseconds later the the URL had refreshed and um I could see payment confirmed. This is a this for me was my aha moment. And I think as the cost of using blockchain has come down um it's opening up the case for for users to transact in a in a cheaper way. And of course that extends to agents as well. >> Okay. Okay. So, you've already uh mentioned the Google Cloud Universal Ledger, GCU. Does that have like a fun uh pronunciation acronym like >> G? It depends on which country. So, um we we just call it the Universal Ledger. Um so, so yeah, I mean the Universal Ledger uh did it was first announced by Terry Duffy, who's the CEO of Chicago Merkantile Exchange earlier on in the year. Um there was some um posts that we did uh informally um from Google Cloud on LinkedIn that kind of caught the the eye of of the media and and users. It's a it's a layer 1 proof ofstake blockchain. Uh it's built to be capable of being public and permissionless um and decentralized although that's not how it will be deployed in the first instance. So it's that's a policy and uh choice. It's not a technical choice. Um but the key two things that I would mention about it uh for for this audience is firstly um I I've you know long been sort of wondering you know is blockchain this technology in search of a business problem and for us uh that was sort of turned the other way around when I started to see the business problem where actually agents who need high TPS uh and also um potentially crossber payments um which I'll mention momentarily um that you need high high performing blockchain with relatively low latency. And so the insight from our perspective was if Google needed to build a blockchain for itself at at our scale, what would that look like? What are the technology breakthroughs that we need? And then what are the business problems that we'd be solving with that? And and largely, what are the business problems we're solving for our customers? So this is very much customer and partnerled. Um the Google Cloud Universal Ledger is sort of a million transaction per second level blockchain. Um, in terms of time to finality, if you want to call it that, it's not quite as fast as some of the other uh modern blockchains, but it's pretty quick. You know, it's um sub 2C, probably under 1 second by the time we finished optimizing, but we're actually not prioritizing that for our use cases. But we think agents will need that level um of of throughput. Now, in terms of why we've opt opted for a private permission ledger, I think it's because, you know, when we speak to our financial partners that use Google Cloud, and like I said, they're the ones kind of driving these these business problems um and partnership with us, they are still um interested in transacting in a safe and known environment. And with the state of um smart contracting today, of course, you have bridging technology. Um you have other capabilities. Um if you want to uh provide say a public um digital asset like Bitcoin from its native chain, you can um clearly hold that in escrow on the native chain. You can use a bridging technology and you can then provide interoperability onto a private permission ledger like GCL so that um trusted parties which in this case will be the banks um governed by uh another um uh institution that will not be Google can join the ledger. they can transact with one another and they can um pay for and buy assets like Bitcoin in a safe and secure environment. And I think the last thing I'll say um is that we we wanted to put commercial bank money on chain natively. So not even like a tokenized deposit, although you can think of it like that. So there's no ERC20 for commercial bank money. It's just money and the protocol supports it as a first party consideration. So you can think of it like a thin ledger that a bank would banks already have hundreds or thousands of ledgers depending on their corresponding banking relationships. So this would be a technology that's fully managed or if they want to they could run a validator in the future that enables them to then increment and decrement accounts commercial bank money as they're regulated to do so today. So it's a multicurrency multiasset ledger that would enable you to do PVP or DVP in two second uh finality. Um we're talking to certain institutions for clearing and settlement. Um, and yeah, if you can get digitally native assets on this ledger, I always think about it on ramp the the digital assets into the fiat world rather than on ramp fiat to to digital assets, then you can create a nice environment for these organizations to transact with one another safely even for digital assets. >> Okay. Well, yeah, I mean this is uh starts to get into almost a philosophical debate within uh crypto is the the permission versus uh permissionless blockchains. Uh what is the role of uh credible neutrality? I know there is some aspect of this that you're looking towards the future of decentralizing in some way and and having a kind of neutral platform is a is a goal but what what what is that what is that can you explain how Google kind of thinks about those issues that make cryptonatives let's say somewhat skeptical of permission to corporate chains? >> Yeah, we're not we're not building to compete with say a cryptonative public blockchain. I think the work that those organizations are doing and those communities are building is firstly very admirable and secondly it's it's quite I mean by design it's to build outside of the existing financial system right like they're trying to build um a set of capabilities that will mean you don't have to work with a bank should you not wish so I think you know that's why we want to be interoperable with but not compete directly with those the reason we've built it as a blockchain is very simple if you're building financial market infrastructure any regulator in the world will say you can't just run that on one cloud you have to be able to show as a bank or a regulated institution that you can run that technology either on prem or on another hypers scale cloud provider quickly and easily and there's many ways you can solve that technically blockchain is just a very very good fit you can deploy another validator outside of Google cloud it will run it will just work and you have the security model that blockchain affords through uh consensus to be basentine fault tolerant and so for me it's a way to make a core payment rail portable to anywhere it needs to be and I think blockchain is a great fit for that um in terms of being credibly neutral I think the context of that statement was in the fact that you don't necessarily want to be a stable coin issuer and own the ledger and own the apps and own the whole ecosystem. We don't want to disintermediate banks from their end customers or their end corporate users. We want to empower them with just managed technology that they can use to build better solutions for their customers. And so that's what we mean about being incredibly neutral. We're known for building great technology. We open source a lot of our work. Um we can run this in data centers all over the world and all corners of the world. uh including um adhering to data residency guarantees. And so when you can offer that type of technical service to a financial organization and say, "Oh, we can get you Bitcoin, we can get you another digitally native asset on this um on this uh payment rail as well." I think it's quite an interesting value proposition and it's different to being a stable coin issuer and running the infrastructure and owning the protocol. >> Yeah. And we we have examples of of uh stablecoin issuers specifically launching chains, payment rails specifically launching chains. Um so it's it's going to be an interesting competitive landscape where who's op where their people are optimizing for and for which customers. Um I want to go back to the the stable coins uh versus other kinds of value transfer such as tokenized bank deposits for example. Um this morning there was a presentation on uh stable coins from city uh report on on stable coins in 2030 let's say and uh one um item that was of interest was uh suspecting that there might actually be a larger market for non-stable coin uh tokenized assets like deposits that you know serve the same function as stable coins practically speaking. Um you alluded to this earlier. I'm curious for your thoughts on that. >> It's a great question. I mean obviously if you're in the stable coin business it's natural that you'll see stable coins as a solution to a lot of different problems and I think in the public blockchain space we don't really have an alternative if you need a stable you know fiat pegged asset I mean by definition uh a stable coin does that but I would encourage everyone who's working in this space to sort of step back and think about what do we like about stable coins is it we love e-oney regulation we love not getting interest on our deposits uh we love them not being funible between issuers do do we like those characteristics or do we like the fact that they're 24/7 seven. They're programmable. They um can settle in 2 seconds. Well, they can reach finality in 2 seconds. I think we love the characteristics afforded by blockchain. I think the e-money regulation of a stable coin as it is in the various countries that um regulate it today, that has a place, but it might not be uh as uh the same the same problems. It might not solve the same problems as commercial bank money already solves today. So, commercial bank money is regulated globally largely in the same way. It's very well understood. The banking systems operated with it for uh many many decades. And so I think city also have an axe to grind here. I think it's natural that a bank who will make money on bank deposits where they can loan out I don't know the exact ratio but you know 15 plus times the amount of what they're storing uh in reserves and make presumably greater than a base rate of interest on that. Um it's in any bank that can make those uh that can take deposits and loan to retain deposits right and it's good you know it's largely I think good for the economy. I think if everything moves to stable coins, banks can't loan against that in the same way. Um the regulation precludes that. And so it has quite an interesting like knock-on effect. If you put commercial bank money, stable coins, tokenized money market funds, and tokenized treasury all on a ledger where it's very very cheap to transact between them. What happens? Does money stay in a bank deposit earning relatively low interest? Does it move to the higher yielding asset? Um and then the assets that are say tokenized money market funds or treasuries what are their regulatory characteristics relative to stable coins and what does that mean for for the movement of value and I think you know banks who are thinking about this deeply know the answer of where that's going but you know as a technology provider we just want to make sure that we're offering the right capabilities for the markets to build what they need on top of that. Yeah, I mean one of the things I enjoy about crypto right now, even today, it's possible to have a stable coin that's earning a yield and can be immediately usable in payments uh you know still today through Visa or Mastercard typically, but uh who knows how long that will be. I think you know from a cryptonative lens um you know one of the appeals of public permissionless systems is freedom and not uh having to fear uh a vendor lock in however that uh may manifest through coercion or through just convenience. Um how does a a giant hyperscaler like Google sort of mesh those two elements as it designs its systems for the future to to be a global platform? Yeah, I mean look, Google obviously operates within the laws and regulations of all the countries where it where it does business, right? And we we will always do that. So from our point of view as a technology provider, we are not in the business of speculating on you know what's going to be successful in terms of public blockchains or private blockchains, but we have customers asking us to build these capabilities for them. And there's a market there for that. And so it's our privilege to be able to work with these organizations like the CME, but but many others to to take that forward. And I don't think that detracts from the work that's happening in the public blockchain space, which is largely solving for a different problem. Um, and yes, they may meet and yes, we should hope to see interoperability, but they're not they're kind of operating in slightly shifted spaces in my mind. >> Fantastic. Well, we're going to have to leave it there, but uh I hope that leaves you wanting more and you can read about it on the pages of Block Works in the future uh for years to come. >> Thanks for having me. Cheers. Okay.
From Protocol to Payments: Agentic Commerce and the Universal Ledger | DAS London 2025 | Day 3
Summary
Transcript
Uh James, thank you very much for joining. Thanks for having me. >> Yeah. Uh we have uh 20 minutes here to to learn as much as we can in that short time about Google Cloud Web 3 digital business from you. Um maybe just by way of context start with a little uh introduction on your background and then how uh Google is now approaching the space. >> Yeah, sounds good. So hopefully you can all hear me uh just fine on the headphones. Um so yeah, my name is Dr. James Strowmans. I've been at Google about six years. Before that I worked at city for about six years and then prior to that I did a number of startups and my own academic background is in AI which is the topic dure at the moment. Um so I worked in AI for a long time from fundamental research during my PhD to about 2019 and always joke that when I stopped working on AI that's when chat GBT and uh the GPT models really took off. So I'll let you know when I'll get out of crypto because um it'll it'll go to the moon. So um yeah so we started the web3 business in Google cloud um about um well 2021 late 2021 uh and in that work we focused traditionally on the uh digital native uh space um indexing data running nodes helping people run validators more recently we run faucets for people to do engineering work to get access to test net tokens um but over the last year we've sort of slowly transitioned the business to focusing on the intersection of web3 technologies and institutional finance and capital markets And so, uh, the two topics that we'll probably talk about today are the universal ledger, which Tim kindly mentioned earlier, which you're right, I think it did, he he's right, it did go under the radar a little bit in terms of the significance of that work. Um, I think a lot of people didn't expect Google to build a layer 1 blockchain. So, I'm happy to share a little bit more about that. And then similarly, the work we've been doing around agent payment protocol. Um I think that's a really interesting um and technology and business agnostic um way to facilitate agentic payments and agentic commerce. So we'll talk about those two today. >> Yeah, let's let's uh start with agenda commerce a bit. Um I get a lot of pitches these days uh on AI agents at Blockworks. Um but uh it means a little bit different um has a different meaning to different people depending on the context. Um so let's just start by defining uh you know what is it? was agent commerce um in the con in the lens of payment the payment landscape um you know we're not just talking about AI agents assisting users but maybe even transacting themselves. >> Yeah. So I think firstly the work we've been doing in this space has been to uh empower in a safe and secure way agents to transact on behalf of consumers and potentially businesses as well. um such that the intent is clear and the agent uh doesn't necessarily hallucinate um uh a false uh capability or intent on behalf of of the the buyer and indeed the the merchant. Um the work we've done in the space agent payment protocol is an extension to our A2A work. Um so there's obviously lots of people are familiar with model context protocol. A2A is a complimentary piece of technology um that I think when combined with agent payment protocol and um blockchain rails are a really great fit to support the future of commerce. But it also supports credit card payments, you know, the heritage payment capabilities that the industry um has today. But I think if we're going to get into the world of micro payments, we're going to get into the world of maybe billions of agents transacting, we need a set of um payment rails that can support that. And in my view, um, if you want to pay for something that's maybe even sub cent, certainly sub dollar or pound, uh, you need a set of technologies that can do that cheaply and efficiently. And so agent payment protocol is is basically agnostic to that. We did some work with Coinbase on the X42 standard as well to link the two. And the idea is that when you want to have agents paying on behalf of a consumer um, with a merchant, you can follow these open- source standards. our partners can contribute to them and it can unlock um potentially new business models uh and certainly uh take us forward into the future of what you know agents might want to do for us. >> Are you seeing this first as primarily a B2B um advantage or a consumer focused or maybe both at the same time? >> I think the obvious like first order implication of this technology as it sort of presents to us as individuals is probably on the consumer side. Um, but I actually don't think that's where we're going to see uh it end. And I think actually some of the most interesting value unlock will observe will be in the B2B space. But of course, it's very obvious to us sat here today, even if you're not deep into AI or of course crypto that you might want to pay for your Nike shoes. Saw a great demo from the Algorand team recently where you would be be able to instruct your agent to go and do that for you. And that's something that you would probably do anyway. And it could be a holiday and it could be something else. and you can understand how you want to provide some guard rails for your agent there um through digital payment credentials and verifiable credentials to make sure that you're signing for something that you you actually intend to and that's you know a consumer concern. But what we haven't spoken about and what people maybe would consider the second order implications are well when businesses can pay for in the moment. So not net 60 you know in an invoice but you know very much in the moment in an untrusted way. So these folks don't have to trust one another but they can use blockchain capabilities to transact for something that is micro payment. You can imagine entire economies that don't exist today forming when you give the technology capability for these agents to orchestrate commerce between one another. And so you can imagine a whole bunch of agents paying each other for information through microtransactions on a blockchain. Um that just couldn't be possible today. And so if it's too expensive to do today and we didn't have agents, what's possible when payments are super cheap and you do have agents? What are the new business models that will come from that? And I think that's a little bit different to just paying for some shoes, right? I think that can be an entire section of autonomous agents paying uh one another for things that don't even exist today. And it's difficult for us to think through all the to get to the second order implications, certainly the third order knock- on effects of that. Um, but that's what we're building towards. You know, we want to make sure that the technologies there to enable that and to enable our customers and partners to build for that. >> Okay, this all sounds very sci-fi here, but uh maybe just back up a little bit if I want to look at a typical user story. Let's start let's go from the simple case first. where is in practice a user benefiting uh you know in the buying shoes example and then I want and then I want to delve a little bit deeper into the future economies that might evolve here as well but >> well I mean I I'm definitely not in the business of predicting the future but in terms of just how uh a consumer may use the agent to agent and AP2 protocol and they probably wouldn't know that they're they're using it right like the idea would be um people build user experiences on top of this technology of course they can run it on Google cloud orever they like um but the idea would be you You might have an agent. You might instruct that agent to say, "Hey, I'm really would like these shoes, but you know, they're too they're out of my price range right now. Um, could you perhaps watch those for me uh and alert me to when there's a price drop?" And we obviously assume that when a merchant or a vendor reduces the price, they're doing it because they want to shift stock or they want to sell uh more sort of simple supply and demand economics. So, in which case, your agent can track that for you. It's already been given permission to buy that for you. So, you actually might not even have the human in the loop. Um, the agent can then go and make that purchase. It's all pre-authorized and they can go ahead and it's exactly what you might expect. So I think that's a a nice sort of simple easy to understand use case that would resonate with with anybody here. >> Well, for the traders in the audience, you're basically setting a limit order on your Air Jordans. >> Yeah, that's exactly right. And uh I think when we think about payments, of course, payments for shoes, you could do payments for forex, you could do delivery versus payment, you could the technology could be scaled to other use cases in the institutional domain. Um I don't see why it couldn't go beyond just businesses buying things from businesses. you could absolutely take it through to agents transacting on markets if if someone wanted to do that. >> Okay. So then in in terms of micro payments um I mean the one that I typically think of is maybe uh compensating uh content creators. This is something that obviously could be applicable on YouTube. Um what what is uh what is this uh technology that is very close to being a practicable um unlock uh that we don't have today? >> Yes. Yeah. So, I think we're sort of maybe transitioning the conversation is naturally transitioning away from just the agent to agent piece. I think the the opportunity of blockchain stable coins is the the topic dure. Um, you know, I I have a well, I think stable coins are a great technology, but I actually think that they're perhaps um being uh conflated with solving problems that um we have already solved with commercial bank money. We can talk about that in a moment. But I think the idea that we can have cheaper rails to facilitate these types of things is only possible with blockchain. um blockchain uh is the is the technology that makes these uh capabilities um affordable so that an agent could do this on your behalf. And so I'm very excited about how um payments which I do think remains the killer app um for for blockchain and web 3. I remember maybe 2014 I bought something online using my Bitcoin wallet and the QR code flashed up. You know, I uh took it with my camera, the wallet autopop populated, clicked pay, and 250 milliseconds later the the URL had refreshed and um I could see payment confirmed. This is a this for me was my aha moment. And I think as the cost of using blockchain has come down um it's opening up the case for for users to transact in a in a cheaper way. And of course that extends to agents as well. >> Okay. Okay. So, you've already uh mentioned the Google Cloud Universal Ledger, GCU. Does that have like a fun uh pronunciation acronym like >> G? It depends on which country. So, um we we just call it the Universal Ledger. Um so, so yeah, I mean the Universal Ledger uh did it was first announced by Terry Duffy, who's the CEO of Chicago Merkantile Exchange earlier on in the year. Um there was some um posts that we did uh informally um from Google Cloud on LinkedIn that kind of caught the the eye of of the media and and users. It's a it's a layer 1 proof ofstake blockchain. Uh it's built to be capable of being public and permissionless um and decentralized although that's not how it will be deployed in the first instance. So it's that's a policy and uh choice. It's not a technical choice. Um but the key two things that I would mention about it uh for for this audience is firstly um I I've you know long been sort of wondering you know is blockchain this technology in search of a business problem and for us uh that was sort of turned the other way around when I started to see the business problem where actually agents who need high TPS uh and also um potentially crossber payments um which I'll mention momentarily um that you need high high performing blockchain with relatively low latency. And so the insight from our perspective was if Google needed to build a blockchain for itself at at our scale, what would that look like? What are the technology breakthroughs that we need? And then what are the business problems that we'd be solving with that? And and largely, what are the business problems we're solving for our customers? So this is very much customer and partnerled. Um the Google Cloud Universal Ledger is sort of a million transaction per second level blockchain. Um, in terms of time to finality, if you want to call it that, it's not quite as fast as some of the other uh modern blockchains, but it's pretty quick. You know, it's um sub 2C, probably under 1 second by the time we finished optimizing, but we're actually not prioritizing that for our use cases. But we think agents will need that level um of of throughput. Now, in terms of why we've opt opted for a private permission ledger, I think it's because, you know, when we speak to our financial partners that use Google Cloud, and like I said, they're the ones kind of driving these these business problems um and partnership with us, they are still um interested in transacting in a safe and known environment. And with the state of um smart contracting today, of course, you have bridging technology. Um you have other capabilities. Um if you want to uh provide say a public um digital asset like Bitcoin from its native chain, you can um clearly hold that in escrow on the native chain. You can use a bridging technology and you can then provide interoperability onto a private permission ledger like GCL so that um trusted parties which in this case will be the banks um governed by uh another um uh institution that will not be Google can join the ledger. they can transact with one another and they can um pay for and buy assets like Bitcoin in a safe and secure environment. And I think the last thing I'll say um is that we we wanted to put commercial bank money on chain natively. So not even like a tokenized deposit, although you can think of it like that. So there's no ERC20 for commercial bank money. It's just money and the protocol supports it as a first party consideration. So you can think of it like a thin ledger that a bank would banks already have hundreds or thousands of ledgers depending on their corresponding banking relationships. So this would be a technology that's fully managed or if they want to they could run a validator in the future that enables them to then increment and decrement accounts commercial bank money as they're regulated to do so today. So it's a multicurrency multiasset ledger that would enable you to do PVP or DVP in two second uh finality. Um we're talking to certain institutions for clearing and settlement. Um, and yeah, if you can get digitally native assets on this ledger, I always think about it on ramp the the digital assets into the fiat world rather than on ramp fiat to to digital assets, then you can create a nice environment for these organizations to transact with one another safely even for digital assets. >> Okay. Well, yeah, I mean this is uh starts to get into almost a philosophical debate within uh crypto is the the permission versus uh permissionless blockchains. Uh what is the role of uh credible neutrality? I know there is some aspect of this that you're looking towards the future of decentralizing in some way and and having a kind of neutral platform is a is a goal but what what what is that what is that can you explain how Google kind of thinks about those issues that make cryptonatives let's say somewhat skeptical of permission to corporate chains? >> Yeah, we're not we're not building to compete with say a cryptonative public blockchain. I think the work that those organizations are doing and those communities are building is firstly very admirable and secondly it's it's quite I mean by design it's to build outside of the existing financial system right like they're trying to build um a set of capabilities that will mean you don't have to work with a bank should you not wish so I think you know that's why we want to be interoperable with but not compete directly with those the reason we've built it as a blockchain is very simple if you're building financial market infrastructure any regulator in the world will say you can't just run that on one cloud you have to be able to show as a bank or a regulated institution that you can run that technology either on prem or on another hypers scale cloud provider quickly and easily and there's many ways you can solve that technically blockchain is just a very very good fit you can deploy another validator outside of Google cloud it will run it will just work and you have the security model that blockchain affords through uh consensus to be basentine fault tolerant and so for me it's a way to make a core payment rail portable to anywhere it needs to be and I think blockchain is a great fit for that um in terms of being credibly neutral I think the context of that statement was in the fact that you don't necessarily want to be a stable coin issuer and own the ledger and own the apps and own the whole ecosystem. We don't want to disintermediate banks from their end customers or their end corporate users. We want to empower them with just managed technology that they can use to build better solutions for their customers. And so that's what we mean about being incredibly neutral. We're known for building great technology. We open source a lot of our work. Um we can run this in data centers all over the world and all corners of the world. uh including um adhering to data residency guarantees. And so when you can offer that type of technical service to a financial organization and say, "Oh, we can get you Bitcoin, we can get you another digitally native asset on this um on this uh payment rail as well." I think it's quite an interesting value proposition and it's different to being a stable coin issuer and running the infrastructure and owning the protocol. >> Yeah. And we we have examples of of uh stablecoin issuers specifically launching chains, payment rails specifically launching chains. Um so it's it's going to be an interesting competitive landscape where who's op where their people are optimizing for and for which customers. Um I want to go back to the the stable coins uh versus other kinds of value transfer such as tokenized bank deposits for example. Um this morning there was a presentation on uh stable coins from city uh report on on stable coins in 2030 let's say and uh one um item that was of interest was uh suspecting that there might actually be a larger market for non-stable coin uh tokenized assets like deposits that you know serve the same function as stable coins practically speaking. Um you alluded to this earlier. I'm curious for your thoughts on that. >> It's a great question. I mean obviously if you're in the stable coin business it's natural that you'll see stable coins as a solution to a lot of different problems and I think in the public blockchain space we don't really have an alternative if you need a stable you know fiat pegged asset I mean by definition uh a stable coin does that but I would encourage everyone who's working in this space to sort of step back and think about what do we like about stable coins is it we love e-oney regulation we love not getting interest on our deposits uh we love them not being funible between issuers do do we like those characteristics or do we like the fact that they're 24/7 seven. They're programmable. They um can settle in 2 seconds. Well, they can reach finality in 2 seconds. I think we love the characteristics afforded by blockchain. I think the e-money regulation of a stable coin as it is in the various countries that um regulate it today, that has a place, but it might not be uh as uh the same the same problems. It might not solve the same problems as commercial bank money already solves today. So, commercial bank money is regulated globally largely in the same way. It's very well understood. The banking systems operated with it for uh many many decades. And so I think city also have an axe to grind here. I think it's natural that a bank who will make money on bank deposits where they can loan out I don't know the exact ratio but you know 15 plus times the amount of what they're storing uh in reserves and make presumably greater than a base rate of interest on that. Um it's in any bank that can make those uh that can take deposits and loan to retain deposits right and it's good you know it's largely I think good for the economy. I think if everything moves to stable coins, banks can't loan against that in the same way. Um the regulation precludes that. And so it has quite an interesting like knock-on effect. If you put commercial bank money, stable coins, tokenized money market funds, and tokenized treasury all on a ledger where it's very very cheap to transact between them. What happens? Does money stay in a bank deposit earning relatively low interest? Does it move to the higher yielding asset? Um and then the assets that are say tokenized money market funds or treasuries what are their regulatory characteristics relative to stable coins and what does that mean for for the movement of value and I think you know banks who are thinking about this deeply know the answer of where that's going but you know as a technology provider we just want to make sure that we're offering the right capabilities for the markets to build what they need on top of that. Yeah, I mean one of the things I enjoy about crypto right now, even today, it's possible to have a stable coin that's earning a yield and can be immediately usable in payments uh you know still today through Visa or Mastercard typically, but uh who knows how long that will be. I think you know from a cryptonative lens um you know one of the appeals of public permissionless systems is freedom and not uh having to fear uh a vendor lock in however that uh may manifest through coercion or through just convenience. Um how does a a giant hyperscaler like Google sort of mesh those two elements as it designs its systems for the future to to be a global platform? Yeah, I mean look, Google obviously operates within the laws and regulations of all the countries where it where it does business, right? And we we will always do that. So from our point of view as a technology provider, we are not in the business of speculating on you know what's going to be successful in terms of public blockchains or private blockchains, but we have customers asking us to build these capabilities for them. And there's a market there for that. And so it's our privilege to be able to work with these organizations like the CME, but but many others to to take that forward. And I don't think that detracts from the work that's happening in the public blockchain space, which is largely solving for a different problem. Um, and yes, they may meet and yes, we should hope to see interoperability, but they're not they're kind of operating in slightly shifted spaces in my mind. >> Fantastic. Well, we're going to have to leave it there, but uh I hope that leaves you wanting more and you can read about it on the pages of Block Works in the future uh for years to come. >> Thanks for having me. Cheers. Okay.