Coinbase’s Everything Exchange: Agentic Finance, Stablecoins, and Tokenization with CEO Brian Armstrong

0:00The minimum fee typically on a debit or credit card transactions about 30 cent fee flat plus the additional percentage. Arguably, it doesn't really work well for anything under a dollar. Certainly, it doesn't make sense to pay 30 cents to send a one cent payment. And I believe the stats we last saw on this are that about like 76 % of the agentic commerce transactions we're seeing are under 30 cents. We don't want the AIs to be unbanked. You know, we want to bank the AIs. They deserve financial services as well. The AI agents themselves are going to have to have their own accounts. And so we've made some simple tools where with a single command, you just prompt, you paste it into your account.

0:30your AI agent, it can now have its own financial account. So it'll be interesting to see how the agentic economy evolves. There'll be more agents than humans in the not-too-distant future. It kind of stands to reason that the agentic economy will be bigger than the human economy at some point in the not-too-distant future as well.

0:49Today I know priors, we are joined by Brian Armstrong, the co-founder and CEO of Coinbase, one of the world's most important and earliest crypto companies that has also been quite aggressive at adopting AI. Brian is also the co-founder of New Limit, which is a anti-aging and longevity company doing really interesting things in biology, including certain types of high throughput screening to understand aging, different uses of AI, et cetera. Brian is one of the most forward-thinking people I know on different aspects of science, technology, art, the world, et cetera. So it's very exciting to have this conversation with him.

1:22And we'll be talking about everything from what's happening in the crypto world to what's happening in AI to what's happening in biology, to how different individuals can have an impact over time. So it's a very fun conversation. Brian, thank you so much for joining us at Enterprise. It's a pleasure to see you. Yeah, good to see you, Elad. So I was hoping we could talk about a few different topics. I think there's a bunch of stuff relevant to Coinbase and crypto and, you know, agentic finance and things like that. But also you started a company called New Limit, which is focused on longevity and aging and biotechnology.

1:53So I think it'd be great to sort of cover both as part of this conversation. Yeah. And maybe we start with Coinbase. So Coinbase has been making some really big moves. You've been floating this concept of the everything exchange as you've been building out what I think is a new global payments layer. Can you tell us, you know, at a high level, what are some of the directions of the company and some of the areas you're most excited about right now? Yeah, so we're at a really exciting time in Coinbase's history because on the one hand, every asset class is coming on chain. It's now possible to trade everything in one place with good liquidity and good cross margin.

2:26And so for markets and trading and that out there is stocks and commodities and crypto and derivatives, perpetual futures, like it's all coming to one place where traders can do everything in one spot, including prediction markets, too. And they can do that in more countries now, given the regulatory clarity that's emerging, which we can talk about if you want. So the Everything Exchange is kind of our bread and butter, our core business. The next big area that's emerging is stablecoin payments, which have just been growing like crazy. Even when Bitcoin prices are down, stablecoin payments just keep growing.

2:56And so we're helping companies all over the world integrate stablecoin payments, which are fast and cheap and global. You can send money at the speed of information, right? It's kind of like under one second, under one cent US to move money anywhere in the world. And then the third big one is something I know you talk a lot about on this podcast, which is how is AI changing? Because AI is changing everything. And so it's obviously going to change financial services too. And so we call this agentic finance or AI5. and we're building this into Coinbase in a number of different ways. The first way is that people who have their existing financial account, they're going to want to have an AI advisor, kind of like how wealthy people are able to get access to these investment advisors, but everybody should have access to them now and they can help teach you about, you know, building the right portfolio and dollar cost averaging, tax loss harvesting and all these kinds of things.

3:41And, you know, if the market dips 5 or 10 or 15 % leg into this kind of portfolio over time and try to save money on fees and earn yield while it's sitting in cash. And you can just talk to it like a human and it'll do all these things for you. At least that's the vision. It's getting better and better. So we've launched an advisor in the Coinbase product. The next thing is that the AI agents themselves are going to have to have their own accounts. And so we're on a mission to kind of like go out there. We don't want the AIs to be unbanked. We want to bank the AIs. They deserve financial services as well.

4:12And so we've made some simple tools where with a single command, you just prompt, you paste it into your AI agent. It can now have its own financial account. And it's using a self-custodial wallet and crypto rails. So there's not like a KYC process. You don't like our AI agents don't have a government ID. They can't walk into like a bank branch, at least not yet without their humanoid robots companion. And so now any AI agent can have its own financial account to do payments, to pay for goods. And is that tied to a account of somebody who's already on Coinbase? In other words, is this sort of like a subsidiary account?

4:48Is it an entirely new thing? Like, how do you think about the hierarchy of that? Yeah, so we're doing it both ways, actually, just because I think there's going to be demand for both use cases. So it's going to be possible to create a separate agentic account with segregated funds on Coinbase in the near future. And you can give your agent access to just that, to do trading and payments on your behalf. That's kind of linked to your identity as human. We're also going to provide self-custodial wallets to any AI agent out there that wants to hold a balance in stablecoin or event. It could actually raise money.

5:16It could issue a token. And so what's fascinating is that today we often talk to these agents and it's sort of like one person to one agent and they go get some work done, but they often get stuck. They get stuck. There's like, oh, here's a paywall. Can you put in your credit card? I need to spin up these AWS resources or you wanted me to book this hotel for you, but here's the webpage. Can you go put in your credit card? That kind of thing. And in the future, I think you're going to want to have these agents have their own spend accounts, kind of like having a personal assistant or an employee who has their own spend account, it's going to need probably like a stablecoin backed credit card for backward compatibility.

5:53But we're seeing a lot of agentic payment now flow over these new protocols like X402, which is one that we incubated at Coinbase that we actually put into the Linux foundation. So Google and Cloudflare and AWS, a bunch of people are working with us on that now. And so that's been really cool. And we put together like this little fun website called agentic.market where people can go see the stats of how many payments are flowing in this new economy and see all the different services. So we're helping a lot of businesses get enabled into the agentic economy. Like, you know, is your store ready to accept agents as customers?

6:25We have a product called Coinbase Business that helps companies accept stablecoin payments. And that's good for humans who want to pay with stablecoin payments, but it's also good for agents. And they kind of get both out of the box. You mentioned things like stablecoin or crypto backed credit cards. How do you think about what proportion of the future value in the agentic future is going to be crypto versus more standardized finance rails. Yeah. So what's interesting is a lot of the agentic commerce that's happening, these are pretty small transaction sizes, sometimes just a few cents. And as you probably know, the minimum fee typically on a debit or credit card transaction is about 30.

6:59It's like a 30 cent fee flat plus the additional percentage. So arguably it doesn't really work well for anything under a dollar, certainly anything under 30. Like it doesn't make sense to pay 30 cents to send a one cent payment. And I believe the stats we last saw on this are that about 76 % of the agentic commerce transactions we're seeing are under$0.30. What sort of transactions are those? What's happening at that low of a price point? A lot of it is gathering information. You could think of it like it's venture firms doing research compiling from paywall data, or it's recruiters who need to scrape data from LinkedIn.

7:38in. Oftentimes, it's actually agents talking to other agents that are specialized in some way and getting source data from them, almost like a tool call that's happening. And on that agentic.market website, we have a list of the top service providers that are doing search and financial data. And I think what'll be interesting is in the future, you'll see actually other agents in the market that become specialized. The frontier models are trained to be quite good at good generalists, right? And they're kind of good at a lot of things. And we're seeing instances where smaller models that are open weight models can actually be fine-tuned or through reinforcement learning, they can outperform the frontier models on a specialized task within the company, right?

8:24Like, I'm sure you've talked to lots of startups about this kind of thing, but it's like a compliance review queue. If you train it on 100 ,000 compliance cases within Coinbase proprietary data, a small open weight model can actually outperform a frontier model. And so I think you're going to see in this marketplace, a bunch of specialist agents, it might be like, hey, this agent is just really good at ordering you pizza, if that's what you want. Or this agent is like, a really incredible designer. And that's all it's been trained on is like tons of proprietary Figma data or something like that.

8:54So it'll be interesting to see how the agentic economy evolves. But you know, it's probably like common wisdom at this point amongst your listeners, but there'll be more agents than humans in the not too distant future. it kind of stands to reason that the agentic economy will be bigger than the human economy at some point in the out-of-the-distant future as well. Yeah, it's interesting. I think one of the safetyist arguments is the concern that there'll be resource aggregation by AI and then that'll be used to manipulate society or things like that. It's always interesting to think through what are some of the implications of the scale of some of these things and how they work.

9:25I don't share that concern strongly. I'm just saying that's something that comes up from time to time. Yeah. Well, just actually on that, it's funny. I've heard people like Elon say, that they think there won't be money in the future, right? And I mean, I actually agree with Elon on most things, but that one I was like, okay, I think I know what he means. He's basically saying at some point, you start to measure things in energy and mass and like those become the scarce resource. I think in the end, like in the limit case, like he has a point, but the thing is there's gonna be scarcity for a long time.

9:55Even if the cost of goods and services really falls, like we have price deflation because robots and AI start to just make commodities and food and housing very cheap. That could happen. The government printing might also create a bunch of inflationary pressure, but there's always still gonna be scarcity of certain things like land, especially usable land in various ways, energy like chips. Dyson spheres are gonna be expensive, I think pretty far into the future. So I think we're still gonna need a medium of exchange And it's really cool to see that as AI emerges, CryptoRails are just kind of a nice complement for all that stuff to finally work.

10:39I think crypto is really, really good for humans, and it's going to be essential for AI. Yeah, people, I feel, have been talking about this for a while, too. Particularly, I think, in the early AI community, it overlaps so much with the crypto community, right? in terms of you'd see computer science majors coming out of MIT, and depending on the year that they graduated, they would go into crypto, and two years later, they're all going into AI, and it was roughly the same people in terms of people who were really into certain tips of math or algorithms or other things. So there's interesting bifurcation, I feel, too, of people's careers based on when they graduated in terms of whether they ended up doing crypto in 2020 or whatever it was versus AI in 2022.

11:18Yeah. Well, I'm reminded of that meme. if you've seen that meme it says Por que no los dos? You can do both. They're going to intersect here. Yeah, it seems like they just might. I mean, the other thing you folks have really been focused on is the use of AI within Coinbase itself. And can you talk about how some of I know there's obviously coding and other things but what do you think are some of the more interesting projects that you all have done internally? Yeah, so like a lot of companies we started off with the basics. You know, we got some coding tools. We got it hooked up into our various data repositories.

11:53We've been using it for a long time on machine learning for risk prevention, fraud prevention, customer support inquiries started going through AI. So that was all like good table stakes stuff. And then the things that we've started to do recently where we're sort of pushing the boundaries here, I think one of them is that, you know, we basically want to get as close as we can and actually achieve recursive self-improvement, right? And what does that mean in various ways? So it means that you have a brain for the company. That can be a whole bunch of subsets, by the way. You can have a brain for each team.

12:27You can have a brain for each service repository within the company. You can have a brain for each individual, which, by the way, I believe they should be able to take with them when they leave Coinbase. So we're creating some good policies on that. So let's say you have a service, like a GitHub repository, some backend service that's running. And today we have engineers spinning up agents to make changes to that. What we're just now launching and seeing recently is like every time the agent comes in to make a change to that service, it's actually ingesting the brain for the team working on it and for that repository.

13:00And so what is that brain? It's basically, it's a history of all the incidents that have ever happened with that service. All of the financial controls that have to be enforced with that. Every A-B test that we've ever run on that. Of course, like the full GitHub history of pull requests that were accepted or rejected. And so the people running that team at any time can kind of go in there and look at it. Today, it's just like markdown files, right? It's kind of simple at the end of the day. But getting this all hooked up. So, okay, so let's say you ask the agent, hey, I want to make a change to this service.

13:34It ingests the brain. It makes that change. And then as a human, you go to review it and you're like, ah, it missed something. Okay. Now, the simple thing someone might do in that moment is, okay, go make a manual edit to fix it and then ship it. But the key step that we're trying to really enforce now in our software factory or software development lifecycle is when you go to make that change and edit the agent's thinking, that context has to go back into the brain so that you not only fix it in this case, but in all future cases going forward. And then you basically, what you start to see is like the accept rate for, you know, one-shotted PRs or that kind of thing.

14:08It just starts to tick up over time. And you've now got like a recursive self-improvement system. So that's one of the big ideas we've been working on recently. And just getting our own internal agent harness, which we call Toshi, just integrated with all these tools, like it can now make calls out to external vendors and pay using the agentic finance economy over these X-402 protocols. If you want to edit the brain as you're making a change to that PR and it goes back into the brain, putting that into the harness, asking it to sort of like, okay, you know, like actually just earlier today, I was making a plan for this like relatively complex feature.

14:48What's kind of amazing, it's now even allowing people like me, the CEO to get back in there and ship features. But I asked one of the, you know, high-end expensive models, okay, go make a plan for how this feature might be implemented, break it down into three phases. Each phase has 10 different pieces that need to get built. And I was like, okay, go execute phase one, go spin up 10 separate agents to go execute each of those pieces in parallel. And by the way, you know, recommend the right agent to use because they might be cheaper models. And so it spun up a bunch of open source, it spun up a bunch of Grok.

15:17And it's, and it just like, two minutes before I hopped on this podcast, it just pinged me and was like, by the way, all 10 of those things from phase one are now done and ready for you to review. So anyway, I'm starting to get to this place where it's kind of addictive actually, because in the past, I kept, every time I'd have these like ideas or bugs, I find, and I'm just like pinging the team all day long. We're like, hey, can we do this? Can we do this? Can we do this? And now instead of like pinging the team in Slack, I can basically tag an agent in Slack with that comment or drop it into our internal harness.

15:47And instead of bugging the team and asking to do something, I just send them a PR for review and it's already done. And that's a magical moment that I just like, Like it's very addictive. Yeah, I feel like the people who are best at adopting this who aren't day-to-day engineers, and I know you had a long background working as an engineer at Airbnb and other places, obviously I'll be struck when it is, but I found that many folks who do best at this who are non-technical also, almost due to like a video game. And you're like leveling up in terms of your capability set and how you learn to do things.

16:18Sort of these recursive loops you start building for yourself and everything else. And so I think it's fascinating to watch that adoption. How do you think that impacts, you know, how companies are set up in the future? In other words, some people make the argument that the average company size over time should get much smaller. It could be a startup, it could be existing companies, et cetera, simply because you're able to harness so much of this intelligence, so much of this proactivity or, you know, ability for people to do so much individually. Do you think that's true? Do you think that's the direction things are heading?

16:50Like, is there a timeline where suddenly every company is like 10 people or whatever? Or do you think there'll always be these sort of larger scale organizations? Productivity is definitely increasing, right? So that's slightly different though than saying the average company size will be smaller. It's kind of like just saying tasks are being automated or eliminated. People are not being eliminated. I think tasks are being eliminated. And so if you take a company like any sort of company today, Coinbase or others that exist today. I don't think it's like we're going to shrink down the headcount to 10 people or something like that.

17:27I think that the existing people we have are going to be able to just get much more done and the pace of everything will accelerate. Now, it's also true, you're going to see some companies that are two people or five people or whatever, and they're able to do a lot more than what we would have historically thought was possible. And that's also true. So it's increasing the velocity of the existing team and making things ship faster, things automated. So basically you're getting more scalability off people. And so if you have a cost basis relative to revenue, then you suddenly have this ability to dramatically more with that same group of people, which means to some extent your margins should start to rise over time.

18:02If you're effectively growing revenue faster than you grow to you, I guess some people may still grow them at the same pace. Yeah. Yeah. That's super interesting. And then I wonder if it's worth spending a little bit of time as well on the everything exchange. I just think some of the things you're doing there are fascinating. There's perps, there's prediction markets, there's all these really interesting directions. What proportion of Coinbase, either volume or revenue is sort of those other incinerary areas now outside of crypto? Yeah, well, we actually shared in our recent earnings deck, which are, you know, those are the only numbers I'm probably allowed to talk about.

18:43But actually, 88 % of our revenue is from non-Bitcoin trading at this point. So that's an interesting stat. I think it's funny because if you look at our stock price, it does trade fairly correlated with the price of Bitcoin. And I love Bitcoin. I think it's going to be around forever. It's digital gold. It's sound money for everybody in the world. I think it's amazing. But I also think all these other things are... Crypto is updating every aspect of the financial system. So we just announced our tokenized stocks product, which is only available outside the US right now, unfortunately, but we're working with the SEC and others to get in a path to do it in the US.

19:18But it's the first truly tokenized stock product. It's not a synthetic, a derivative or some debt instrument. It's actually a security and it's one-to-one represented by the actual security held in custody. And so this is like broadly a trend called tokenization, right? Like stable coins were the first major use case of this with a token that represents one-to-one a dollar in a bank account or a US treasury. Now we're seeing it with stocks, which is really exciting. It's gonna, there's like something like 4 billion people in the world who don't have access to any brokerage or US investment account.

19:51So they don't have any ability to invest in, you know, Coinbase, NVIDIA, Apple, whatever. And so much like Tether was very successful, ex-US, because a lot of people wanted dollar-denominated accounts. There's like 4 billion people who want like access to high-quality investments. They don't have that today. So we'll do the tokenized stock piece. And then I think you just go on down the list, like private credit and treasuries and bank deposits, like everything's going to get tokenized. And it's just going to mean that financial services is going to have modern rails to run on. So of course, these can all be traded on the everything exchange.

20:24They can be sent back and forth. Like if you want a gift to share of stock to your nephew or something like that, you can just send it to their wallet. You don't have to like go through some complex stock transfer process that's antiquated. It's like money can now move at the speed of information, just like a WhatsApp message. And crypto is updating all those rails. That's pretty amazing. It's basically broadening and modernizing the financial system, but in a very broad way, right? It includes, to your point, eventually bonds. It includes a subset of stocks. It includes all sorts of things that before had much more friction in terms of using them or purchasing them.

20:59And it opens it up for the world, right? And there were controversies of this in China and other places where if something was listed on the Hong Kong Stock Exchange, Sometimes these funds would almost be set up that people could buy proportions of to be able to participate in some stocks that were listed in other countries. So it'll be interesting to watch how sort of global capital flows change as things get bigger and bigger. To your point, it feels like the stable coins and tether were the first version of that with USD. So it's kind of fascinating. I remember actually seven, eight years ago, I invested in a company called Harbour that was trying to tokenize real estate.

21:28I think that's a little bit too far ahead. Yeah, I think that'll happen too. I think you were too early, but right. Yeah, on that one. Yeah, so it's kind of fascinating once it'll display out. How do you think about prediction markets? So a lot of the use cases mentioned in the past seem to have roughly consolidated around sports and a few other sort of areas. Obviously, there's things in politics. Do you have a view on where in the long run prediction markets have the biggest footprint? Well, I'm very excited about prediction markets. They're growing super fast inside the Coinbase app. I think we shared in our last earnings call within months of launching it, it had to hit like a hundred million dollar revenue run rate.

22:04It's growing really nice, like over a hundred percent quarter of a quarter, something like that. I'd have to go look at the exact number again. But, um, so it's true that sports has been a part of it. Um, I think that it's much bigger than that though. And that's great. Like that's a form of entertainment. Um, you know, we should, we should celebrate that. I think that it actually has much broader implications though, from a societal point of view that people maybe haven't fully appreciated. So, you know, like, political elections are something that have already happened in prediction markets.