00:42:57 · Brian Armstrong
[00:42:57] S0: Yeah. Well, this was after we put out the Mission First blog post where we said, we're gonna focus on work at work. And if you wanna do political activism stuff not related to the mission, that's fine, but just don't do it at work. You can do it outside of work. And, yeah, I mean, the SEC or not the SEC. Sorry. The The New York Times, at this point, it was just I guess that was just, like, against their worldview. And so people told me later, what happened internally is they said, basically, just they put a team together to go write negative articles about Coinbase. So they sorta they had the they had the headline written before they'd even asked anybody what the news was because, you know, in my view, The New York Times is more of, a political propaganda arm than it is doing any real journal journalism at this point. So that was frustrating at the time. I was like, oh, man. These negative articles are coming out. But very quickly, I realized, it just didn't matter. You know? First of all, most of our customers don't read the New York Times or any traditional media, to be honest, or if they do, it's very rare. Like, I'd say people under 50 years old are mostly doing new media now, which would be, like, podcasts like this. They're reading blog posts in Substack. They're looking at x feeds. And so I think all companies should really, own their own distribution and, like, just publish directly and talk to new media. That's what I do with 8080% of my time. 20 of my time, I still spend with traditional media because there are people who are over over 50 or over 60, mostly on the East Coast, who still read traditional media. And so from, like, a deep like, if I wanna influence, like, policymakers in DC, I'll probably go talk to Politico or something. But, for most of our customers, the traditional media, it's dead. It's just not relevant anymore. And it took getting a negative article or two, which was scary at first to realize it just didn't matter. And it freed me up to think about, okay, what do I actually wanna create in the world if I don't have to, worry about being tact or, you know, and then it I think I got more creative and more confident.
00:44:56 · S2
[00:44:56] S2: Today's episode is sponsored by VCX by Fundrise, the public ticker for private tech, allowing investors of all sizes to invest in venture capital. Learn more at getvcx.com. Some of you may not have heard this yet, but our sponsor Public just launched something called generated assets, and it brings AI into investing in a way I've honestly never seen before. Here's how it works: You type in an idea like AI powered supply chain companies with positive free cash flow or defense tech companies growing revenue over 25% year over year. Publix AI then dispatches a swarm of agents that scan every single US stock, evaluates them, and instantly builds a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S and P 500, so you're making decisions with real context, not just guessing. Beyond generated assets, Public lets you invest in stocks, bonds, options, crypto all in one place. They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com/sorcery, paid for by Public Investing. Full disclosure is in the description. Enterprise AI runs on Merge, the AI infra platform for integrations, agent tooling, and model orchestration, so your teams ship product, not plumbing. Mistral, Dropbox, and Drata already trust Merge in production. Start building at Merge. Dev. Founders scale faster on Deal. Set up payroll for any country in minutes, hire anyone anywhere, Get visas handled fast and get back to building. Visit deal.com/sorcery. That's deel.com/sorcery.
00:46:41 · Molly O'Shea, Brian Armstrong
[00:46:41] S1: I mean, it is pretty interesting, like with all, with all the FinTech companies wanting to be the everything app or the super financial app, there's also the competing headlines of, like, Anthropic and OpenAI being the everything company. Like, they're gonna do everything. Where do you stand in that? Because some people are like, oh, I don't especially with some investors, like, the VC side, they're like, they're having an existential crisis because they don't know if they should be investing in more things or if Anthropic's just gonna build all the companies. Where do you stand in that? [00:47:18] S0: Well, I don't think OpenAI and Anthropic are gonna build all the companies. I think they're mostly focused on building foundational models, which are broadly applicable. But for instance, don't think anyone thinks they're gonna get into regulated financial services. It's a very complex and typical thing. Most tech companies don't want to have anything to do with that. And then I think there's actually gonna be specialized models. Like, actually, I think Coinbase has a real opportunity to go build a frontier model for investing, for instance, right, based on the datasets that we we have from our customers. So, yeah, I'm not worried about them eating lots of different companies. I think the best companies are just gonna get even more efficient, like Coinbase is doing. And if there's companies that fail to integrate AI, then they could be they could get left behind, but the best companies in every industry are just gonna be even more productive.
00:47:18 · Brian Armstrong, Molly O'Shea
[00:47:18] S0: Well, I don't think OpenAI and Anthropic are gonna build all the companies. I think they're mostly focused on building foundational models, which are broadly applicable. But for instance, don't think anyone thinks they're gonna get into regulated financial services. It's a very complex and typical thing. Most tech companies don't want to have anything to do with that. And then I think there's actually gonna be specialized models. Like, actually, I think Coinbase has a real opportunity to go build a frontier model for investing, for instance, right, based on the datasets that we we have from our customers. So, yeah, I'm not worried about them eating lots of different companies. I think the best companies are just gonna get even more efficient, like Coinbase is doing. And if there's companies that fail to integrate AI, then they could be they could get left behind, but the best companies in every industry are just gonna be even more productive. [00:48:12] S1: Who do you think your biggest competitor is?
00:48:15 · Brian Armstrong
[00:48:15] S0: It depends, in what category because Coinbase does a handful of things. Think, you know, if you look at, stablecoins, like, we've partnered with Circle and USDC, the biggest competitor to that would be Tether, right, which is offshore. I think, on the exchange like, on the crypto exchange side, our biggest competitor is probably Binance, right, which, again, is offshore. I think if you look at The US market where some of these, fintech apps are integrating more crypto, you know, and we offer stock trading now, so Robinhood would probably be for our retail apps specifically, like, Robinhood is probably the most, clear comp. And then, with Coinbase developer platform, which is what it's allowing all these businesses to integrate stablecoins, you know, Stripe is probably the best competitor we have there. So, yeah, we also run the largest institutional business in crypto. I actually don't think there's, like, a close competitor for that, at least not one that comes to mind. So, yeah, we have a number of different products and, competitors in each of those categories. Oh, by way, one other maybe interesting take on the AI, usage. You you know, you're asking about anthropic and all that. Yeah. So one interesting phenomenon, like a lot of companies, we were seeing our AI spend go up, like, exponentially. Mhmm. And I went with a team and I kind of looked at this and I said, the open source models are about three to to six months behind the frontier models. Right? But they're 99% cheaper for inference. And I was like, what percent of our prompts are we routing to open source models? Because if you're asking a really complicated question, you might wanna use the frontier model. But if you're just asking like, you know, who's Molly's manager, you know, or whatever, you're trying to look up something, why are you why are you using the the most expensive model? So we actually made some very simple changes, like some more intelligent routing of prompts to, depending on the complexity of the query to the right model. We started caching some requests, just some sensible defaults. When certain employees were about to cross their budget threshold, we would just say, are you aware this is about to happen? And you can keep going if you're doing something useful. We don't wanna slow you down, but just even making them aware of that.
00:50:30 · Brian Armstrong, Molly O'Shea
And what was really cool is we are now seeing our usage of AI tokens is still continuing to grow exponentially, but the cost curve has really started to flatten. [00:50:39] S1: Really? [00:50:40] S0: Yeah. I think that this is something that a lot of companies are gonna do next, and we're getting this intelligent middleware to know where to route the query to the right level of model. [00:50:51] S1: Did you, you built that internally? You're not using another company for [00:50:55] S0: There's an open source project that we based it on, and then we modified it. I think the open source project was called Light LM is one of the ones we use. There's some other in house stuff that we used as well. [00:51:06] S1: Oh, wow. [00:51:07] S0: Yeah. So I, my guess is that within twelve to eighteen months, 80% of our workloads will be going toward models that are 99, 99% cheaper, but 20% of it will still go to these frontier models where you need to be, you know, IQ maxing or whatever you want to call it. [00:51:22] S1: Yeah. IQ maxing. Okay. So we've had a really fun and eventful background to us with everyone getting ready for tomorrow. I do want to go back to a couple of the announcements that you have coming up. What is the one that you're most excited about?