Timing in the IPO market is a perpetually tricky dance. This week, we've seen a number of companies push off their listings, raising the question of what that could mean for Anthropic. Few people know that dance better than Circle CEO Jeremy Allaire, who listed the company last year. I want to bring him on for some of his thoughts on all of this. Jeremy, welcome back to the show. It's great to have you here.
Yeah, thanks. Happy to be here.
Just remind me, Jeremy, how many companies have you taken public now? Is it three is the count?
Three companies. Yes. Exactly.
All right. So, you know, kind of Web 1.0, 2.0, 3.0, etc. Okay. And maybe we'll see another one from you. Who knows? I mean, we don't know.
Okay. So, you wrote this piece last week about the considerations that AI companies might be making as they look towards the IPO window. And look, you led with, you know, four potent words: 'Take the leap, Anthropic' is what you wrote. So, walk me through, you know, why you're encouraging them to go.
Yeah, I mean, it's been interesting. Like, I think there's been a lot of media coverage about these potential IPOs. There's been a lot of media coverage about, you know, are they ready? Are we still waiting on, you know, does everyone need to figure out what regulation's going to be, or how they deal with safety, or can they prove that they have a business model? Like, there's all these sort of questions that are out there, and some of that is sort of throwing shade on whether they should go public or not. And then I think you sort of have had, you know, different leaders of different companies sort of pacing not the frontier, pacing their own IPOs.
And so, you know, I looked at it and kind of reflected a little bit on what we had gone through. And yeah, there are a lot of people who thought, you know, we shouldn't go public, that the markets were so volatile and uncertain, or, you know, there's still a lot of unresolved regulatory uncertainty. Laws weren't clear. You know, no investor was going to want to invest when there's this uncertainty that exists. And I think, you know, the reasoning that we had, it sort of made sense to me that it's sort of very similar for a firm like Anthropic or its peers, right? Which is, you know, essentially this infrastructure is becoming important infrastructure for the world. And, you know, when the infrastructure that you create is something that people are going to really depend on in society — in our case, it's sort of this is monetary infrastructure, like, is this money real? Is this, you know, is this trustworthy, you know, etc. Public markets create a real opportunity for that transparency and, you know, the system of public disclosures of risks, of operations, of other things. It sort of holds firms to this higher standard, and it's possible to do that. And I think that's advantageous for society.
And for markets, right? Markets themselves, society more broadly, not just for investors but for the rest of the world. I mean, I think, you know, what we certainly saw post the Circle IPO is that that amount of disclosure and accountability means that counterparties around the world, they can have a better understanding of us. Whether they're going to buy or sell the stock is totally irrelevant; it's just like they can have a better understanding. And so I think something as significant as these, you know, frontier AI companies, these kind of foundation model companies, I think there's a very similar argument to be made, right? Just take the plunge because, you know, someone's got to go first, and transparency is good.
Yeah. So, I hear you. It was a good point you made. I mean, the one thing that you pointed out, though, in your piece is you wrote, 'An IPO cannot substitute for fit-for-purpose public policy, and regulation cannot substitute for the discipline of public markets,' and you talked about them being complementary. So it sounds like the way I read that was that it will help with transparency, but it will not substitute the need for additional regulation around AI.
And we've seen that exact same thing, like in our own world, which is like public markets are great. And there's the discipline, you know, that comes from that in terms of evaluating your business, your prospects, like the decisions you're making, other things as well. But like we still needed the regulation, right? And so, you know, I think they're serving different functions. Both are actually forms of public governance, right? There's the public market governance, and then there's the public sector, you know, kind of policy governance. And again, I think for these really important firms that are providing infrastructure for society, both are obviously critically important. What I think is clear just from on AI policy is that there is not AI policy today. And while there's people talking about AI policy, there is not in fact any AI policy. And so, but I think that this kind of public disclosure and public market governance is actually very important. So, you know, AI policy is going to take a little while, maybe unless there's like emergency actions and things like that.
So what then do you make of the Jensen view that we have all the policies we need? You know, under existing policy, if an agent goes rogue and hacks into your system, you can't do that under existing policy. I mean, it sounds like you're saying we do need more government regulation around this tech.
So, I mean, look, I'm not an expert. And so, but I certainly do have a point of view, and I don't agree with Jensen's view. My own view is we're in a place right now where the labs themselves, who are at the frontier of these new AI supercomputers, where they themselves do not have an understanding of the fundamental safety alignment of these systems. And even in their attempt to have that, they're having to use the AI supercomputers themselves to even attempt to understand the model safety and alignment. And it feels very precarious. And so I don't agree that, you know, this is just a liability question. And I think the stakes are extraordinarily high as well for society. And it feels like, you know, we're at this event I was at, you know, last week, there was some, you know, kind of high-level discussion on some of these themes. And I think there's different worldviews here. I think, you know, one worldview is sort of this is almost like species evolution, and we don't understand the new species, and the species has, you know, cognition, reasoning, and coordination capabilities that exceed ours. And yet we're trying to say, hey, our human cognition, reasoning, and coordination capabilities are good enough to make sure we can control this new species. But the species may have already, you know, I don't know what the right, you know, flew the coop, whatever the right phrase is. And so, we are definitely dependent on the private sector right now to do the regulating, to get to the bottom of this as best as possible. We also need third parties to do that as well. I think we also need, you know, self-regulatory structures, you know, also. But it's, you know, I am of the view that, you know, we're dealing with technologies that themselves are kind of weapons-grade systems. And so when you're dealing with weapons-grade systems, and those are weapons that can be deployed in a variety of different ways, and we have significant degrees of uncertainty, that's a policy environment. That's a very real policy environment.
