00:54:22 · Will Manidis, John Coogan, Jordi Hays
[00:54:22] S3: be applying for union benefits very shortly. [00:54:25] S1: OpenAI. Yes. Yes. Yes. That's for sure. [00:54:30] S0: How's week been? Yeah. How's week been? Know this has been in the works for... [00:54:33] S1: Are you signing autographs yet? Are you getting recognized on Print the [00:54:37] S3: copies are en route. Mhmm. Commanding-heights.com request a physical copy. We have a thousand of them going out today. [00:54:44] S1: Cool. [00:54:45] S3: Leather bound? Leather bound, hardback, hand sewn, not America. You can't hire American folks anymore. [00:54:52] S0: Me too. Wow. There you go. Interesting. What... Talk about the process. [00:54:57] S1: Yeah. [00:54:57] S0: We've been hearing about it for for months now. Lighter's been... [00:55:01] S1: Well, it's been quiet.... Couple of things. Because for a while you were on a generational run of these essays that would go mega viral and it felt like chapters in a book and then you went quiet for a while. That was the book writing process? [00:55:13] S3: That was the book writing process. Joe and [00:55:15] S0: I have
00:55:15 · Jordi Hays, Will Manidis, John Coogan, Tyler
[00:55:15] S0: I have [00:55:16] S3: been convinced that AI is real. Right? Scaling laws are going to continually uphold us correct about almost everything. Mhmm. And we've been kind of shocked at how little discussion there's been about what this means for currencies and what those means for nations and what this means for populations. So about a year ago we said, what is kind of... What does the East Coast have to say about the the coming scaling laws that are going to happen? What is the finance take on everything that's going to happen? And we set out to write this manuscript. Joe Roberts and I, Farrell Gregory sat down, spent the last year drafting this thing and published this week. Very exciting. [00:55:48] S1: And what is the what is the TLDR on the knock on effects? Because everyone knows the bottleneck constraints like the the the situational awareness thesis has been digested. You're saying AI is real, but what are some of the knock on effects? [00:56:02] S2: Yeah. I mean, I think we need to
00:56:04 · Will Manidis
[00:56:04] S3: move past bottlenecks and start to think about who is uniquely advantaged as a result of this technology. Permanently. It's staggering to me. Permanently. Yeah. Right? Like, it's staggering to me that you can use instinct every day at work and not process of like millions of people's jobs are within the capabilities cone of the models we have today. And if take the epoch AI scaling curve seriously, I think we should, these models are going to get so expensive that it gets to a point where there's only kind of one or two states that can continue to maintain a sovereign tier, meaning models that are kind of world class and continue to be trained. Yeah. So if we have a world where the best models are held inside of American skiffs, where the national security community is given Fable 5.5, right? Maybe some unreleased TBPN episodes, really kind of high security stuff. Yep. And the Chinese are able to maintain their own sovereign frontier in the East. Then what does this mean for the populations that are within the diffusion net? And we introduced a couple of new ideas as part of this PDF. The sovereign frontier is the big one. What I think is a really obvious idea that you will not soon be able to buy the best models, that labs will become something much closer to the defense industrial complex, that they will have models that are available to nation states for national security purposes. You will not be using them. Know, harkens back to the GPT-two release when Anthropic said, you know, it's too dangerous to release, but much more seriously. And then two, more importantly, as we think this will manufacture huge amounts of demand to transact on American payment rails. We introduced the concept of the neo dollar. Yeah. Which is this idea that if models are transacting neo dollar was going on. Yeah. If these models are transacting in the dollar, if all of agents are transacting in the dollar, if the thing that is backing US sovereignty is dollar denominated models, then the rest the world is going to dollarize quite quickly. And we introduced kind of two competing financial systems in
00:57:50 · Tyler, Will Manidis, John Coogan, Jordi Hays
[00:57:50] S2: the RMB and the dollar Yep. [00:57:52] S3: As two kind of different abilities to produce huge amounts of dollars to train the sovereign frontier. [00:57:57] S1: So how do you take that thesis into some of the more low risk mundane AI enabled technologies? I'm thinking of just like website builders. Like that's something like are you gonna have to go... Does Lovable have to go to the government for their inference or can they just use current models? [00:58:13] S0: These use cases are like pretty saturated. [00:58:15] S1: They are saturated. Yeah. But do they continue? Okay. [00:58:19] S3: Think like if you think like the the you know, the the the vast majority of workloads in The US Mhmm. Very soon will be kind of Chinese open source models. OpenRouter makes this data very obvious. [00:58:28] S0: Mhmm. [00:58:30] S3: A bigger and bigger share of Frontier Lab revenues are just going to be from USG. It's the only possible way you can kind of finance these training runs with this So cost and the if you imagine kind of a a Chinese Open Frontier powering most US commercial operations, at the same time you imagine incredibly proprietary, incredibly close frontier intelligence powering kind of the the tip of the spear for US hegemony. I think you end up with a pretty stable market equilibrium. Like, I think it's all going to work. Right? Like, most discussions of this as a bubble, think, really miss the fact that this is the central economy of The US at this point. The entire empire is hinged on this. We got to make this work.
00:58:30 · Will Manidis, John Coogan
[00:58:30] S3: A bigger and bigger share of Frontier Lab revenues are just going to be from USG. It's the only possible way you can kind of finance these training runs with this So cost and the if you imagine kind of a a Chinese Open Frontier powering most US commercial operations, at the same time you imagine incredibly proprietary, incredibly close frontier intelligence powering kind of the the tip of the spear for US hegemony. I think you end up with a pretty stable market equilibrium. Like, I think it's all going to work. Right? Like, most discussions of this as a bubble, think, really miss the fact that this is the central economy of The US at this point. The entire empire is hinged on this. We got to make this work. [00:59:07] S1: So what are some of the other knock on effects? I mean, by dollars that be beyond that? [00:59:14] S3: I mean, dollar dollar devaluation seems really obvious, right? We will have to print huge amounts of dollars in order to fund those build out. Mhmm. One of our very strong convictions is that we will dollarize the entire world at the same time the dollar is inflated to basically zero. [00:59:28] S1: Okay. [00:59:29] S3: You probably want to own equities, right? A long equity is short to the dollar. Mhmm. And we've become particularly obsessed with the class of businesses we call commanding heights, is Yep. You know, where the title of the piece comes from, which are businesses that get better as intelligent scale. [00:59:44] S1: So you can
00:59:44 · John Coogan, Will Manidis, Jordi Hays
[00:59:44] S1: So you can [00:59:44] S3: think of it as kind of a mixed sleep shared economics or share of economics shared to the customer. But in the context of continuing AI improvement, what kind of businesses get better and better with size? [00:59:55] S1: Mhmm. And what are some examples of those types of businesses that should get better in the AI era? There's a article in the journal today saying that despite Google being down, Alphabet can still win in an AI pushback. How are you thinking about the different markets, the different tiers? [01:00:12] S0: Yeah. Thought I thought it was interesting that that this week we had commanding heights and we had Cliff's episode Oh, post episode on invest like the vest which was like somewhat complementary Yeah. In some ways, talking about the comp... What are the compliments of intelligence Mhmm. Which results... [01:00:27] S3: Yeah. I think Cliff and I are both like giant meta bowls at this point. Right? Like it's really hard to imagine a business that is better fit to diffuse intelligence to the global economy than Meta. All right. You have payment rails, you have WhatsApp which reaches the entire global South, you have frontier capabilities through Meta Super Intelligence Labs. Right. It is just an unbelievably well positioned business to both oversee the dollarization of the world and oversee the intelligence frontier across the world. Mhmm. We really like meta. In the piece we also support hyper...