Anthropic's $30T Assumption & OpenAI Confirms IPO | Why Customer Service & Robotics are Overinflated

Let me tell you one thing I guarantee: if you get 15x on your failures, eventually you'll die a rich man.

I think you have to believe we're less than a third of the way through this cycle.

It's all about code. That's the only sales that matter.

Jesus Christ. If 30% of my company leaves to go work for Harvey, I'm dead in the water.

Silicon Valley forgets every three years that the average American is not trying to be efficient.

The facts are correct. I thought, first of all, I thought it was an excellent letter. I mean, I read it and I reread it, and there were some really good phrases in there. One of them I want to pick up, and it said, you know, 'We have found ourselves on the right side of prediction in a market which has scaled exponentially in terms of capital intensity,' which is nicely. What they're basically saying is, we were right three years ago that there was a market for a US open-source model, and we've built that model. We've done everything we said we'd do, and the capital intensity for the next turn of the model crank is, just as you said, way beyond us, right? So nicely phrased, right? And so a couple of things here. You could look at that and go — I mean, you could look at that and go, 'Oh, they did something, they took on a task they weren't able to do with the capital. Oh, that's a negative.' On the other hand, they made money for themselves and all their shareholders. And I think the lesson here is, and we talked about last week, in a market that's exploding, sometimes bets that on a standalone basis really just can't get to a positive DCF — you know, they just couldn't make the math on a standalone basis — still have pretty significant value to the acquirer. Because from the perspective of Nvidia, they looked at exactly the same facts and said, 'Well, you know, we have capital. It turns out we have access to GPUs because we make GPUs, right? And you've kind of carried the ball down the field this far. We'll take it from here, and frankly, we'll give you a pretty compelling return.' I mean, I know you're a small Poolside shareholder, Harry. So all those guys made really good money, and the company gets to go standalone and continue on. So I thought the lesson here is — you're right, Jason, in one kind of negative way. The lesson is it's almost impossible to compete now at the frontier. And by the way, that has positive implications for OpenAI and Anthropic, which we should talk about in a second. So the negative lesson for everyone else is it is almost the next smartest people who were really going for it just hit the capital wall. So probably all the other people behind them are going to hit the same wall, right? That's the negative implication. And the positive implication is sometimes trying and moving the ball in a hypergrowth market, you can still get a very compelling acquisition, right? You know, in a different market, if you'd run out — quote unquote — run out of money, if you had reached the next generation where you can no longer finance the company in a time when the capital markets were oppressed or there was some feeling that the overall, you know, overall kind of buoy around equity — sorry, around AI — wasn't as strong, you might have had a very different outcome there because your fate was outside your control, right? But I think in this market, it pays to take risks. It's what we were talking about last week in Cursor. And I'm sorry if I'm rambling a little. Even when you have this kind of, 'Oh, on a DCF MBA basis, it's not quite great,' it turns out if you're moving in the right direction and you execute and build a product that's valuable right now, you're getting great exits. And I think there's going to be a ton more. I think I said last week there's going to be a ton more of these kind of exits, and we'll be talking about them. So yeah, I read that and I thought, as an investor, if all the time you're saying is sometimes doing things where there is a fair amount of capital-raising risk can still pay off in the nut market — in a very different market it wouldn't have, but here it did.

Absolutely. Pretty amazing. I think I was very uncrisp a few minutes ago. When you're in the direction of travel, even when you're wrong, you can make a ton of money because you've moved, you know, because you've created something of value to the acquirer. Different market could have been a very different outcome.

Well, there's two questions within that. One — and I'm sorry — one is, is this additive to Neotron? Then the second question is, does Neotron even — even if it is competitive, does it move the needle? But if you zoom out a level, right now there's a whole bunch of Chinese open-source models that are getting a lot of the token volume, which means they're doing a lot of the compute, even if they're not getting a lot of the margin, right? And if I'm the maker of compute, if I can get an open — it's awesome for Nvidia if there's a viable US open-source model running on Nvidia chips taking market share away from the frontier models at the margin. So I totally get why they're doing this. This is — yeah, an open-source model is a complement in the economic sense of the word for Nvidia, because the more open-source market share there is, the more money goes to chips relative to the money that goes to financial model builders. So they're like, 'Yay, team.'

I don't think 15x is good enough for a seed investor in Poolside. If that's the number — I mean, listen, first of all, let's step back. Seed has a weird definition today. Seeds could be at a billion or two billion pre or post, right? But not to get off track, but just — 15x sounds good for a later-stage investor, and I know for scales, that's a good outcome, right? But for a true seed investor, it's not — you're not going to get a fund return out of a 15x. Listen, if you're a personal investor, it's fine. But if you've got a seed fund and you're investing it, just a reminder: if you're investing at some of these valuations, it may not be enough. You're not going to get your 50 or 100x out of 9 billion.

But Jason, play it back. I hear you mathematically, but what we're really saying is you made a bet and it didn't work. It was not viable, and you still made a 15x, right? My point is —

No, I'm saying 9 billion isn't enough of an exit to proceed investments today.

But the point — I agree, but —

It's not enough. 9 billion doesn't clear the bar for seed investing in 2026. That's the irony.

So there's two things in that. I'm going to push on it because I often use this as a way to refine my own thinking. One, it is weird that you can have an exit at 9 billion and only make 15 times your money, which by definition would imply 600 million pre-money, right? So yes, it turns out if you do a seed deal at 600, not 60, your return is —

Well, so the dilution is epic. Good point. Effectively, it's what your point. It's ex post facto. Ex post facto. The price was — the effective price was 600 because you took all the dilution. You're exactly right. Jason, I'm going to go back to —

Is 15x a great return on your best deal for a seed fund? No. But if you're doing a series of bets, one of them is, 'Hey, I think I can build an American open-source model,' and you can raise enough capital to do it, and it turns out that thesis is not correct — you can't raise capital to meet the capital intensity — and you still get a 15x. Let me tell you one thing. I guarantee if you get 15x on your failures in venture, you'll die a rich man.

Look, we don't have to spend much more time on it. It's just — I guess obviously, if you can get 15x out of your — Poolside, of course, was not a failure. It's a big win, right? But if you get 15x out of your failures, you're going to be a wildly successful investor. Just the only point, then let's move on, is it's just interesting. I do think unless it's a hyper-concentrated investment, your best investments still need to do 50x as a seed investor to make the math work. So just if Poolside returned 15x with further dilution, what exit price would it have to be to be a good seed investment, right, to return 50x? Help me do the math, Rory. I think it's about 7x. So it would need to exit at about 63 billion with dilution.

If you think about it, Jason, the answer to the question is actually very noble, right? You're playing in the foundation model — you're playing in the frontier model game. The two winners in that game in closed source are both worth a trillion. The two or three kind of ostensible open-weight winners in China are each worth 50 to 100 billion. So it was a rational act. The win, if you could have done it, was in fact large enough to be a — you could have had a 100-to-1 return on a seed investment if you'd been able to be — possibly if you'd been able to be equivalent to the Chinese open-source players, and definitely if you'd been a winner like OpenAI or Anthropic. So my point is this: the potential was there for that bet, which is why, you know, at the time it was a rational bet, and then the capital markets were such that you couldn't get it, and you still got a 15x. That's the movie. So it was a good bet because it had — it was one of the few businesses on the world — I mean, there's only going to be one or two trillion-dollar outcomes per decade, and likely going to be all concentrated for now in, you know, frontier models. And this was a player to frontier model. And basically the aha is: if first prize is a trillion dollars, turns out fifth prize is still 9 billion. It's like the guys, you know, in golf on the last day when you don't win, you miss a few putts, you don't even come in second in the US Open. You come in seventh, and everyone goes, 'Oh, poor you.' And then you think to yourself, 'Oh, it's 5 million bucks. I'll take it. On to next week.' That's what happened here. You basically placed, you know, high but not in the top three in the US Open, and you get a bunch of money.

I land in the bunker and I don't get out. So,

You land in the bunker and you chipped out. You got out. You got out. You took two extra shots, but you — you remember this is a win.

There's an investment bank called Croll. I'm embarrassed to say I don't even know them. Do you know them?

No. Oh, I thought they were a detective agency.

No, different one. They published a bunch of — they published their report this week looking at all M&A and big transactions over the last six months of the year. And I guess this is not that — just a micro point to your point, but the analysis they basically made is gross margins above 30% don't matter anymore. You don't get any benefit in M&A or other exits like for an agentic stock. They're not looking — there is a penalty if you look at all deals below 30, but there is no penalty. So, you know, if you can value Merkar at classic multiples for 80% gross margins and growing, it's, you know, it's not expensive. I don't know why Nvidia would do it per se, but it seems like Nvidia's tact — they've got a lot — they've got massive strategic goals here, right? But the tactical basically kind of seems to me: we'll spend all our free cash flow on our ecosystem.

Yeah.

Right. Not that this is our budget. Jensen says there's a budget. Our free — I don't know, it's going to be 70 billion this year. Maybe I've gotten that wrong, right? And we're going to spend it all on our customers and ecosystem. And the team — the strat team and the top VPs probably get around a room and they decide what's our best ideas. And there's some guy that thinks data labeling is important, and his best idea was Merkar. And so they put a few billion of the 75 billion. And you think I'm kidding, but I'm confident they go around the room and everyone has their best ideas. And the budget's 75 billion. And, you know, OpenAI and friends are going to get a big chunk of it, and there's going to be off-balance-sheet guarantees. But I think they've decided to just spend it all one-for-one. And I would do the same if you can get away with it. You know, adding cash to the balance sheet other than being defensive does nothing for you as a CEO of a profitable company, right? If Wall Street let you get away with spending it, I would spend 100% of my cash too.

First of all, I think you're totally right, Jason. It is as simple as that. It's like, you know, because — and by the way, Nvidia will have reported between the time we make this video and the time we distribute it on Thursday. So there could be a data point that makes us look like total buffoons by Thursday, and that's just life, right? But on the basic assumption that the trend continues apace, which I think it does, which is, you know, strong growing quarters — you know, 60 or 70% up last quarter and similar quarter last year — wildly profitable because the demand signals are still strong from the hyperscalers. Let's assume on Thursday the hypothesis as of Tuesday continues correct. Then Jason, you're right, they're kicking off — I mean, I just looked at it, you know, they make huge operating margins. You know, it's like — it's like 34 — I mean, one out of 100 billion of gross cash flow. And then you have — they do a lot of buybacks, we'll talk about that in a second. So they have yunks of money to invest every year, right? And you're right, they just make a list of what can we do to move the ecosystem along right now. I'll make two negative comments on that — you know, two comments on that. One is it's worth remembering that only four years ago their cash flow was 10x — was one-tenth of what it is today, right? They're making, you know, and I think the free cash flow after capex and everything is something like — and I think some of the capex is really investments, so it's a little misleading — it's kind of gone from 4 billion to 50 billion, right? You know, the gross profitability is well over 100 — it's wildly profitable. But, you know, you say to yourself — and you do, you're right, Jason, you get no points for cash on the balance sheet — but they only have 50 billion of cash on the balance sheet, right? Cash and investments. There's a little part of me that says I might keep more for a rainy day, right, than just doing share buybacks. But I agree, you do have to do something with it. And they're clearly — the best use for it is to spend the money with people who will enhance the ecosystem, which is why, you know, which is why OpenAI gets a big watch for their money all the time.

If you invest right, it's brilliant. You get a customer, you get — not only do you get circular revenue, but you ensure the success of your ecosystem, the viability. You get a twofer out of it. So you just have to play the game really well. But if you play the game well, it's a lot better than making 3%.

The Poolside acquisition is time expansion by buying an adjacent product. The Merkar acquisition is, I think, as you say, Jason, straight in. It's kind of just, hey, it's a related space, we know some — here's some money. And then something like a Perplexity investment or OpenAI investment is literally — especially the OpenAI — is vendor financing, where you as the vendor of the chips are choosing to give your customer money, right? And you're right, that's — it's not like that's nefarious, because a lot of, 'Oh my god, it's circular.' It's not like that's nefarious. It's just exactly what you said, Jason: you got to get it right. If you overextend credit and underwrite projections that aren't realistic, that stuff comes back to you and you look like an idiot in two years. And we all remember the telecom crash in 2000. Well, we don't all remember, but some of us remember the telecom crash in '02 where all those '99 deals unwound. And the bet that Nvidia is taking here is it's not going to happen this time. You know, you're not going to find that the OpenAIs and the Anthropics don't suddenly — you know, if they don't suddenly need a hundred billion worth of compute. You might regret some of this vendor financing, but right now it looks smart.

I think it's a good investment. I consistently regret not putting more money in overtime. I think your largest data providers will be $200 billion companies. If you think about OpenAI and Anthropic being two- to five-trillion-dollar businesses, is it crazy to think that the data provider that provides them their core data assets will be 10% of their market cap? I don't think so.

I don't know. I don't know. I mean, I think you have to think about it in terms of revenue rather than market cap. And then you start saying to yourself, you know, what's the training budget for the frontier models at scale? If they're doing, you know, like — I mean, what's this — and let me give you the negative math. We know that, you know, OpenAI is running at 18 billion H1s, call it 25 billion, 30 billion a year. Anthropic's at 60 billion a year. Together, let's call it 100 billion, round up, right? You know, most of that spend goes on — half of that spend goes on compute and inference. What's the training budget? Is it 10% of revenue? Is it 5% of revenue? Yeah, it's a 5-to-10 billion market, right?

Way more. Way more. You've got Surge. Oh, I know it's way more because Surge are doing three, three and a half. Merkar are doing two and a half. Handshake are doing one. Micro are doing half.

The question — you're right. So my point is, so it's predicated on if that 100 billion of spend this year goes to four or five hundred billion, then you're right. Then 5% of that is 25 billion divided four or five ways. I don't think you get 10% of the market cap of the frontier models, but you still get a healthy 20 billion a year revenue spend. Divide that up three ways, and then as Jason said, actually it turns out the most important question is the one Jason asked, which is what multiple do you attach to that? Do you attach the AI multiple or do you attach the lower gross margin multiple? That's really what will swing it.

The AI multiple.

Yeah. The related thing I was thinking on that Croll data, you know, if you go back to Cursor, right, 60 billion we talked about last week. This was one of the classic ones where the VC-isms were right. It worked itself out. You started off with something with negative gross margins. You started off with something that radically subsidized its cost. I forget what Cursor cost initially — 200 bucks a month for unlimited use, right? But it didn't have — but it didn't — it had no way to defend that, right? And so then it had to cap it, and then it had to stop doing it, and then it had to develop its own models, and then it had to do — it had to do — and but the VC-isms: these are some of the smartest kids ever. They have a strong market position. The wind is at their sails. The kids will figure it out. The negative gross margins — and more or less they did, and they got to the 60 billion. And so it's kind of freeing for models like Merkar that we maybe made fun of, right? Okay, this is a commodity, low-margin business. But the kids will — the kids are figuring it out, man. The kids are figuring it out. So guns are blazing.

Yeah, no, I think though there is a bit of survivor — first of all, I agree with you, and I think it gets to the — when as a venture investor you're investing in something that has troubling gross margins, it is rational to say for certain bets the gross margins will show up, will come right over time, and you add Cursor as an example of that. Just in the interest of completeness, I don't know why I'm the Debbie Downer today — like, not every negative gross margin company makes it. And in the end, I think you want — and strong gross margin companies are the best investments. So don't just look at the sample of deals where you started with negative gross margin and it all worked. I can think of plenty of deals, including some we've done, where you started with shitty gross margins and you ended with shitty gross margins and you were wrong. Right. I think, Jason, to your point, the thing is having negative gross margins is not a reason — clearly not a reason not to do a deal. I mean, if every part of the financials were pristine, they wouldn't need venture capital because they'd be profitable, right? It's a question of in which cases is it rational to underwrite massive improvement. And it has been for the foundation models themselves. I mean, went from 911 to positive 30 in a year. It has been, as you say, Jason, for Cursor. It'll be interesting to see what the sustaining long-term margins for something like Merkar are. I don't know if they have the same dynamics in terms of the ability to improve those margins that, say, Cursor did, because the training companies only have three to five big customers, whereas Cursor has literally hundreds of thousands. So I'm not as convinced those margins go as well as the Cursor ones. But you're right, Cursor beat the margin drop. And Rory, and as I tell my partners and subordinates on Monday meetings: what if it all goes well? What if it all goes better than we expected? What if it all goes — and I challenge my team: what if it all goes right, guys? That's how we think about it. And the entry price for any of these winners doesn't matter. What if it all goes — what if it all goes right, guys? That's — and I've coined that term, and I've noticed many, many of my colleagues in the industry have copied it. What if it all goes right? You know what I mean?

No, I love it.

That optimists make money and pessimists are right. We often say —

And the only reason I say it is because I admit that I can be naturally a pessimist. So I actually — I'm trying to learn to retrain the model to do that more. And Jason, I love the sentence, 'What if it all goes right?' And —

The first two have said it.

I know, I know that, of course. But the ancillary point, interesting, going back to Poolside, is it turns out even what if it all doesn't quite go right, but you're in a great market, it turns out that can be okay too.

If enough goes right, you can all do fine. There's been a lot of cynicism, skepticism around Arav and Trinyas and Perplexity. It's a company that actually people like to dislike for whatever reason. We've done shows with him. We're a small investor in the company, so Nvidia is actually propping up most of my portfolio.

Thank you.

But that level of growth relative to expectations would have disappointed massively, everyone, including all the people planning to sell them chips at them for a much higher growth rate.

I have no clue, Harry. But the most important point you have to say is this: it will be lower than the other guys. Now, I mean, at the high level in terms of report card, you know, there's a concept in math, I can't remember what it is, where you can't do accurate grading, but you can rank things, right? The big picture fact is the ranking has changed and you're now number two. So you will go out later and at a lower price than the other guys unless you change the trajectory massively. What that is, I don't know. I just think it'll be interesting the rest of the year, by the end of the year, to see where OpenAI is positioned in the platform, the enterprise versus Anthropic, because at the beginning of the year there were two choices, right? There was Anthropic and OpenAI for the most part, and you had a default choice. And then people wanted to have multimodels, they wanted to have at least two, but you'd often use the cheaper version, you might use Sonnet and Opus or whatever it was. Now people, we definitely want different LLMs. Everyone wants multiple LLMs in their stack one way or the other. But if Anthropic is number one in platform, which it clearly is, you can't argue with the numbers. Peri's point. Now the number two could be, there could be five choices. So OpenAI's position in a sense, and we'll see. There's so much change, guys, but it could be, no matter what the numbers are, it could be getting perpetually weaker because there's so many choices for number two. There's so many choices. There's so many open-weight models where the performance is close. And obviously we've talked about OpenRouter. I've talked about other routers. You can use 78 models if you want. But it's tough when there's seven choices for number two. It's just tougher. It's just tougher, right? Especially if you're a premium product.

I think there's more pressure on them. I mean, for merely becoming number two, there's more pressure on you from open-weight and open-source competitors than there was otherwise. You're just battling out for that second spot, and you can't compete on price. You can compete on brand and security, but man, you want to at least be plugged into every workflow, right? You want to at least have that. That would be the nice thing if you had Claude and OpenAI, and then you just get a nice oligopolistic bake-off, right? Sales team, then you hire a bunch of folks from Salesforce that walk in for the oligopolistic thing and they put up a PowerPoint slide and it's just us versus them. They know how to sell that. It's Anthropic versus us. You don't want the crazy guy. You want the guy the government trusts. That's us. But with 11 competitors or 12 competitors running on open source, running inference on new platforms, man, it's a hyper-competitive world for number two.

Going back to the question, because I mean, I think that even if they are number two, I don't think they get pushed into that, you know, compared with all the open sources. I don't think it's quite as dire as that. But I do agree there was a push from underneath from that. But I think to your question on what quote unquote...

Agreed. Yes, I think that the vast predominance of tokens processed will be open-weight, and obviously, as I keep reiterating, the vast majority of revenue, the significant majority of revenue, will still be frontier state-of-the-art because it can command more value than just the price of inference, right? So yeah, and so I suppose at some level Jason is correct because if there's five or six, and I don't see them as being peers, but I think if there's five or six open-weight companies dragging down everyone's gross, you're basically trying to steal gross margin away from the closed frontier models, especially if one of those open-weight companies is now funded by Nvidia. So it gets rid of all the 'oh, they're all Chinese' comments, right? Jason is right. In a situation where a low-cost competitor with nearly equivalent functionality is attempting to enter your industry, you would far prefer to be number one than number two. Because number one can say, 'You got to just buy us because we're number one.' But you're right, Jason, number two has to say, 'Please buy us as well and don't buy the cheapo guys.' So I'm recounting my position and agreeing with Jason, right?

It would be bad enough to be number two in an industry you invented, right, which is where they are now. But to be number two with a whole bunch of ankle-biters on top that you have to deal with, that's a total pain in the ass, right? Which gets back to your question, Harry. I will answer your question on what price do they go out in 2027. I actually can answer it. They'll go out whatever price they get in '27 because they can't wait any longer. It's as simple as that, right? If Anthropic goes out this year at the scale they're talking about now, then I can't imagine a world where OpenAI says, 'We're going to hunker down and get cash flow,' right? And then we'll go out in 2028. They have to go out. So to some extent, again, it's a destiny outside your control, right? If Anthropic trades at two, then they might get one and a half. If Anthropic trades at one, they might get 700. But they'll do what they have to do, right? And I mean, there's a great quote in what was it? Come on, the book, the Sebastian Junger's book, The Perfect Storm, right? Super book. But when the guys finally realized they're going to die at sea, right, and it said something like, 'If danger can be described as the absence of choice,' they were now in danger, right? And I think for OpenAI, what you're seeing here is the absence of choice is starting to pile up. The other guys are ahead. The other guys are profitable. You know, correctly, the CFO says they're not going to write the only narrative. But to some extent, your narrative is getting written. You've got to get profitable because the other guys are profitable. You got to get public because the other guys are profitable. So they have less options than they did a year ago by a lot. And that will translate as in 2027 we're going to go public, and if the markets are slightly down this year, we'll take our medicine and keep moving. So that's what I think.

Super rider, kind of sad obviously, but wow, yeah, that line always stuck with me. Absence of choice, that's how you know you're in danger.

I think, listen, we can move on to the next topic. The other thing, when I'm listening to Rory on the IPO, the other thing, maybe it's minor, we can move on, but I'm getting confused today what OpenAI's differentiated mission is. Like, why OpenAI, right? I mean, we can all look at evals and we can read what Ramp says and what Rippling's report says, and we can view this as just an LLM, right? Paying top of market, right? Paying eight figures per engineer. But these were companies on very specific missions, right? When we started this pod, I don't know what OpenAI's mission is, right? I think Dario's is nutso a lot of the time, right? Apparently interviewing folks today asking them if they'd be happy joining Anthropic if it all went to zero, right? I mean, cool question actually, but nutso. I think Sam is a much more approachable CEO now. He got through all the Sam Altman crap and he's got the sweaters out and he's the more likable guy. But I don't know what is the differentiated mission of OpenAI today that I would rally around as an employee, as a shareholder, or is it just a piece of infrastructure plus some software? I just don't know what is special about the mission vis-a-vis Anthropic or now all these strong competitors. I just don't know. Or is it just an eval? Is it just three lines on an eval? I honestly don't know. And these have been very mission-based organizations from inception, right? The most we've ever seen in our lifetimes, right? These crazy missions.

For me, the astonishing thing is the consumer brand that ChatGPT has and the penetration it has in audiences that no other LLM has. To most of the actual general population in large majorities of the world, AI is ChatGPT. I am in awe that Sam is not going, 'We are the next Google. Our business is going to be advertising and we're going to see Jevons Paradox like never before when we have a consumer hardware device that actually partners with consumers and you see usage.'

But that was the plan. He just got outraised. That's it. He just got laughed at. It was a good plan. And Sora was in it and cool videos. It just was not the highest ROI for limited compute. It just wasn't the best use of it. Up until then it was all babbling stuff.

Yourselves, guys, but Jason nailed it. It turns out, I mean, you're right, OpenAI is the name and ChatGPT is the name everyone associates with AI, right? They have massive consumer market share and at some point intuitively say to yourself that turns into a pretty big business. We can circle back on how much, but Jason's right. It turns out, again, I repeat, it turns out that it wasn't the highest ROI use for compute. And I often, I believe this is sometimes when you look back at outcomes you realize, oh, only one sentence matters, right? And if you just internalize that sentence, you've been rich. The example I always use is if you've been on the board of Yahoo and for 10 years all you've done is screamed it's all about search, you could have made them a hundred billion dollars, right? Today what Anthropic is, it's all about code. That's the sentence. That's the only sentence that matters because to make it concrete what Jason said, coding is the fastest adopting market. It's the highest ROI market. It's the mother lode, right? And you know, it's as simple as OpenAI was focused here and Anthropic focused there. Does that...

And it could, but it could end up being even worse. We'll see in the financials, right? But the problem with the consumer businesses for OpenAI and Anthropic is they're, you know, for power users, they're massively subsidized, right? You can spend $200 on Anthropic or a hundred and some odd dollars on OpenAI and get 8 to 12,000 worth of tokens and that's fine for when ChatGPT was a proof of concept for a platform. Like we talk about OpenAI being a consumer company, but it's not where it started. ChatGPT was just a proof of concept app, right? And Claude was just the same, but it worked much better for Anthropic, right? Anthropic can lose, you know, a couple thousand dollars on some consumers and it won't impact them. But it's tougher for OpenAI. It's tough. It's a crappy business that consumers selling $10,000 worth of tokens for $200 is one of the worst business models of our lifetimes, right? If that was the only business, these guys would be dead in the water. It's a pretty bad business. And you say to yourself, I mean, look, on the other hand, I'm just going to argue Google is one of the best businesses on the planet because the cost to serve is low. And it may well be over the next five years that if you can continue to be the dominant consumer brand in AI as the cost to serve goes down, as you manage that cost to serve, as you build an advertising business, I could totally see a business, you know, plus or minus, you know, 50% to the same size as the Google consumer business, maybe over the next decade. So it's not like it's nothing, right? It's just that the S-curve in the adoption sense for coding was super high. The S-curve of adoption for consumers was super high for ChatGPT, but unfortunately the propensity to pay was almost zero, you know, relatively zero, right? Whereas on the coding side the propensity to pay is high. And you know, I'm just going to say it because I just saw the thing come true as we were talking here. You know, just to dump on the other side, I just saw on topic in the Wall Street Journal they believe their TAM is 30 trillion. And then I say to myself, 'Oh, I remember everyone when they're doing really well gets slightly delusional.' Your TAM is, you know, the entire US GDP. Thanks. Thanks a bunch, Dario. Good to know, right? And it's one of those overreaching statements that you get at this time of the year.

I mean, he needs it to be a pretty big number. He's got 2%. So...

Let's not do Hugging Face because I'm not smart enough to understand why anyone would pay 13 billion for it. I just don't get it. I'm just not smart enough.

I talk for precisely that reason. I agree.

I think it's kind of a muchness with, frankly, the pool side and the OpenRouter thing, which is everyone's... I mean, look, I think it all goes together. Everyone's looking at a world where, I mean, I might think it's delusional, but OpenAI and Anthropic are claiming TAMs that are larger than the entire US GDP. And if I'm running an IT company in the US, I'm saying, let me get this straight. They're claiming they're going to take everything. I better get me something, right? And therefore I want to be relevant in models that aren't closed-source frontier models. So you get in this whole trend for enterprise having their own models, you starting with open-weight models and then adopting, and Hugging Face is the place where you can access loads of those models. You know, revenue is relatively light at the moment relative to a $15 billion outcome. I think it's roughly 150 million. But if you think of strategic assets that an IT company might want to own if they were trying to build a counterbalance to the closed-weight frontier models, this would be a super interesting asset. Now, I don't know if it'll sell at that price. I can't make head nor tail. I'm with you, Jess. I can't make head and tail the price. But if you think about assets, you know, if you were Microsoft, IBM, you'd love to own to be relevant. This would be one. Absolutely one.

I don't get it. Other than it seems intuitively to me if right now, this is OpenRouter 2. Right now is the moment in time to benefit from the lift of open weights, right? The demand is so strong. So just like Eli Gil said, sell if you have an AI asset right now, right? I think even better if you have an AI product that's benefiting from the transition to open weights, there can't be a better time to sell than plus or minus 90 days from today. I agree.

It's just a phase transition and your numbers are going to look amazing. Like they said, OpenRouter was growing 15% at 150 million, right? The Information said when Stripe bought it and breathtaking if that 15% accelerates and scales, right? But it might not, right? This is the moment when all of a sudden open weights and these models went from experiments to mainstream. Sell, baby, right? If you can get north of 7 billion, 10 billion, I'd probably sell. I mean, even if I only got 15x, I'd probably sell.

No, it just...

It's a moment. It's not going to last. This moment in time, this transition is not going to be a transition anymore.

I agree. It may well be that. Remember, you have to... the founders of Hugging Face also have mission objectives beyond financial enrichment. So they may choose not to sell because they may have angst about that. But yes, from a valuation peak perspective, anything to do... I mean, it started with the Satya comments on every enterprise needs to have its own knowledge and not give it up to the frontier, to which you want to say, no, fuck you, Satya. Well, thanks for funding OpenAI for three years. But yes, everyone in IT has woken up and realized that these two frontier models could steal a lot of their TAM and everyone is saying we better have a different story. And the enterprises are saying it. Talent here is saying it. And you're right, Jason, if you are an enabling technology for open-weight models, now is peak moment.

On the Hugging Face thing, I don't think it's for what it's worth to Rory's point and the mission thing. I would say one small thing: if someone does buy Hugging Face for whatever reason, the deal has got to be you don't touch it because if you touch it, you break it.

Yes.

Right. If you promote, you know, it's a much bigger version of the TPN challenge. If it becomes an OpenAI commercial, TPN has no value. I know we're probably the only people that are going to compare TPN to Hugging Face, but if you mess with this marketplace for 10,000 models, you know, even if you put a little ad at the top for, you destroy it. So it look fun to get acquired, but I'm almost confident if anyone actually spends three billion, let alone 13 billion, they're going to 95% leave it alone for 24 to 36 months, right? They're going to...

Too obscure, too obscure, because as Jason correctly says, no one other than us is tracking that anymore.

You know my favorite news from the publics was Ken Griffin's Citadel unwinds 80% of Leopold Aschenbrenner's sold book. I'm like, man, never fight with Ken Griffin, man will come out on top. 80% sold already. Again, I don't think it's surprising, but...

No, it wasn't. I mean, they're not in the... I mean, yes, they're not in the business of holding those kind of assets long term. They're a market maker and a short-term trader, and this was a great short-term trade. Exactly. It's funny because about a month or two ago, he had kind of did something about starting to add to their kind of stable of investment options, you know, longer term, you know, multi-month holds. And clearly that was a conceptual idea, but it turns out when you buy a bunch of stuff at 10% below market and then that market jumps an extra 5% or 10% just because you've put the assets out of weak hands into strong hands, then the correct response is to take your money and run. No. I mean, exactly.

Look, it's great. It's just you can't do that every month.

No.

So to me, it's not that it's impressive. It's incredibly impressive. It's just not interesting because you just have to have the balance sheet and the cojones to wait so that every couple years, like Warren Buffett used to do, you could pounce on one of these special situations, right?

Yeah. Every few years someone gets confused about how leverage works in the public markets. They screw up and you're ready to price and buy and yes on top of your nice business which is still earning good money. Every three years some idiot gives you three or four billion of free money and you politely take it and put it into real estate in Miami. It's good to be Ken, right? Exactly right, Jason. That's how we read. There's nothing surprising in there. And you know, by the way, it does get to the... it's kind of... I'm going to circle back to Nvidia and all their investment and vendor financing because in both ways, both of situational awareness and Nvidia, the aha here is you when you're dealing with money and leverage, you don't just have to be right in the long term. You also have to be right every step along the way, right? If you don't have leverage, all you have to do is be right in the long term and hold, right? And it's probably a situation where it was right in the long term. But when you put four-to-one leverage on it, you have to be right every step along the way. And the same is true about vendor financing. You know, if you just sell people, as long as you're in the business of selling chips, all you have to do is be right in the end people want to buy chips. If you choose to lend against those chips, then you're basically saying you got to be right all along. The company's got to grow next year. They got to pay their debt back next year. So leverage does that. It raises the return from being right and raises the importance of being right all the time. And situational awareness just got the other side of that. Ken takes because remember he doesn't try in his business to be right all the time. He's not trying to make five-year bets. He's like stocks are worth 10% more today than yesterday. We should sell. Moving right along.

The other thing on the other side of the stocks for what area, I know this is Captain Obvious, but if we look at Kospi as sort of like AI on steroids, right, with risk or Korean exchange, it's still up 56.46% for the year. So I'm not a day trader. I pull up my Goldman and Morgan Stanley accounts and look how they're doing. I'm still feeling pretty like I'm some genius in my public market stock because overall plus 46 is pretty good. It's just boy, whatever Leo got just trapped in a dagger when I look at the chart, right?

All Korea.

Memory.

Rounding error now. Korea now equals, you know, two memory providers with a bunch of other stuff attached.

Right.

Yeah, so...

And it's a very volatile market, so it's on steroids, but it's still just like NASDAQ is tech on steroids. Kospi is like 46%, all the components of AI on steroids, right? And the peak was 9,000 in June and then boom, poor Leo, the dagger, 5,600 in July 29th and the guy had a generational loss that if it were even bigger might have brought down our financial ecosystem, but it's rebounded 20-some percent since then, is up 56% of the year. Cry me a river if it's up 56.46% a year. I mean, you know, you got to be a day trader or whatever to not love being up 56% a year. It's okay. But all these headlines are like, 'Oh, Kospi is down 6% today.' And it's just hypervolatile, right? The growth and margins, we've never seen margins like this in semiconductors. So the volatility and expectations...