Gavin Baker

So that is like saying that people get younger every year because there's a depreciation on this stuff.

So when Gavin Baker, and I'll be covering this, like when Gavin Baker goes in there and says, do you realize that everyone that owns these chips running basically a data center The compute prices today that they can sell at those things is higher than they were a year ago.

And these chips are costing more.

So if you want.

And they're able to market these at higher prices because all the other restaurants right now are charging higher.

So.

If people don't understand what that means, and Gavin Baker talked about it, he believes that the cloud providers, so the guys that are spending all the money and taking all the debt out, he said on Patrick O'Shaughnessy's podcast a month ago, and that was a month ago, H-100s are higher than even when they were back then.

that you'll be shocked at how quickly the cloud providers will start to have profits coming through, not just from new data centers coming out, but from the fact that the compute prices that they can charge are higher and their costs are actually lower on the existing stuff they have.

OpenAI and Anthropic, how much revenue are leading companies like this going to be able to generate? Anthropic, as part of their IPO process, has pointed towards exiting 2020 at about $200 billion in what you call annualized revenue, ARR.

Some of their VCs have said that's closer to 250 billion and some third party VCs like a Gavin Baker, who I've talked to on this podcast, have claimed the number could be 400 to 500 billion.

Where these companies hit on their lofty targets will say a lot about how big this market can go.

Because again, if an open AI and an anthropic are hitting numbers beyond 250 billion in terms of their revenue by the end of next year, they have essentially become companies the size of Microsoft.

Like, just wanted to make clear also on technical side of things, you can directly connect with Nvidia GPUs even today through standard, you know, ethernet and, and InfiniBand connections, right? So, so this is not like this doesn't stop you from doing it, but obviously you wanna get closer.

And then, like, look, you mentioned the people on our, on our board, right? You know, Gavin is on our board.

He, he's very, very clear-

Gavin Baker at 80s.

Yeah.

Correct.

Gavin Baker at 80s.

He's very clear about, you know, like, look, we have to...

And what I find pretty interesting too, is this comment, uh, that fireside alpha put out.

And it's, uh, this interview with Gavin Baker.

And he says, you know, this is kind of how I look at it too.

He goes, uh, It's impossible for, you know, all these people on X are saying that space-based data centers are going to fail and are stupid.

That's just the warm-up act.

You made reference in the last few days to the All In podcast, Gavin Baker on the All In podcast.

There is a body of opinion that he didn't actually express, because I did listen to it back, that he was recounting, and I've got no dog in this fight, but I don't think he is an anthropic shareholder.

In fact, I think he's an Elon guy, and he said he was, but he doesn't.

It's massive.

So based on Gavin Baker's recent analytical breakdowns, some very intentional language from Ilya himself and that monumental NVIDIA investment, we're going to look at the mechanics of building an AI that mimics the, quote, overlooked aspects of how the human brain functions.

Yeah, that that phrase overlooked aspects of the human brain.

That is the crucial piece here, because for years, the entire industry has basically been sprinting in one direction.

One is... uh, by, uh, Danny Penny, uh, who is, uh, and ex military guy who was recruited, uh, to, uh, come into a 16 Z and, uh, to, to work on their American dynamism fund.

And the other ones by, uh, Gavin Baker, who is the, let me make sure that I get this right.

I think Gavin is the managing partner and CIO of a treaties management, uh, uh, I think it's right to call it treaties, a hedge fund.

And both of them took a positive view on data centers.

So this idea that your data is yours, you want to have a company like Snowflake or Palantir sitting between you and the model provider so they're not training on your data, this is something Alex Karp has talked about.

Is the growth that we're seeing here in Snowflake up over 20% today, Palantir up 7%, the result of that thesis? Is it people rotating from hardware to software as we've talked about? Because one of the problems that I have is, hey, I'm glad any position in the portfolio is doing well, but like Gavin Baker talked about a while ago, How can it be mutually true that Snowflake and Palantir are well-positioned for this multi-year run of token management and agentic operations, and Broadcom and Credo are at this cyclical build high? They can't both mutually be true, right? Because one feeds the other.

So...

What's going on with Snowflake's earnings? Is this data sovereignty, the Alex Karp argument? Is it just a really, really good print that nobody expected? Is it the rotation from hardware into software? Some combination of all of them.