Jordy Vandeput (1:29)

So I think, I think, you know, everyone freaks out. Oh my God. Putting stuff in space is expensive.

Lex Fridman (1:33)

Yeah.

Jordy Vandeput (1:33)

But if you look at like starship launch cost and you know, they keep falling, you're like fine. Right? Like I think that's not, you know, by the end of the decade, the cost of space launch will be fine. The heat dissipation, I mean, it's a challenge. But you just put a massive, massive. Effectively radiator.

Lex Fridman (1:47)

Yeah.

Jordy Vandeput (1:47)

And it's, it's fine, right. By the end of the decade, like you'll be good. I think, I think the big challenge is that chips are just really unreliable. Right. And so, so how do you deal with like a couple of things? Right? Satellites can only be so large before they like you start needing a lot of support and structure before they tear themselves apart. So when you look at like the, the launches. Right. These, these things are shooting out satellite, tiny satellites and many of them. Okay. So you can't have like a big fully connected cluster of chips and then like on top of that. Right. How do, how do you deal with any random error on, on earth you have text running around the data center unplugging stuff, putting in spares, things like that. What do you do in space? You RMA it to the factory where they like might unsolder it and re solder it and then like test it and it works and go back out. Sometimes it is just trashed. But like you know that's the challenge.

Lex Fridman (2:36)

To me is that, is that I feel like maybe the pattern we should be looking at is like how often do the Tesla self driving chips need to get serviced? Because that's like the team that would probably be building or like bridging the gap there. Like the Starlink satellites certainly go down but like the service works like you're just relying on some sort of like you know, 90% uptime stuff's coming down. But most people that are in a waymo like the chip keeps working. Right. Most people that are in a Tesla self driving like they're not. Like you don't hear about Tesla owners being like I love FSD but I'm constantly in the shop getting my, my custom silicon chips unseated and receded. Right.

Jordy Vandeput (3:15)

Well I mean it's also a function of like the complexity of a chip. Right? Sure. You know if, if a chip is twice as fast. Yeah. And let's say the bit error rate. Right. Like how often a bit flips. Sure is the same. Then it's erroring out twice as often.

Lex Fridman (3:28)

Yeah.

Jordy Vandeput (3:28)

But let's say the chip is 10x as big. Right. And so when you look at like a Tesla FSD chip, very, very good, very, very efficient. Very still like relatively inexpensive and cheap compared to you know, a big old GPU or TPU or whatever. Right. Those things are extremely large. And you know again like if, if the error rates are the same then it fails 10x more. But in fact the error rates are a bit harder, higher because they're pushing these things to the absolute limit. Whereas Tesla does have some level of. Well first of all the Tesla car has two chips sort of redundancy already built in. Right.

Lex Fridman (4:01)

Maybe you do that on the satellite, but then there's more power. More.

Jordy Vandeput (4:04)

Yeah. Right. So the whole allure. Right. Of it is effectively power is free. Right. And solar panels, you look at the cost curve of solar panels, you look at the cost curve of satellite launches, you're like, this is free, this is great. But power is less than 10% of the cost of the cluster.

Lex Fridman (4:19)

Sure.

Jordy Vandeput (4:20)

Right, yeah. So. So, like, it's that 90% you're not saving anything on.

Lex Fridman (4:24)

Yeah, yeah.

Dwarkesh Patel (4:25)

And insofar as much as potentially 100 times the hassle.

Jordy Vandeput (4:29)

Yes, yes.

Lex Fridman (4:31)

Yeah.

Jordy Vandeput (4:31)

There's this whole, like, you know, like, if you look at Nvidia GPUs, right. When you first turn on the cluster, about 10 to 15% of them fail RMA in the first two weeks. Wow. And then that's fine. Like, you have to reseat them, whatever. And like, the industry knows how to deal with this. Right. And over time, like, hopper's now at 5%, but Blackwell's still 10 to 15%. Right. Actually started out higher than that.

Lex Fridman (4:53)

Sure.

Jordy Vandeput (4:53)

And when a new generation comes out, it's gonna be higher than 10, 15%. It'll have its curve gradually decline down. But, you know, who, who's gonna. Are you gonna test it and burn it in on the ground, or are you gonna say 5% of my chips or 10, 15% of my chips are trashed because someone can't go up there and, like, do these things? Or am I saying, oh, I need robots who can do all this stuff in space? And now that's like an additional engineering problem when sacks of meat are actually very cheap.

Lex Fridman (5:15)

Yeah. Yeah. Speaking of Nvidia, we haven't talked since the Groq acquisition. What does that look like in the bull case? Like, if it's, if it's a good. If the next version of Groq is a great chip, is it sitting next to the, you know, H200, H1 hundreds in the, in the rack GB200? Like, how does it fit into the actual, like, what Nvidia deploys? Is it just a separate chip?

Jordy Vandeput (5:38)

I think it's. I think it's a big vibe shift from Nvidia. Right before they were like, all right, I got this big gpu. Everyone's going to use this gpu. Software ecosystem of the GPU is so good. It's one size fits all. Everyone's trying to make all these specific point solutions, but we've got the thing that's good at everything. Then they had a vibe shift. They launched this thing called cpx, which is a chip made for pre fill prompt processing, creating a KV cache, and also good at video generation and image generation. And that's coming out later this year's release.

Lex Fridman (6:08)

They were really talking about video generation.

Jordy Vandeput (6:10)

Well, so, yeah, You've got like cpx, you've got like the standard GPU now, you've got the Groq chips and they all fill a different niche, but really it screams, oh, crap. We don't really know exactly where AI is going, which I don't think anyone does. Right. I mean, it's moving so fast. The software is the model architectures, et cetera. So we're just going to, like, engineer solutions that are along multiple points of the Pareto optimal curve and then one of them will win. Right. And I think it's like sort of like a big vibe shift from Nvidia. Also, they just knew OpenAI was going to do the Cerebrius deal, so they.

Lex Fridman (6:38)

Freaked out, but got it. Yeah, yeah. Get me up to speed on what makes Cerebras important in the ecosystem right now.

Jordy Vandeput (6:46)

So. So, you know, you have people thinking like, oh, latency matters in terms of where a data center is. It doesn't matter at all. What matters is, you know, as we've moved from, you know, chat applications which were like. Or search response immediately chat applications, let's Say response takes 10, 20, 30 seconds. You've got agents, you know, I don't know. My cloud codes are working in the background for a long time. Right. Doesn't matter where the data center is. But what does matter is that these streams of inference take 30 minutes versus 10 minutes versus 5 minutes. And for a lot of people, I'm fine to spend 10x the price on something that completes 10x faster. And so Cerebras sort of just makes a ton of sense there. So OpenAI, they've got these long horizon. There's like Codex 5.2 extra high thinking or whatever. It's terrible. Can you guys teach them how to market? You have to sponsor this podcast.

Lex Fridman (7:38)

Yeah, yeah, yesterday. And I did actually ask him, like, I had the Codex app pulled up on my desktop and I was like, there are six different models and then there's another button that I can pick.

Dwarkesh Patel (7:50)

Well, how many different products are called Codex now?

Lex Fridman (7:52)

There's a lot.

Jordy Vandeput (7:53)

Now there's an app. Yeah, yeah, we actually have another guy.

Lex Fridman (7:57)

On just to do branding. Lexicon Branding came on the show yesterday talking about the. All the naming architectures. Naming architectures. It is complicated, but hopefully you could.

Dwarkesh Patel (8:05)

Tell he's just blood's boiling because, like, all the AI companies just have the most chaotic anthropic Claude. Claude code. But also you can use Claude code for other stuff.

Lex Fridman (8:16)

Yeah, but yeah, I mean, with Cerebras, it seems like there is a value to it. But are they constrained on the supply side? Like can they actually scale up to you know, a colossus style data center that could actually speed up Codex not just for like one user, but all.

Jordy Vandeput (8:32)

The users reverse can speed up multiple users for sure. The question is sort of like where you use it and that's where they have to like figure out where within Codex, right? Yeah. Because there are times where Codex is running for like 10 hours and sometimes you don't mind. Right. Like, screw it. I've put up this nice prompt, gone work on it, refactor my code, do this thing, do this task. Other times I want this iteration feedback loop. So, so how do you expose it to the user without saying, hey, actually there's another toggle. So your permutation is 18 times.

Lex Fridman (9:00)

Hopefully like a really robust model router. But it feels like that's been a process.

Jordy Vandeput (9:05)

Yeah. So the OpenAI deal is like for 750 megawatts. It's not that much capacity on the order of like what OpenAI has talked about. By the end of 28 they'll be at like 16 gigawatts of that cerebral.

Lex Fridman (9:17)

Is like the absolute cutting edge. The most price insensitive customers in that specific use case of this is the type of prompt that you need to return fast. Then you'll get the speed up potentially.

Jordy Vandeput (9:26)

Right, Right. And they've got to figure out how to do it from a product exposing to the user, et cetera. But it's, it's clearly something where there is demand, right? Like, I don't know, like Andrej Karpathy doesn't care if he's spending a thousand bucks per agent per second or whatever. Right. Like, you know, whoever it is, this, these like super cracked engineers don't care at all. And then obviously there's like a long tail of like actually cost does matter for most people. And so all along that curve they've got to have solutions. Right.

Dwarkesh Patel (9:52)

When did you first think that XAI might end up at another Elon company?

Jordy Vandeput (9:59)

I mean this has been rumored for a long time, right? Like people are saying Tesla, Tesla, Tesla. For the longest time.

Lex Fridman (10:04)

It's harder with a public company.

Jordy Vandeput (10:05)

Yeah, yeah. And then a few a bit ago, people were like, oh, SpaceXi. I'm like, wait, this makes no sense.

Dwarkesh Patel (10:11)

Very coordinated narrative podcast hump.

Jordy Vandeput (10:14)

And then the space data center at.

Dwarkesh Patel (10:16)

The end of last year, it was almost perfectly telegraphed.

Jordy Vandeput (10:20)

Well, there's a bet right between basically the head of compute of XAI and the head of compute of anthropic and the bet is what percentage of worldwide data center capacity is in space by the end of 28? And the bar is 1%. Oh, wow. And so the Xai guy is like, really bullish and the anthropic guy's like, yeah, yeah. But it's a really interesting bet. I take the under on 1% by 28, because that's a gigawatt in space. But it's actually not that crazy. Right. It's roughly 150 Starship launches. We'll get them to 100 to get them to a gigawatt in space. So, you know, starship hasn't worked yet fully.

Lex Fridman (10:58)

I was looking at the energy draw of the current starlink fleet, and I think they're at like, what is it, 200 kilowatts or something like that. So you get a thousand of those 200 megawatts and like, you're starting to be in the territory something.

Jordy Vandeput (11:14)

Yeah. So the V2 Starship satellites, I think, are the only ones they've launched. Maybe they've launched a few V3s, but the V3s are coming soon. And those are. Those are like 100x more bandwidth each, right? Yeah.

Lex Fridman (11:23)

And more power and just more power. And so when I'm just thinking of like, can you scale this thing up at all? It's like, are they two orders of magnitude off? Are they three orders magnitude? It feels like they're like one order of magnitude off being something that looks like an H100.

Jordy Vandeput (11:35)

I think the metric is like 50. It's either 50 kilowatts a ton or something like this per satellite for V3.

Lex Fridman (11:41)

Yeah.

Jordy Vandeput (11:42)

If, let's say from V3 to whatever the compute, they double it again, get to 100. I think the V2s are like 25.

Lex Fridman (11:48)

Yeah.

Jordy Vandeput (11:48)

So if you get to 100 kilowatts per ton for launch, it's only 150 or so Starship launches.

Lex Fridman (11:53)

Yeah.

Jordy Vandeput (11:54)

I think that's so reasonable. Maybe not 28, maybe it takes 29, but like, you know, it's so reasonable. The question is cost and reliability and, you know, what happens when the chip fails, how do you service it, that kind of stuff. How do you. How do you deal with having clusters be much smaller instead of like, you know, these big clusters? Even for inference, big clusters are useful.

Lex Fridman (12:12)

Yeah. Yeah. How do you think about Google's response to Grox or Eris TPUs? Obviously very successful. But are they forking that project to eat more of the Pareto curve?

Jordy Vandeput (12:26)

Yeah. So for the longest time, Google's had One main line of TPUs, right. All made by Broadcom. And then sort of next year they've diverged it. Right. Where Broadcom makes a TPU and MediaTek makes a TPU, these two TPUs are focused at different things and they're fabbed at different. They're both fab tsm. Everything at the end of the day goes to Arrakis. Right.

Lex Fridman (12:45)

I want to go there next.

Jordy Vandeput (12:46)

But so FABTA simply regardless. But both of these TPUs are focused on different things and they've actually got a third project for another kind of tpu. They also see this need to proliferate along the curve of like, hey, do I care a lot about super high amounts of flops, not that much memory? Do I care a lot about super fast on chip memory only do I care about 3D stacking memory? Do I care about, you know, this sort of general purpose middle ground AI chip which is what, you know, an H100, a Blackwell, a TPU looks like today? You know, they're sort of like, oh, we need to hit the entire Pareto optimal curve. And it's like, okay, within this there's training versus inference differences and like what numerics you want and all these other things. There's so much complexity there. Everyone, everyone sort of is diverging their roadmaps once they're at a sufficient scale, I think.

Lex Fridman (13:34)

Yeah. Or is Google still way ahead on cross data center training?

Jordy Vandeput (13:38)

Yes.

Lex Fridman (13:38)

Are the other labs, like, is that important to the other labs to catch up there or is it something that will just naturally happen because everything sort of commoditizes? Or do the other labs need to sort of marshal some Herculean effort to like crack the code on what it takes and what Google's doing?

Jordy Vandeput (13:55)

Yeah. So it's a couple of things. Right. In 2023, everyone thought that scaling was pre training. Right. You know, more parameters, more data. And that's very difficult to split across data centers.

Lex Fridman (14:05)

And has Google been able to do that?

Jordy Vandeput (14:07)

And Google's been able to do that to an extent. Right. So what they've done is they've got, you know, they don't have the largest individual data center campus, but what they do is they do these like regions where it's like, hey, each data center is roughly 40 miles apart from each other.

Lex Fridman (14:20)

Sure.

Jordy Vandeput (14:20)

So in Nebraska and Iowa and then in Ohio they've got like these complexes and now they're building one in Oklahoma, Texas, you know, these complexes where there's all these data centers pretty close to each other.

Lex Fridman (14:29)

So it's not really across data center, across, across the world.

Jordy Vandeput (14:32)

Right.

Lex Fridman (14:32)

Just cross like region.

Jordy Vandeput (14:33)

Yeah. And then that, that makes a lot of the difficulties a lot easier. Flip side is we've also moved to rl. Right. And majority of the time of the chips is spent generating data. Right. Only doing forward passes through the model. And then you only send the final tokens that you verified sort of back to train, on to the training, to the train, to train. Right. So then you end up with like, oh, instead of in pre training scaling, you need to like synchronize all the weights every 10, 20, whatever seconds when you're doing these rollouts. And especially as things get more and more agentic in training, you might not only need to send not the entire weights, but just the tokens that are relevant. So way smaller amount of data and way less frequently. Right. Minutes at a time instead of seconds at a time.

Lex Fridman (15:14)

Yeah.

Jordy Vandeput (15:14)

And so you've got this like now, now it's become like reasonable where. Oh, actually multi data center training is completely reasonable. And people do this. People do multi data center, multi chip training.

Lex Fridman (15:23)

Sure.

Jordy Vandeput (15:24)

Right. You know, you do your inference on one set of chips and you do your training on another set of chips. So like anthropic does this. I don't know if Google does this, but Google's kind of already got the cards. Yeah.

Lex Fridman (15:35)

Okay, got it.

Dwarkesh Patel (15:36)

Let's go to talk about Arrakis.

Lex Fridman (15:39)

Just there's this debate TSMC risk. Is that the bottleneck or is energy the bottleneck? I was doing back of the envelope calculations. Seems like we're using maybe like 1% of global energy production or Western energy production on, on AI, specifically workloads. And then we're using like 50% of leading edge fab capacity on AI workloads. And so that feels like. Okay, well even if we all agree and we say as a society we're going all in on AI, we can only double the AI chip capacity before we need to build more fabs. That takes years. Whereas we could say, everyone turn off your air conditioning. We're sending the electricity to the data centers.

Jordy Vandeput (16:16)

Right.

Lex Fridman (16:17)

Like we have the ability to generate without creating new.

Jordy Vandeput (16:24)

Turning off the ac, turn off your to eat heat strokes for all the grandmothers. I need my cat dancing videos to feed cloth.

Lex Fridman (16:33)

Right, but, but, but, but seriously, like there's this debate over, you know, is TSMC the main bottleneck or energy the bottleneck? How are you feeling about that?

Jordy Vandeput (16:40)

Yeah, yeah. So, so sidebar before I answer the question, because I think it's fun. Yes. You know, in the US it's insane to say turn off your AC for AI.

Lex Fridman (16:47)

Yes.

Jordy Vandeput (16:47)

Right. And the general public hates AI already. But in Taiwan, they've had droughts before and they've turned off water to entire cities. They're like, oh, you get to, you get water three days of the week. Whoa. And then the FAB still gets supplied water. It's like this is, you know, you've got to understand the mindset. We are not ready as weak Americans to do this. No, but at the end of the day, right, like water, water. Water and power are certainly less big of constraints now. Now you've got to imagine like, you know, semiconductor industry is used to, hey, doubling the amount of transistors made every year or two. Part of that is more slow, part of that is more capacity. Whereas the energy industry in America wasn't. And so like initially people were like, not creative. They're like, let's do, let's do these kinds of gas plants. It's like, well, no, now we've realized, you know, yes, there's three main manufacturers of turbines and then you've got for a dual combine cycle, then you've got like igts, but you've also got like medium speed reciprocating engines, right? Like, turns out Cummins can make like a million diesel engines a year and like those can make electricity like if I don't give a fuck and I put it in West Texas, easy. So now it's more of like a regulation thing, a supply chain thing. Power is not a constraint in insofar like that much, right? I think it certainly is a constraint still. Today it was the biggest constraint in 24, 25 data center capacity, power. Because the industry was not ready. People have woken up, they've sort of been shocked to the system. Now you've got tens of gigawatts being deployed. Next year, 30 gigawatts are being added and we think the power is there for it.

Lex Fridman (18:14)

What was this year?

Jordy Vandeput (18:17)

This year is, I think it's like 18ish. 10ish. Okay, 15 to 18ish.

Lex Fridman (18:21)

Sorry. So almost a doubling?

Jordy Vandeput (18:22)

Yeah, almost a doubling, yeah.

Lex Fridman (18:23)

Wow.

Jordy Vandeput (18:24)

And when you look at, when you look at TSMC and the crew, right there is not really, oh, this random, you know, there's 12 people making medium speed reciprocating engines that you can now convert to make power at some random data center. No, no, no, there's like, there is Arrakis, right? There is one set of spice, like, you know, there's, you know, that's it. Right. And so, and Then the flip side is like, okay, when you have 12 vendors, everyone's got a little bit of slack capacity, you know, there's more likelihood. You know, you can. People are like, oh, turbines you can't get. You can call a broker and you can get a turbine. You might be paying 50% more, 2x more, but you can get a turbine. Yeah, right. Like it's.

Lex Fridman (19:00)

You can't get a 3 nanometer fab.

Jordy Vandeput (19:01)

You cannot get a 3 nanometer fab. Exactly. And so when you talk about what's the, you know, the. The baton got passed from semiconductor shortages in 23 to power and data centers in 24, 25, 26, we're swinging the pendulum, but it will fully be semiconductors again in 27. Right. And so we see this across the entire space of the ecosystem. It's not just tsmc, it's also memory Both, because both of them have built at a certain pace. Now, TSMC has been expanding at some rate. The memory makers, in fact, have just not expanded capacity. Basically, they've not built new fabs since 2022 because their cycle is so undulating. Yeah. And so when you look at it, it's like, oh, even if they wanted to double capacity, they need to build the fabs. Right. And building the fabs, it is the most complex building humans make. Right. It's the entire air of a clean room. Circulates itself every 1.5 seconds. What? And you don't even feel it when you're inside, really. It's like that. And it's like parts per billion of particles. Right. Like, it's actually insane how you could get coughed in the face by someone who has Covid and not get Covid.

Lex Fridman (20:07)

And so it gets circulated so fast, it doesn't even hit you.

Jordy Vandeput (20:09)

It's like, it's like that meme of.

Lex Fridman (20:10)

Like, the spraying when someone's talking, and then it's just circulating.

Jordy Vandeput (20:14)

So another. Another sidebar is everyone knows Covid, like, really popped off in Wuhan. Yeah. Right. Wuhan also is home to China's largest memory company, ymtc. And so when they were like, welding people into their homes, the people who worked in the fab still went to work.

Dwarkesh Patel (20:30)

Wow.

Jordy Vandeput (20:31)

It was because it's, you know, one. It's a national, like, it's national importance. But to, like, these people are getting sick. This fab is, like, way too clean. Yeah.

Lex Fridman (20:40)

Sorry, Jordy.

Dwarkesh Patel (20:41)

I want to talk about Oracle. They put out a post this morning that said our partners financing for the Dona Ana County, New Mexico, Shackelford County, Texas and Port Washington, Wisconsin. Data centers are secured at market standard rates, progressing through final syndication on schedule and consistent with investment grade deals. Obviously they were fast following their posts from yesterday where they said the Nvidia opening ideal has zero impact on our financial relationship with opening I. We remain highly confident in OpenAI's ability to raise funds and meet its commitments. And obviously everyone was looking at this being like, give me a cigarette smoking, like it's like bank run language. I haven't seen posts like this since, like the ftx.

Lex Fridman (21:19)

Is it just bad comms or is there something worse?

Jordy Vandeput (21:22)

It's terrible comms. Yeah, like, like I told my Oracle contacts, like, who the hell is in charge of the Twitter? Like, what are you doing? Nvidia did something similar last year when the whole TPU mania was going on.

Dwarkesh Patel (21:34)

It was, yeah, it was like we're thrilled with Google's progress with the tpu that said Nvidia chips are the only. You know, it's like no one asked you to comment. I mean like, I'm sure a handful of people in your DMs and random, but that doesn't mean, it doesn't project confidence.

Jordy Vandeput (21:52)

It's sort of the lion shouldn't concern themselves with the sheep. And like, okay, Nvidia's line, maybe, maybe Oracle is a little bit more bumpy, but I think Oracle is like fine. People are just freaking out because, you know, OpenAI is peak. You know, people are peak negative on OpenAI right now because of how good anthropic's been killing it. Yeah, I think it's just like kind of silly, like they need to hire someone to do comms, like a lulu or something. Right. Both Nvidia and Oracle, because what are you doing?

Dwarkesh Patel (22:19)

How did you process yesterday in general? Jensen was clip farming. He was like, I don't know why he does these street interviews. Right.

Lex Fridman (22:28)

No other CEO does those where they just stick 25 microphones in your face and the paparazzi flashing. It's a great vibe.

Jordy Vandeput (22:35)

It's, it's, you know, Jensen's not been as famous as other CEOs for as long and yet he's so important now. And if you've, like, if you know of Jensen, how he is in meetings, there's, I feel like there's two Jensen, right there is like pr, like good at pr, just good at talking, good at like making people hyped up and believe what he's doing.

Lex Fridman (22:56)

He's great at standing on stage, holding up the chip.

Jordy Vandeput (22:59)

And then there's the real Jensen, which is Like a business killer.

Lex Fridman (23:03)

Yeah.

Jordy Vandeput (23:03)

And like actually just knows about every like aspect of the supply chain. Right. All the way from like niche semiconductor, you know, design and manufacturing stuff, all the way to like energy power, data center, like and then doing the business deals too. Right. And so like you've got this whole Pareto like of a whole thing, whole range of things that he's good at. And he's a killer in. And clearly he's like he was in a meeting where he was being a killer and like negotiating like supply contracts or something.

Lex Fridman (23:30)

He walks out and then he walks. I mean that's hilarious. Yeah. Theory. But I like it. Yeah, that's awesome.

Jordy Vandeput (23:36)

And that's why he was like, you know, like, he was like still killer. Like, no, we never said we committed to 100 billion. You know, like, and it's like, I.

Dwarkesh Patel (23:42)

Don'T know, where do you even get the hundred billion dollar number from? It's like, well, you did go on. People would assume that it was. But, but they did say in the press release. I remember these are early talks.

Lex Fridman (23:57)

Yeah.

Dwarkesh Patel (23:57)

But they just kind of jumped the gun. This was the height of the press release economy.

Jordy Vandeput (24:01)

Yeah. Yeah. What's funny is Oracle stock peaked like just like a week after they announced OpenAI deal. And so like the press release of like, hey, OpenAI is going to do this humongous deal. Stock peaks. Same happened with a couple other vendors who announced deals with OpenAI or Nvidia. Like sort of a lot of these, like, like they all peaked to that. And then it's sort of been like Nvidia OpenAI trade has been going poorly and sort of like the TPU anthropic Google Amazon complex has been doing well. It's quite interesting.

Dwarkesh Patel (24:30)

Good, good energy back at home with the roommates.

Jordy Vandeput (24:35)

What's going on?

Dwarkesh Patel (24:39)

Yeah, I got one more thing. So yes, over the weekend it was sort of drowned out by all the justice department stuff, but have you guys.

Jordy Vandeput (24:48)

Talked about Elon saying you can smoke a cigar in the fab? No. Yeah, yeah, yeah. Okay. I was gonna say that's this is, this is, this is part of the whole thing.

Lex Fridman (24:55)

I didn't realize that was related yet.

Dwarkesh Patel (24:56)

I mean, yeah, indoor heaters. We have indoor heater technology. No one's taking advantage.

Lex Fridman (25:01)

Yeah. What does the fab look like if you have no humans inside? Like, that's probably his long term thing is like, yeah, there will be an optimist.

Jordy Vandeput (25:07)

But no one knows. Like the number of people who work in a fab is like irrelevant.

Lex Fridman (25:11)

Like, but, but is it is it irrelevant because there's all these things you have to do when a human's in there because they sweat and they breathe. And if you don't have to do that because it's a robot walking down, even if it's puppeteered or teleoperated, you. You might be able to have different considerations. I don't know if that actually affects.

Jordy Vandeput (25:26)

Well, it's like a nesting of, like. It's a nesting of, like, cleanliness. Right. For example, you've got this wafer. You've put like, down. Let's say you put down copper.

Dwarkesh Patel (25:32)

Yeah.

Jordy Vandeput (25:32)

And now you're moving it from one area to another. Well, it needs to be stored in a vacuum. But easiest way to store a vacuum, like or, or an inert gas.

Lex Fridman (25:38)

Yeah.

Jordy Vandeput (25:39)

And that's like the thing that's being transported in. But then around that, you want it to be super clean as well. You, you, you. If you don't, then the copper starts getting oxidized, it affects our yields. All this sort of stuff happens. And so, like, you kind of want it to be a nested layer. Nested layer of like, well, this chain, this thing inside the EV tool super clean. And then the thing feeding it is super clean. And then the thing it sits in is super clean. Because that's how you get to like, there's zero particles. Because, like, you know, in the, in the, in the foup and the transportation devices, like, parts per, you know, trillion and maybe foup. It's called F O u p front operated, front opening. I don't know, something pod.

Lex Fridman (26:12)

Pod, but it's called a foup.

Jordy Vandeput (26:13)

It's like the thing that moves and it carries the wafer. Sure, sure, sure. And then, and then the fab is like parts per billion. And, you know, sort of like, you know, you've got to like, got this nesting relationship. So everything is super clean. You know, I'm bullish on robots. Like, super bullish on robots, but only for, like, not for tasks that have like TSMC Arizona Fabs or, okay, let's say TSMC Tainan, which I think produces like, you know, indirectly hundreds of billions of dollars of global gdp. Even directly. It's like still tens of billions of dollars has like 5, 10,000 people in it. Like, it's like, irrelevant in terms of the number of people who work there.

Dwarkesh Patel (26:48)

In terms of the overall economic value that's created.

Jordy Vandeput (26:50)

Right. It's like, it's like, it's like how many people fold laundry or how many people wash dishes or how many people, like, do construction work. Like, these are way bigger markets for robotics.

Lex Fridman (26:58)

Yeah. Speaking of China, what are you making of the Dario essay? Or I guess his comments at Davos about selling chips to China is equivalent to, you know, nuclear weapons these days. The Ben Thompson line was something like, he's, he's okay selling chips because he wants dependency on the Nvidia ecosystem Cuda, but he would ban lithography tools from going to China. And I'm always, I've been wrestling with this idea of like, I don't know if China would accept this, but wouldn't there be a different world where you want them dependent on American LLM APIs and you don't even send them the chips? And you say, yeah, you're, you can have as much AI as you want as long as you're paying, you know, OpenAI and anthropic API.

Jordy Vandeput (27:43)

Yeah, I think it's, I think it's like a curve of, like, what they will accept. It's, it's, it's, you know, one, you, you, you push someone to the corner, they're going to start swinging. Right. And I'm like, very concerned that China does this. Right. Do they, do you, do you push them too far into the corner? Do they say, screw this, we're going to start being a lot more aggressive. We're gonna, we're gonna, you know, do more military actions or.

Lex Fridman (28:04)

Military actions or, or even just invest twice as well much in, in global supply chains.

Jordy Vandeput (28:09)

Take over Africa more than they already have, like Latam, like, etc. There's, there's, or, or just take over Taiwan chips. Yeah, right. Because if I can't have the chips, what value is there in Taiwan existing?

Lex Fridman (28:17)

Sure, sure.

Jordy Vandeput (28:18)

In its current state. Right. So there's like, there's this like, game theory aspect at the same time. You don't want China to be able to, like, you know, if you believe AI is going to be, do what I think many, at least in San Francisco think it's going to do, which is like, completely revolutionize humanity and cause GDP growth to accelerate? Do you want to have China also own that technology and, and all, you know, their ability to integrate that into their military and all these other things much faster, you know, so there, there is like these competing, like, you know, interests.

Lex Fridman (28:46)

Yeah.

Jordy Vandeput (28:47)

Where, where is the, like, right line? And some people think it's like, hey, yeah, sell them AI model. Well, I think Dario would say, don't even sell them AI model access.

Lex Fridman (28:54)

Don't even sell them tokens.

Jordy Vandeput (28:55)

Yeah, I think so. I think, like, I think anthropic does not sell AI access to China. Yeah, they get, they loop it through and you can see, see this in the traffic data. They go through Korea and Japan, other places, but like. And so they get it.

Lex Fridman (29:04)

Yeah.

Jordy Vandeput (29:05)

And then the other, other side is like, sort of like, I think like the Ben Thompson view, which is like. And I think I'm more sympathetic to that, although I think I'm not exactly aligned with that, which is like. And we've been saying like, don't sell them equipment, don't sell them equipment, don't sell them equipment. And my, my argument is like more economic in the sense of like, if you sell them like tens of billions of dollars equipment, they can make hundreds of billions of dollars in AI value or chips with that equipment. Whereas if you sell them AI model access and it costs them this much to get the economic, you know, they're not able to.

Lex Fridman (29:32)

You're capturing more of the value.

Jordy Vandeput (29:33)

Exactly. And so that's sort of the question that is at foot here. Right. Do you want them to capture all this value of the supply chain in equipment or by buying the chips or using the models. Right. And services. And we've seen, you know, across many, you know, stacks. China refuses to accept, you know, using American ecosystem. And they'll wait many years before they develop their own. Whether it was like, hey, they didn't use Windows, they figured out a bootlegging economy, or they didn't use Visa and eventually they came out with like Alipay and WeChat Pay or whatever it's called on. And like, these things are way better than Visa in fact. Right. Lower transaction cost and higher volume.

Lex Fridman (30:11)

I use Red Star Linux. It's North Korea's Linux distribution.

Jordy Vandeput (30:14)

Wait, really? Yeah.

Lex Fridman (30:15)

If you, if you don't, if you put it on a network, it'll immediately call home. So you have to put it on a firewall network or else it just like steals everything. And immediately.

Jordy Vandeput (30:25)

I'm a fan of Templeos, you know.

Lex Fridman (30:29)

Is Doug o' Laughlin suffering from a case of Claude Code psychosis?

Jordy Vandeput (30:35)

Okay, yes, yes. So I think everyone's like, claude code is for coders. It's like, no, Claude code is for people who don't code now. Right. And that's the big realization this year. We've got a couple folks now in the firm who have psychosis, but Douglas o', Laughlin, who is like semianalysis number two. He's president. He's my boy. He's the one, in fact, he's the one who encouraged me to make a substack a long time in the go. A long time ago. What were you doing before? I had a WordPress blog, and I was, like, consulting on the side, but I was like, okay, let me do a substack now. Because I saw him making money off it. I was like, this is shit. Like, why are you paying for this? There were multiple times where he wrote something. I was like, I could do way better. And, like, obviously, like, it was good because we both taught each other a lot of things and we've been great friends. And eventually he joined semianalysis. But, like, you know, he, his background is he was a hedge fund analyst, and then he decided to do a substack slash walk, hike the Continental Divide trail for, like, six months, walking from Mexico to, you know, and then, and then, you know, came back to doing substacking.

Lex Fridman (31:34)

Tried to do a fun six months of touching grass. And then he was like, I'm ready to lock it on clock code.

Jordy Vandeput (31:38)

Yeah. And so now he's, you know, like, he's never been a software developer, right? But he's been on a generational run. Like, he's, he's, he's not coding anything, right? He's just telling Claude to do stuff. And, like, it's to the point where it's like our, our, our, like, head of data. Head of it is like, oh, can you send me that? And he's like, how do I do that? And then he, like, he zips the whole thing and sends it to him. It's like, local host. He sends him a leak once. It's like, local hosts. Like, it's like, bro, that's not smart.

Lex Fridman (32:02)

But, but yeah, no, I, I, I've talked to some folks who vibe code, and they'll be like. And I'll be like, why'd you choose Node js? And they're like, what's Node js? That's a very specific someone. Yeah.

Jordy Vandeput (32:15)

We went on a little tour of a lot of our clients. Like, you know, roughly, like, half our business is. Or 40% of our business is, like, hedge funds. So we went to New York a week, two weeks ago, and we went to all of our clients. And, like, part of it's like them asking me, is OpenAI fucked? And I'm answering like, no, I think they're fine. And then, like, some, like, actual ideas. And then, like, a lot of his. Doug's just telling them cloud code is like, they're like, you don't have to hire any junior hedge fund analysts anymore. And they're like the junior hedge fund analyst and then he's explaining, you know, what can you do? It's like, well, like you can just do like financial models and perform a financial models and like everything in cloud code without ever opening Excel. And you can generate charts and like, you don't need to know how to code, you just need to know how, like how this stuff generally works and you can just do it.

Dwarkesh Patel (32:57)

How many hedge funds are just trying to copy trade situational awareness now?

Jordy Vandeput (33:02)

I mean, I think everyone who's. I think, I think a lot of hedge funds obviously believe in AI. I think there's a lot of them who don't believe in it. Right. To be clear. But a lot of them that have done the best.

Dwarkesh Patel (33:12)

Why are they selling software everywhere?

Lex Fridman (33:15)

Oh, you mean selling software stocks?

Dwarkesh Patel (33:16)

Yeah, yeah, yeah.

Lex Fridman (33:18)

Why the sell off then?

Jordy Vandeput (33:19)

Yeah, I mean, I mean, of course it's like an incremental thing. Right? But anyway, so, so these hedge funds, like, and then, then the question is like, okay, if you believe in it, how do you manifest that trade? And so when you look across the ecosystem, I would say almost all my clients sometimes think our two years out numbers are too high. But like there's, there's like Leopold's, like, your numbers are too low. And so it's like, it's like, it's like in general. Right. And I think, I think, like, if you think about how much do you believe in AI and what's your access to information of AI? You know, there's not many hedge funds who live in San Francisco and like fully breathe and live and understand it then. And then depending on how much you believe in AI, how do you manifest that trait? Right.

Dwarkesh Patel (33:58)

Are you surprised that more hedge funds wouldn't, like, even just smaller shops wouldn't say, like, hey, this I think, seems like it's going to be big. Maybe we should set up in San Francisco or higher.

Jordy Vandeput (34:08)

There's, there's a number of people, right? So we're, you know, we're getting an office together. Leopold, myself, Dwarkesh, and then a client of mine, another hedge fund, and they have one analyst here. And it's like. And there's like a number of other hedge funds that are like hiring analysts here. But, you know, being plugged into the AI ecosystem does not mean you're just in San Francisco because you can just walk around and talk to like doofus, like startups and VCs and like, not actually, you know, see what's coming down the pipeline. And you have to combine it with all sorts of information, right? You have to have A good tune with like what's going on in Asia supply chains. You have to have a good tune with what's going on in New York. Sure. You have to good tune with like what's going on like in the, in the financial markets. Right. And then like what's going on in credit markets and what's going on in all, you know, the data center, energy, blah, blah, blah, all these different industries. And so it's, it's, it's actually not like so simple to like be in tune with what's going on in AI. You can easily get like head faked. Right. You know, for the longest time people were thinking, you know, Adobe is an AI company and like, and it's like first for a bit like Adobe was going down on AI and then they like launched a few features and the stock skyrocketed and then now it's going back down again because people realize, oh wait, no, actually it's not an AI company. Like I think it's the manifestation and thought of like, what is actually going to the world going to look like if anthropic 3x is its revenue again this year, OpenAI 2x's its revenue again this year or you know, by the end of the year. How many people even believe by the end of the year AI startup revenue is over $100 billion. I think that's an insane statement for a lot of people, but that's what it's going to be. Right? And who believes that number? Right. It's like very few people. And then you, you, you draw the continuation. It's like who believes? You know, and when Anthropic says in their funding like, hey, we're going to have $300 billion of revenue by the end of the decade. And it's like, actually I think that number is too low because, because the economic value of what they're going to create is going to be insane. Yeah. And, and you tell people, oh, Excel, you know, OpenAI is going to have 18 gigawatts or 16 gigawatts by the end of 28 and they're going to be able to pay for it. And that's like, well that's $300 billion to spend. How they going to pay for it? It's like you, you sweet summer child, don't worry, Sam can raise. They're going to blow up on revenue. They're fine. Right. Like, it is like a bit of a vibe thing. It's a bit of like, you know, irrational exuberance almost. Right. Like Leopold's in his, you know, mid-20s. Like, I'm 29. Like, we, we are irrational, right, because we have not lived through, you know, you've got these, these PMs who, like, never been.

Dwarkesh Patel (36:29)

You've never been that humble.

Jordy Vandeput (36:30)

I, I don't know. Like, we almost, my family almost went bankrupt in 2008. Like, you know, because we lived in a motel and we almost foreclosed and we actually did foreclose on one motel. It's like, pretty bad. But like, yeah, I mean, I was still a kid, right? Like, it was like, yeah, yeah. I've never been humble.

Dwarkesh Patel (36:41)

It's good to, I mean, it's good to live through that, understand how things can go wrong. What are you expecting out of Zuck and Meta this year? We've been big Zuck defenders especially. I mean, there's this pressure of like, oh, Meta is spending so much and yet they haven't created, you know, any, any AI product that's super compelling or that's really working. And our stance has, has generally been met as making more money from AI than almost any company in the world outside of Nvidia. So it's like, of course Zuck should be justified in saying, hey, this is real, it's big. Like, I'm going to like back this truck up and, and go all in.

Jordy Vandeput (37:19)

Yeah, I mean, it's, it's clear. If you look at the most recent earnings, I think their CPM went up 9% when the consumer's weak. Which means, like, if you were to like, try and strip out like, what is consumer spending increasing for CPM of ads versus what is the effectiveness of their algorithms or. Algorithm got better by double digits in one quarter. Yeah, right. It's like actually insane how good the algo is getting right, at serving you the slop in the ads, right? So, so, so in that sense, like.

Lex Fridman (37:46)

The pig sound the trough.

Jordy Vandeput (37:49)

I gotta hear slop for the slops.

Dwarkesh Patel (37:53)

We're going all in on the farm.

Jordy Vandeput (38:00)

I love it. So, so, you know, if you, if you think about it, right, like, okay, Metas, where are they gonna like win, right? You know, I think if you have the Galaxy Brain take, it's like, well, they've got the best like wearables coming down the pipeline. They're going to put AI on it. Apple won't be able to put good AI on their wearables, so they'll seed it all to like micro Google or the other thing.

Dwarkesh Patel (38:20)

People, people have had this narrative, oh, as AI gets better, the value of real world experiences will increase. And I Think that's a cool theory, but if you actually play it out, AI getting better means more content that's more like effectively crafted for you, more personalized, 100 times more thousand, a million times more content. That would imply to me that people will just use digital products more, which means more time on site, more time in the app for Meta.

Jordy Vandeput (38:49)

So I don't know, I mean, I'm with you entirely. But I think, I think like the Galaxy Brain take is that you're just going to have a wearable and that's going to have AI assistant. OpenAI is trying to make it wearable. You know, you know there's, there's, you know, everyone's trying to make wearables. Google is, et cetera, et cetera. I think metal will actually execute and then they'll have a good AI and then you stack on like a few things. Right. How do they get users? Well, we've seen at least if you look at the user metric charts, Google's use, you know, OpenAI's users were growing, growing, growing. They were going to hit a trillion by the end of the year, 800 billion. Why did they not keep growing in the last quarter? It's because Nano Banana came out and they took all the incremental users. Right. And likewise, if you go look at like, you know, Gemini 3 didn't actually make Google grow that much. It was Nana Banana and Then Pro or 2 or whatever it's called, right. Those are the ones that made them really grow. Meta's licensed all of Midjourney's code data models, right. 1, 2. They're like actually just like focusing hardcore.

Dwarkesh Patel (39:46)

Was that a billion dollar plus deal?

Jordy Vandeput (39:49)

The number is undisclosed. Midjourney still exists as a company.

Dwarkesh Patel (39:54)

No, it felt, it looked to me like effectively a massive exit. But the best case scenario where they can just keep kind of being artists.

Jordy Vandeput (40:02)

I think if you had me guess, I would bet it's over a billion, right?

Lex Fridman (40:05)

Every deal that Meta did was over a billion. Basically, whether it's an employment contract, a licensing deal, an acquisition, everything had a B after it.

Jordy Vandeput (40:14)

Well, so the interesting thing is that.

Dwarkesh Patel (40:16)

Is you're missing a zero again. Don't never miss the zero again.

Lex Fridman (40:20)

Every discussion was how many billions are we spending on hiring this person buying this company?

Jordy Vandeput (40:25)

Well, Meta interestingly has gone down market for compute because they, there's not enough compute in the big size deal. So they've actually gone and like bought like small clusters. Oh, because it's like, well, I want.

Lex Fridman (40:37)

More from like Long Tail Neo clouds.

Jordy Vandeput (40:39)

Yeah, just like yeah. From a longer tail.

Lex Fridman (40:41)

Okay.

Jordy Vandeput (40:42)

Because that's the only place they can get the compute they need because, you know, they've already like went out and signed big deals with Google and Core Weave and so on.

Lex Fridman (40:48)

And so is cluster max 3 going to be a smaller shark because of consolidation in the industry?

Jordy Vandeput (40:51)

No, it's. There's more.

Lex Fridman (40:52)

It's going to be bigger.

Jordy Vandeput (40:53)

It's going to be bigger.

Lex Fridman (40:53)

Bigger.

Jordy Vandeput (40:54)

But, but, you know, so, so metal. That's ominous. It's ominous. So I think Meta will, you know, capture consumers through Generative. If there's more content, people are just going to go to the content marketplace. Right. The creator of the content captures less value as there are more content creators and more diversification of content. Right. And so I think Meta just wins by being a platform. Right. Google does too. And bytedance does too. Right. But like those three win by having the platform and then the real question is, can they get in the assistant productivity game? Right.

Dwarkesh Patel (41:30)

And I think this is important through that effectively search. Like if you're an assistant, it means that you can like there's some commerce happening.

Jordy Vandeput (41:36)

Well, they span out and poached a bunch of people from Google. Yeah. So this wasn't in the media much, but like they actually poached Google search people with similar size deals as like these crazy researchers.

Dwarkesh Patel (41:46)

Yeah. And I always, I always, I, you know, demoing, demoing any of the, the, the wearables you can imagine. Like Meta wants you to walk around in the world and see like, oh, what are those headphones? And like while we're talking, I just hit my little thing and buy it. Right. And it's like you didn't even necessarily know that it happened. But like of course Meta is going to want to.

Lex Fridman (42:02)

Everyone knows those are the Sony MDRX SU2 7 2S462.

Jordy Vandeput (42:07)

Dude, I've been, I've been screaming about them like doing some proper marketing.

Lex Fridman (42:11)

Branding.

Jordy Vandeput (42:12)

It's literally like they're over here is like WH1000XM5 and then their in ear is like WF1000XM1000. It's like, dude, just call them like Bravia Buds and Bravia like headphones or some shit.

Dwarkesh Patel (42:25)

Well, China just bought Sony. Yeah, yeah.

Lex Fridman (42:29)

Bravia brand's actually a Chinese company now. Sony sold their TV and PlayStation Buds. Yeah, yeah, yeah.

Jordy Vandeput (42:36)

PlayStation Walkman. Oh yeah. Come on. Something, something for sure.

Lex Fridman (42:43)

Anyway. Anything else, Jordy? No, this is great.

Dwarkesh Patel (42:46)

I'm excited for this weekend.

Jordy Vandeput (42:47)

Yeah, yeah, super excited. You guys.

Dwarkesh Patel (42:49)

What are some plays that we don't watch a lot of sports.

Lex Fridman (42:53)

Football guy, right?

Jordy Vandeput (42:54)

Yeah, yeah, yeah. Rural Georgia. So I like football. High school football was the thing. College football was the thing. I think NFL is a little less soulful.

Lex Fridman (43:04)

Sure.

Jordy Vandeput (43:04)

But, you know, now college football has. Has. Has the nil. And so it's also soulless to some extent. It's fine. We. We enjoy it. You know, primal desire of seeing heads clash.

Lex Fridman (43:15)

Yes.

Jordy Vandeput (43:15)

You know, and sometimes that manifests in, like, you know, Twitter drama, and sometimes that manifests in real football.

Lex Fridman (43:20)

Yeah.

Jordy Vandeput (43:21)

All I can say is fuck the Patriots.

Lex Fridman (43:22)

Okay.

Dwarkesh Patel (43:23)

Whoa. Okay.

Jordy Vandeput (43:24)

Okay.

Dwarkesh Patel (43:25)

I'm.

Lex Fridman (43:26)

We're.

Dwarkesh Patel (43:26)

We're going to. We're going to. Since we're going to be at the game, we're not going to really get the great experience seeing the ads.

Lex Fridman (43:31)

I'm going to be, like.

Dwarkesh Patel (43:32)

I'm going to be, like, glued to my phone. I want to see all the AI. The different.

Lex Fridman (43:36)

Well, don't worry. I got some more ads for you. Thank you so much for coming.

Jordy Vandeput (43:39)

Thank you so much. Great segue.