BUY COMPUTE

“We're doing well more revenue than that. Second of all, Brad, if you wanna sell your shares, I'll find you a buyer if you don't feel I I just enough. Like, you know, people are I think there's a lot of people who would love to buy OpenAI shares. People who talk, with a lot of, like, breathless concern about our compute stuff or whatever, that would be thrilled to buy shares. So I think we we could sell, you know, your shares or anybody else's to some of the people who are making the most noise on Twitter, whatever, about this very quickly.”

“Used to be 1,400,000 smartphones were sold a year. Now we're at, like, 1.1. But our projections are we maybe get down to, like, 800,000,000 this year. And next year, like, 600 or 500,000,000. They see Xiaomi and Oppo are cutting low end and mid range smartphone volumes by half. Today, you already see all the memes, like, on PC subreddits and PC gaming PC Twitter is, like, cat dancing videos. And it's like, this is why memory prices has doubled, and you can't get a new gaming GPU.”

“It's a giant CapEx boom in in in and I think check this out. This is slide 47. This is Tesla, expects their CapEx. So that's everyone's basically forced because before the competition was, I wanna raise debt and buy back my stock and shrink the equity. Now it's like, hey. We have this frontier that we're pursuing. And if I don't raise a lot of debt and pursue it, I'm gonna be so left behind because this is like a divergent technology.”

“Allbirds, which basically went bankrupt, just decided to execute a $50,000,000 convertible financing facility to pivot from direct to consumer freaking shoes to becoming an I AI compute infrastructure company. Look. I've been I've been hitting the whole we need more compute thing and talking about, like, bidding the stuff that that is a tailwind to that. But, oh my god. Like, when I saw this this morning, I was like, this is, I mean, you know, the the the crypto crew knows all about this because back in 2019 or whatever it was, we started to see, like, you know, Kodak was gonna be a a crypto company.”

“So, you know, one of my one of my, I guess, worries, although it's also an insight into into, you know, in into kind of what's happening is that, you know, some very critical decision will be will be some decision that, you know, someone just comes into my office and is like, Dario, you have two minutes. Like, you know, should we should we do, you know, should we do thing thing a or thing b on this, like, you know, someone gives me this random, you know, half page half page memo and is like, should we should we do a or b? And I'm like, I don't know. I have to eat lunch. Let's do b. And and, you know, that ends up being the most consequential thing ever.”

“And that's where I thought we did the right thing to give them flexibility to go procure that from others while maintaining, again, a significant book of business from OpenAI, but more importantly, giving ourselves the flexibility with other customers, our own one p. Right? Because sometimes, you know, Sam may say, hey. Give me build me a dedicated, you know, big, you know, whatever, multi gigawatt data center in one location for training. Makes sense from an OpenAI perspective. Doesn't make sense from a long term infrastructure build out for Azure.”

“The market's just a giant derivatives trade. It's what happens when everything gets so centralized and everyone everyone's asset management looks the same. If you have, you know, this beta neutral book, that's generally a lot of hedge funds. That's that's the guys moving the books around. And it's it's a lot of, like, you know, they they they do things in unison as much as they think they're original thinkers.”

“Methos is potentially the biggest step up in model capabilities in two years. I think that's really, really an important detail that it's so good that they're, like, don't wanna release it even though they already announced the price to their people that they did a selective release for cyber for, and it's 5 or 10 x the token cost. They just don't wanna release it because they're worried about the impact on the world. And they're releasing a worse version, Opus four seven, to us, and they explicitly said on the card, hey. We actually Preferably. Made it worse at cyber.”

“There is this person who leads our network operations. And the amount of fiber there, the AI ran, and what have you. It's just crazy. There are, I think, 400 different fiber operators we are dealing with worldwide. But the person who leads it, she basically said to me, you know what? I there's no way I'll ever get the headcount to go do all this. Not forget, even if I even approve the budget, I can't hire all these folks. So she she did the next best thing. She just built herself a whole bunch of agents to automate the DevOps pipeline of how to deal with the maintenance.”

“AlphaGo was a turning point because it showed that AI could not only reach human performance but discover entirely new ways of thinking that humans hadn't considered.”

“Austin's really interesting because we actually expanded the housing supply here massively. So my home value has gone down quite a bit in Austin since the peak, which is actually really good for the labor pool and job market. So it keeps the homeostasis of the ecosystem working. But in in San Francisco, you you know, my buddy's like, I don't understand how this housing market does not crack. It's like regulatory crap shit piled on top of each other. No supplies getting built.”

“I've said to people my view is that we'll look back ten years from now and the S&P 500 will probably be around the same level it is now. ... but right now the S&P 500 market cap is more than two times the size of the economy, a record going back all the way to the Great Depression. I'm not saying that's speculation. I'm saying that private sector has suffered dramatically.”

“One, I write this thing called the DVQ, Dario Vision Quest. I wasn't the one who named it that. That's the name it it it received, and it's one of these names that I kind of I tried to fight it because it made it sound like I was, like, going off and smoking peyote or something. But the name just stuck. So I get up in front of the company. Every two weeks, I have, like, a three or four page document, and I just kind of talk through, like, three or four different topics about what's going on internally.”

“None of that incremental capacity really gets here until the second that they've decided to do in addition to the typical 20 to 30%. They can stretch a little bit, but, really, the true incremental supply doesn't come till '28, which is a very unique thing. Even if they wanted to build as fast as possible, it doesn't come till '28, late twenty seven at best. So the result is memory prices have gone through the roof. And guess what? They're gonna double and triple again, at least on DRAM, especially.”

“What AI is is the ability for a company to fire people and still have the ability to make growth. So it becomes at the expense of labor. ... QE is gonna be for companies that exist, an ability for them to get rid of people in the form of AI. So if they can replace every person with a bot, that's great.”

“I I hope for very small scientific discoveries in 2026, but if we can get those very small ones, we'll get bigger ones in future years. That's a really crazy thing to say is that, like, AI is gonna make a novel scientific discovery in 2026, even a very small one. This is, like, this is a wildly important thing to be talking about. So I'm excited for that. But, yeah, my personal bias is if we can really get AI to do science here, that is I mean, that is super intelligence in some sense.”

“ASML makes the world's most complicated machine, I e, an EUV tool, and the selling price for those is $304,100,000,000 dollars. And currently, they can make about 70. Next year, they'll get to 80. Even under very aggressive supply chain expansion, they only get to a little bit over a 100 by the end of the decade. ASML's commented their supply chain has over 10,000 people in it.”

“So it's not like, for example, the per seat versus consumption. The reality is agents are the new seats. And so you can think of it as, the enterprise monetization is much clearer. The consumer monetization, I think, is a little more murky.”

“Well, I mean, I think the the cycles of demand and supply in this particular case, you can't really predict. The secular trend is what Sam said, which is at the end of the day because quite frankly, the biggest issue we are now having is not a compute glut, but it's a power, and it's sort of the ability to get the bills done fast enough close to power. So if you can't do that, you may actually have a bunch of chips sitting in inventory that I can't plug in. In fact, that is my problem today. It's not a supply issue of chips. It's actually, the fact that I don't have warm shells to plug into.”

“I have enormous respect for Warren Buffet and what he's done. He's the goat of goats, but the returns post reg FD are they are just materially worse than they were for reg FD investors. I pointed out one unavoidable fact, which is his returns are bimodally distributed pre and post reg FD. When you have to follow the rules of disclosure, everybody's returns got kneecapped. The reason why Nancy Pelosi's returns are so consistently good is reg FD does not apply to people in congress.”

“I'll put my neck out on the line here and say that the next time I see you, Bitcoin will be significantly higher. ... I believe we've had a massive distribution of which makes sense. And I think eventually when it when it breaks higher this time, I don't think it's gonna stop.”

“There there will come a glut for sure. And whether that's, like, in two to three years or five to six, Satya and I can't tell you, but, like, it's gonna happen at some point, probably several points along the way. So, you know, if a very cheap form of energy comes online soon at mass scale and a lot of people are gonna be extremely burned with existing contracts they've signed, it if if we can continue this unbelievable reduction in cost per unit of intelligence, let's say it's been averaging, like, 40 x for a given level per year. You know, that's like a very scary exponent from an infrastructure build out standpoint.”

“I don't know how we're supposed to comply with that California sorry, Colorado law. I would love them to tell us, and, you know, we'd like to be able to do it, but that's just from what I've read of that, that's like a I literally don't know what we're supposed to do. I'm very worried about a 50 state patchwork. I think it's a big mistake.”

“But what we've seen from from the beginning, you know, at least if you look within anthropic, there's this bizarre 10 x per year growth in revenue that we've seen. Right? So, you know, in 2023, it was, like, zero to a 100,000,000. 2024, it was a 100,000,000 to a billion. 2025, it was a billion to, like, 9 or 10,000,000,000.”

“I think the yen is the clue to watch what happens because if the yen really breaks $1.60, the dollar wrecking ball kind of if it breaks with volume and power, I think the dollar wrecking ball merges and causes credit credit problems. Look back at every time the yen crashed down from 106 or, you know, or the dollar, weakened from 160 USD JPY, and the Nasdaq usually peaks right there or around it.”

“They have you can either make housing and mortgages more affordable by getting bond yields lower and therefore mortgage rates lower, but you can't do that at the same time as having stock markets higher because you need disinflation. You need to get a handle on inflation, which means some sort of hawkish reaction function, which means lower stock prices to get those lower mortgage rates. So you have to decide. And he you know, we don't wanna see housing prices go lower. So that doesn't matter as much. So they decide to pump the stock market instead.”

“AlphaGo was a turning point because it showed that AI could not only reach human performance but discover entirely new ways of thinking that humans hadn't considered.”

“The OpenAI growth rate's been around three to four x a year. The anthropic growth rate has been around 10 x a year. So they went from, let's call it, 1 to 10,000,000,000 of ARR last year. And by the '1 this year, they were already at 30,000,000,000 of, again, that's called their revenue. And they're on their way, you know, like Brad Gerson was saying on our podcast, they're gonna end this year at 80 to 100,000,000,000, at least on the current trajectory. So if it's taking anthropic, let's say, one year to 10 x and it's taking OpenAI two years to achieve a 10 x, then it's obvious which one's gonna win.”

“There's an interview with the CEO of NVIDIA, and he said the biggest bottleneck we have right now with AI data center build outs is we can't find enough plumbers for the data centers. Like, that's the the bottleneck is in the semis. It's it's literally just the plumbers to be in there setting up the plumbing at the data centers.”

“Methos is potentially the biggest step up in model capabilities in two years. I think that's really, really an important detail that it's so good that they're, like, don't wanna release it even though they already announced the price to their people that they did a selective release for cyber for, and it's 5 or 10 x the token cost. They just don't wanna release it because they're worried about the impact on the world. And they're releasing a worse version, Opus four seven, to us, and they explicitly said on the card, hey. We actually Preferably. Made it worse at cyber.”

“I think there will be a large scale protest against Anthropic and end up at AI. People hate AI. AI is less popular than ICE, less popular than politicians. With Anthropic adding so much revenue, that's gonna start causing business changes downstream. People are gonna get more and more scared of AI. They'll start blaming more and more of their own problems and things that are global, have been deep seated problems for a long time. Those will bubble up and be blamed on AI.”

“Growth is king right now in this in this world, in this segment. Growth is the whole damn thing. And if Anthropic is growing faster than OpenAI by a significant clip, the investors right now are gonna play it forward. You start to get network effects around compute, network effects around the number of tokens you're pushing out for various customers, enterprise or consumer. If you believe there's network effects with the scale of data that you have and the scale of customers and the revenue, that's this cash that's coming in that you redeploy into compute. I'd be very worried, if I'm OpenAI and seeing somebody growing faster at the same size.”

“If you ship out all the process engineers and assuming it's, like, hot enough that you destroy the fabs, no one has all the fabs in Taiwan now, which is a big risk. These tools actually use a lot of semiconductors, which are manufactured in Taiwan. So it's like a snake eating its own tail sort of like meme because you can't make the tools without the chips from Taiwan, which you can't use without the tools in Taiwan. Just shipping out all the engineers and blowing up the fabs means China has a stronger semiconductor supply chain than the rest of the world.”

“I think there will be a large scale protest against Anthropic and end up at AI. People hate AI. AI is less popular than ICE, less popular than politicians. With Anthropic adding so much revenue, that's gonna start causing business changes downstream. People are gonna get more and more scared of AI. They'll start blaming more and more of their own problems and things that are global, have been deep seated problems for a long time. Those will bubble up and be blamed on AI.”

“And so you sort of have this, like, dynamic that is quite interesting in that. An h one hundred is worth more today than it was three years ago.”

“After we get through the next CPI print in early May, it will be 3.6 or higher. So we're not far from four. And the following month, I think what rolls off is either a zero or a point one, which means all you'd need is the kick in from inflation, from plastics, from diesel. I mean, this country, anything shipped to you is on a truck.”

“Huawei is arguably the only company in the world that has all the legs. Huawei has cracked software engineers. Huawei has cracked networking technologies. That's, in fact, their biggest business historically. And they have cracked AI talent. But furthermore, beyond NVIDIA, they actually have better AI researchers. And furthermore, beyond NVIDIA, they have their own fabs. It's very arguable that Huawei, if they had TSMC, would be better than Nvidia.”

“I mean, it's sort of one of those interesting things, which is everything is a commodity. Right? Compute, storage. I remember everybody saying, wow. How can they give a margin? Except at scale, nothing is a commodity. And so, therefore, yes. So we have to have a cost structure, our supply chain efficiency, our software efficiencies, all have to kinda continue to compound in order to make sure that there's margins, but scale.”

“I actually believe it could be the case, is that is that dictatorships become morally obsolete. They become morally unworkable forms of government, and that and that and that the the the the crisis that that creates is is is sufficient to force us to find another way. I just wonder if it will motivate new ways of thinking about with with with the new technology, how to preserve and protect freedom.”

“Dude, when I walk to work, there's just empty office space everywhere. And it I'm like, there there's this, like, below our work, there's a a jewelry store on South Congress, which is like it doesn't even have jewelry in it. It's just sitting there. Some guy probably bought it in, like, 1960. And, like, it says Cougar Jewelers, and it just, like, there's no there's no jewelry anymore. It just sits there, and it acts like a jewelry. It's it's all like a front. We're living in, like, a dystopian.”

“There's people who have argued GPU's full lives are less than five years. Complete nonsense. There are clusters now resigning. Three or four year old Hopper clusters resigning for three or four more years. There's a 100 clusters that are resigning for another couple years. So the useful life is clearly not five years. It's maybe even seven or eight years, arguably. We don't know yet. We'll see when Hopper gets there, but it's clearly not five years. So useful life is extending, and the prices are going up on that renewal.”

“This year, the spend has just skyrocketed, and it really started in late December with Opus. ... We signed an enterprise contract with Anthropic, and it's gone to the point where now I think when I last talked to you, it was 5,000,000 spend rate. It's actually 7,000,000 spend rate now. ... We're north of 25% of spend on cloud code as a percentage of salary. And if this trajectory continues, then, you know, we'll spend more than a 100% by the end of the year, which is a bit terrifying.”

“Probably one of the worst things the Trump admin has done is just, you know, kinda disintegrated the integrity of capital markets in in The US and basically allowed white collar crime to be legal insider trading everything to to just an extreme extent. I mean, it's just not good for because short people shorting the markets and people long in the markets both do a do a deed by, you know, steering efficient allocation of capital.”

“Anything that was created on code in the digital economy has been getting killed. Salesforce.com, Adobe, all of these things which had moats around their businesses were quickly unwound. ... I think we're in the stage, the, I think, the most important stage for Bitcoin where it no longer is considered a software, a piece of of code. It is considered something that is scarce.”

“Negative negative real we're living in a negative real yield world now, which is, like, global flows talks about this too. But, like, this is the most important thing to realize is, like, if inflation's really at, you know, 5% and yields are at 3%, you should lever up and buy whatever is growing faster than that. And that's that's pretty much how they're gonna grow their way out of this.”

“What used to matter a lot was execution was very, very fucking difficult, and ideas were cheap. Now ideas are cheap and plentiful, but execution is very easy. So, really, only the good ideas are the ones that can justify the spend on super cheap implementation. ... If implementation costs continue to tank, which they are, we don't even have Meetos yet. ... What now comes to the world, it's a complete reordering of how economies work.”

“I've got a side quest where I'm just investing. I have agents investing and betting on Calci and Poly and, you know, these other places. And it's silly how dumb the agents are, even their best agents, to be honest. We had to spend a lot of time with our investing agents, getting them on board with the idea that if you wanna make money investing, you can't be on both sides of the same bet.”

“This is a really good one from the Fed a Fed research paper that came out last week, which answers the question that we've all been wondering, which is how much of tariffs is causing inflation, namely in core goods because that's mostly where you would see it. And so they came out with this one that just shows, like, before the Iran war, we were seeing core goods inflation accelerate pretty significantly. If we didn't have the tariffs, we would have had deflation in core goods. Instead, we have two and a half percent core PC core goods inflation.”

“Space data centers effectively are not limited by, you know, hey. We have this energy advantage. It's actually just limited by the same contended resource. We can only make 200 gigawatts of chips a year by the end of the decade. So what are we gonna do to get that capacity? It doesn't matter if it's on land or in space. Elon doesn't win by doing, you know, 20% gains. Elon never wins that way. Elon wins when he swings for the fences and does 10 x gains.”

“What used to matter a lot was execution was very, very fucking difficult, and ideas were cheap. Now ideas are cheap and plentiful, but execution is very easy. So, really, only the good ideas are the ones that can justify the spend on super cheap implementation. ... If implementation costs continue to tank, which they are, we don't even have Meetos yet. ... What now comes to the world, it's a complete reordering of how economies work.”

“Obviously, the cost that we just talked about is inflation, but, like, who who cares about inflation if the stock market's going up? Right? The policy is pump stock market. That's what it is. And, like, every time they try to actually, like, target another pocket of the social, like, strata, that derails the stock market.”

“I believe an iPhone has 12 gigabytes of memory. Each gig cost used to cost roughly 3 or $4. That's $50. But now the price of memory is, like, tripled. Let's call if it's now, it's $12 per gig for DDR. So now you're talking about a $150 versus $50. A $100 increase in cost on Apple. Apple has some margin. They're not just gonna eat the margin. So now that's a $100 cost increase. The NAND also has the same sort of, like, market. So in fact, you know, it's probably a $150 increase on the on the iPhone. That means the end consumer is paying $250 more for an iPhone.”

“None of that incremental capacity really gets here until the second that they've decided to do in addition to the typical 20 to 30%. They can stretch a little bit, but, really, the true incremental supply doesn't come till '28, which is a very unique thing. Even if they wanted to build as fast as possible, it doesn't come till '28, late twenty seven at best. So the result is memory prices have gone through the roof. And guess what? They're gonna double and triple again, at least on DRAM, especially.”

“My anecdote on this is back in December when it was first rumored that Swalwell was gonna run for governor, I started making some calls to various folks to be like, hey, you know, what do we think of this guy? I spoke to several people who independently told me that there's knowledge about this guy sending pics to employees and that this guy has a bunch of stuff that's gonna come out about him. I largely kind of dismissed it because I was like, if this is true, this would have all come out already. And then everything that I had been told started to come out in the last week. The striking aspect of all of this for me was how much knowledge there was about these various incidents with the guy, how so many people had this knowledge, and how no one had actually brought the knowledge to bear.”

“I think diffusion is very real and and and and doesn't have to, you know, doesn't exclusively have to do with limitation limitations on the AI models. Like, again, there are people who use diffusion to to, you know, as kind of a buzzword to say this isn't a big deal. I'm not talking about that. I think AI will diffuse much faster than previous technologies have, but but not infinitely fast.”

“Yeah. So it actually happened in a pretty simple way, which is we had our own, you know, we had our coding models, which were good at coding. And and, you know, around the beginning of 2025, I said, I I think the time has come where you can have nontrivial acceleration of your own research. And then, you know, this thing, I, you know, I think it might have been originally called Claude CLI, and then the the name eventually got changed to Claude Code internally, was the thing that kind of everyone was using, and it was seeing fast internal adoption. And I looked at it, and I said, probably we should launch this externally.”

“But only but JCal, only if it's a pia de terre. So what it means is that the most elastic part of the market is what they're targeting for this tax. In other words, people who don't live in New York, who just have it as a second or third home, who could buy that property anywhere are now being taxed the most. So what do you think that's gonna do? It's gonna have a massive impact on demand for second homes in New York, which will crash the whole market.”

“In isolation, you almost always wanna go with a smaller model that gets RL'd faster and gets deployed into research and development. So you can build the next thing and get more compute efficiency wins. And then this compounding effect of, oh, I made a smaller model that I r l'd more that I then deployed into research and development earlier, and I spent less compute on the training itself because I was able to allocate more compute to the research. This compounding effect of being able to do the research faster and faster is potentially a faster takeoff.”

“It's not what we thought. And as I said to somebody, it's not like when we first invested our billion dollars that, oh, this is going to be the 100 bagger that I'm gonna be talking about to VCs about, but here we are. But we are very thrilled to be, an investor and an early backer, and and it's a great and it's really a testament to what Sam and team have done, quite frankly.”

“And so I could buy a trillion dollars. Actually, it would be, like, $5,000,000,000,000 of compute because it would be a trillion dollar a year for for five years. Right? I could buy a trillion dollars of compute that starts at the 2027. And if my if my revenue is not a trillion dollars, if it's even 800,000,000,000, there's no force on earth. There's there's no hedge on earth that could stop me from going bankrupt if I if I buy that much compute.”

“This is the ESG population. I I was dying when I saw this just because when we saw all these, you know, giant BlackRock cares about, you know, oh, the environment and social good. And they they convinced all the pension funds. You know what's interesting is it died right when inflation returned at the top of 2022. So so so the extrapolation is ESG was just a zero rate phenomenon. And as as soon as real life costed something, it went away.”

“There's people who have argued GPU's full lives are less than five years. Complete nonsense. There are clusters now resigning. Three or four year old Hopper clusters resigning for three or four more years. There's a 100 clusters that are resigning for another couple years. So the useful life is clearly not five years. It's maybe even seven or eight years, arguably. We don't know yet. We'll see when Hopper gets there, but it's clearly not five years. So useful life is extending, and the prices are going up on that renewal.”

“Google sold, I think, a million was it the v sevens, the Ironwoods, to Anthropic? DeepMind people are like, this is insane. Why did we do this? But then Google Cloud people and Google executives saw a different, like, thought process. They saw this dislocation. They negotiated a deal, and they were able to get access to these to this compute before Google realized. We saw capacity on anthropic or sorry, on TPUs go up by a significant amount over the course of those six weeks. Google even had to go to TSMC and explain to them why they needed this increase in capacity because it was so sudden. But that a lot of that capacity increase was for selling to Anthropic. Because Anthropic saw it before Google.”

“The insight is is like Trump's weather vane is a stock market. We see Volo's up. We see the VIX is really high, but you know what? The S and P is trading in this band that's actually fairly tight given the crazy shit that's going down. He does not let the S and P go down too low. He gets people nervous. He makes moves, and then he comes back to sort of reality and gets things done as practical and probably the better parts of what he might do well. And people are pricing that in.”

“I'll tell you the one thing I think you're missing, which is that most people in America really are starting to really hate rich people. And there's no physical space that better represents the wealth in America, the wealth creation that's happened that a lot of people feel left behind from than the data center. It is the temple of the wealthy. It is the mechanism, the tool, the machinery of the wealthy. It is the way that the rich, elite tech, kind of political connected billionaires that we're obviously all attached to are taking from the poor, getting themselves ahead, shooting themselves to space, leaving everyone else behind.”

“The whole, housing thing is complete cap and utter bull because if you move to a place like Austin where you are in Nevada or Florida, you see what happens when you allow people to build units. In Austin, three years in a row, rents and housing prices have gone down while net migration has gone up. Austin has, like, roughly doubled as a city over the past decade, and yet the rent for, you know, whatever, a one or two bedroom apartment's gone down. So in other words, if you let people build to satisfy the demand, you won't have this problem.”

“OnlyFans' interest has been going down, and interest in church has been surging. And then I've been seeing these I don't know if you've seen them, but these videos of, every Sunday in West Village in New York, all Yes. Everybody's there's a huge the the hottest place to go right now in West Village in New York is going to Catholic church.”

“I think, not the buried not the bury the lead here, but the biggest thing is that he said he hated Ford guidance. He wants to get rid of Ford guidance. I was like, dude, it's just a podcast, man. Why are you so upset about it?”

“So so on the ten years, I'm, like, you know, 90%, which is about as certain as you can be. Like, I think it's I think it's crazy to say that this won't happen by by by 2035. I have a strong view, 99, 95% that, like, all this will happen in ten years. Like, that's I think that's just a super safe bet. And then I have a hunch this is more like a fifty fifty thing that it's gonna be more like one to two, maybe more like one to three.”

“Allbirds just pivoted from ugly sneakers, discretiad, to AI and the stock has ripped. Talking about peak bubble behavior, podcasts getting bought by frontier models and sneaker companies, pivoting to data centers. This reminds me of the late nineties where all you had to do was change your name to whatever.com and you get a huge pop in your valuation.”

“Dario, when he was on your podcast, was very, very, like, conservative. He's like, you know, I'm not gonna go crazy on compute because if my revenue inflects at a different rate, at a different point, I don't wanna go bankrupt. But in reality, you know, he's definitely missed the pooch in terms of, like, going, like, OpenAI, which was, let's just sign these crazy fucking deals. And OpenAI has kind of got way more access to compute than Anthropic by the end of the year. And so what does Anthropic have to do to get the compute? Well, they have to go to lower quality providers that they would not have gone to before.”

“This year, the spend has just skyrocketed, and it really started in late December with Opus. ... We signed an enterprise contract with Anthropic, and it's gone to the point where now I think when I last talked to you, it was 5,000,000 spend rate. It's actually 7,000,000 spend rate now. ... We're north of 25% of spend on cloud code as a percentage of salary. And if this trajectory continues, then, you know, we'll spend more than a 100% by the end of the year, which is a bit terrifying.”

“This is the put to call ratio. A lot of people have been calling calling this out. So, like, now everyone went balls along, puts previously at the Lowe's. This is what giving retail these products, it's really it's good in some ways, but it exacerbates things because they always you know, they push things in both directions. But now they went balls long puts, and now they unwind it. And now they're going long calls.”

“The thing that was being voted on is we're going to ban all state regulation of AI for ten years with no apparent plan to to do any federal regulation of AI, which would take congress to pass, which is a very high bar. Given the serious dangers that I lay out in adolescence of technology around things like the, you know, kind of biological weapons and bioterrorism, autonomy risk, and the timelines we've been talking about, like, ten years is an eternity.”

“You you know what happened to me was I started in the business in 2007, and, you know, they always say, like, that's your first impression of of things. And I was working with these guys that were caking it. And then they had no idea the market was built off of house of cards, and then 2008 came. So I was bearish from that point on for, like, ten years, and it's taken me to unwind, like, okay. Like, if you actually let a free market happen, like, a lot of this leverage would go. But now it I had to learn the hard way that, like, they're gonna back this thing up no matter what.”

“The other side of the value capture for us is going to be incorporating all this IP. Not only we have the exclusivity of the model in, Azure, but we have access to the IP. I mean, having a royalty free, let's even forgetting all the the know how and the knowledge side of it, but having royalty free access all the way till seven more years gives us a lot of flexibility business model wise. It's kinda like having a frontier model for free, in some sense. If you're an MSFT shareholder, that's kinda where you should start from is to think about it.”

“Every week every week, it starts with a gigantic Gemini produced or or nano banana produced visual of scarcity and abundance. Everyone who's watched it knows. It's like get out of anything that is abundant and only be invested in things that are scarce. That's it.”