The Compound and Friends - The Four Horsemen of the AI Apocalypse with Ed Zitron

Episode Date: August 28, 2026

On episode 257 of The Compound and Friends, ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Downtown Josh Brown⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ and Michael Batnick are joined Ed Zitron to discuss: the ultra-bear case for AI, Nvidia’s explosive growth, the economics of OpenAI and Anthropic, whether AI demand can justify the massive hyperscaler CapEx boom, the data center buildout, CoreWeave and the neoclouds, Oracle’s AI bet, private credit and debt financing, the warning signs that could finally break the AI spending cycle, the “rot economy,” whether AI is actually improving corporate productivity, and much more! This episode is presented by Fidelity Investments and the all-new Fidelity Trader+, Fidelity’s most powerful trading platform yet. Learn more at http://www.fidelity.com/TraderPlus Sign up for The Compound Newsletter and never miss out: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠thecompoundnews.com/subscribe⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Instagram: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠instagram.com/thecompoundnews⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Twitter: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠twitter.com/thecompoundnews⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ LinkedIn: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠linkedin.com/company/the-compound-media/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ TikTok: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠tiktok.com/@thecompoundnews⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Investing involves the risk of loss. This podcast is for informational purposes only and should not be or regarded as personalized investment advice or relied upon for investment decisions. Michael Batnick and Josh Brown are employees of Ritholtz Wealth Management and may maintain positions in the securities discussed in this video. All opinions expressed by them are solely their own opinion and do not reflect the opinion of Ritholtz Wealth Management. The Compound Media, Incorporated, an affiliate of ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠Ritholtz Wealth Management⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠, receives payment from various entities for advertisements in affiliated podcasts, blogs and emails. Inclusion of such advertisements does not constitute or imply endorsement, sponsorship or recommendation thereof, or any affiliation therewith, by the Content Creator or by Ritholtz Wealth Management or any of its employees. For additional advertisement disclaimers see here ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://ritholtzwealth.com/advertising-disclaimers⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠. Investments in securities involve the risk of loss. Any mention of a particular security and related performance data is not a recommendation to buy or sell that security. The information provided on this website (including any information that may be accessed through this website) is not directed at any investor or category of investors and is provided solely as general information. Obviously nothing on this channel should be considered as personalized financial advice or a solicitation to buy or sell any securities. See our disclosures here: ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠https://ritholtzwealth.com/podcast-youtube-disclosures/⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ Fidelity Disclosure: Fidelity Investments and The Compound are not affiliated. Views, opinions, products, services, and strategies discussed are not endorsed or promoted by Fidelity Investments. Fidelity Brokerage Services LLC, Member NYSE, SIPC Learn more about your ad choices. Visit megaphone.fm/adchoices

Like, I ran a PR firm until, like, last year. And I was, like, doing that as kind of a contract labor thing. And then I started writing this in 2020. And I just kept going and going and going and going. And then initially I was writing about, like, management theory stuff. And how much I liked remote work and how evil the articles were about it. And I wrote about crypto and I wrote about Elon. And by the end of those two things, I was so thoroughly depressed with both.

I was just like, what am I going to write about? Well, with the topics or with writing? With the topics. The topics were so depressed. Like, every week following what fucking Elon was. Musk does. And then the metaverse happened. That was kind of fun to write about, but that was before that. But then Open AI got, sorry, Sam Altman got fired from Open AI.

And I saw on Twitter that there were journalists who were like, oh my God, I hope they bring him back. They were like talking about him like a rock star. It was like, oh, then everyone's, there was someone like, oh, I'm crying. Because how beautiful it is that everyone's shown support for Sam Altman. Josh cried. Journalists. Yeah, journalists. It's crazy. I saw this, and I just, I didn't even judge them for it.

I was like, fine, whatever feel. But I was like, this is weird. Why are you acting? Like, I was like, what does this mean? Why are you acting this way? And so I kept looking really intently, and I just immediately saw this thing where everyone was saying, hey, oh, it's the biggest, most hugest, best thing ever. It's inevitable, blah, blah.

But I couldn't find any revenues. Like, I couldn't find, and every time I was like, okay, great, but how much does it make? And everyone was like, I couldn't possibly say, how much does Microsoft? They're spending billions of dollars on CAPEX. How much this is this? I can't tell you. None wants to tell you. But you kept having these articles every earnings season saying

Microsoft's AI bet pays off. Google's AI bet pays off. Even though they never mentioned their AI revenue. And I just kind of taught myself economics from there. Like a lot of this, like everything's self-taught, which to the chagrin of my many haters, is just like stuff I learned. Yeah.

But you can learn a lot by reading books and such. Okay. So you, all right. So your, I guess, It's a substack or? Ghost. Okay. So what you're doing is being read everywhere.

People are sharing it and bowls on AI are reading it because I think, so my view of the market is most professional investors are intelligent enough to listen to both sides. Yeah. And they don't have to agree with everything they read in order to want to keep reading it. Right. I don't think that's the same is true for retail investors, but we're not going to concern ourselves with that. I think professionals do like to hear both sides, especially if they have a lot of money on the line.

That's actually been my experience talking to some of the banks and the institutions. Like they want to, even the ones we're very pro-out. I agree with that. Yeah. All right. We're the same way. So we've been doing this show. We tape every Thursday going back to the summer of 2021.

And for five years, the majority of our guests are bullish. And so far, so good for the most part. Yeah. Okay. We have had people who are bearish. They are non-specific bearish. So we've had people like Jeremy Grantham, who's become famous for, he takes issue with the perma bear label, but he's almost always bearish. So we've had people like that on the show.

And I always find that fascinating, but we haven't really had specifically an AI bear. Right. Somebody that, like, specific, you know, most of the people that come on portfolio managers or their analysts covering the space. And you know all the incentives to. Yeah. see the brighter side. So this will be an interesting test of all of the things that we've done on the show to hear the other side.

So we're really excited about it. I'm pumped. Okay. Do Bulls think you're an idiot or do they generally like what you're saying to them? So it's got to a weird point. So if we're mixing in every kind of bull here, so I imagine a lot of them are retail bags. They are driven, because I used to be in PR, they're like, he's just a PR boy. He's just a PR guy.

He shouldn't talk about. Stop talking to him. You could say something as if financial experts have never got anything. wrong before, there are some of, like, what's weird is like, there are some of them who are quite bullish on semi-caducts, like, who follow me on Twitter, like, Bubble Boy, who's really great and very, very smart on semis. I'm not sure he agrees with me on everything, but very kind, and friendly, and, like, there are lots of them who are. Some of them are vile and horrible, but so say, many such cases. And so, the thing is what the Bulls is, the ones who are like,

I love this tech, I have some crazy dream, what it was. will be, but there's a bubble. I respect those. If you can like love a, I still think there are massive environmental and social problems that they're not thinking of, but at the very least, you're living in reality. I think the ones who are like anthropic and open AI are going to grow to $284 billion in revenue. That's what open AI is projecting in 2030. The people who read that and they're like, yes, sounds good to me. Those people are not living in reality. And I don't need a degree in phone. Because of number? Yeah. Open AI is going to become the, going to become bigger than meta in three and a half years, if you believe their projections.

And they're going to do that while spending more than twice of meta's OPEX. I think Microsoft's OPEX is like $150 billion. And Open AI is going to spend $200 something billion or more on compute in 2030. To get to $380 billion in revenue. $284. Are you outright bearish or you think the balls are smoking crack? Or am I saying the same thing? Possibly the same thing.

I think the, after 2022, when everything got really great, I'm in tech. I'm done if you remember, like, Nvidia had a flat year, fiscal 23. Of course. Really rough time. And everyone was kind of blowing hard. Microsoft, Google, Amazon, meta, cranked up prices, changed ad auction stuff. Amazon actually made a surprisingly healthy ad business. They had to find ways to grow revenue. And also chat GPT happened.

So they went, oh, we'll buy a bunch of GPUs. And the market immediately was like, all of this revenue growth is coming from AI. And because the companies never disclosed their AI, revenues and journalists don't have the teeth to bother them, or analysts, especially, people were like, oh, it's AI. Every time they spend this money, it's AI. And this was really good at help their stocks pop crazy style. Like, one fair criticism in my work is if you traded off of my work, which I've never said to do, I do not give financial advice.

Yeah, you probably... Well, but if you end up being right about this, they're going to put a gun through your head and say start a hedge fund. I mean, oh, Christ, I don't even know how to do that. I'm just telling you that. It sounds fun. But the point is, I like writing a lot. I really enjoy actually getting into it's genuinely intellectual.

actually fascinating. But the point of making is up until about 2025, it was mostly just buying GPUs and hoping Open AI and Anthropic grew into him and also hoping they'd become profitable. But then what happened is the hyperscalists have now become financially dependent on the growth for Open Airon Anthropic. Analyst expectations from UBS, Barclays, and Wells Fargo have $440 billion of cloud revenue across Google, Amazon, and Microsoft coming just from Open Airon Anthropic, two unprofitable startups who need to constantly raise money. And also, the big companies need to build the data sense to make the money. So there's all of these very improbable, if not impossible, things that need to occur. And also, Open A&A&Athropic need to have 10 times the demand they have right now.

It's not even, however you may feel about these companies, they are not big enough, they are not generating enough cash right now to even get close to covering their 1.1 or more trillion dollars in commitments. And on top of that, they're taking up 90% of AI infrastructure. This is creating an illusory demand signal because right now it's difficult to get GPUs. People are saying, oh, oh, it's because there's so much demand for AI. There's a big Barry Bond's asterisk at the top of that.

There's so much demand from AI from two companies, pretty much. Meta as well, but they're not doing anything. They're just rolling in their filth. So what's happening is everyone's going, oh, there's tons of demand for AI. I'm going to build a bunch of data centers. even though when you, so in fiscal year 26, Microsoft, they made Kuntar Bloomberg about $34.33 billion on AI, $24.1 billion of that is Open AI, which means the OpenA. Microsoft software and GPU rental services are single-digit billion dollar businesses. And they have 260 plus billion dollars

in CAPEX in this. It's a disaster and it's not something that gets easily fixed. I don't think it gets fixed at all. And on top of it, AI GPUs are. pretty specialist. They're useful for like AI and data analytics and 3D modeling, which is a very small business. There's no dot-com bubble thing after this. The electricity will be just as expensive, if not more. Oh, wait, you're on a roll. I want to give people the chance to close out their long positions and we're going to introduce the show properly and then we're going to dive right back in. Who's doing my countdown? Miss Nicole? This message is brought to you by Fidelity Investments. When timing is everything, you need powerful tools and research that can meet

in the moment. With the all-new Fidelity Trader Plus platform, your charts and preferences show up consistently synced up across all your devices so you can act fast whenever and wherever you're trading. You can save an order on your desktop at home, get a mobile alert when you're at work, and complete the trade in the Fidelity app without starting over. And with the downloadable Fidelity Trader Plus desktop platform, you have more control with multi-monitor views, enhanced tools and customization options, and integrated screen sharing with Fidelity trading specialists. Try Fidelity's most powerful trading platform yet at Fidelity.com slash Trader Plus. Fidelity investments and the compound are not affiliated.

Views, opinions, products, services, and strategies discussed are not endorsed or promoted by Fidelity Investments. Fidelity Brokerage Services LLC, member NYS-S-I-P-C. All right. Ladies and gentlemen, welcome to the world's greatest investment. and podcast. I say that without a trace of irony. I really mean it. Today you're in for a very special treat.

We are having a conversation that we have not really had here before. What is the ultra bare case on AI? Why are people clapping? It's crazy. It's your retirement on the line. What's going on? Guys, calm down.

All right. Joining us today, Ed Zittron. Ed is the founder and CEO of EZPR. a research and media analysis firm he launched in 2013. He also hosts the Webby Award-winning Better Offline podcast with IHeart Radio and Coolzone Media. Ed writes a newsletter called Where's Your Ed at that he launched in 2020, where his 2023 rot economy essay, I'm going to get into that, arguing tech firms have chased growth at the expense of innovation and users has gone

viral. Today, he's one of the loudest and most cited AI skeptics focused on the unit economics of open AI anthropic and the hyperscaler KAPX buildout. Ed, thank you for being here. We appreciate it. Thanks for you for having me. All right.

All right. So we have to start with Nvidia. Yeah. I feel like they're proving everything that you've said to be not correct. I know, but just... No, no, no, no. Hear me out.

They just, they just came out. Here's what they did last night. And Michael will jump into the numbers and then we're going to give you a chance to comment on it. But they just said for the first time ever, we're actually going to give a full year's worth of guidance. They were going quarter by quarter. They actually gave guidance to 2028, fiscal 2028, which is not calendar 28. But February, February 237 often Woods. So they're saying, like, we're going to grow revenue by about 70%. We've got the sales locked in for GPUs and Vira CPUs and et cetera, et cetera, we know that like 90% of their business is selling to data centers.

A huge chunk of that is selling to the same four or five customers, hyperscalers. They are talking about broadening that out. They're talking more about now automation, robotics, but the data center business is the business. Okay. So is he going to be completely wrong in his own forecast? Or will they make their own? number, but that's part of the problem because of who the buyers are and what their motivations

are. Let's give people the nuanced take on exactly what you're saying, because Nvidia really is at the heart of the entirety of the AI story. I just want to be clear as well, Nvidia made a bunch of money. Like, I'm not questioning that. I'd be crazy. Of course. However, 16% of their latest quarter revenue was one customer.

44% if their first half of fiscal 2027 was three customers. five customers make up 70% of their accounts. Can I jump in right there? They're good customers. Whoa, well, can I jump in right there? Are they? Because they don't name them. Well, we know who they are. Do we?

Well, we might not know who's 16 and who's 12, but we know the batch of companies. But hear me out, that's their immediate customer. Those customers have millions of customers. And that's what this is. In other words, you don't have a hyperscaler spending on Nvidia chips to serve. itself, except in case of meta. Right.

They're serving Fortune 500 businesses, governments, sovereigns, etc. Those are their customers. But I just, the point I made about Microsoft's AI revenue is where this comes in. So, fiscal 26, which just ended, $34.33 billion of AI revenue. 24.1 billion of that is OpenAI. That means that selling GPUs to anyone else other than their large fail son is a single-digibillion business. It's, Microsoft is the apex predator of software sales.

They have hundreds of thousands of resellers. They have tens of thousands of salespeople. And they can't scrape together more than single digit billion dollars of sales for co-pilot and all the stuff. And this is on top of the fact that GitHub co-pilot, one of their only successful AI products, has now gone token-based billing, which means people before able to spend $40 a month, spend $5,000 of tokens. Now the token-based bill.

Now it's usage. Gone. Usage-based. That business is dead. And on top of all of that, wow. $260 billion in Kappex and all of that Fortune 500 demand, like whatever it may be, is what, $10 billion a year? Saty, a dollar would say, give us a couple of years.

Why do I have to give, sorry, but the point is, is like, we've given him a few years. And this is the best he's got for us. We have a light, Open AI is a liability now. Like, that's, they are material enough that Microsoft actually has to, I think, what's like 27% of them? It's actually a deeply load-bearer company. They have to consolidate their financial.

Yeah, yeah, yeah. Yeah. Yeah. Yeah. And the other thing is, is Microsoft's only getting paid if OpenAI can keep raising money. And Open AI, it has a shrinking amount of people to raise it from. Their last round, $122 billion. Only $12 billion of that came for venture capital and the very same asset managers who were funding data centers.

110 billion was Nvidia, 30, soft bank 30, which put them into financial diastrates, and they're still there. And Amazon, $50 billion. that's not going to happen multiple times. Or if Nvidia invests again, they said they wouldn't. They literally said the words. If they invest again, it's bad times ahead. So Microsoft's growth is dependent on open AI and now anthropic spending. Outside of that, they don't have significant demand.

This isn't even an opinion. This is a Bloomberg article that spelled this out. And if you go and look at... Was it Bloomberg opinion? No, it was Brodie Ford, I think. But it was, and it was also based on the actual earnings documents. Microsoft doesn't have the demand for its own AI native products from its own customers. At the scale that they would need to justify, like, I'm not saying there's zero demand.

They have found, but this is also after them harassing their customers and the massive sale, I must be clear, Microsoft sales teams are, there's only one level higher of more annoying, and that's Oracle. And Oracle is a whole other problem. But the point is, is, if this was a case where 30% of their revenue was up in AI, I would be like, all right, maybe my book was kind of falling apart. 70% of their AI revenue being Open AI is existentially bad. They have $250 billion of Azure spend committed to.

That's so much, Microsoft was $331 billion, I think, last fiscal year. But now they've divorced themselves from the exclusivity agreement with Open AI. Which only makes it worse because now Open AI needs to spend 138 billion over eight years with Amazon. 20 billion with Cerebrus, 22.4 billion with Corwe, 300 billion plus with Oracle. They're spending, this is the greatest thing. Steven Drew from UBS, he had an analyst note. They're spending like $12.5 billion a year at Google. I didn't know this until very recently.

It's weird that's not being reported very well. But the thing is, is Open AI, Sam Altman has one real talent, and that's signing his name. My man loves signing contracts. And the thing is when the problem that everyone has right now is you can only rethink a quarter, two quarters in the future. You can't, because even like, oh, it's going to happen in 2028. Nvidia will get to 600-something, $674 billion, I guess it would be, based on consensus, for fiscal 28.

So, Nvidia will get there because right now, Nvidia makes lots of money. It's not a problem now. To get there, Nvidia is going to have to basically take three to five customers. They're going to have to find way more debt, like way more, because the price of debt is increasing, but also, Nvidia just bumped up prices 17%. The cost of memory skyrocketing. And on top of that, you have a shortage of talent, shortage of electrical grade steel. There are all sorts of things stacked against Nvidia here.

They could pull it off, but I don't know how they're going. The actual money that needs to get raised is astronomical. So, Daniel, chart five. Ed, if somebody said to you two years ago that this is going to happen, you would be saying the same things that you are right now about how implausible it would be. So what we're looking at for people that are listening is the trailing 12-month net income. for Nvidia, just screamed past Apple. Uh-huh.

But how much of that is equity gains, though? What does that mean? As in their investments in Anthropic, OpenAI, they've invested in like Corweave, Nebius, Iron. They have a bunch of equity investments that help bumping that up. To be clear, I was wrong back in 2024. And why?

Because I was fucking naive. I was like, the market's a sensible place. Microsoft and Google, Amazon, Meta wouldn't spend hundreds of billions of dollars for no goddamn reason. Wait, so you were wrong because the market's stupid? I was wrong because I was stupid. I was wrong because I was naive about how the world works

and I've had to take a harsh... You didn't think they would take it as far as they've taken it. Exactly. And I didn't think that the debt system would support it. So yeah, I... So what do you think the market is getting wrong right now? Because I think everyone knows that Open AI is a potential problem. Look at Oracle stock. It's down 70% still on the mat. Even as software stocks are bouncing, Oracle has caught no bounce.

So what do you think the market is misunderstanding that you're seeing? The thing I said about the centralization of data center demand. Open AI, I think, is like 60% of all AI data center demand. Without open AI, there is not another open AI sized spender. And the only reason there is an anthropic or open AI sized spender on AI compute is because of the availability of venture capital dollars. So without that, you can't, like, the demand isn't there. The market isn't seeing that because it hasn't happened yet. The quarter to quarters think they're doing where they're like, well, the money's still coming in.

Because what, I kind of said this already, but the amount that Anthropic and Open AI has to spend has to increase by like 10x. Well, their revenues are increasing. And those are real revenue dollars. And they're both going public in the next, call it six months. Well, you just made a face. Anthropics doing $60 billion. No, they're not.

Run rate. Go, go, go. Run rate. This is the biggest scam of them all. A, what, ARR run rate as opposed to last quarter. That's the first part of the scam. So ARR used to mean annual recurring revenue, referring to if I have 10, 100 million contracts, I've got like a billion dollars a year. Easy, peasy.

That's annual recurring revenue. Wow. That's SaaS. I probably f***ed up the math there I realize, but whatever. Moving on. And how they're booking their revenue you probably don't like. Well, now they're calling it run rate. And run rate can have, Anthropics has never defined this.

Even Bloomberg in the 65 billion story didn't define it. It can mean four weeks times 12 or four weeks times 13. The problem is, is Anthropic and OpenAI are both annualizing token spend, which is not a, it's not a software subscription. It's not recurring. A customer could spend $1,000 this month, and then spend $50 next month, especially if they're moving to open source models. Sorry, that's 100% true. But every publicly traded corporation that says anything to Wall Street about its spend, the only thing you ever hear is we already ran out of our compute spend. and need to spend more.

Yeah, that's because Anthropic and OpenA are taking up most of the compute infrastructure so there's not enough. There is more demand than there is right now. The question is how much more? Because I don't think that, I mean, I kind of estimated in a newsletter a week ago, I think there's about $22 billion of compute demand.

Sightland climate said back in February that it was 190 gigawatts in planning, that's about $1.5 to $3 trillion a year in compute demand. they'd need. There is a complete economic mismatch, and it's stark. But you don't like ARR run rate because it's just an extrapolation. It's a snapshot of a period in time. And then pulling it out to like a year or worth. Yeah, so we also don't know what the period. We don't even know what it means. They never define it.

Anthropic has used it several times in their own announcements. They never define it. It could refer to, hey, we just released a model and everyone's trying it. hey, we just did a new feature and everyone's buying a new subscription. Oh, on this day, we lost a bunch of subscriptions. So we'll pull the period over here so we look bigger. It is not a trustworthy measure of a business. And the fact it gets accepted is an insult to investors intelligence. So who's the idiot here?

Is it the CEOs and the boards for all of this CAPEX? Is it the equity investors? Is it the people loaning the money to build these data centers? Like who is not seeing past their nose right now? Lenders, all of the investment in AI data centers terrifies me to my core. It's so bad. These things are like miniature cities. They condense city of Bristol in England, using this in the carpal, about 1.25 billion square feet,

takes about 800 megawatts of power. Stargate Abilene, Open AI Data Center, 1.2 gigawatts of power in 998,000 square feet. So they're condensing a city's power into a thousandth of the space, and that thing is way behind schedule. each one of these, even the smaller ones, are like... Is that the Oracle one in Texas? Yeah, it's one of the ones in Texas. And all of these data centers are very ambitious. And the bigger they are, the more ridiculous they are.

Gigawatt data centers are a new idea. Also, a gigawatt data center is a campus of smaller buildings, just to be clear. But they're new. They're brand new. And they're trying to build so many of them in a power grid, kind of put together with like staples and tape and such. Which is, I mean, and the thing they always say is, oh, it's helping us upgrade the power grid.

No, it's not. it's strategically placing power in a place that you can maybe rent it to someone. And the problem is, is the for everything to be... But is it nefarious, or do they just not understand the demand as well as you do? Little, little... Here's the thing. Because it seems like they're the ones taking the risk. It's nefariously ignorant.

I think that their ignorance is so harmful. Stargate Abilene, the information had a report about this. The Blue Owl agreed to invest in 10 minutes. That's how the due diligence is going, boys. We're just 10 fucking minutes. Come on. It's just like, and that is for a project that will cost tens of billions of dollars. And it's just, why is no one putting the time in? Wait, hold on.

There's no way that's true. It's, I will glad. 10 minutes? Maybe it was 15 minutes. No, no, I can get you the article right now if you'd like. I believe that the article is real, but I just don't believe that that is actually real life. Have you read anything about Blue Owl? I don't believe it in two seconds, man.

I'm sorry. Like, having seen what Blue Owl has. does. Yeah, I would believe that. And also, this is a mania. And number go up. Hey, the biggest the justification makes sense if you don't think for a second. The biggest companies in the world wouldn't spend a trillion dollars for no reason. I got to build one of these things. I will make so much money. It's, it's gambler logic. Let's go through, let's go through some of these numbers. Daniel, you want to do chart one? So back on Nvidia, just because this is like, well, I call Jensen Wang the chairman of the AI Federal Reserve.

Basically, he's the person that comes out and makes everybody feel better. Pretty much every quarter. We get the hyperscaler earnings and then there's a three-week lag and then Nvidia comes out at the end of the earnings season and ratifies everything that we, okay. So just like on its face, $96 billion in revenue. We know that the cost of a data center is approximately 50% on chips. Yeah, more now thanks to memory, but yeah. Okay.

That's 106% revenue growth at 75% gross margins. Yeah. Invidia right now, I mean, this is real money. Yeah. So your issue is not whether or not the spending is real. Your issue is the source of where the capital is coming from to buy from from Nvidia is really what's in question. Yes.

And just to be clear, the source of the revenue and the money into all this stuff, real. The money going into AI software, which this is meant to prop up, very small when you remove the subsidized stuff. Every AI startup is subsidized. Every single one loses money. Up until fairly recently, Open AI and Anthropic let you burn thousands of dollars of tokens for $200 a month. I think you still can but not as much.

These companies, when their customers are exposed to the real costs, they shrivel away. And just very simply, if they thought they could charge the real cost of their services, they charged a real cost of their services. They wouldn't do this weird dance. So the problem isn't finding, well, I mean, actually it is becoming one now. This, a lot of this was, I think that the one big customer, the 16% one, I think it was SpaceX. So remember, SpaceX, a bunch of debt.

All of this, bunch of debt. Other than Microsoft at this point, they're all having to take on debt. And this is the thing. Everything, it's not actually about real demand because they can't pay this out of cash flow. They're clearly not making enough money to pay this out of cash flow. and they would say, oh, it's the new industrial revolution or some such wank. Well, it's clearly not going very well. I'll work out in the future.

When? Where's the real demand? Where is it? And so they're increasing. And because everything's getting more expensive, both the debt and the things the debt's buying, they're going to need more and more and more money at a time when the system's kind of straining itself, but the only thing the system wants to invest in is the dumbest thing possible.

It's actually, more I spell out the worry, it gets. When did Anthropic first on their revenue? I think it was March, 23. What? When was, when did they want? When did they first book their dollar, first dollar of revenue for Anthropic? It was sometime in 2020. Yeah, maybe.

They found it in 2021, but I can't remember when they actually. But you say that like you're sick and tired of saying where, when it's going to come. You don't, you don't have any open minds in this to the fact that it genuinely is early. How much of the demand that we're going to see are we seeing now? Because I think it's a tiny, tiny fraction. Well, well, people all of a sudden, stop spending on this? Customers.

So there's a guy called Nick Suresh, who has a really good blog called A.I. is eviscerating global decision making. And he's talked to, he's a tech consultant, gifted software engineer, talks to a lot of CEOs. A lot of this is, you ever see the death of Stalin?

Okay, so a lot of people doing make one. Was he on Netflix? No. Well, maybe, actually. But nevertheless, there's a bit at the beginning where they're sitting around Stalin's corpse

and going, he's so, he looks great. He looks very healthy. Oh, yes. A lot of this is people saying, oh, AI is making me 100x more productive or whatever. It's not true. Most AI integrations fail. All right.

So I want to give you a counter to that, though, because I listen to, I'm going to say 20 calls every earning season. And usually the stocks I own, but sometimes just stocks that make a big move. And I just want to learn, like, why all of a sudden are people really bullish or really bearish? So one of the stocks, and I happen to have been talking about this today, so it's fresh in my mind, is Airbnb. be. And Chesky comes out and says AI is the best thing that ever happened to us. And then proceeds to cite specific examples, one of which is that last quarter, 45% of all their customer support calls were closed by AI with no human involvement

whatsoever. Okay. It boosted cash flows. It reduced expenses. Did they say by how much? They gave the numbers, yeah. That's good. So, it's good.

They also moved to open source. they very, Chesky very specifically Not paying the open-on. He was actually one of the early ones. So for me, that's the thing that I'm most bearish about. And we'll get into that a second. I'm not, so here's where I don't agree with you.

Yeah. I really don't think this is going to stop on a dime as an AI user. It's going to slow down. At a company, every week we're finding new things that we can plug into AI and do better than we did the week before. I'm curious. We have to manage 90-some-odd employees, 4,000 client households. There's a constant orchestration of operations activity that has to take place between Salesforce and our portfolio accounting software and the trading that we're doing.

And the general idea from our standpoint is the customers don't give a shit about our AI use. This is operations. How do we run the firm better? Right. How do we do things faster, more efficiently? and that's what I think every company in America right now is trying to do. And we are increasingly hearing from them one by one on Q&As with analysts saying, why did we have an upside surprise?

Well, glad you asked. We invested in AI probably in 23 or 24. And now all of a sudden we're able to tell you specific things that we're doing with AI that are improving our earnings. I don't think that part's going to stop. So I think it will start. happening as much because I think a lot of it is make-believe AT&T being the worst one. Well, if they're lying then...

With AT&T, here's the clever thing. So they had this article in the journal saying 80%, 90% savings on using open source. Then you go down the article. It says 80 to 90% savings in some fee, some functions. And then the next paragraph is AT&T uses, has thousands of things. It has plugged into AI. So it's like, oh, so some things in some things may have saved you some money. this is the problem.

A lot of this, the reason I brought up Nick Suresh's article is because he makes a point where a lot of people are doing this because they're afraid of how it will look if they don't or indeed the management is so cooked that they must see AI because it's the thing they're... I agree with that.

I definitely agree that there's a lot of that comment. And here's the thing. There is a point we can come to, which is even with that being the case, is that going to be the revenue sufficient to keep open AI and Anthropic alive, who are both unprofitable?

Or is that going to move on to, I don't know, a hosted GPU system with an open source model. There are so many ways, like, Anthropic and Open AI must keep growing, and they must keep growing so much faster. This is not even close to where they need to be. They need to be able to afford $440 billion in the next three and a half years. So a lot of your bare case rests on that idea that the two most well-known foundation models are being relied upon by the S&P 10 to deliver.

over there, no. Okay. I don't dispute that. But what about Gemini then? What do you do with the Alphabet Gemini situation where they do not require OpenAI or Anthropic to spend with them? I think Sundar Pichai's asking that question based on all the press I've seen around Gemini. Gemini is just another LLM company. The problem is, is that as much as people love to pretend, oh, OpenAI Anthropic is special, maybe they like Sol, whatever. For the most part, people don't really give a shit which model they use. And each model, the differentiation between them is kind of, why do I need this? And then there's a new better one.

You try this. And every time you move it, there's a bunch of harness changes and prompt changes. The actual functionality of this stuff is so malleable and volatile that I think it's exhausting to engage with on the regular. And actually talking to pro-AI people, the few that I talk to, they experience this. There's an exhaustion around it. You constantly have to keep up. And they kind of rationalize it like any bad relationship. Oh, the worst waste.

Like you're saying, Open AI releases the newest model. I must insigrate this. Yeah, yeah. And especially if they run a company that has models facing the customer, they've got to have that immediately, which bumps up their cost. Because the thing is, the cost of the tokens may stay around the same, but the amount of tokens being burned changes arbitrarily between models. So for models, the same price, but it's burning more tokens.

It's like having a car, the miles per gallon is the same, but you're driving 200 extra miles. So my thing is, I could see a world where there are LLMs in the future, but I don't know how profitable is to, well, I don't know if it's possible to make these things profitable, even with custom silicate. I don't know if there's a future for these companies at all, or indeed where the hyperscalers will still offer these services at the current prices. They may offer them as a really expensive thing for certain businesses.

They may get that. That might happen. I think that's realistic. Gemini, for example, could become that. This might make you more bearish. Okay. There is too much, these things are too big to fail. There is too much riding on the success of maybe not just,

these two companies, but this whole idea not imploding, and bearish investors have learned over and over and over again, the system is rigged to work out and to go up. Now, that might make you more bearish, but I'm curious to hear what your take would be on that. So, the problem with this is people talk about a bailout, and I actually think private credit is the warning sign there, Mr. Walter to the operating room. Like, it just like, ugh. A lot of bad things happening with private credit. With AI and data centers, you've got multiple problems.

you have Open AI and Anthropic. You bail these out? Okay, but that doesn't fix the fact that they won't be able to afford to pay $440 billion in three and a half years. It doesn't fix the fact that it's very clear that Microsoft, Google, Meta, and Amazon, their overall growth is slowing. Otherwise, they would have never done this. They are slowing down, and 7% of Microsoft's fiscal year 26 revenue was from Open AI. I have to challenge that. Their cloud revenues are accelerating this quarter.

Based on Anthropic and Open AI's spent. But you're saying they're slowing. but they materially are not. When you remove the open A& Anthropics. But why would you remove? Why would you remove them? No, I'm saying you have. Sorry, you're right.

Okay. In the future, you can't count on it. I should have put it like that. You can't count on it in the future. When I say that 440 number, that is exponentially larger. Because, yeah, they are growing right now because Open Aion Anthropic, their checks are clearing. Well, all of the hyperscalers, all of them reported way better than expected growth numbers for cloud. In the quarter, they just reported second quarter numbers.

Other than better. So meta is a whole other basket case. Yeah, but I want to ask you about. This is actually a point, though. Okay. But it is true that AWS and Azure and Google Cloud are seeing accelerating growth. You're not questioning that. You're saying there's a big asterisk.

It's two companies that are ramping up all that spend. So what also happened in Q2? The massive token maxing. The massive spend on cloud compute. It times directly with them. Their numbers went up when Anthropic and Open AI. And that money is real. Some of that money came directly from the hyperscalers themselves. The reason I brought up meta was not, not to be argumentative, is you'll notice that

meta didn't grow like gangbusters. Oracle did. Microsoft did. Meta's not in that business. But that's the point of making. You'll notice that AI is not making meta grow massively. It's just the ones who get the money from Anthropic and Open AI. And this is the problem. The problem is that for now, everything, you'd look, a quarter, two quarters in the future. Wow, money not, money good, money here. Well, the meta bulls, and I'm not one of them, but the meta bulls would argue they're not rent and compute. They think they want to prioritize it to improve reels. They could at any time do an about face and say, oh yeah, by the way, now we're a data center and we're going to rent compute.

Yeah. If they do that, I think you would agree with me. That day, at least, the stock goes up. Yeah. Okay. They haven't done it yet. They've rumored, though. They haven't done it yet.

But they're a core business, which is. serving ads is in better shape than it's ever been. That's not, they had their first decline in users. No, no, no, no, no. The core ad business, like the profitability of their ad business. Right, but I'm saying that if the users are declining, though, that will decline too. Again, it's, I understand what you're saying. I'm saying, for now, it looks good.

And this is the eternal problem I've had. It's like, at some point this goes wrong and the reason I'm no longer giving timelines, because I was wrong. At some point, everything goes wrong. But this is the way this. goes wrong is astronomically bad because it's not just getting back to the bailout thing. It's not just, oh, Open AI and Anthropic have to survive. They have to become 10, they basically have to become the size of Google or Microsoft and spend more than they make

every year just on those two or three companies. So you can't bail that out. On top of that, if we build, let's say they cut back the process and they only build a quarter of the data centers. We're still talking $600, $700,000, $700 billion of data center demand that needs to exist that won't because most AI startups are unprofitable. I mean, all AI starts are unprofitable, so where's the money going to come from? Do you bail out the data centers? Do you bail out all of them? What happens to them afterwards? How do you bail out the fact that the point I'm making around cloud growth, which is most of the growth for Amazon, Google, and Microsoft, though Amazon has a surprisingly good ad business comes from cloud, and most of that growth

comes from Anthropic and Open AI. Without that, their growth story ends. And they actually think we'll eventually... Do you think they'll be able to come public? I think Anthropic will. I think Open AI is going to have the problem of Anthropic going public and having... I think their economics are going to be bad. I've seen Open AIs.

I've seen their actual numbers. They're bad, bad. Bad meaning, yes, there's a lot of revenue growth, but the expenses are ramping up even faster. Yep, $13.07 billion of revenue, loss of $20,000. $1.9 billion. If it does, okay, SpaceX losing a lot of money too. Yeah. If it does manage to go public and get a reasonably warm reception on Wall Street, either or both of those companies, and they do raise what will probably be record-setting IPO of, you know, share sales, they do raise that kind of money. Does that alleviate some of your concern on the sources

of all the data-centered demand going forward? I mean, in the short term, because $100 billion, So it does push it out, though. It pushes it out. Because they raise capital and they can keep spending. But the thing is, with these cloud compute agreements, you have to do prepayments. So that money's going out the door. Anthropic is rushing this IPO because they had this first two-quarter window where token maxing was happening. Everyone was saying burn as many tokens as possible.

And they'd moved all of their enterprise customers, so businesses over 150 people from paying 200 bucks a month or 150 a month to paying put tokens. Their revenue exploded. It was great. And this was at a time when the CEOs had all got concussions, they'll spend as much as possible. So Anthropic needs to go public, ideally before they have to really have the books for Q3, but they can still show the growths there,

even though most companies are now cutting back. They're talking about open-source. So you think the data center's customers, the Fortune 500, will actively pull back on their AI spend now that the pricing has changed, and they'll look at it and say, whoa, whoa, whoa, whoa, whoa,

we need to have more of the handle of the budget on what we're letting our people do on AI. Yeah, and I think that they've hit a ceiling. I think there are some that was... You think it's now? I think they've already hit a seat. I mean, Sam Hortman said on stage, he was saying, yeah, it's a huge issue for our customers.

He's so good at this. The price, and Anthropics Fable... It's a huge issue for our customers, not for us. It's just like, when you need a go to reassure you, just going out there. But the thing is, Fable, so the expensive, the most expensive Anthropic model, the one that was so that had the marketing of being banned by the US government for being too powerful.

Not true, but still, they had it, has kind of petered out according to RAM at 11% of market share. And that's bad because that's the only way you really grow revenue now, because it's clear that there are some businesses that will spend a lot of money. Metas talked about spending $10 billion. If they're spending $10 billion a year on Anthropic, the story came out today, that's also really bad because Mark Zuckerberg changed the name of his company to META mere years ago, just dump that shit in 2023. Like, this man does not care. He has complete control. He will stop doing something on a whim.

That's a really bad situation. Can we do a little bit of meta stuff? Yeah, sure. Okay. So I know this is, let me ask this question. My understanding of your story is that you started out representing technology companies. Yeah. Helping them with PR, helping them with messaging.

And while you're doing that, you start to become. disillusioned with their intentions, the way they think about their users, the rapacity with which they are willing to monetize their user bit. So that's like a lot of your origin story. You're almost like the guy who saw too much. Yeah. And I would say I was lucky in a lot of my clients because I would, I've been saying since like 2013, like, journalists don't like profitable companies.

And I've tried to work with profitable companies. Just a really easy start. Or if I work with unprofitable ones, I need to hear. were a very convincing, tangible story as to why. Don't do it anymore, but still, that was the thing. But yeah, when I would see the companies that got press, and they were just these big pieces of shit like clinkle. Klingle, you ever hear?

Klingle. I'm still bullish on Klingle. I'm a Klingle head. No, but that was one where it was like, oh, it's going to use Sonic waves to send payments. Died within a year. It's a big loss. Color, an app where you could share photos

with the people you're closest to. Dead. All these things kept dying. They kept raising a bunch of money dying. They didn't have a business model. And everyone just went, ah, Pobody is Norfolk. Like, moving on, and I saw this again and again and again.

And really looking, what actually really got me was Clubhouse. So I don't know if... Oh, I remember that one. Yeah, because everyone was gooning out over Clubhouse. They were like, oh, this... Well, we were all stuck in our houses and looking to talk people. A16Z was big in that, right? Yeah, and the A16 did the classic con.

If they got all the people they knew, did that Samuel Jackson or something? It was so strange, but everyone was at home and they were like, this is going to replace radio. I'm like, holy crap, you don't respect your radio hosts. Because these people, go, yeah, so, yeah, I'm here from a startup. But everyone's like, this is the biggest thing. Everyone needs to have a clubhouse strategy. I'm like, you people, you people. That was weird.

Do you know that that era coincided with the SPAC boom and so Clubhouse and selling SPACs to people? Oh, yeah. It became synonymous. All of the financial conversations on Clubhouse were like, you know, a 28-year-old who just managed to launch a SPAC who's never experienced the cycle or anything. There was actually a show called What the SPAC and they invited me on it and it was a huge mistake for them to have done that because in a prior career as a retail stockbroker, I sold 100 spacks.

So I knew they were all going to zero. Oh, God. And I told everybody in Clubhouse and I was never invited back to What the SPAC. All of the people involved in what the SPAC all went to zero. Not to alleviate everybody from their responsibility, but that was a very bizarre. All of our brains were a little bit broken. Sure, but I think it was more the media that really got jokified, because I just saw people being like,

Clubhouse is going to be the new thing. Everyone's just going to listen to Clubhouse. Everyone's going to do Clubhouse, it's going to do all court. And I was like, are you people? And then the Metaverse happened. They thought it was like Snap or Instagram or it was just going to take off. Then the Metaverse happened. I'm like, holy fucking shit, you're either describing something that doesn't exist or video games.

What do you mean? This looks like, this VR thing looks like shit. What are you doing? And people were like... We were laughing at that in real time. And that's the thing. Oh, yeah, of course, 2020. They were selling virtual land.

I said, I can't. Virtual real estate. I saw that. Virtual real estate. I can't have this. And I saw it being written about like it was real. I'm like, oh my God, what it is like having the glass system they live. Jesus fucking crap.

And so every week I would see on the new outcome like, Jesus, and then crypto happened. And then NFTs happened. And I would read things in the newspaper telling me about crypto and NFT. And I went, there's a reason people keep being. There's a reason now why bubbles keep happening. and it is the media. It's a mixture of other things. They're not the only problem.

There was bubbles before the media. Yes, but the media still help them. Social media. The Fed. Let me finish. The point is. Everybody's in. But everything is a social contagion, ultimately.

And social media, partly from being at home to your point, allowed it to exacerbate these bubbles and speed them up. And AI was special because of that 2022, 2022, 23 depression. Plus, it gave everyone something to do. You could become an AI consultant. You could have an AI consultant. you could put AI in your software. You could sell AI infrastructure.

Everyone, the bankers had something to sell. We had new things. Are you not at all impressed with the products? Not for the amount of cost. Okay. So if they were, so an open weight model that you pay almost nothing for it and it gets close enough? I mean the cost to get here. So a trillion dollars.

The infrastructure cost. Infrastructure and the environmental cost and the social, the horrifying things the media has done scaring people about jobs. Disgusting. I agree with that. But the thing is, you can't divorce the technology from that. What it does today is unremarkable compared to its cost,

and it's about what I would expect. And remember, this, just even if you take out the infrastructure, this still costs tens of billions of dollars. The Kimmy K-3s of the world, they still distilled from, they used the same training data, they still used a bunch of compute, they still used innovation from other people, and I don't care about anyone copying anyone in AI,

that's kind of the point. But the thing is, there is this massive cost to make this thing that, great, it's kind of helpful, I guess, somewhere, in a specific thing, but you can't really rely on it. You can't really trust it. You have to look at everything it does, or you have to double check it didn't muff something up, which doesn't sound particularly intelligent.

So back to matter. So the reason I went into your backstory is, because I want to ask you, do you hate these people? Sounds like it. I absolutely hate it. You personally. I think Mark Zuckerberg is a scumbag. I think he's a piece of shit.

He cares, he does not give a rat fuck about his customers. He has turned Facebook and Instagram into a hellhole. He has deliberately, I've seen documents, I've published them, where Mark Zuckerberg requested 12% perpetual growth of all the numbers. They at one point for years were growing, I think, time spent on app, not realizing that engagement was dropping because they were measuring the wrong thing. They have made that product, like all of these companies, worse to increase growth.

Google search, Google search worse because Prabagar Ragavan, the VP of ads at the time, did a coup to take over from Ben Goams, who, in emails with the Department of Justice, published these as part of the antitrust trial, said like, hey, this is going to make the user experience worse because you just want to increase queries, which just means people searching more times to keep them, to show them more ads. This is why Google sucks now. And Microsoft... You've called this the rot economy.

Exactly. So you wrote this big piece in 2023. Corey Doctor wrote talked about in shittification. Yeah. Your version of that, it seems like, is like this idea where all of these products start out good with, like, the right intentions. let's create a searchable index of the entire internet. Let's connect everybody's family members with photos.

Like, it's a great idea. And also probably if we get enough people using it, there's some money. And then your point about the rot economy is like five years goes by. The numbers aren't numbering. Yeah. All right, we have to shear the sheep now. The sheep have gotten fat. They're all on the platform.

We charge them nothing. We subsidize this. now we're at this like critical mess of people let's turn them upside down and dump their pockets out. And what's crazy is for a few... But that's how you see it. Yes. Because they had ads on Google Search.

They had ads on Facebook and Instagram. Now that's the whole first page. And now it's everything. And you can't have a default chronological feed. And everything is like a Nashia painting trying to find stuff. They move the buttons around. Because they're experimenting on their users. There was a Guardian article eight years about it where they took 700,000 people

and they just started emotionally abusing them with the notifications they gave them, the stories they've heard them to see what would happen. That should be illegal. But these companies, it's not just they want to grow, it's that they have to grow perpetually, and they will do anything. It's why, honestly, they've run out of ways to squeeze other than price increases at this point. And that's why Microsoft has done multiple price increase the last few years, same with Google, Meta.

I think it was Google and Meta changed how ads work around how the first bid thing. I think that they basically have a way of juicing more money out of it. raising the bid. The point is... It's sort of capitalism, though, also. It is, but at some point it eventually goes completely haywire and explodes. Like, that is what... That is the ultimate thing.

You eventually get to a point where you've chased out anyone at the top who cares about product or good stuff or happy customers, and all they're there is for growth. Andy Jesse, NBA, Sachi Nadale, MBA, Sondapashai, McKinsey, and MBA. Mark Zuckerberg, not MBA. Cheryl Sandberg, NBA. And the original growth team,

know had another one at Schmarth, they had Naomi Gleit, they had Javier Olivan, who's now the CEO. So, but like philosophically, these companies have shareholders. The shareholders are there because they expect the cash flows to grow, some portion of that to be returned to them in the form of dividends, buybacks or both. They want to see the earnings growth. That's like sort of the job of the people running these things. It's not to, it's not necessarily to be caretakers for the population of users. You mean, you mean provide a good service at a fair price? Yeah. So that's the thing.

Well, Instagram would tell. So meta would say, well, here's Instagram. How much do you pay for it? Zero dollars. They make 55 bucks a user. You pay for it an attention. That's what they do. Sure, but there is a way of doing that without making it horrible, which they have.

The point I'm making is, however you feel about this ethically, it has a natural endpoint where you change the company from making a product that you sell and you turn it into a growth engine. You turn it into something that's only about expressing growth. You hire people with that explicit intent. You direct them with that explicit intent. Okay. So what's stopping the users of Google, Amazon, and meta from saying,

I no longer like this and will not use it? Monopoly. No, I'm with Ed. I'm with you. I think what you're describing is how a lot of us feel. I hate Instagram. I love it so much. I'm addicted.

I have to use a physical brick device to stop myself from using it. I think we all feel that these places, are toxic and disgusting. I don't like the way that children were onboarded. That's where you and I are probably agree the most. There's just, to me, there's something different about a 12-year-old than a 42-year-old. I'm not worried about a 42-year-old. Yeah.

I want to sit there and scroll your phone for nine hours. It's not my problem. But I don't like the addiction of the, and there was a settlement this week. Yeah, but over 10 years. How much money was it? It's 12 billion and then maybe 17 billion. So is that 1% of Mehta's market cap, basically? I thought it was going to be an upfront payment,

and the settlement people are always such. 28 states, all settled at once? Yeah, apparently. Yeah, but what will this do? So, meta will enforce, the two-hour daily limit on Instagram and Facebook for children as part of a record claim. That's great news.

That's awesome. Now, I love those changes, but they should have made them pay up front because that would have been most of their income from the last quarter. But there is a point to this as well. Taking away the ethical concerns. You eventually engineer

your company around growth and you stop learning how to build new things that customers like. You find ways to twist customers' arms, but you don't really find ways to win them over. You eventually get into a mindset where you're like, what if I could just spend money and get money back without really thinking about it? AI is the perfect vehicle for that, in theory. If you were a company that hasn't really thought about user experience, you've got this magic box you can ask questions to. You could detach that to anything and you could answer any question.

Shit, you don't even need to come up with a product. It will come up with a product, right? This is the theory they had. It didn't work. But when you're not really thinking user-friend, anyone who's thinking, wow, I want to make a user-friendly experience, LLMs are a fucking insanely questionable product. Some people get value out of them.

Some people get no value. Sometimes it just goes wrong. Sometimes it makes stuff up. That's a crazy fucking product. I'm sorry, that's insane. The fact that people are so defensive and they say hallucinations have been sold when they haven't is insane. But all of these companies have become,

around growth, and so they went, wow, I can just buy these GPUs, and I'll keep growing, right? Meta was horribly wrong, of course. Well, you think the products suck? I love using Cloud. I mean, why? I find it to be miraculous, truthfully. I use it all day. And I'm not the only one.

But this is the thing, like, the fact we have such divergent experience is so interesting to me, because this is meant to be the new big thing. I, to be clear, love technology. I really do. So here's the what for it. writer like you. Yeah. I will never have AI write my words.

Because I think the only reason people read me is because I have a very specific way of talking and a very specific point of view. However, I don't necessarily want to use the blue links on Google's front page to go hunt down information. Oh. When I'm in the flow of writing, I'd much prefer to have, whether it's Gemini or Claude, have a co-pilot where I say, while I'm busy writing this, go double check these numbers for me

and cite the sources before I actually press publish. I use it as research. You use it. That's describing search. Search and research. And now what Gemini has done, or what Google has done, I should say,

Google saw this a year and a half ago was extinction level event if they didn't cannibalize their own search and put the AI result up top. I don't. I know they thought the way. I don't know how I agree. Well, I don't know what else they could have done

if people's searches started to move to chat GPT, they couldn't just sit there. But the thing is, you're describing search. But do you not find that? You're a great writer. You must find that valuable. Here's the thing. I use my terminal, for example.

BQ, not having to write BQL is fucking great. I can't do that. I don't know what that is. So with the Bloomberg terminal, they have this thing called Ask B. So if you need to look something up, it would run basically Python, their coding language.

There's BQL. And instead of doing that, you can go look up this, and it will generate the code and run it. So you can actually try it, take it, test it yourself. That doesn't impress you? It doesn't impress me for the cost. For like a trillion dollars and tens of billions of dollars to get to a better search?

But you think this is the end state? You don't think that... Yeah. I think that we're going to get iterations from here. There's a reason they move to cybersecurity. They're going to iteratively get better at code, which doesn't necessarily replace anyone or even necessarily help software engineers past a certain point, because it gets more multicatively, I can say that word correctly,

more complex and more hard to manage. And I think the big problem with LLMs I have is the distinction between when you are outsourcing work, like searching for something and thinking. Oh, I totally agree. And the thing is, I always do on that. Tons of people are doing the thinking one, which is a problem.

And no, I'm not that impressed because I used it to fix a thing with my kids' Minecraft game recently. Because I think, ethically, I should try and use this. And I have. And it's like it took half an hour of chasing its tail. It eventually got it done. and I probably couldn't have done that, or maybe it would have taken.

But it's just kind of like, okay, how many mistakes did it make? If I was paying on a per million token basis, I'd be pissed because I'd have probably spent $15. Are there workflows that you've been able to automate utilizing AI where you've said, okay, I still hate these people, but this is literally saving me time from some of my repetitive tasks? Outside of the Bloomberg terminal, no.

Okay, so we're in the other boat. Like, there are skills that I've been able to create on Claude, where I used to have to go to these five websites, copy, paste, do this whole thing. And now I can say to it, all right, it's Friday morning, you know what I need, go get it. I find that to be like, so I'm not a tech person. I'm a finance person. But for me, it's like miraculous. I also want to be clear about something.

LLM's in a vacuum when you remove the social cost, the environmental cost, the fiscal cost, all this. Interesting technology. I think if they'd have called them like library models, no one would be. getting a rock hard about it. But it's just like, they are, clouds, they're automation. Like, fine, great. I don't know why we have to talk about it. I don't mean here, but it's like, I don't know why everyone's sinking a trillion dollars into this. That's crazy. Because, well, there is a thing where if they think if they don't, you've probably heard this a million notes. Like the existential threat to their own companies, like if we miss this wave,

our existing data center business is worthless. If we don't rip out all these CPUs and replace them with GPUs and be able to do these calculations for these models, then AWS will go to zero and Google Cloud will win. They feel that way. I get that. And I think what's happened is maybe they've realized, but they now realize they have to double down because everyone else is doubling down. And I think when one of them pulls back.

Some cost. And they already spent too much. And that's the thing. Yeah. When they stop spending on AI CAPEX, everyone's going to go, great, you're done with this. Why did you do it? Like, that's the thing. And the moment they do it, it's just like everyone's going to like, brilliant.

What were you doing the last few years? Have you got enough AI now? What are you going to... And the spending literally can't stop because the chips burn out. Yeah. Well, that one, it's like, again, they obfuscate all this data. But I actually had a question around your workflows. Do you pay on a per million token basis?

No, right now, everything that I'm able to do is very cheap because I'm not Jane Street and I'm not running a billion calculations a day. We are working on Gemini. we're working on Claude. Are you paying a monthly fee? We're paying a SaaS-like, we're paying a SaaS-like fee at this point. Yes, but you're not paying.

And I know it could change. And that's kind of the point I'm getting to. It's like, also, this is an insane way to run a business. GitHub copilot had two million customers. And on June 1st of this year, Microsoft just went, pulled a rug out from under them. That is an insane thing to do to customers.

That kind of, it's the most rock economy. Why did they do that? Because they were allowing them to burn $5,000 of tokens for $40. But the thing is, why did you do that? And they would argue, oh, we thought it would get cheaper. Where'd you get that? Because it's never got, nothing has got cheaper. The cost of intelligence has come down, but the models spend more tokens.

And the models may be able to do more stuff. They should have to, considering the tens of billions. Like, I'm not saying they haven't improved, but it's like, what will be the first sign, do you think, if you had to guess? And I know you're not timing it. And I know you're not telling people to trade on this. what would be the first sign of the apocalypse for all of the spending and like the comeuppance or like when everyone goes uh-oh because here's why I ask you that question I genuinely know enough to know that I will never be the person that spots it but this is the new parlor game

people say one of these days meta is going to throw in the towel yeah and say that capex projection we gave you last quarter, don't rely on that. It's actually going to be lower. And then that will set off a chain reaction where the semiconductor stocks instantly lose 30% of their value. The utilities fall like penny stocks. And all of a sudden, like multiples compress, other companies pull back. That creates a panic.

Okay. So that would be like my guess is it has to come from one of the hyperscalers. None of them is showing any signs that they have any intention of pulling back yet. But is that the thing that you're looking for? That's an obvious one. I agree. Like, it would be a KAPEX pullback. I've been saying that for years.

Who knows when they eventually cry uncle? I actually think there are a few different pale horses. One of them is if CoreWeave cannot raise debt. CoreWeave as a Neo Clouds at the rent AI GPU compute. And their customers are Open AI, Microsoft for OpenAI, Google for Open AI, Anthropic and Meta. And they buy a lot of GPUs. And they are horribly unprofitable.

And they need, I think they've raised billions of dollars three times this year. They need to do that last time they paid over like 9.2 or something, like some ridiculous rate on it. And hey, I don't think things are getting cheaper in that realm. And also the chips are more expensive. So CallWeve needs to perpetually raise debt. In fact, everyone does. Ira Nebius, N-scale, who may be idea.

Or the neoclouds. They all need to raise debt. So they don't have a Google search-like business to fund this. They don't have a business at all. They just have, like, the Findon that calls them and ask them for $3 billion every quarter. So one of those companies, doesn't raise money. It's going to be they try and they fail.

I could also see a hyperscale of bond sale that gets even interest. Or like just above even. We just had an Amazon one that only got 1.6x oversubscribed. If that comes to like 1.1 and 1.2, if it gets like real thin, people will write articles like the enthusiasm for data center bonds are drying up. And I thought that was happening because the Amazon one happened, then Google did one that was 4 or 5x oversubsubs. I'm like, okay.

But the thing is, those are the kind of signs because the big naivety I had was like that some sort of ethical or social concern would change things or that the companies wouldn't keep doing something because they all knew that they had to and if one of them stops all of them can stop but the first one has to first no it comes down to money will the money be available and that's the thing because invidia's GPUs are so expensive and they just got more expensive and the cost of RAM and construction are going up you necessitate being able to raise billions of dollars that's actually a fair thing small amount of companies that can do so.

Most of them are bankrolled by Jensen Huang or dependent on Open AIA or Anthropic. So it's the reason, do you look at Oracle's Twitter account? I know this is weird. I'm going to go somewhere. No. So every time something goes wrong with Open AI, so I haven't seen it. So Oracle, when that story came out because Nvidia was going to invest $100 billion in open AI, it didn't happen. When that came out, Oracle did a tweet that was saying something along the lines of,

We have no concerns of our financial situation with Open AI. Oracle will just randomly post stuff, be like, Stargate Abilene's going great. And you can basically just take the opposite of whatever it's saying. That's because... What do you think runs out of? It's not SafraCats sitting there. God, no.

Saffircats, funniest thing I've ever seen. Oracle signs this massive deal with Open AI. Like a week later, she's like, see ya. Yeah. You're going to have two CEOs now. Let these guys deal with this. I'm going to go sit back here. Right.

And so Oracle also has to keep raising money. Oracle is still, I think, in the process of doing their mass share sale. But you must be heartened by the fact that the stock market is treating Oracle differently than it's treating Alphabet. Yeah. And the bond ratings agencies are certainly looking at Oracle. Yeah, but they don't have the stones to downgrade. They don't have the stones to turn. It's that chunk.

It's a notch above junk. No, they're notch above. They don't. Will they have the balls to make them for an angel? Hell no. that the fucking ratings agencies rated core weave debt investment grade because it was connected to a hyperscaler. Unbelievable.

But that's the thing. It comes down to when the money runs out. If Open AI fails to raise, if a major AI startup goes insolvent. So, have you heard about Plexity? What's going on? I haven't heard much about them in a while, actually. They seem to a feed of the way.

That's actually really good to know because Nvidia is considering investing billions at a $30 billion valuation. Why? Because Poplexity is one of the few companies that actually spends any real money. on GPU compute. The Plexi's death, what it's hundreds of millions,

but still, in this industry, that's a lot when no one else is spending much more than millions or tens of millions. So, Invidia is keeping them alive, because if they fail, well, everyone will notice. So the pale horse could either be, a major hyperscaler decides it wants out of the race, a major hyperscaler has its debt downgraded, a la Oracle.

Yes. Probably the most likely. More than certain. A major startup runs out of money that, like, everyone associates with the AI trade, like a perplexity. Okay. Or what was the fourth one? If CoreWeave and CoreWeave can't roll its debt. Basically, anyone who can't race morning.

Oh, also. Or an IPO bombing. IPO, of course. SoftBank. That would be a big one. If an IPO did not go well. If SoftBank can't take a company public, SoftBank's liquidity situation is really bad. They are having to raise 20 billion in bonds because they raised 40 billion.

That guy's crazy. I love him. I love his goosema. His goose math rocks. Ed, where's the mania, though? Because I feel like at least in the equity market, a lot of investors are not treating these companies as if they're Teflon. Because Nvidia's 70% profit margins, stock trading 18 times forward earnings, meta 16 times.

Like, where's the mania? The media is in the data center construction of the companies themselves. And it's an activity bubble as much as it is a financial bubble. of the amount of these deals. Yeah, there is momentum that still exists that keeps everyone thinking this is normal. Oracle shouldn't trade it, whatever. If it's anything above $40, it should.

Oracle's revenue has been flat for 15 years when you adjust for inflation. Like this dog of a company, I'm surprised. They are the only one of them I'm certain would get bailed out. Stocks down, like 65%. It should go lower. It sucks. Oracle is a horrible company.

They also fired a bunch of people, and I know there's some sources, the deal with contract renewals, which is like the number one most important thing at Oracle. And if they're firing those people, they're just kind of disorganized. The data center thing that makes me the most bearish, I made the huge mistake of rereading a book called 1873 about the build out the boom and bust of the railroads. I don't think that this looks like dot com. And I don't think that this looks like the great financial crisis.

I think that is the best analog because of how physical the demand is. The build of the railroads was like extraordinarily physical in the way that whatever mega-wad data center you just described, like, it's that type of a infrastructure build-out bubble

to me. And I should not have re-read that book because it made me want to sell everything. GE Capital is the other comparison I'd make with Invidia. Invidia is acting a lot like Welch. They're doing some... Vendor financing. Vendor financing, but also with the

G-CAS, the airline side over there. Yeah. They kept... airlines alive constantly. They were a massive leaser. They were America's largest equipment leaser at one point to their customers. There were airlines that could only survive as long as G could lease them stuff. But another thing is as well is, this is a bit of everything.

Because the difference and the problem with data sensors, these things are not really useful for other things. GPU compute is not really that useful outside of this. And it's very clear that the only people that will pay a lot for it are unprofitable startups. if real businesses are not buying it at the scale they need to, and there doesn't seem to be a reason. In fact, there seems to be more reasons to buy less, open source models. But you've got Nvidia...

Sorry, self-driving cars, automation in general, humanoid robots. None of these... We're saying that non-linear processing is not useful for those things. I got to go to the bathroom. Carry on. Not at the scale they built them. If they were building $30 billion worth of these things, very different.

Okay. But they're building... what will require $700 to a trillion a year in income for something that doesn't need that. Look, we've had Tesla and Volvo and all sorts of other companies that have self-driving cutters, and what is it, Waymo. You have all of these things, and they didn't need all these GPUs. And indeed, having more of these GPUs has not changed, like there hasn't seemed to be a breakthrough in them.

They're getting better. I fucking love autonomous cars. I think they're fascinating. I love getting in a Waymo. I think that there are social problems we're going to have as a result of taxi drivers. I think that there's going to be real problems there. Yeah. I wish we actually do some socialism and actually help people.

I'm not exactly sure what all those drivers are going to do. But I also think... The only get out of jail-free card is this could take place over 30 years, not over three years. I think it's decades and decades and decades because the problem with autonomous cars is not the 99%, it's the 1% issue. And that means you have to expand slowly. But the point is, all those GPUs might be useful for that, but we've built so much of this. We're building so very much of this. And per Nvidia's own earnings, we're at a point where there are like three...

three to five companies that can afford this at any kind of scale. And Nvidia has to make so much more money. Everyone is in this situation. Open AI and Anthropic have to grow like crazy. Invita has to grow like crazy. I think Nvidia may have said the 70% thing because they're desperate. And I know that sounds crazy because they're doing mad numbers. But remember, they've been down.

People have been a little bit worried about Nvidia. They needed something exciting. Also, they got some concerning shit in that 10-Q. There's some really weird stuff. That statement they made about investment-grade partners that can pay between three months and one year, it's really weird. So you don't like seeing them become investors and in some cases creditors of their own customers. It seems like it's the classic sign of illusory demand or artificial demand, I should say. So to close this out, this is the question my, at this point, my entire audience would be screaming this at us.

Yeah. This is the thing they would most want to hear from you in closing. is there any chance that you're wrong about how this turns out? And if so, what would have to happen for you to change your mind and say, I probably got a lot right, but I missed the thing that actually mattered. And I was wrong. What are those concrete things? Because it's not just sentiment shift.

Just promise me the next time you have an AI bull, you ask them the same question. It'll be like next week. Please do. Please ask them the exact same question. What if you're wrong? But so, I don't have to be 100% right for things to be really bad. Oh, I agree with you. Really, like, Open AI Anthropic could become much smaller businesses

that hang in there through some messy accounting and some wing of Microsoft or Amazon and Google or some consortium and the LLM industry could be much smaller but not disappear. Those are feasible. There is no breakthrough coming that's going to magically change this. The AGI story is dead. Even they are trying to move away from it. Nobody can even define it.

Exactly. Jensen Wong said it was here and it's just like, oh, shut up. Jensen, come on, mate. But the thing is, the hyperscale growth story is so difficult for them to me. It's so great. Like, even what Jensen, if anyone can pull it off, it's Jensen, but to do $670, 690 billion next year, probably from the same customers, means that Amazon, Google, maybe Microsoft

definitely meta, are going to have to take out a bunch of debt. SpaceX, 100%. Amazon says they're buying 2 million GPUs. Really? How? How much money? you're going to have to raise your cash flow negative. All these companies are going to see their economics and I think meta and Amazon both guided lower than expected in Q3. So we're already

seeing some slowdown, how interesting. But the point is they are going to have to spend more and more and more and more money and more, even if they're buying the same amount of GPS they planned to six months ago. They're going to have to spend more because they're more expensive now. So there's no chance you're wrong is what I'm hearing. I don't see how this goes on as long as they, they expect. to be making hundreds of billions of dollars in compute revenue just from Anthropic and Open AI, I don't, even in the dumbest, craziest world, I don't see how the money exists to do that because that requires them to likely spend half a trillion to a trillion dollars in CapEx and have those data centers actually open.

How would you answer that question? Which part? How would you be, how would you know for sure that Ed ends up being wrong? What would be like, would it be the profitability of Anthropic and Open AI? happens in 28? Even then, are they profitable enough to be able to afford all the compute they've committed to? So even that wouldn't be enough to change your mind.

1.2 trillion is so... What would you say? I don't know. I don't know how you could disprove that or prove that he's wrong. Are you betting against any of these companies? Do you have any money in the line? Nope. I don't...

Here's the thing. I don't have cash in the market. People go, oh, he doesn't have skin in the game. I have emotional skin in the game. I don't know what to say. Oh, no, I agree. You have become the face of AI skepticism. you have as much in reputational risk as other people have in dollar risk.

But the other thing is, is the market is you don't invest in the market now based on good sense and reading about companies. You invest based on reading the smoke signals. And last year in September, there were like three or four open AI announcements that manipulated, SKHinex, Samsung, AMD, and Nvidia, four announcements where nothing happened and the stocks popped. I can't play in a market like that. I can't play in a market where companies do fake announcements.

Do you talk to hedge fund managers that are making active bets against these companies or they're long but they want to hear like your views? I occasionally do. Without like naming people. What kinds of conversations are you having? Similar to this? Yes, but I never get near the actual trades. No one ever talks what they're actually investing, which is great.

But no, but they might want to hear your train of thought. Like why shouldn't I have a billion dollars in fucking semiconductor stocks right now? And that's the thing. When it comes to just equities, I'm kind of like, if you are willing to ride the crazy train, go ahead. Like, if you want to read the smoke signals, if you're capable of seeing where the market will go and want to influence the market, believe for you, this is not my expertise. But if it comes to a long-term thing, especially if it involves investing in, like, the debt underlying a data center, I will just explain how the demand is illusory. There's not no demand, but it's a teeny tiny amount compared to how much we're building. It's definitely a hype cycle.

Even the Bulls would have to admit. There's a lot of hype. There's nothing after this. Nothing. There's no dot-com bubble-style fixer here. AIGPUs are not useful for other stuff. And also, if a data center is left incomplete, unbuilt, it's going to cost just as much in three years to finish.

AIGPUs are going to cost just as much to run in electricity, if not more. And if the power isn't finished, if the power never made it, or if you were running it using gas turbines, so nothing permanent, it's going to be even more expensive, depending on the cost of gas at the time.

And there's just not a, with a dot-com bubble, There's a bunch of server hardware that went out that was useful for building the future with like Amazon Web Services, for example. Amazon Web Services. The dark fiber was useful later. But energizing that fiber, using that fiber wasn't anywhere near as expensive. Amazon Web Services, by the way, between 2003 when it was created in 2015 when it became profitable, Amazon's total CAPEX was $29.7 billion adjusted for inflation. And that's the thing that I try and tell people because, oh, Amazon Web Services cost a lot of money. Yeah.

this costs way more. And that was all of Amazon's CAPEX, not just AWS. This is nothing, people, and I do the same thing. People want to compare things, people want to schemata they can look at and say,

this looks like this and that and the other. I just challenge people to say, stop relying on that. If you don't agree with me, then at least don't rely on their being a happy ending here and prepare accordingly. Because I think that people may expect

open AI today. They may not be surprised when it dies. I think some will. But they won't be, like there will be some people went, yeah, I kind of saw it coming. The data center's not being used and the data center debt not being paid. That is going to be what shocks people. Because people, because we're talking hundreds of billions, trillions of dollars, regular people can't think of that.

Like, that's an unknowable amount of money. It's almost to the point that it stops meaning anything. And you come to this logic of, these are the smartest people in the world. Why do you think they're smart? So the, all right, so the implications, though, of what you're saying, you realize we're not talking about a dot-com meltdown where the NASDAQ declines 80%, but the rest of the economy is fairly unscathed, which is actually what happened. Yeah. What you're describing, because the numbers are so much bigger, and because the private credit people are all in, and they are not NASDAQ stocks, these are banks and insurance companies. And all right, so what you're describing does have the ability to literally take down the economy in the way the dot com cap-x blow off top could not have.

And the problem is it's hard to quantify how bad it will be because of the private part of private credit. There's also stuff we don't even know that people are doing. And that's the thing. There are these kind of like guys just like, yeah, 10 billion, whatever. I don't give a shit. Blackstone, pissing money up the wall. And they're funding it with insurance money, private credit money. Everyone's talking about this Mark Ward situation.

cool. I wrote about this three months ago with the private credit situation. Like, I think a sixth of insurance annuities are private credit funded. We've got public pension funds. We've got private pension funds. We've got private pension funds. Everyone's in. Yes. I don't know how widespread the damage is. I truly don't. I don't want to be like, it's going to, but it seems really bad, and people are alarmed about it now. And I mean, Pablo Tori's done a great job. And Nick Nemeath has done an amazing job pulling out the private. Yeah, Nick's fucking wrong. Oh, it's public of this world.

Nick's really great. Have done a good job really illuminating this. And Sam Copleman over at Hunterbrook, legend. We know. We know, Sam. And they've done a really good of elevating this. Still, people don't connect that to data centers like they should. And they don't connect it to the fact that open AI and anthropic don't even need to die.

They don't even need to stop growing. Like, they could grow. They won't. But, like, they could grow like five times as big. And they still won't have the demand for these things. And what happens if the third. if the thing doesn't get built or runs over... When have you heard of a building project that didn't run over budget?

And these are the most ambitious infrastructure projects of all time. And so you've got private credit with their wonderful underwriting quality, where most just people winking at each other, and you've got them funding these big, unstable, difficult, complex projects that can get locals getting in the way. They're cutting the tax breaks in, Arizona, Illinois, in somewhere else, I can't remember. Texas, Pennsylvania, both showing opposition to them.

And they need these things to finish and the customer to exist to make any money on the debt. And with all the weird senior tranches and all that, who knows where they'll actually be at the end of them. And they don't know because they're investing in a Blackstone or what have you, the Blackstone Infrastructure Fund. And it's just very dangerous in a way that makes me really angry because it could have been avoided. It could have. But everyone got a might bit too greedy. and Jensen as well has done so well on this.

But here's my real Nvidia question. And I've been waiting to bring this up. It feels like the right time. How and when does Nvidia realize revenue? That is a very interesting question when you actually look at their earnings. Those accounts receivable have grown. Nobody special fund is 55% sequentially.

Like, what's going on there? Why are your accounts receivable? Why your daily sales standing growing? Right. We like to think of them as backlog. But that's the thing, though. Why is this extending as you're making more money? Is that because you're reaching the edge of when you can get money quickly, or at all?

When does that? How much more affordance are you going to have to make your customers to grow to $690 billion? Also, technically on an accounting level, you can ship something and you can count the revenue and no money can pass. Oh, no, maybe you're feeling generous with Call Weave. You really have to disclose that. You might eventually, but Collick Crest might just not send it out as quickly. They're going to push because we don't have an essential. right now. They're going to push the absolute limits of accounting shenanigans here.

It's already looking really weird. Inventories growing as well. Invidia has $30 billion of cloud compute agreements, as in to rent back their GPUs. They have $25 billion of data center lease agreements. Why are they... You sell to the data center, Jensen. Why are you also renting the day? You're just building a data center to feed yourself money? and this is the ultimate point of the root economy. This is eventually they wanted to create something

where the only thing was just handing yourself money, but the money increased by draining debt. Right. And there are limits. And I don't know when we hit them, but we will. And I worry that when we do, it's just going to be so abrupt and horrifying. And the era afterwards, because...

Do you want to be wrong? Do you hope you're wrong? Yeah, because you do see... Like, you do seem to... I don't want to say enjoy wheeling off all of these problems that you have with it. But it does seem like you're enthusiastic to see this thing.

I feel like you're going to be cackling if this all comes down. No, so there's that it wasn't a bit in the big short where it's like unemployment drops, 1% and 30,000 people die. Right. I'm not going to, I might be satisfied. I'm not going to be happy.

Like, I don't know if it's come across. You're not going to take personal pleasure in seeing you, but you will get some intellectual satisfaction. Yeah, I mean, sure. Because out of having put all your personal capital on the line, so to speak, and been right. But the thing is, is like,

regular people are going to suffer on this one. Retail investors, though, are barreling into this what we're at the highest point of retail leverage and history. Like, we're seeing a lot of people, look at what's happening with the Cosby. It's going to be like there in America. Like, there's something very scary about that. That chills me. And I don't know.

I know I sound enthusiastic, but I am like, this comes from a place of, like, worry. Because we could have stopped this. We can't now. And now it's just how big is the system going to let itself grow before the system itself runs out of cash. Because there's enough cash, there's enough, but there are limits to how quickly you can mobilize it

and where you can put it. And everyone's asking for more. Everyone is asking for more. And more and more and more. On that uplifting note, did you at least have fun on the show today? I would love to come back. This is one of the most fun are added in podcasts.

Listen, we feel that these conversations need to be had. And there will be people that are fans of our show that will listen to this and say, don't ever have that guy on. Don't ever have that guy on again. But what if not you're right in the world comes to an end, but what if there is a minor catastrophe and people listening to the show were able to kind of understand why it's happening in real time

because they listen to it. So we do put bear. We've had Sam Coppulman on the show. I wouldn't call him a bear, but he is skeptical. He did the blue energy piece back on. And I've had Nick on the YouTube live stream. So I think these conversations are important to have. I don't have to agree with everything you.

say, but I am learning from you and I hope you've learned from us. I have. Okay. All right. You learned how to do a proper podcast. There we are. All right. Dude, this has been, this has been fun. Such pleasure. Fun might be the wrong. This has been intellectually stimulating. There we are. Yeah. Is that cool? Nice use for us. All right. Guys, thank you so much for watching. Thank you for listening. Where can people go to learn more from you, Ed? Tell people the URL. Where's your ad. Dot at for the newsletter, subscribe to the premium. And of course, betrothline.com for links to the podcast. Dude, you are a virtuoso at this. You crushed it, so thank you.

I love podcasting. All right, thank you. Thank you guys for see you soon. Two and five Canadians will hear the words, You have cancer. That's why every step and dollar raised matters. On September 19th, join thousands in Toronto for the Princess Margaret Cancer Foundation walk.

Challenge yourself, friends, and family to walk 21 kilometers in support of life-saving research. Together, we can carry the fire and help create a world free from the fear of cancer. Register today at pmcf walk.ca.ca.