"This Is A $10 Billion Industry Pretending To Be A Trillion Dollar One" — We Had To React

You know those wild tech headlines that make you wonder if the whole AI boom is smoke and mirrors, or if we’re on the verge of a revolution? Today’s episode is for every single one of you feeling FOMO, confusion, or flat-out dread about where tech is going and how the world will look in a few years. We’re digging deep into the BEAR CASE for AI—what’s fact, what’s hype, and what might bankrupt half the industry before our jobs even change. We break down the no-nonsense take from Ed Zitron (the guy calling B. S. on AI hype), paired up with hot-off-the-press financial receipts, juicy tech drama, and a reality check from Palantir CEO Alex Karp. We’re talking margins, runaway spending, why enterprise buyers are majorly sketched, and how every new tech revolution leaves a trail of broke first-movers in its wake. If you’ve got money in the market or just fear the next bubble pop, you NEED this convo. Welcome back to Impact Theory with Tom Bilyeu. In today's episode, we dive into the heated debate around the future of AI, spurred by Ed Zitron's bearish perspective that claims the industry is overhyped and built on shaky financials. We'll explore the staggering costs and mounting debts at the heart of today's AI giants, the skepticism around their business models, and whether major players like OpenAI and Anthropic can ever become profitable. We'll also examine Palantir CEO Alex Karp's insights on how enterprises are reacting to AI's current shortcomings and the necessity of proprietary solutions to protect company data and foster innovation. Get ready for a deep-dive into the economic realities, risks, and road ahead for artificial intelligence—where technological promise collides with financial reality, and only time will reveal if we're witnessing the birth of a true revolution or the next big bubble. Quince: Free shipping and 365-day returns at https://quince.com/impactpod Whatnot: Download the Whatnot app today and get free shipping on your first order. ATT Business: Switch to AT&T Business at business.att.com http://business.att.com Shopify: Sign up for your one-dollar-per-month trial period at https://shopify.com/impact Incogni: Take your personal data back with Incogni! Use code IMPACT at the link below and get 60% off an annual plan: https://incogni.com/impact Quo: Try for free PLUS get 20% off your first 6 months at https://quo.com/impact Paleovalley: 30 for $36 https://bit.ly/PaleovalleyIT Netsuite: Right now, get our free business guide, Demystifying AI, at https://NetSuite.com/Theory https://netsuite.com/Theory Sign up for my AI Masterclass: https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=you https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=youtube&utm_medium=social tube&utm_medium=social https://tombilyeu.com/ai-masterclass?utm_campaign=TBS-Livestream&utm_source=youtube&utm_medium=social Check us out wherever you get your podcasts: Spotify:https://open.spotify.com/show/1nARKz2vTIOb7gC9dusE4b?si=a8daffd2bf1f48fd Apple: https://podcasts.apple.com/us/podcast/tom-bilyeus-impact-theory/id1191775648 Do you need my help STARTING a business? Join me here inside ZERO TO FOUNDER: (https://tombilyeu.com/zero-to-founder) Get the exact systems, mindset shifts, and principles that built a $1B brand delivered straight to your inbox every week. Subscribe for free (https://tombilyeu.com) https://tombilyeu.com Check out our Video game - Project Kyzen: (https://projectkyzen.io/) https://projectkyzen.io/ Catch Me Streaming on Twitch - (https://twitch.tv/tombilyeu) Link to IT discord:https://discord.gg/TZKJ2etPbT Tom's Favorite Things List: https://amzn.to/41Ftt7e For Business inquiries: [email protected] mailto:[email protected] FOLLOW TOM: Instagram:https://www.instagram.com/tombilyeu/ Twitter:https://twitter.com/tombilyeu YouTube:https://www.youtube.com/@TomBilyeu See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.

• 0:00Speaker 00

  Let's talk about the moment your side hustle becomes a real business. Now, I see way too many founders stay stuck running their business like it's a hobby. And then they wonder why nothing ever feels real. It doesn't feel real because it's not. I've helped thousands of entrepreneurs build real businesses through Impact Theory University. And the first step in the path to building a legit business is registering it. And Taylor Brands will help you do just that in a few clicks. An LLC, a limited liability company, is an official business structure. It's inexpensive to form and very simple to maintain. It's how you open a business bank account. It's how customers know they're dealing with a real company. And Taylor Brands doesn't stop at the filing.

• 0:45Speaker 00

  Their dashboard tracks what comes next based on what your business actually needs. So EIN number, registered agent, licenses and permits, bookkeeping, business banking. I tell every founder the same thing. Just make it official. Click the link in the show notes to check out Taylor Brands and get started on your business today. This is a paid advertisement.

• 1:07Speaker 01

  The most effective people at work aren't working harder than everyone else. They're working smarter inside better systems. Superhuman Go, from the makers of Grammarly, is the AI chat that works inside every tool you already use. Always ready and already aware of what you're working on. It's a teammate whose only job is to help you be better at yours. With Go working with you, you can show off what you do best. See what Superhuman Go can do at superhuman.com. That's superhuman.com.

• 1:37Speaker 00

  This one is crazy. So if you guys don't know Ed Zitron, he's a bear. So he is very negative on AI. Thinks that basically the industry is a big nothing burger, that people have blown it all out of proportion. This is one of the more interesting ones. We're going to go deep on this, so.

• 1:53Speaker 02

  Ed, thanks for joining us here. So, you know, we don't really know at this point in time, right? What the actual financial picture of these companies look like. And they wouldn't be the first with bad financial profiles to go public.

• 2:08Speaker 03

  Well, they'd be the first to be this bad, other than WeWork. And even then, this is so much worse than that. OpenAI burned $20.9 billion in 2025. That's their auditive financials that the FT and I reported. And the problem with these companies is their margins are getting worse. And they actually, their costs increase linearly with their revenues. There is no proof that they can improve their margins. No amount of specialist silicon or supposed Vera Rubens will bring these costs down.

• 2:31Speaker 00

  So here's the big thing that people do not understand about what's going on in AI. Now, listen, the final conclusion that Ed makes, I actually disagree with. I don't think the industry is hopeless in the way that he does. I think it is the revolutionary technology. I think it is literally going to change the fabric of society. I think even if it asymptoted right now, just the things that it can do in the medical industry alone are transformative. The amount that it's impacted our business is transformative. But he is absolutely right in terms of the mismatch between the revenue that's coming in and how much money these guys are losing because of the infrastructure build out is absolutely insane. I've done deep dives on this topic, man. If people do not wake up, they're going to get blindsided. I am nervous about owning tech.

• 3:14Speaker 00

  I will just tell you that right now. That's just a me thing. You do you. But people are not being honest about the historical pattern that repeats. We've seen this over and over. I'll talk more about this as we go. But we've seen this pattern play out before. So he is getting people to look at something that is so important to look at. The mismatch between revenue and debt accumulation, the losses that they're taking is terrifying. And people should be very worried about the rate at which their revenue is increasing.

• 3:42Speaker 03

  It's clear that people are wising up to the problem of generative AI, which is there's not really a business there.

• 3:48Speaker 02

  So that was something that was discussed yesterday on Squawk Box. The comments from Alex Karp, the Palantir CEO, making their rounds. I think we have a clip teed up. Take a listen.

• 4:00Speaker 04

  The general way these things were sold, and these people are, Sam and Dario. There's nothing more fun than debating Dario in private. So I'm not throwing shade at them, but something has gone completely wrong. And the basic view among enterprises in this country is I'm going to chill, lax and waste my time with tokens. I'm going to get no value and they're going to get my IP.

• 4:25Speaker 00

  We had talked about this previously, what Alex, Alex is banging the drum. So for people that don't know, they're Palantir. Palantir has created like a middle layer is maybe the right way to think about it. I believe it's called ontology. And their whole idea is, okay, wait a second. I'm a gigantic company. I've got all this proprietary data. I want to take advantage of AI. And when I put it into AI, you're talking, your token costs are absolutely ridiculous. So I'm already getting hammered there. And so you see corporations pulling back. And on top of that, the companies have learned basically the business model of these companies that are using it. They're training on all the data that they are supplying the agent and then launching competitive companies.

• 5:09Speaker 00

  They did that against Figma and they're doing it now in medicine as well, in medical. And so it's like Dario is screaming to basically the creators of AI and saying, you guys need to understand, you've got a terrible reputation now with the people actually using this technology. You've got to get your shit together. And so while I think that feeds in a little bit to what Ed is saying. Dario, not Dario, excuse me. Alex from Palantir, Karp, is making it clear there's a there there. But these guys have got to understand, they've got to put this obfuscation layer in the middle where companies can feed their data to something that they trust, isn't storing it, retaining it, going to mimic them.

• 5:54Speaker 00

  And then they can basically change the model to match their weights. Now they've created something proprietary. This is huge. I don't want to go on too long about this. But understand the future of AI, I guarantee you, is you have to somehow, some way, be able to turn that AI into something that is proprietary to you so that gives different answers. With AI, we're really trying to mimic humans. And just like I can't put anybody random in Drew's seat, you can't put anybody random in my seat. It fundamentally changes the nature of the show. AI must be the same. The second you can just swap it out because they're all the same, you lose. And so that's the point Alex is trying to make.

• 6:30Speaker 02

  Enterprises are balking at this idea that they're going to be paying all this for tokens. They're not going to be able to control the data usage and how the data is used. And they're not able to really measure ROI. What do you make of those comments?

• 6:42Speaker 03

  I mean, I do agree with him, but he's right. These companies, Anthropic and OpenAI, they encourage waste. They, he actually said that these companies don't charge on outcomes. They don't charge on success because you can't with large language models. They're inherently hallucination prone, as proven by OpenAI themselves. And so they encourage waste. They want you to spend a lot of money and then they try and pilfer your ideas. Anthropic already tried it with Claude Design with Figma. And I believe they have a relationship with the company and Dylan Field even said he was shocked by it. And what's crazy is as well is their copies of these companies not particularly good. They don't make great software. You'll notice that both Anthropic CEO Dario Amade and Sam Altman have both said, we can't wait to see what you build with this. Well, that's because they don't know what you can build with this.

• 7:25Speaker 03

  They want everyone else to do their innovation for them, spend as much as they can on tokens and then take whatever's left.

• 7:30Speaker 00

  In what universe would you expect the telecoms who built out the internet to know what's going to be born of this? You look at the technology and it's a question of what does it allow you to do? But how that's going to play out over time, when the internet was first born, I certainly didn't think, oh, somebody's going to come up with the iPhone and then somebody's going to come up with Uber. I never would have seen those jumps. And then it's going to be, of course, a question of, well, in what specific way are you going to execute against that thing? And everything is going to be different and many of them will fail. And it takes, you know, one person to just hit it right in terms of what the world wants at that moment. And so the fact that Sam Altman doesn't know what you're going to do with AI or Dario Amade doesn't know what you're going to do with AI, obviously, as it should be. But these guys have to start thinking about themselves.

• 8:14Speaker 00

  These guys being the AI companies have to start thinking about themselves differently. And so one is intelligence is rapidly going to become a commodity. So where am I going to get my intelligence? And so again, this goes back to Alex Karp, which is such a banger thing. We talked about it when it first came out, but it was a really brilliant drum that he was beating. You must create that layer in the middle where people can make that unique to them. That's going to be the thing that makes this work. That's the thing that I think that Ed is missing is just because the fundamental business model of what AI is today isn't going to be the thing, it's not going to survive in the same way that the railroads went bankrupt, the canals went bankrupt, the internet went bankrupt. And then it was the inheritance generation of investors actually built on top of it.

• 8:59Speaker 00

  It is almost certainly going to play out like that with AI. But that doesn't mean that AI isn't real.

• 9:05Speaker 02

  Then what do you think should happen? Do you believe that these are companies that should be nationalized? As we read in the intro, there's discussions reportedly that the administration and potentially some sort of sovereign wealth fund would take a 5% stake in OpenAI. I mean, is that a situation that you would like to see happen or do you think that's more disruptive?

• 9:23Speaker 03

  No, I don't think there's any need to. They talk about, I think Sam Altman said, well, we can share the benefits of AI and the profits of AI. What profits? What returns? That 5% will have to get congressional approval. The sovereign wealth fund is still an idea at this point. But fundamentally, large language models are not the future. The only reason big tech is investing in this is that they've run out of hyper growth ideas. They don't have a next iPhone. They don't have a new Google search. So they've put over a trillion dollars with trillions more to come.

• 9:50Speaker 00

  So when people say that LLMs are not the future, basically what they're putting forward is the argument that pattern recognition is not going to be the game. I don't know that that's true. And this is really going to come down to whether or not, because what AI has already done, this is the part that I really want to get people on board with. Whether AI will make the initial investors their money back, probably not. That would be ahistoric. Okay? So when you look back, when revolutionary technologies come on board that have massive infrastructure build outs, they almost always bankrupt the first generation of investors. And it's what's called the inheritance generation that comes along. That's like, oh, cool. You built out all the fiber. Oh, cool. You built out the railroad network. Got it.

• 10:31Speaker 00

  And then they build successful businesses on the back of that because they don't have to deal with the debt, right? Same thing, US right now, under all this debt weight, we've got to get out from under it somehow. You're sure as hell not going to pay it back. So. AI definitely faces that problem. You have a cost problem, but China is already showing us that you can drive the cost down dramatically. Even some of the companies themselves, I think Elon has talked about driving the cost down. And so there, you can look back at history and say, when these technologies come on board, in the beginning, people don't know how to do it well. They don't know how to make it very efficient, but over time, these are the exact kind of problems that get solved. And so we can expect that on a long enough timeline, it is going to get cheaper.

• 11:13Speaker 00

  And so if you just sort of go, eh, we may lose, you know, a lot of money to people that have gone and invested in this, but the technology has been invented. You're not going to put that toothpaste back in the bottle. And even if we, even if LLMs asymptote, meaning they, they stop getting better. So if they stop getting better right now, today, the capabilities are already extraordinary, and they don't seem to show any signs of stopping. I don't know about you guys, but these things keep getting better. There, you're not going to see just like this only up phenomenon where without any either building things out bigger breakthroughs in the way that they actually do the algorithms

• 11:57Speaker 00

  that make these things more efficient. It's not just going to go up to artificial superintelligence in the next six weeks. But in terms of the things that we use here at the company, like new things keep coming online. For us, it was a big step forward with the ability to edit blueprints. The amount of our C++ now in the game that's being written by AI, it's just like, I keep seeing that we're using it for things. Now, it falls down, there's no question, you hit limits and it's like, oh, it sucks once you get to this part. But I would say that makes it great. This is the, what I call the golden age of AI, because you're still necessary. If it ends up completely erasing us as being needed, that isn't necessarily going to be fun, but that's a problem for another day. All right.

• 12:40Speaker 02

  So you think AI, LLMs overall, it's a dead end. You see no business model here. You see no utility from AI whatsoever.

• 12:48Speaker 03

  I see this as in the future as a boring hardware-based business, the kind of the Oracle licensing hardware model. I think this is a 10 to $30 billion TAM industry pretending to be a trillion dollar one. And because everyone's propping it up, because they have no hypergrowth ideas, everyone's just kind of pretending it's the kayfabe of the tech industry that really has run out of ideas.

• 13:06Speaker 00

  I think that's the, the sharpest thing that he says is that. That's really going to be the question that plays out. Like, are, I really think that the people investing in this are out over their skis in terms of, you're pulling, in the case of SpaceX AI, if I remember right, you're pulling forward like 100 years? I mean, it was some insane number, maybe I'm misremembering that, but it was an absolutely ridiculous number of years at their current revenue rate that they were pulling into today. I did a deep dive on this, if you want to watch it. I think that one is crazy. I don't think that that's going to make people their money back in anything approaching near term. And we'll see if they end up surviving all of this. This is a debt game.

• 13:50Speaker 00

  I think what you're unfortunately going to live through is because this is following the traditional trajectory of a revolutionary new technology with a very expensive infrastructure build out. What you're going to see is we're going to run the 2008 playbook where we try to hide all of that debt. The banks are going to diversify. The banks are going to diversify into your insurance, they're going to diversify into your index funds, they're going to diversify into your retirement program. That's going to be the way that the banks like mitigate the risk on their side. It's entirely possible that that debt gets packaged up and given a high credit rating, but it's certainly going to be, it's already being sliced up and hidden in other places in the economy. And so that's the part that worries me is making sure that people understand what they

• 14:35Speaker 00

  own, because I think we're in for a rocky ride in terms of, this is, this is a super known pattern that these kind of technologies go through. And so it doesn't mean that the tech doesn't end up continuing to have incredible application, but his idea of like, this is going to be tied to the hardware, that is very interesting. Now, I think ultimately, it is, it's going to be tied to the hardware, because that's how they build these gigantic brains. But I think that we've already seen the evidence from the US and especially from China, that you can make these things more and more and more efficient, which drives the cost way down. So I think tokens are going to get a lot cheaper. I think the ability to store more tokens in memory at once is going to get more massive, which creates more applications.

• 15:17Speaker 00

  I think we're going to find that LLMs don't work for everything, but it's so, remember, it's so good at coding that when they dropped Methos, they treated it as a weapon system that couldn't be exported to our allies, because it could hack basically everything. So these are things that have extraordinary capabilities, and it just becomes a question of can we make them more efficient over time? And so Ed is putting his finger on something that I think is very important. Myself and countless other people have not been paying enough attention to how can this go wrong? I started to a few months ago, just looking at the investing and the debt and being like, oh, there's a potential problem coming here. I think Ed is maybe a little unjustified in just how bearish he is, but he's making

• 16:01Speaker 00

  people face something very, very difficult.

• 16:03Speaker 05

  One more thing about this, he said he's completely right. Is that when he was talking about the total adjustable market is only 10 to 30 billion at this trillion dollar thing?

• 16:10Speaker 00

  I don't remember when I said that, but there's a couple of things that Ed is very right. One of them is that this is currently very much tied to hardware. He's very right that right now, the rate at which they're spending money, just losing money outright versus the rate at which they're bringing in capital or revenue is, is so insane. I don't know what people are thinking. Like, listen, as a, a very paranoid investor who does not think he sees the future well enough to go all in on one thing, I'm captain diversify and I totally understand this. Not how you get rich. The way that you get rich is to have a concentrated bet and it pays off. It's all true. That's also how you go broke.

• 16:50Speaker 00

  So we're hitting pause for a moment, but there's plenty more ahead, so don't go anywhere. There's a reason Quince sells cashmere for $60, and it isn't the cashmere. It's real 100% Mongolian cashmere. It's the same stuff luxury labels charge a fortune for. Quince just works directly with ethical factories and cuts out the middleman. So there's no need for a markup and no logo tax. The $60 goes to the actual sweater, not the fancy name on the tag. That's why everything they make runs 50 to 80% less, and it's not just the cashmere. I ordered a few things myself, their fleece joggers, a couple of their tees, which I wear obsessively.

• 17:33Speaker 00

  First thing I noticed was the quality, soft, well-made, built to hold up. They make down jackets and wool outerwear now too. So upgrade your stuff to things you'll actually wear. Find your next fall favorites at Quince. Download the Quince app for app exclusive offers or go to quince.com. Get free shipping on your order and 365 day returns, now available in Canada and the UK too. And when Quince asks where you heard about them, let them know it was Impact Theory. It's the best way to support the show. Let's talk about what it means to actually be visible in today's day and age. When you build something public, whether it's a business or an audience, anything on social media, you become a target.

• 18:16Speaker 00

  This is something that legitimately keeps me up at night, because our data isn't just on the obvious broker sites, it's on old news portals, public forums, obscure directories, places you would never think to look. That's where Incogni's exposure scanner comes in. It hunts down your personal information across the wider web, not just the 420 plus broker sites. Incogni already removes your data from automatically. They can't harm you if they can't find you. Go to incogni.com/impact and use code IMPACT for 60% off an annual plan. Try it risk-free for 30 days.

• 18:55Speaker 00

  That's incogni, I-N-C-O-G-N-I.com/impact and use code IMPACT. Let's talk about your sales pipeline reviews. If your CRM is half-empty fields and notes nobody finished, every review turns into guesswork. I've seen it so many times in my own sales team, and that's where today's sponsor, Pipe Drive, comes in. An intelligent AI-powered sales CRM that is loved by growing sales teams. Pipe Drive just launched Meeting Intelligence built right into the CRM. Before the call, it pulls the deal history and past conversations into one brief. During the call, it records and takes notes. After, it drafts the CRM updates for your rep to review and approve.

• 19:40Speaker 00

  The record stays complete and you see what's actually happening. Switch to a CRM built by salespeople for salespeople and join the over 100,000 companies already using Pipe Drive. My link gets you an exclusive 30 days free instead of the usual 14-day trial, so make sure you use it. No credit card or payment is needed. Just head to pipedrive.com/impact to get started today. That's pipedrive.com/impact and you can be up and running in just minutes.

• 20:14Speaker 00

  Thanks for sticking around. Let's get right back into the action. I'm like always super paranoid. I assume there's something I'm missing. Things are not going to go the way that I think they're going to go. And so paying attention to that discrepancy between the extraordinary levels of debt that these companies are taking on. Man, I don't get how people aren't like, yo, this is like really scary. And then another thing that he said that I think I won't say that he's completely right. I'll say, ooh, you got my attention. I want to look more at that. Is are these tech companies out of ideas? Like, where's our new iPhone? The thing that hits like the iPhone hit, right?

• 20:58Speaker 05

  Yeah, we're going on 20 years.

• 20:59Speaker 00

  Yep. So is that going to be robotics? That's like when you start getting into Elon territory. So Elon is thinking big. You can say that he's a fool, but you certainly cannot say that he doesn't have a vision for how this all ends up paying off. So he's switching over like if you want to think about, okay, wait a second, these hyperscalers are out of ideas. It's like, no, they're not. We're just at a this I'm channeling Elon here. No, they're not. We're just at a different part of that curve. And so now we're leveraging AI to build things like self-driving cars. I have ridden in a Tesla that took me like, I don't know, five or six miles up into the Hollywood Hills without the driver ever once touching the steering wheel. I could not believe it. So I was like, whoa, this stuff is getting good. It's going to keep getting better. They just need more training data, more training data, more training data.

• 21:45Speaker 00

  Elon almost always misses timeline, but ends up being right in terms of what the technology can do. And what he's saying about robotics, man, listen, the technology may not come as quickly as we want. But the robotics is being driven by AI. And so it's like these things, one of I don't think he is this over, is that the?

• 22:03Speaker 05

  No, we got a little bit more.

• 22:04Speaker 00

  Okay. So I don't remember if he talks about it here, but he has done interviews where he talks about how the guys building the big data centers, he was like, if they had ideas, they would just use the compute themselves, but they don't. They're starting to like lease it out to other people. And so the big data center that Elon built, he is like leasing that out to Anthropic. And so he's like, bro, that tells you like, nobody needs this stuff. These guys built this all on hype. They, everybody thought this was the next big thing. Capital is always seeking a return. And now this is me. When you inflate the currency, the way that you're doing it, you force VCs and stuff to go, we've got to have somewhere to put money. And so they'll, they, he doesn't say this explicitly, but he's intimating this

• 22:49Speaker 00

  that the VCs will trick themselves into thinking, no, no, no, this is real. We just got to like, keep putting money, keep putting money, keep putting money. And because of the abuses of government deficit spending, you are making the number go up. So you look like a genius. And so it's like, oh my God, this is all like incredible. This is going to go up forever. Then you get someone like Ed is like, wait, hold on. What are you actually building with this again? Like, remind me, you just built out a massive data center. Elon, you're saying you got to go to space and you got to build more, but you're leasing your unused cycles to another company. Like, why the hell would you expect me to believe that there's really a there there like?

• 23:26Speaker 00

  So I think it is very important to be able to step into Ed's shoes and other bears and articulate why they're so nervous. And if I were going to put a really fine point on it, it's really two beats. You've got they're leasing out their own intelligence, if you will, because they don't have anything to do with it. They're not pointing at anything. We're out of ideas. And then you've got, bro, the cost versus return is so insane. You can't do that. You can't do that. It is guaranteed to hit a wall of problematic debt before the revenue reaches something that makes it self-sustaining.

• 24:02Speaker 06

  Quote, "Some of our customers, OpenAI, may be highly leveraged. We may experience risks of non-payment in our dealings with such parties. Are we going to see this from a market standpoint first in shares of hyperscalers?"

• 24:15Speaker 03

  I think we might, but Oracle is a particularly scary one because they are building 7.1 gigawatts of capacity just for one customer. And they even said in their annual report that the risk was they might not get paid. OpenAI only loses money and I think I estimate it's like $75 billion of revenue annually that they will have to pay for the full Stargate data center project in annual compute revenue. OpenAI can't afford that. And if they can't, Larry Ellison can't afford to pay back those bills. And Oracle's stock will be in jeopardy along with the margin loans that Mr. Ellison holds. It's genuinely dangerous and it's dangerous across the board for Neo clouds, for hyperscalers, for every associated party of the LLM industry.

• 24:52Speaker 00

  Not only, not only is it dangerous, it is historically to be expected that we're going to obliterate. I mean, Drew, is it in the trillions? It's hundreds of billions of dollars of investment that is likely to get wiped out because they won't be able to make the payments on the debt.

• 25:11Speaker 05

  Is this, is this some of the reason why like X money might be a thing that Elon is taking like, all right, I need to kind of hedge my bets. See if I can do some type of, I'm just thinking as anybody, that's the first new idea in the X AI in the space as I was talking about.

• 25:23Speaker 00

  Well, that's definitely not a new idea. China's been handing us our ass on that for God knows how long.

• 25:27Speaker 05

  I'm just saying, as an AI company with all the four of them together, they're going this way. Meta is now making do it yourself, video game apps to kind of come up with a new Instagram for video games that's built with their compute. So they're trying to do these side quests, I guess. I'm just saying, is that something that they're trying to do to kind of fill the coffers a little bit to help the revenue or is it?

• 25:48Speaker 00

  Yeah, everything that they do is trying to make a little bit of money, but the, the numbers here are staggering. These guys, the odds that they'll be able to make their money back by building a thing themselves is effectively zero. That's why I think right now the right way to think of this is I'm a hyperscaler. I build data centers and I build the intelligence itself. And then the intelligence is a commodity that we sell to other people. And the bet that we're making is that everybody is going to want that intelligence in the same way that people want electricity. So if you think of, of the, the actual I in AI as that's what we sell. We sell intelligence and we're neutral. We're not, this is Alex Karp trying to convince people, stop.

• 26:32Speaker 00

  Don't be a company that actually builds the business. Create the intelligence, create an obfuscation layer so that the people can go in and make this their own thing. That their AI is unique to them. It gives outputs that nobody else can get. And you're not trying to pilfer that. You're like trying to be a highly trusted, always on provider of intelligence. And when I look out into the world, basically everything will be imbued with intelligence. There's no reason to have a dumb chair if you can have a smart chair. There's no reason to have a dumb TV if you can have a smart TV, right? So it's just going to keep getting better and better and better in terms of integrating into your daily life.

• 27:13Speaker 00

  But right now there just isn't enough appetite for what AI can deliver. So to me, this just feels like a timing thing. So the base assumptions that I write on the back on are you're not going to get the revenue fast enough. So the early investors are going to get destroyed. By the way, that's going to be terrifying to the global economy, but I'm setting that aside for now. So that's one thing. The second assumption is that people will figure out in the same way they figured out how to integrate the internet in ways we never would have thought of in like the late 90s. Forget it. We didn't see this coming and we're at that stage of AI. We're in the late 90s of AI compared to where we're going to be, you know, 2007 becomes the iPhone moment. When does Uber come along? 2014, 2015, something like that.

• 27:58Speaker 00

  So it's like, you got a lot of years of people being like, well, what about this? No, that doesn't work. What about this? Well, that doesn't work. And then slowly it's like, oh, wait, it's AI plus this like, oh, now that my microwave has it, now I can make, right? And it becomes like, it needs to start permeating. It needs to get out there in order to be a thing. But one of the things that, one of the base assumptions that Ed has is that the LLM as the progenitor of the intelligence is just always going to be dumb. Now that's me putting words in his mouth, but I think that's pretty close. So he thinks that's going to tap out. And if he's right, and that this is like as good as it's ever going to get.

• 28:38Speaker 00

  And it's like, yeah, you get some efficiencies and but you know, you can't just turn your whole company over to AI and walk away. I get why he would think AI is cool, but it isn't this multi-trillion dollar industry. I don't share that assumption. I think that even if you can't just keep making it smarter and smarter and smarter and smarter that just integrating what we already have, just making what we already have more efficient and cheaper will transform things. I don't know if he's not close enough to the medical side and doesn't understand things like protein folding. Be very interested to ask him that question because just seeing what's happening in medicine

• 29:16Speaker 00

  based on some of the things that are coming out because AI can just look at an unimaginable number of patterns. And so the fact, I forget how many tens of thousands of proteins it has like folded in its mind to know this goes with this, goes with this, goes with this. That will play out in drugs and all kinds of things in ways that I don't think we yet fully understand. And it's already happened. That's the thing. You don't need another breakthrough.

• 29:43Speaker 03

  I think it could be the Neo clouds because with this rumor of Meta selling their AI capacity. I think companies like Cor even especially Nebius and Iran and cipher mining and all of them, Terra Wolf as well. They are all very, they're basically outgrowths and they're subsidiaries of Nvidia. Nvidia is now, according to the information, going to be paying them to rent back their GPUs when they install them in the data center. This is the, this is something that only happens in an industry without diverse and real demand.

• 30:11Speaker 07

  Meta is an interesting case in this instance because they don't as yet have the third party cloud services revenue propping it up. But people point to their own top line, having accelerated in the first quarter of this year from like 23% to 33%. Presumably because they're implementing these AI tools or somehow making their platform more productive along those lines. And at the same time, it shows you the fallback option. We're creating all this capacity. Even if it's not a business model, the capacity is there. We've gotten the kind of the power and the compute in place. Maybe that helps society down the road.

• 30:46Speaker 03

  Well, the thing is with matter is I don't necessarily agree. AI is not driving Meta's revenue growth. The fact that they have an effective monopoly in social media is Microsoft, Google and Meta and Amazon are all doing a funny little, I don't want to call it a scam, but it's a trick. Where because their other businesses are still growing, but they never disclose their AI revenues. Everyone conflates that with AI driving their growth. In reality, their other businesses are growing and AI is losing the money across the board. You'll notice that neither Microsoft or Amazon, who both share their run rate of AI, will share the actual AI revenues. That tells you that these companies are afraid, public companies love good news. If they had good news, why wouldn't they share it? That's because they've only got bad news here.

• 31:26Speaker 02

  So what do you think then is the straw that ultimately breaks the camel's back? You've got. Trillions potentially at stake in terms of just overall CapEx investing through the years. I mean, it seems like there is a collective desire for this to work. What ultimately provides the challenge or presents the challenge in your mind that kind of reverses all that?

• 31:48Speaker 03

  So I heard a Goldman analyst say recently that the first hyperscaler to pull CapEx will get rewarded by the markets. I think that CapEx pullbacks are, they're the sign. I also think any financing falling through for OpenAI or Anthropic would be a sign, but I think we're going to start seeing AI companies kind of start falling out of favor and not being able to raise money. But the big thing is debt. When data center debt stops being issued, that will be when it's bedtime for this industry. Because even if they think AI is going to win, we've got 100 gigawatts or so of data center capacity allegedly under construction or in planning. That's trillions of dollars of money that needs to come from somewhere. And we are tapping out the debt markets. We saw that with Google raising that $85 billion equity raise.

• 32:30Speaker 00

  I'll be very interested to see what ends up happening because you also have the arms race aspect of this. If Fable 5 really was as scary, because the version that ended up getting released to the public was nerfed so hard, but the version that made Anthropic go to the government and say, hey, listen, we can hack basically anything. There's all these legacy bugs that people didn't realize. We're taking advantage of them. We're shutting it down. And that made the US government go, okay, whoa, we can't even let our adversaries have access to this. There's a big hubble loop about OpenAI and Google selling technology to Chinese companies via Singapore, I think. So, and the government's saying technically it's legal, but they don't like it and blah, blah, blah. So they are treating this as

• 33:15Speaker 00

  a weapon system. And so I haven't heard Ed talk about that, but it's an interesting angle on all of this. Like, will the government be able to stop themselves from going, all right, we've got to keep these guys going some way or the other, because from an arms race perspective, this becomes incredibly important. All right. I don't think we need to keep going on this. We get the idea. But I have a feeling that we, we are not going to see less of Ed in the future. I think we're going to see more. We are all going to have to contend with what the bear case is for this. In terms of if we're going to survive AI and we're really going to get to the other side and we're going to figure out what this is going to be in our lives, we've got to have a long runway. There's

• 33:58Speaker 00

  got to be time for this to play out, for innovators to try things and fail, and the company just completely goes under, and then the next person comes along and tries something new. That whole process, the creative destruction of a new technology, we have to go through. And I think that the internet is our guide. And I think that we have expected, AI came in like a storm and improved so fast. I think that people really, and we needed that next big thing from an investment perspective, because again, you've got all this economic pressure due to inflation, forcing people into the markets. They want to know where they're going to get their returns from. And so that, because there weren't just

• 34:37Speaker 00

  a endless bevy of ideas, that money was just going heavier and heavier into AI. And so now you've got yourself in a position where the debt is accumulating so fast, the revenue is coming in slow, it's systemically important to the entire economy, and certainly to any individual investor, you've got to really pay attention. And I think where, what Ed said about the debt is the real thing. And if you think about this as the market starts to pull back and says, ooh, like when DeepSeek first launched, I think we lost almost a trillion dollars in value in a single day. So people were like, wait, I thought

• 35:20Speaker 00

  America was going to win the AI race. Now all of a sudden, maybe it's going to be China? Yo, investors got spooked. So what's going to be the story that spooks the market? That I don't know. But if the market starts getting spooked, that's when you start seeing people say, I'm not willing to pay for that debt anymore, which then is going to force companies to pull back on the CapEx. Now, capital expenditure for people who don't know. So it's you investing in your business, you're going to build a data center. That would be a capital expenditure. So when somebody says CapEx, that's what they mean. Ed is saying the first person to not do it because they're getting pressured, or not do it because they failed to raise debt, where they just go, we don't need anymore. We're good. We're not going to keep expanding, that that's going to be the winner. I think they might

• 36:01Speaker 00

  take a bigger hit. I think people might go, uh-oh, you're in trouble. You're the first one to admit problems. And they'll probably say, no, no, no, this is an industry-wide thing. But it becomes a question of whether investors buy it or not. And so now, I think it will be difficult to be the first mover, because you know, if you say, eh, we're all out of ideas, but we're the first ones backing out, maybe people buy it, or maybe they just go, you're the only one that's out of ideas, because investors are still incentivized for this to be the next big thing, to keep pouring capital in. And so this one is going to be a bumpy ride. I've been saying this now for a couple of months. So people need to, yeah, keep your wits about you. I'll leave you with one last thing, and that is,

• 36:45Speaker 00

  where is the debt hiding? Where is all of this debt? The banks are already diversifying. Is it hiding anywhere in your portfolio that you're not aware of? It's certainly worth digging in and taking a look. Let's talk about a pattern that is guaranteed to be killing your progress. You know what you need to do. You need consistent nutrition. We all do. You need vitamins, probiotics, greens. We all know that we should be doing more of it. When your morning gets chaotic, you skip it. When you travel, you skip it. When your routine breaks, everything tends to break, and that inconsistency compounds against you every single day. AG1 is designed to solve the execution problem. One scoop, eight ounces of water, and you're done. You're

• 37:26Speaker 00

  getting 75 plus ingredients, vitamins and minerals, pre and probiotics, nutrient-dense superfoods, everything that used to require six, seven different supplements, and perfect planning now happens in one drink that takes about 30 seconds to make. Right now, AG1 is giving you $87 worth of free gifts with your first subscription. You get a welcome kit, travel packs, vitamin D3 plus K2, and flavor samples. Click the link in the show notes or visit drinkag1.com/impact to claim this offer.