# Ray Dalio: America is Heading Into Very Dark Times!

## Метаданные

- **Канал:** Investor Center
- **YouTube:** https://www.youtube.com/watch?v=XnNCoJitrHI
- **Дата:** 01.05.2026
- **Длительность:** 14:30
- **Просмотры:** 4,012
- **Источник:** https://ekstraktznaniy.ru/video/51323

## Описание

Ray Dalio, the billionaire hedge fund founder who predicted 2008, just declared we're in stage five of economic decline — and pointed directly at the AI bubble as the crime scene. US tech companies are spending $675 billion annually on AI infrastructure, yet China is deliberately making frontier-class AI free and open-source. DeepSeek trained a frontier model for $5.6M versus $112M+ for US equivalents — then released it under MIT licence. The "enemy" is not AI itself; it's the broken profit model that assumes US companies can charge premium prices in a world where competitors are commoditizing the technology. Dalio's most quotable line, "AI is eating everything and it might eat itself," sits at the heart of his warning.

## Транскрипт

### Segment 1 (00:00 - 05:00) []

What happens when the biggest companies on Earth pour $675 billion into technology and another country decides to give that same technology away for free? Ray Dalio built Bridgewater into the world's largest hedge fund. His net worth sits around $20 billion. He predicted the 2008 financial crisis before most people knew what a subprime mortgage was. And last month, he sat down and said something that should worry every single person holding a stock market index fund. He named the exact stage of economic decline we're in. Warned that most AI companies are already walking dead and pointed at the one country that could destroy the entire AI profit model overnight. I think you were moving toward the that war. We're in what's state what I call stage five of a cycle, okay? In the book, I describe the pattern that's happened over and over again. And when you get to this position when there are a bad finances combined with large wealth and values gaps and irreconcilable differences and you have external threats as well as domestic threats you have this dynamic. I think that's where we are I'm like a mechanic. My goal I'm not ideological. I'm just a practical guy trying to make money in the markets and trying to describe things and that's what it looks like. I think when we look at the bubble question on AI, what a lot of people don't realize in bubbles is that through all technologies, they think that they are betting on the technology when they buy the stocks in the companies. That's not true. Okay? There's a giant difference between the behavior of the companies and technologies. And that the norm is in these is that a lot of companies won't survive in the start. Very small percentage and they'll all fight and so on, but the technologies will go on and it'll be great. The technologies will. So, I want to emphasize to people that dynamic and I can go on and describe, you know, what it's like. Of course, we've seen it to some extent with the 2000 bubble in the technologies and what went on, but even if I describe what it was like in the late '20s but, you know, it's just it was unbelievable, but the technologies will go on, but the companies won't necessarily go on. So, Dalio just dropped a framework that covers centuries of economic history and then pointed it straight at your brokerage account. Here's what backs it up with hard numbers. Take a look at this chart. The Nasdaq peaked at 5,048 on March 10th, 2000. By October 2002, it had crashed 77% to 1,139. Not slowly, but in waves of liquidation that felt like blood on the screen for anyone watching in real time. Over half of every publicly traded dot-com company gone by 2004. Just gone. 431 tech IPOs happened in 1999 and 2000 combined. Fewer than half survived 5 years. And here's the part that should make your stomach drop. Amazon, the company we now think of as unstoppable, lost over 90% of its stock price during that crash. It survived, but most didn't. Cisco hit $80 a share in March 2000. 24 years. That's how long before Cisco saw that price again. Or maybe it's simpler than that. It's what happens when an entire thesis gets right. The technology thesis was right. The internet did change everything. But the company thesis that was a massacre. Dalio's saying we're watching the same pattern repeat. And this time, the numbers are bigger. Now, let me show you how much bigger. US tech companies are projected to spend over $675 billion on AI infrastructure in 2026 alone. Amazon, 200 billion. Alphabet, 180 billion. Meta, 125 billion. Microsoft, 120 billion. Oracle, 50 billion. Goldman Sachs projects a cumulative $1. 15 trillion in hyperscaler capital expenditure from 2025 through 2027. And here's where it gets sharp. The tech sector may need $1. 5 trillion in new debt to finance it. New debt, not

### Segment 2 (05:00 - 10:00) [5:00]

retained earnings, borrowed. I want you to sit with that for a second. When was the last time an entire industry took on a trillion and a half in borrowed money to build something that hasn't proven it can pay for itself? That question has a very specific answer. 1999. But then, Dalio said something that changes the entire calculation. He started talking about a country that's decided the technology should be free. And so, when I'm looking at that has big implications. Right now, it looks to me like AI basically is eating everything and it might eat itself. And and what I mean by that is not produce adequate profits. We can't take just a domestic view of that. We have to look also what's happening in China and make interesting distinctions there. You know, there's a difference in philosophy that's carried through in the economy of how the economies of the United States and China work in that we have basically primarily a profit-based system. They have a system in which they might believe that profits are a second consideration. They're not necessarily needed in order to achieve the best results. For example, in China, they would say usage of AI is fantastic. So, it should be like electricity or something and let's make it free for everyone. open source for everyone. Okay? And they might get much higher usage and they'll get their productivity gains through the usage. And we have a profit system to pay back. Okay, well, now we're in one world. How do you compete in that world? What do you do with that? In other words, just imagine that their technologies are almost as good as ours cuz they are. They're not far behind. And and then but that you could get them for free. Open source. Okay, now you got to pay it back. Okay, so I just want to emphasize that these are also systematic risks that enter into the picture of AI. AI is eating everything and it might eat itself. When I first heard that line, I assumed he was talking about competition driving margins down. Standard stuff. But then the China detail landed and it rewired my entire mental model. Here's the evidence. In January 2025, a Chinese lab called DeepSeek released a frontier class reasoning model. Training cost $5. 597 million. The leading US equivalent, over 112 million. That's 18 times more expensive and the DeepSeek model performs in the same ballpark. Think about the speed difference there. China achieved frontier performance at a tenth of the cost. They released it under an MIT open source license. Free for everyone. And it worked. Chinese AI models went from roughly 1% of global market share to 15% in a single year. Alibaba's Qwen model passed 700 million downloads on Hugging Face by January 2026. The velocity is what matters here. China didn't build something marginally better. They built something faster and cheaper and released the playbook. So, here's the cold math Dalio is pointing at. If China treats AI like electricity, a public good free at the point of use, and their model costs 95% less to build what exactly is the revenue model that justifies $675 billion a year in US spending? It's not rhetorical. A betting company analysis says AI needs to generate $2 trillion in annual revenue by 2030 to justify current infrastructure spending. That means AI revenue has to grow roughly 10 times in 4 years. Meanwhile, an MIT Sloan study found that 95% of companies report zero measurable return on their generative AI investments as of mid-2025. Not low returns, zero. Nothing. The spreadsheet is still blank. Nvidia right now trades at a forward PE. That's the stock price divided by next year's expected earnings. Basically, how much you're paying per dollar of future profit of about 43. That assumes the growth train doesn't slow down. It assumes the revenue materializes. And it assumes nobody undercuts the

### Segment 3 (10:00 - 14:00) [10:00]

price to zero. Every one of those assumptions is now in question. The strongest bull case comes from Tom Lee at Fundstrat. His argument, these aren't venture-backed startups burning cash like in 2000. Apple, Microsoft, Nvidia, they're the most profitable companies in human history. They're spending their own earnings, not borrowed money. And they're already seeing real productivity gains from AI. That's a reasonable point. Nvidia posted $130. 5 billion in revenue last fiscal year. Up 114% with gross margins near 75%. That's a real business printing real money. But two things undercut Lee's argument. First, and I want to be precise here. Goldman Sachs estimates the tech sector needs $1. 5 trillion in new debt for AI infrastructure. Quote, spending their own profits only tells part of the story when you're also loading the balance sheet at that scale. Second, and this is the part that keeps me up at night. Lee's entire framework assumes US companies can charge premium prices for AI. If China's strategy works, that pricing power evaporates. The bull case works in a world where AI stays expensive. Dalio is warning about a world where it doesn't. So, what does this mean for you personally? Let me introduce you to Marcus. He's 42, runs a portfolio in Seattle, and in January 2026, he bought an S& P 500 index fund. He thought he was diversified, thought he was safe. Then, he ran the numbers. If you hold a standard S& P 500 index fund, roughly 32 to 36% of your money is concentrated in the 10 largest technology companies. Apple, Microsoft, Nvidia, Amazon, Alphabet, Meta. Every one of them is making massive AI bets. Marcus had 67,000 of his $200,000 portfolio sitting in those 10 names. A dot-com style correction in AI alone could drag the S& P down 15 to 20%. For Marcus, that's 30 to $40,000 at risk on an assumption he never articulated. Most of that risk, invisible. Here's the personal test I'd run, and it's the same one Marcus ran too late. Can you explain specifically how the AI companies in your portfolio will generate enough revenue to justify $2 trillion a year by 2030? If you can't, you might be doing exactly what Dalio just warned about. Betting on the technology when you think you're betting on the companies. Let me pull the threads together. Dalio's framework says we're in stage five of economic decline. The late phase where bad companies meet internal division and external threats. That's the macro backdrop. Lay on top of that, you've got the single largest capital expenditure cycle in history. Aimed at a technology that a rival nation is trying to make free. AI will transform the world. I truly believe that. The technology works, it's getting better, and it's not going away. But the profit model that $675 billion in annual spending depends on, that's the part under siege. If Chinese open-source models keep closing the performance gap, and the evidence says they are, the revenue ceiling for US companies may be dramatically lower than anyone is pricing in. Now, I want to flag what could prove Dalio wrong. China's open-source advantage could stall. Regulation, export controls, or technological breakthrough that US labs monopolize. Washington could impose tariffs or restrictions that protect domestic AI margins. The uncertainty is real, and anyone who tells you that they know exactly how this plays out is selling something. But the concentration risk, that's real right now. And most people holding an index fund have no idea they're carrying it. If you enjoyed this video, please give it a like and check out this video next where Bill Ackman reveals what may be the biggest investment opportunity in 70 years. I will see you over there.