Transcription
AI might be the biggest financial bubble in history. The AI boom might be heading for a bust, where people are worried specifically when it comes to comparisons to the dot bubble of uh assets that have vastly exceeded their value. Yeah. So, you know, a lot of people would lose a lot of money if if the AI if it is a bubble and it bursts, and the bubble will pop much like it did during the dot bubble in the late '90s. I think you can say that we are pretty much in the advanced stages of a bubble. It is a bubble. The question is when will the bubble pop. So the valuations just don't make sense in terms of the income that these companies are making uh and the enormous amounts of infrastructure that they require. It it's just not economically sensible.
If if everyone is talking about it like it's a bubble and everyone thinks it's a bubble, why is everyone continuing to act like it isn't? And why is the uh is the pace of investment not slowing? Tech CEOs have been telling you that it can solve all of the world's problems. Cure cancer, healthcare, eliminate poverty, abundance, new kinds of energy, manufacturing, incredible new jobs, AI will actually discover new science, you know, new cures for diseases. People's individuals are going to get much much richer. All the all those things. But even though they want you to think that AI can fix everything, behind the scenes, there's one thing AI companies can't seem to do. Run a profitable business.
The core issue at the heart of the AI bubble and all these flawed businesses is the fact that AI is incredibly expensive to run. Open AI alone is planning to spend $600 billion in computing costs by 2030 just to run its business. And that's after they significantly cut their spending plans after investors expressed concerns about unprofitability and said, "Show me the money." Originally, Open AI was planning to spend $1.4 4 trillion before investors pushed back, which is just a mind-boggling number. At the heart of all of this reckless AI spending is Sam Altman, OpenAI CEO, who has explicitly said that he does not care how much it costs to build AI. Whether we burn 500 million a year or 5 billion or 50 billion a year, I don't care. I genuinely don't. As long as we can, I think stay on a trajectory where eventually we create way more value for society than that. And as long as we can figure out a way to pay the bills, like we're making AGI, it's going to be expensive. It's totally worth it.
This insane level of spending would be one thing if Open AI was bringing in ridiculous amounts of revenue. But guess what? They're not even close. Open AI may be a private company, but we have a pretty good idea of their revenue and total losses based on what they've said publicly and from Microsoft's earnings reports. Since Microsoft is one of Open AI's largest investors and owns roughly a 27% stake of the company that's worth approximately $130 billion, the performance of Microsoft's investment is something we can see. And looking at the performance of OpenAI through Microsoft's financials, we can see that OpenAI managed to lose $12 billion in just three months. That's one of the biggest quarterly losses for a tech company in history. And when you compare that to OpenAI's revenue, they're estimated to have brought in just $13 billion for the full year in 2025. You do the math there. They lost nearly as much money in one quarter as they did in revenue for the full year. Doesn't seem like a great business model to me.
So, as you might guess, Sam's gotten quite a few questions on why the company's revenue and investment commitment numbers are worlds apart because the whole arrangement makes no sense when you think about it for more than 2 seconds. Sam Altman was called out a bit on his questionable math and asked in an interview how OpenAI could possibly try to commit to $1.4 trillion in infrastructure when it only makes $13 billion a year and well Sam snapped a little bit. So I think the single biggest question I've heard all week and and hanging over the market is how you know how can a company with 13 billion in revenues make 1.4 4 trillion of spend commitments, you know, and and and you've heard the criticism, Sam.
"First of all, we're doing well more revenue than that. Second of all, Brad, if you want to sell your shares, I'll find you a buyer."
Now, maybe Sam woke up on the wrong side of the bed. But it doesn't take a genius to figure out that Altman probably snapped at the interviewer because it struck a nerve on a problem a little too close to home. And that's because the economics of AI are a fantasy right now. AI's costs are simply too expensive for it to be profitable anytime soon. Sam Altman has admitted that OpenAI loses money on its $200 monthly chat GPT subscription because even that is not a high enough price tag to cover the computing costs that subscribers use on it. And the reality is that if you're losing more and more money with each additional customer, you've got a fundamentally broken business model. And Open AI and other AI companies have that in spades. And that's if these companies are even lucky enough to find paying customers in the first place. Consumer AI revenue is expected to be 12 billion in 2026. Not bad, obviously, in a vacuum. But if you compare that to the $500 billion the industry is expected to spend that same year, that's basically nothing.
So, you're probably wondering, how have these unprofitable AI companies like Open AI even been keeping the lights on since they seem to just be massive money pits? Well, with private equity investment money, of course. Big tech companies like Nvidia, Microsoft, Amazon, Oracle, and the countless venture capital firms have been funneling insane amounts of money to AI startups in the hopes of hitting on the next big thing. Perhaps the biggest reflection of this irrational exuberance and the growing AI bubble is that there are now more than 1,300 AI startups with valuations over $100 million. And if that's not insane enough for you, there are now 498 AI unicorns, which is just a fancy way of saying a private company worth over $1 billion. The non-stop investment venture capital money has been pumping these companies to insane valuations, even though so many of them are terrible businesses that are incinerating cash every single day.
But it gets worse. A ton of this money that AI startups are getting is being accessed through circular financing deals. And circular financing is just a fancy way of saying that it's basically a giant game of musical chairs with cash. Nvidia and Open AI have been the ring leaders of this circular financing that looks suspiciously bubble-like. What do these strange deals look like in practice? Well, picture this. Nvidia says that they'll invest up to $100 billion in Open AI. But guess who OpenAI's biggest supplier is? Nvidia. So after Nvidia gives OpenAI the $100 billion investment, Open AI just turns around and spends it all on Nvidia's chips. They're just passing huge sums of money back and forth in a circle, hence the name. You throw in a bunch of other players and middlemen like Oracle and Microsoft into the equation and you've got an immensely complex web of arrows pointing every direction, making it look like everyone is getting rich when they're really just swapping the same dollars back and forth. But if you're just looking at the funding news headlines and not looking at where the money's actually coming from, these circular financing deals could fool you into thinking that these companies are doing much better than they really are.
In many cases, Nvidia seems to be paying its customers to buy more of its products by buying equity stakes in these customers. It's not like this is a new playbook either. Similar deals were also around during the dotcom bubble when telecom equipment makers like Lucent Technologies gave loans to their customers so that they could buy their products. Of course, when these companies went bankrupt, they also dragged the telecom companies down with them. So, right now, everything looks great because of the circular financing. And Nvidia and Microsoft's revenue numbers have never been higher as these unprofitable AI companies keep buying their chips and cloud computing. But what happens if these top AI customers are never able to turn a profit? Then not only will Nvidia and Microsoft lose some of their biggest customers to bankruptcy, but they'll also lose the hundreds of billions of dollars that they've invested in these unprofitable AI startups. If their customers don't succeed, these AI infrastructure providers will also be in a world of hurt.
So why should you care? Well, it affects the entire stock market and economy now. Unfortunately, the S&P 500 stock market index has never been more concentrated than it is now. Just five mega tech companies that are heavily tied to the future of AI make up about 30% of the S&P 500. Now, that means that anyone that buys the S&P 500 is investing heavily in the future of AI, whether they know it or not. And with 401ks often heavily invested in the indices, your retirement is likely heavily tied to the AI bubble, whether you like it or not. And one of the worst parts is that many tech billionaires recognize that it's a bubble and straight up do not care about the risk for anyone else's finances. They argue that this is a good bubble that's creating innovation that will eventually be good for society, that we should just stop worrying about bubbles altogether or the fact that some people will lose a lot of money. Jeff Bezos went on the record saying that he thinks the AI bubble is good because it's an industrial bubble and not a banking bubble. Check this interview out.
"But the great thing about industrial bubbles, this is a kind of industrial bubble as opposed to financial bubbles. And I'll tell you what I mean by that. If you go back like the the '90s had a biotech bubble and there were a bunch of uh pharma startup companies that were designing drugs and using new techniques and the world got very excited. The investment world got very excited as a group. They all lost money but we did get a couple of life-saving drugs. A bubble like a a a banking bubble the crisis in the banking system that's just bad. That's like 2008. And so that's you those bubbles society wants to avoid. The ones that are industrial are not nearly as bad. It could even be good because when the dust settles and you see who are the winners, society benefits from those inventions. They still get those life-saving drugs. And that's what's going to happen here, too. This is real. The benefits to society from AI are going to be gigantic."
Of course, these venture capitalists and tech billionaires are going to be happy inflating the bubble even further because it's making them money in their investments and they will sell before the bubble bursts. It's regular people who will suffer from the AI bubble collapsing. And the only way this AI bubble doesn't pop is if these AI startups are able to start to turn a profit. It's super clear that the ultimate success or failure of AI companies hinges on whether they are able to significantly boost economic productivity for their customers. And the productivity boost has to come at a reasonable cost. If AI can't be productive at a cheaper cost than what it would take to hire an employee to do the same job, companies aren't going to be interested or care. Low-quality weird AI slop isn't going to get businesses on board. It's about return on investment at the end of the day. And that's a problem for AI startups because artificial intelligence is not providing value to most companies right now. A recent MIT study found that 95% of corporate AI initiatives are showing zero return. Zero.
AI is often incredibly expensive compared to what a human can do. Goldman Sachs noted in a report that while AI was able to update historical data and their company models faster than what humans were able to do, it also did so at six times the cost. Because of the large price tag involved with AI computing, one MIT professor Goldman Sachs interviewed for its report estimated that the amount of tasks that AI can actually profitably impact over the next 10 years is just 5% total. So things look pretty bad for businesses on the AI software side of things. And this market dynamic means that all the real profits surrounding AI are actually being generated by semiconductor companies like Nvidia and other hardware and infrastructure providers. Big tech companies and AI startups are throwing hundreds of billions of dollars into Nvidia's AI hardware. But software companies and the startups that are actually trying to build AI applications are not making money on these hardware investments yet. And that can only go on for so long. This is not a sustainable market dynamic for anyone involved. If AI startups and software companies can't actually start to make money from AI applications, they will eventually be forced to stop investing money into NVIDIA GPUs and other AI infrastructure.
Right now, everyone is buying up GPUs to build AI models with the assumption that they'll be able to make money with them in the future. But that's far from guaranteed because another major issue facing these companies is that increasing competition and a lack of differentiation between AI models means that artificial intelligence could become a commodity that lacks real pricing power. The differences between many large language models and chatbots are relatively insignificant, and this is a fact that many AI founders are keenly aware of themselves. The differences between OpenAI's GPT model, Google's Gemini model, or Anthropic's Claude model are arguably pretty minor when considering the use cases of the average person. The fact that so many AI companies have similar capabilities and offerings means that they have to compete with each other primarily on price. This lowers profit margins for everyone involved and will make surviving difficult or impossible for many AI startups. And that's not even taking into account the fact that there are many free and open-source AI models that already exist in the first place. It's hard to sell a product when there's a free option available to everyone.
All of these factors make the current AI investment boom eerily similar to the dot bubble of the late '90s. AI as a technology is probably here to stay, much like the internet was from the dot bubble. But similar to the unprofitable dot companies of the late '90s, AI startups lack a viable business model at the moment. And the technology is still very early. Because of this, it's very possible that many of these unprofitable startups won't exist 5 years from now. And like all of the past investors who piled money into startups that ended up in the dotcom bubble graveyard, many current investors trying to pick individual winners in AI could get seriously burned. Even Cisco, the dominant stock that was the backbone infrastructure provider of the internet, fell apart when the dotcom bubble burst, and its profit margins declined due to increasing competition. Cisco finally hit its first new all-time high since the dot bubble earlier this year, but it took over 25 years for the stock to recover.
Looking at our present day, Nvidia currently has gross profit margins around 75%. But with increasing GPU competition from companies like AMD, Nvidia's current growth rate and margins also seem equally unsustainable and their pricing power will almost certainly decrease as the competition catches up. As Goldman Sachs, head of global equity research said, "Overbuilding things the world doesn't have use for or is not ready for typically ends badly." The NASDAQ declined around 70% between the highs of the dotcom boom and the founding of Uber. A big part of what has been driving this current AI investment cycle is also big tech essentially investing in artificial intelligence simply to protect their own market dominance and to prevent any disruptions to their core businesses. Mark Zuckerberg basically admits that's why Meta is spending so much on AI in this interview right here.
"No one knows when super intelligence is going to be possible. Is it going to be 3 years? Is it going to be 5 years? Going to be eight years? Whatever. Is it never going to happen? But but I don't think it's never going to happen. I I I'm more ambitious or optimistic. I think it's going to be on the the sooner side. But let's say let's say that you weren't sure if it was going to be three or five years. Like in a conservative business situation, maybe you'd like hedge building out your infrastructure because you're worried that if you build it out assuming it's going to be three years and it takes five, then you've lost, you know, maybe a couple hundred billion dollars or something. I mean, my view is that..."
"That's a lot of money."
"Well, no. Well, I was going to say in the grand scheme of it is it is objectively a huge amount of money. Yeah. Right."
"I mean, didn't you just tell Trump you were gonna spend like 600 billion?"
"I did. Yeah. Through 2028, which is..."
"It's a lot of money."
"It is. And and if and if um and if we end up misspending a couple of hundred billion dollars, I think that that is going to be very unfortunate, obviously. But what I'd say is I actually think the risk is higher on the other side."
"If you if you um build too slowly and then super intelligence is possible in three years, but you built it out assuming it would be there in 5 years, then you're just out of position on what I think is going to be the most important technology that enables the most new products and innovation and value creation in history. So um so I don't know. I mean, it's um I don't want to be kind of cavalier about it. I mean, obviously, these are very large amounts of money and we're trying to get it right, but I think the risk, at least for a company like Meta, is probably in not being aggressive enough rather than being somewhat too aggressive. But, but part of that is like we're not at risk of going out of business or something like that, right? if you're one of these um you know companies that like an open AI or an anthropic or something like that where you know they're raising money as as the way that they're funding their buildout and you know there's there's obviously this open question of to what extent are they going to be able to keep on raising money."
Big tech companies like Google, Meta, and Microsoft are continuing to invest heavily in AI infrastructure simply because they are worried about the risks of underinvesting in AI if it does take off. But it's very possible that revenue growth in the AI market simply won't materialize at the rate many people are expecting. And if that ends up being the case, big tech companies aren't going to continue feeling like they need to hedge their bets with even further investment in AI computing capabilities. And when that happens, we will very likely see the AI investment hype fade.
Everyone knows that the oldest trick in the Silicon Valley playbook is to take investor money and focus on sales growth and capturing market share by any means necessary. These AI startups have been doing that for a few years now and don't mind burning cash while they focus on sales growth in the hopes that they can become one of the dominant players in the space. But this strategy only works if there's a pot of gold at the end of the rainbow and the company can eventually become profitable. These startups have to be able to make a profit eventually and it's looking like it needs to be sooner rather than later. Investors won't fund a company's operations forever with no return on investment in sight. And with AI's current economics, turning a profit is very much in doubt.
This desperation for profit is leading AI companies to do essentially anything to bring in money. Open AI recently found itself in a huge controversy by opening the door to the US government using its AI for mass surveillance of US citizens. It's very shocking and scary stuff and you can check out my full video on that controversy linked here. And be sure to subscribe to the channel for more finance and business content. And let me know what you think of the AI bubble in the comments below.