Torsten Sløk, chief economist at American asset company, Apollo Global Management, has warned that the AI companies and their stock prices are more over-inflated than the dot-com companies of the early 2000s, suggesting that an even bigger crash could be coming. He highlighted 10 of the top-performing AI companies, then suggested that the only real difference between AI businesses today and the dotcom companies of the late 90s and early 2000s is that AI businesses are even more overvalued (via Gizmodo).
The dot-com crash around the turn of the century saw companies rushing to adopt and take advantage of the internet. A relatively new technology and phenomenon at the time, but one that venture capitalists saw as having earning potential. Over the last five years of the 20th century, they invested trillions of dollars, and stock prices for publicly traded internet entities soared, only to come crashing down when the bottom dropped out of the market.
By the early 2000s, many of the companies involved in the boom had gone bust, and even now, industry giants like Amazon have lost huge portions of their investments, earnings, and market capitalization.
That’s what Sløk argues is coming for the major AI firms. That’s Apple, Microsoft, OpenAI, Meta, Google/Alphabet, Amazon, and a range of other companies. He highlights how these firms have seen huge upticks in their valuations and stock prices in recent years, driven by investments in AI ventures.
This is completely out of whack with the earnings potential of these companies, Sløk argues, and suggests the majority of the gains made to the stock market during this AI boom have been because of the overperformance of these top stocks.
That, he claims, is not going to last, and because the boom is bigger this time, the bust could be even worse. You can expand the tweet below to see the comparison of valuations.
Torsten Slok: “The difference between the IT bubble in the 1990s and the AI bubble today is that the top 10 companies in the S&P 500 today are more overvalued than they were in the 1990s” pic.twitter.com/OEervHU4WGJuly 16, 2025
Although Sløk doesn’t present a timeline for when any such bust could happen, it’s clear even for us non-economists that the money being thrown around by some of the major tech companies is difficult to sustain. OpenAI recently accused Meta of offering $100 million signing bonuses to new AI talent. That’s after it invested $14 billion in ScaleAI (only to see 200 employees fired from that company). CoreWeave is investing $6 billion in a new AI center, Amazon might be investing an additional $8 billion in Claude maker, Anthropic. Not to mention Nvidia’s push to drive $500 billion in investment in “AI Factories,” all while hoping that name catches on.
“Buy the rumor, sell the news,” is not what’s happening here.
Even if this wasn’t quite to the overhyped levels of the dot-com crash, there are plenty of recent examples of tech trends seeming to encapsulate the future and then failing to deliver.
Meta invested tens of billions in the Metaverse and then promptly pivoted to AI as if it never happened. NFTs and blockchain were going to change how art, finance, and investing work, and that still hasn’t happened yet.
Where each of those was constrained to a single company, or a niche market, though, AI is already everywhere, even without much of a profitable product to show for it. AI holds a lot of promise, perhaps a little like virtual reality, the metaverse, and blockchain technology, but does it warrant the hundreds of billions of investment it’s getting already?
Sløk argues no, and that when the world catches on, these companies, which have ridden high on a wave of hype and investment, may find their ephemeral silicon empires melt away into sand.
What comes after that is even more speculative than a potential downfall, but if the dot-com analogy tracks, we could see huge consolidation, with many of the top companies surviving, but scaling back their investments dramatically. Speculative AI companies would likely fail, while those with more robust revenue streams, like Amazon, Google, and Meta, would likely survive, albeit diminished in their influence.
Just as the dot-com crash didn’t kill the internet, or all of the websites on it, the AI bubble bursting wouldn’t kill AI, or make the technology less useful. It will continue to be developed and is likely to remain a useful, functional tool. However, perhaps we wouldn’t need to see “AI” as a buzzword on every product we buy over the following few years.
Follow Tom’s Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.