Klarna Co-Founder Sounds Alarm on AI Infrastructure Frenzy—Despite Heavy Investments
AI hype hits a wall—Klarna's visionary warns of bubble economics while writing checks.
Subheader: When the skeptic is also the backer
The payments pioneer—whose firm bankrolls dozens of AI startups—just compared infrastructure spending to 'gold rush dynamite sales.' Translation: The real money's in selling shovels... until the mine collapses.
Active verbs only: Venture capitalists dump billions into GPU farms. Startups pivot to 'AI-native' buzzwords. Consultants rebrand legacy analytics as machine learning.
Provocative closer: Next time someone pitches you an 'AI infrastructure play,' ask who's buying—and who's just stockpiling TNT.
Sebastian questions trillion‑level AI construction bets
Sebastian said the numbers being spent today are far beyond what he thinks is necessary to run advanced AI systems.
He pointed to the combined $112 billion in third‑quarter capital spending announced by Alphabet, Amazon, Meta, and Microsoft, and said he is unsure whether the size of these investments will make sense once the industry stabilizes.
He also noted that OpenAI, which still runs at a loss, has made $1.5 trillion in total commitments to secure future access to computing resources.The scale of that figure is the Core of his concern.
Sebastian said the popularity of tools like ChatGPT proves AI demand is global, but added, “[But] that’s a different thing than asking myself ‘is it worth putting a $1tn worth into servers.’ I am concerned that piling that kind of money into data centers may turn out to be not worth it.”
Sebastian argued that future AI models will likely need less power, not more, because they act like what he called the “most effective compression technology ever invented,” and he believes the systems will run more efficiently over time.
He pointed to outside examples as well. In January, Chinese AI group DeepSeek shook the market by unveiling low‑cost, power‑efficient models that compete with U.S. systems at a fraction of the cost.
Investors reacted quickly, and the past month has seen sell‑offs in several U.S. companies tied to the AI build‑out as doubts grew over whether the spending levels make sense.
He said he has raised these concerns privately with the leadership teams of the companies he invests in and suggested the internal mood is much more cautious than public statements imply.
“People have an incentive to say I’m wrong,” said Sebastian. “And I feel, behind doors, people are more concerned about what I’m saying than they are in public.”
He warns that everyday investors are being dragged into the AI bet
Sebastian said the rally in AI names is now so large that index fund flows are pushing pensions into the trend whether people like it or not.
He mentioned Nvidia, now valued at about $4.5 trillion, as an example and said the amount of market wealth tied to AI worries him. “That makes me nervous, because of the amount of wealth that is currently automatically allocated into this trend, without some more thoughtful thinking,” he said.
Sebastian added, “You can say ‘I disagree with the fact that Nvidia is worth that much and I don’t care, some rich people are going to lose some money.’ But the truth is, because of the index funds and how this works, your pension right now is going into that theory that it is a good investment.”
He also noted that well‑known hedge fund manager Michael Burry, famous for predicting the 2008 housing crash, shut down his fund Scion Capital last week after saying the market had lost touch with reality.
Scion had short positions against Nvidia and Palantir. Sebastian said, “I partially agree with Michael Burry. The question again is about timing because he’s betting against the whole market.”
If you're reading this, you’re already ahead. Stay there with our newsletter.
