In August, the founder of hedge fund Praetorian Capital Harris “Kuppy” Kupperman penned an essay on the absurd finances behind AI data centers. While the tech industry has likened data centers — or more specifically, the expensive semiconductor chips that power them — as the “shovels” of the AI gold rush, Kupperman’s napkin math found that AI data centers have an impossibly short runway to achieve profitability.

In short, this is because data center components age rapidly, either made obsolete through rapid advances in technology, or broken down over years of constant, high-powered usage.

After publishing his initial findings, Morningstar reports that Kupperman got an earful from anxious professionals in the data center industry. Thanks to those conversations, the investment manager realized he made a crucial mistake — and that because of it, his grim prediction may not have been cynical enough.

“I clearly hit a nerve in the industry, when judging by the number of individuals who reached out to chat,” he wrote in an followup blog post. “In total, I’ve spoken with over two-dozen rather senior people in the datacenter universe, and there was an interesting and overriding theme to our conversations; no one understands how the financial math is supposed to work. They are as baffled as I am, and they do this for a living.”

Kupperman’s original skepticism was built on a guess that the components in an average AI data center would take ten years to depreciate, requiring costly replacements. That was bad enough: “I don’t see how there can ever be any return on investment given the current math,” he wrote at the time.

But ten years, he now understands, is way too generous.

“I had previously assumed a 10-year depreciation curve, which I now recognize as quite unrealistic based upon the speed with which AI datacenter technology is advancing,” Kupperman wrote. “Based on my conversations over the past month, the physical data centers last for three to ten years, at most.”

In his previous analysis, Kupperman assumed it would take the tech industry $160 billion of revenue to break even on data center spending in 2025 alone. And that’s assuming an incredibly generous 25 percent gross margin — not to mention the fact that the industry’s actual AI revenue is closer to $20 billion annually, as the investment manager noted in his previous blog.

“In reality, the industry probably needs a revenue range that is closer to the $320 billion to $480 billion range, just to break even on the capex to be spent this year,” Kupperman posited in his updated essay. “No wonder my new contacts in the industry shoulder a heavy burden — heavier than I could ever imagine. They know the truth.”

Kupperman called that gulf between tech industry spending and actual revenue in 2025 “astonishing.” However, it doesn’t even begin to scratch the surface. For example, how does it all shake out when we account for 2026, when hundreds of new data centers are expected to pop up?

“Adding the two years together, and using the math from my prior post, you’d need approximately $1 trillion in revenue to hit break even, and many trillions more to earn an acceptable return on this spend,” he writes.

“If the economics don’t work, doing it at massive scale doesn’t make the economics work any better — it just takes an industry crisis and makes it into a national economic crisis,” he concludes.

Overall, the pessimists broadly agree: it’s no longer a matter of if AI is massively overhyped, but when the whole thing comes crashing down.

More on AI hype: Data Shows That AI Use Is Now Declining at Large Companies