Here’s a thought that should keep Big Tech executives up at night: Mark Cuban thinks a lot of the massive data centers they’re building right now will eventually become… pickleball courts.
The billionaire investor dropped this gem on the latest “All-In” podcast, and honestly, it’s the kind of joke that lands because there’s real teeth to it. Cuban’s point is simple but devastating: if AI models and data centers get more efficient—which they almost certainly will—then all the computing capacity being built today becomes yesterday’s news. Obsolete. Useless. Perfect for converting into recreational facilities.
Now, Cuban isn’t saying the hyperscalers like Meta and Alphabet are wrong to build aggressively. They’re probably right that AI usage will explode. But here’s where it gets spicy: he thinks technological breakthroughs will make AI so much cheaper and power-efficient that the current buildout will look like overkill in hindsight.
He points to fiber networks as the historical parallel. There was a massive rush to build fiber infrastructure in the late ’90s and early 2000s. Everyone thought we’d need it all. Then connection speeds got faster, efficiency improved, and suddenly? No bandwidth crisis. The infrastructure that seemed essential became redundant.
“I don’t see how we don’t get the same price-performance improvements on the AI side,” Cuban said. Translation: we’re probably building way too much, way too fast.
The kicker? Cuban is the co-owner of the Dallas Flash pickleball team, so he’s not exactly joking about the pickleball courts part. He’s genuinely suggesting that’s where some of these data centers end up.
This isn’t just Cuban being a contrarian for clicks. He’s sounding an alarm about Big Tech companies doing something genuinely risky: borrowing heavily to fund massive AI infrastructure spending while pledging to keep the cash flowing for years. That’s planning for perfection, and as Cuban puts it, “nobody can predict that well.”
The good news? Cuban thinks this isn’t the dot-com bubble 2.0. The blast radius is smaller. You’re not seeing startups go public at “crazy valuations” with zero revenue, and you’re definitely not hearing cab drivers pitch their favorite AI startup to every passenger. The mania is more contained.
But here’s the bad news for venture capitalists and private equity firms: they’re “going all in,” and if Cuban’s right about efficiency gains, a lot of them are going to get crushed. The pain won’t be spread across the general population—it’ll be concentrated among the people who bet the farm on this buildout.
Cuban’s not alone in this thinking. Michael Burry (the “Big Short” guy) has warned that Big Tech is overinvesting in chips and data centers that’ll become obsolete. Jeremy Grantham called AI “obviously a bubble.” These aren’t random Twitter accounts—these are people who’ve made serious money by spotting when markets get ahead of themselves.
So while the AI boom might not crater the entire economy, it could absolutely crater the portfolios of the people who believed the hype a little too hard. And somewhere down the line, someone’s going to be playing pickleball in what used to be a $500 million data center.
