Mark Cuban said that many AI data centers currently under construction may eventually be repurposed as pickleball courts [1].

The prediction highlights concerns regarding a potential bubble in artificial intelligence infrastructure. If the rapid buildout of these facilities exceeds long-term demand, the industry could face a significant surplus of specialized real estate.

Cuban shared these views during an episode of the All-In Podcast [1]. He said that the current surge in AI-driven construction could lead to an oversupply of space. According to Cuban, when the demand for these high-tech hubs wanes, the resulting large, flat spaces would be ideal for recreational uses [2].

"A lot of data centers will be turned into pickleball courts," Cuban said [1].

He linked the potential for this conversion to the growth of the sport in the U.S. Cuban said that the architectural nature of these facilities makes them suitable for such a transition. He said that when the AI boom cools, those massive, flat roofs are perfect for a sport that is exploding across the country [2].

This perspective mirrors historical warnings about the dot-com bubble, where over-investment in physical infrastructure preceded a market correction. Cuban suggests that the scale of the current AI buildout may create a legacy of underused buildings that must be adapted for other purposes to maintain value.

While the AI sector continues to expand, the possibility of widespread facility abandonment remains a speculative risk. Cuban's comments emphasize the volatility of emerging tech trends and the physical footprint left behind when those trends shift.

“A lot of data centers will be turned into pickleball courts.”

This prediction reflects a broader debate among investors regarding whether AI infrastructure is being built for sustainable long-term growth or a temporary speculative peak. By comparing data centers to potential sports courts, Cuban is pointing to the risk of 'stranded assets'—expensive infrastructure that no longer serves its original purpose and lacks a high-value alternative use.