ALL INPredictions

Mark Cuban on the AI Bubble: Who Actually Gets Wiped Out?

Mon, 20 Jul 2026 18:51:00 +0000

Back to episodes
Mark Cuban Too Early attribution: medium 00:02:56 aitech

Because AI compute price-performance improvements will outpace power-scaling, many AI data centers currently being built will end up underutilized and repurposed for other uses in the coming years.

“there's going to be a lot of data centers that are going to be turned into pickleball courts” View on YouTube
Explanation

This is an explicit multi-year prediction ('over the coming years') about AI data centers becoming underutilized and repurposed. Only weeks have passed since this episode aired (published July 20, 2026); there is no meaningful data yet on data center utilization trends to score this against.

Mark Cuban Too Early 00:10:51 aieconomy

AI will not cause a 50% loss of white-collar jobs, contrary to widely repeated predictions; overall employment will keep growing even as AI adoption increases.

“AI is not going to you know take away 50% of the jobs. There's just so much it can't do that regular people need it to do.” View on YouTube
Explanation

The specific '50% of white-collar jobs' claim being referenced (from Anthropic CEO Dario Amodei, made in 2025 with a 5-year horizon) has not yet had its full timeframe elapse. Early 2026 data is mixed but leans toward supporting Cuban's skepticism so far: modeled AI-attributable job losses are estimated at roughly 200,000-300,000, far short of a 50% white-collar wipeout, and many CEOs still expect to grow entry-level and senior hiring. At the same time, some research warns a 'Great Recession for white-collar workers' remains plausible, and youth employment in AI-exposed roles has already fallen meaningfully. Too early to score right or wrong.

Mark Cuban Too Early attribution: medium 00:28:08 aipolitics

As AI/LLM adoption grows, more people will turn to large language models rather than social media as a trusted source for political decision-making, because LLMs prioritize truth-seeking answers over engagement.

“I think people, as they become more and certain with their politics, are going to go more and more and more to large language models and say, 'Who should I vote for?'” View on YouTube
Explanation

This predicts a gradual, multi-year adoption shift in how people form political opinions (turning to LLMs over social media). Too little time has passed since this July 2026 episode to observe a meaningful shift in behavior at this scale.

Mark Cuban Too Early attribution: medium 00:31:44 politicsgovernment

US political polarization will subside and the country will return to a more 'normal' political climate after the 2026 midterm elections and the next (2028) presidential election.

“I think as we get past this next the midterms and the next presidential election... I think we go back to normalcy. I think we'll get back to normal.” View on YouTube
Explanation

This prediction is explicitly conditioned on events that haven't happened yet: it points to political normalcy returning after the 2026 midterms and the 2028 presidential election, neither of which has occurred as of this review (August 2026, with midterms still months away).

Mark Cuban Too Early 00:39:11

Under the NBA's current second-apron roster-construction rules, no team will win three consecutive championships (a 'three-peat'), though back-to-back titles will still be possible.

“No more three-peats... There can be back-to-backs.” View on YouTube
Explanation

This predicts a multi-season structural outcome (no NBA team achieving a three-peat under the current roster-construction rules). Far too little time (well under a full season) has passed since this prediction to assess it.