Tech leaders debate claim that CEOs are prone to "AI psychosis"

In short: A TechCrunch podcast episode discussed a debate over whether some tech CEOs are overconfident about AI, even when they do not use it day to day.

What's going on

Box CEO Aaron Levie sparked conversation this week after posting that tech CEOs are “uniquely prone to AI psychosis.” In this context, he was not talking about a medical condition. He was criticizing what he sees as unrealistic beliefs about what AI can do.

On TechCrunch’s Equity podcast, the hosts said Levie is not rejecting AI tools. Instead, he argues that leaders should actually use the tools themselves, rather than relying on presentations and big promises. The idea is simple, like a restaurant owner who sets the menu without tasting the food.

The episode also pointed to signs of a growing split in public opinion about AI. The hosts mentioned students booing when AI is brought up at graduation ceremonies, and ongoing frustration around tech layoffs. They also discussed a separate TechCrunch report that DuckDuckGo said its app installs rose 30% after Google announced more AI features in Search.

Part of the discussion focused on Google’s challenge. The hosts said Google is trying to add AI while still keeping the classic list of links many people want. They also noted recent examples where Google’s AI gave wrong or confusing answers, including basic spelling-related questions.

What to watch

If more people look for search tools that feel more predictable, smaller services may benefit by keeping AI separate or optional. Another open question is whether companies will adopt AI from the top down, or because employees find it genuinely useful in their daily work.

Source: TechCrunch AI