AI Psychosis Among Tech CEOs: Risks of Overestimating AI

Colleagues, I want to share an observation from the AI world: tech executives often overestimate model capabilities.

Aaron Levie noted that CEOs are distant from the "last mile"—they see prototypes but not the bugs and integration nuances.
In 2026 roughly 115,430 people have been laid off versus 124,636 for all of 2025 (Layoffs.fyi); many companies cite AI as a reason. ClickUp cut 22% of its workforce after deploying ~3,000 agents in a bid to build a "100x org".
Research (UC Berkeley, MIT, HBR) shows AI's productivity impact is contested: agents do not yet replace humans fully, and managers are becoming the bottleneck.

Why it matters: overestimating AI can cause organizational chaos and flawed staffing decisions.

What should leadership change to avoid this?

#AI #technology #management #layoffs