At a recent Meta town‑hall, CEO Mark Zuckerberg admitted the company’s recent layoffs failed to accelerate progress. He said the trajectory of AI agent development over the past four months has not met expectations, and that Meta’s new structure has yet to deliver results. The speech followed a Reuters report by Katie Paul and Courtney Rozen.

Zuckerberg warned that early optimism around tools like Claude Code from Anthropic had not translated into speed. He blamed a culture that relies on “vibes” instead of data to steer decisions, citing the failed metaverse venture as evidence of misdirected focus. AI agents, designed to automate user tasks, remain under‑developed.

In response, Meta placed chief AI officer Alexandr Wang in charge of the company, a move that led to thousands of layoffs and the replacement of human programmers with AI. The company also mandated spyware on remaining staff to train an agent for content moderation, a strategy that many critics say is impractical and ethically fraught.

The reckoning shows that Meta’s reliance on vague intuition over data‑driven strategy has cost the firm both talent and credibility. Employees now face uncertainty, beta the company’s attempts to automate core functions have stalled. Meta must confront the gap between its ambition and its execution if it hopes to regain industry trust.