Meta's AI Unit in Turmoil: Employees Lash Out at 'Soul-Crushing' Work

Morale at Meta's Applied AI unit has hit a new low, with employees openly expressing frustration during a company-wide livestream this week. One participant interrupted a technical presentation with an expletive-laden outburst, demanding that organizers tell a specific Meta AI executive he's "a piece of shit." The presenter covered their face in response, while the meeting leaders quickly moved on after asking attendees to mute their microphones.

The incident underscores deeper dysfunction within the 6,500-person Applied AI team, formed in March to support Meta's Superintelligence Labs. Three current employees, speaking anonymously, described the unit as a "gulag" where engineers once building social media apps now churn out puzzles to test AI models. They say the work feels menial and disconnected from their skills. "You have zero purpose," one employee said. Another called the tasks "soul-crushing."

The unrest isn't confined to Applied AI. Meta's recent restructuring, which cut 8,000 jobs, has left remaining staff across divisions like data center engineering and Instagram overwhelmed. Over 1,600 employees signed a petition protesting a new program that monitors keystrokes and clicks for AI training data. During an all-hands meeting, chief product officer Chris Cox acknowledged the "brutal" environment, comparing it to "running a marathon in a hailstorm."

CEO Mark Zuckerberg admitted in a Friday memo that Meta has "made mistakes" in its reorganization and pledged no further mass layoffs this year. He described Applied AI as a temporary assignment for "very talented people" who will eventually move to other roles. But engineers feel trapped: joining the unit is mandatory, or they must leave the company. Many describe themselves as "draftees."

Zuckerberg's vision for Meta remains ambitious: personalized Instagram feeds, AI-powered glasses, and virtual assistants that work around the clock. But for now, the path to those products runs through a workforce that feels undervalued and overburdened.

Source: Wired