Inside Meta's roughly 6,500-person Applied AI unit, morale has cratered to the point that an employee hijacked a livestreamed, internal presentation this week with an expletive-laden tirade, telling attendees to inform a senior Meta AI executive that he was "a piece of sh_t." The outburst, reported by Wired, captured what engineers describe as a deeper revolt brewing inside the three-month-old team, where staffers say they were forcibly reassigned with no real alternative but to join or quit. Many now refer to themselves as "draftees," and their daily work, generating puzzles and coding challenges to train Meta's AI models, has earned the unit a grim nickname. "It's literally the gulag," one employee told Wired. "Most people find the work soul-crushing," said another.
The discontent has been building for weeks. A Business Insider report last month revealed that many employees learned of their transfer through surprise emails, a process one self-described draftee later called on Reddit "quite random." An internal announcement from April, reviewed by BI, acknowledged that Meta's AI models still lacked the sophistication to outperform humans on technical tasks like coding. "For agents to understand how people actually complete everyday tasks using computers, we need to train our models on real examples," the post read, according to the outlet.
In a leaked audio recording from an internal meeting that same month, CEO Mark Zuckerberg laid out the rationale for tapping Meta's own engineers instead of outside contractors. Alexandr Wang, who sold his data-labeling startup Scale AI to Meta for $14.3 billion before being named chief AI officer and tapped to lead the newly formed Meta Superintelligence Labs, reportedly argued that Meta's average employee has "significantly higher" intelligence than third-party vendors, making in-house training data more valuable. The strategy has so far produced an anxious workforce and at least one viral meltdown, with more than 1,600 employees reportedly caught in the upheaval.