source: techcrunch ai: meta’s months-old ai unit is a soul-crushing gulag, say the engineers stuck inside it
level: business
meta's three-month-old applied ai team is in turmoil after an employee hijacked an internal livestream with an expletive-laden rant, reflecting widespread anger among the roughly 6,500 engineers and product managers forced into the group. many call themselves draftees, saying they were given no choice but to join or quit. their work involves generating puzzles and coding problems to train ai models, which employees describe as soul-crushing.
the unit was formed to support meta's ai research by using its own engineers instead of outside contractors, as ceo mark zuckerberg believes meta employees have higher intelligence than third-party data labelers. this follows meta's acquisition of scale ai for $14.3 billion and the appointment of its founder as chief ai officer. however, the sudden reassignments via surprise email have caused distress, with over 1,600 employees also protesting a program that monitors their clicks and keystrokes for ai training data.
chief product officer chris cox addressed the brutal environment in a call, and zuckerberg acknowledged mistakes in an internal memo, promising to address the issues. the applied ai team, led by a 12-year meta veteran, originally had up to 50 employees reporting to one manager. the unrest highlights the human cost of meta's aggressive ai push amid ongoing layoffs and a shift from metaverse investments.
why it matters: forced reassignments and employee monitoring for ai training data show the ethical and operational risks in scaling ai development, potentially affecting model quality and company culture.
source: techcrunch ai: meta’s months-old ai unit is a soul-crushing gulag, say the engineers stuck inside it