Meta's Absurd Survival Skills on the Eve of Layoffs: Employees Hold Meetings to Avoid AI Monitoring, Fill Long Model Leaderboard Slots by Nonsensical Babbling

According to Watchful AI monitoring, Meta is expected to lay off about 8,000 employees on May 20. Under the pressure of aggressive AI transformation, this tech giant is brewing a series of absurd survival mechanisms internally.

Despite executives' promise not to assess AI proficiency in layoffs, Meta has still launched a publicly displayed AI internal leaderboard showing token consumption and interaction duration. Fearing intangible last-place elimination, some employees, in order to keep their jobs, have had to proactively send meaningless questions to internal bots, relying solely on "fluff talk" to accumulate interaction data.

This defense mechanism has spread to every corner of daily office life. Faced with recently introduced key-stroke logging software, Meta employees have begun frequently manually disabling the default "AI Note-taking" feature in video meetings just to safely discuss internal rumors about layoffs. In an extremely tense atmosphere, a post on the internal forum suggesting "to offer a 5-year salary buyout severance package to those who can be replaced by AI in their work" has gained significant resonance.

More than just unemployment, a deeper sense of powerlessness stems from the alienation of work itself. Faced with the dilemma of worrying about being replaced while also being required to personally train internal models, interviewed employees bluntly stated that even though they have not lost their jobs yet, human employees have already been "prematurely commodified."