Meta means formally What amounted to an tokenmaxxing The incentive program is a way to reward employees. The social media company announced this week that performance evaluations will no longer depend on how much employees spend. used AI toolsThree employees received this message and told WIRED. Meta workers say they are consuming more tokens as Hatch, a new AI-agent tool is being tested.
Meta said almost a full year ago that employees would be graded based on their ability to work. “AI-driven impact,” This meant, in reality, assessing the extent of their use of chatbots and agent to improve their work day to day. reported at the time. Employees receive labels like “AI Native,” “AI First,” The following are some examples of how to use “AI Enabled,” According to a lawsuit Meta filed a lawsuit in July against about twenty employees who claimed that the company violated US laws on antidiscrimination when they laid them off in may. Meta is disputing the accusations.
This week’s new guidelines for employee performance appraisals replace references to criteria used in evaluating employees. “usage of AI” They are a group of people who have a lot to offer. “AI Native” The designation is a looser term that includes caveats. “these outcomes can be supported by AI or other means.” Meta now grades employees more on impact and less on how they achieve it. Some Meta staff tell WIRED they find the subtle changes to be welcome and that it frees them up from being pressured by their managers into using AI where it makes no sense.
Meta spokesperson Tracy Clayton tells WIRED that the updates this week aim to emphasize what has always been the case—Meta evaluates employees based on their contributions. He says that labels like “AI Native” Performance evaluation is never based on the use of these. This week, engineers across the entire company received a message that they will no longer be able to use the performance evaluation system. “will not use AI adoption dashboards or token counts to evaluate impact,” Information reported On Wednesday,
For months, some Meta employees say, their colleagues prompted AI tools repeatedly—or what they viewed as frivolously—to maximize internal usage measurements, like the amount of tokens. or units of AI, that they consume. When one Meta employee created an internal leaderboard to track how employees use AI and labeled them, for example as a “Token Legend.” The details are below. dashboard leakedThe employee took it In April, Meta’s downtime began. Meta was discovered a couple of months after the April date. moved to ration employees’ AI usage.
More recently, Meta has been encouraging—but not requiring—employees to use its most advanced AI experiment, an agentic tool that can autonomously carry out actions on a computer. Known as Hatch Similar to the viral sensation OpenClawIt is able to browse the Internet and use other applications. Hatch has been available to employees for several weeks, on corporate devices.
According to two Meta employees, who have experience with this issue, some employees are hesitant to link Hatch up to their non-work email, calendar and other digital accounts. They worry about privacy concerns, but also the risk that AI will mess things up. Meta had previously run a project to track employees’ keystrokes on work computers and collect data. This project was run by Meta has since pausedSources claim that the AI tools of the company have eroded the trust many employees had in them.
