May 12, 2026
Bot till you drop
Amazon employees are "tokenmaxxing" due to pressure to use AI tools
Workers say Amazon’s AI race is turning into a weird office game of fake busywork
TLDR: Amazon workers are reportedly inflating use of an internal AI tool because heavy use is being tracked, even if the extra activity isn’t useful. Commenters called it a costly version of fake productivity, though some argued employees simply need to get with the times and learn the tools.
Amazon’s latest workplace obsession has the internet howling: employees reportedly feel so much pressure to use the company’s internal AI helper, MeshClaw, that some are allegedly cranking up pointless activity just to make their numbers look good. In plain English, workers say they’re being nudged to show they use AI a lot, even if that means automating little things that don’t really matter. Amazon says the tool helps people handle repetitive work and insists these usage stats aren’t supposed to decide anyone’s job fate — but commenters smelled a classic workplace trap immediately.
The loudest reaction? This is just “keystroke counting” with extra cost. One commenter compared it to smashing your keyboard for an hour to look productive, except now every digital “smash” can cost real money. Another summed up the whole mess as Goodhart’s Law: once a metric becomes a target, people start gaming it. That idea absolutely took off.
But not everyone was buying the panic. One skeptic argued this might be more of a grumble than a full-blown crisis, saying a new tool naturally needs real-world use and adding that avoiding AI entirely in a software job now is like refusing to use modern drafting tools. Then the jokes arrived: commenters fantasized about bots endlessly checking their own work in a useless loop — maybe with a Tarot reading thrown in for chaos. Add worries that an AI assistant with permission to act for you could go rogue, and the community verdict is deliciously divided: part workplace satire, part warning sign, part meme factory.
Key Points
- •Amazon employees are reportedly using the internal AI tool MeshClaw to automate non-essential tasks and increase visible AI usage.
- •The reported behavior follows Amazon targets for more than 80 percent of developers to use AI weekly and internal tracking of AI token consumption.
- •Amazon has told employees token statistics are not used in performance evaluations, though some staff believe managers still monitor the data.
- •MeshClaw, inspired by OpenClaw, can connect to workplace software to perform actions such as code deployment, email triage, and Slack interaction.
- •Multiple employees cited security concerns about giving AI agents permission to act on their behalf because of the risk of errors or unintended actions.