The corporate AI race is slowly starting to feel less like innovation and more like performance art. Companies desperately want employees to “embrace AI,” employees desperately want management off their backs, and somewhere in the middle, everyone is now apparently automating tasks nobody actually needed automated in the first place.
According to a new Financial Times report, Amazon employees are using the company’s internal AI tool called “MeshClaw” for unnecessary tasks simply to inflate their AI usage scores and appear more aligned with the company’s growing AI-first culture. For context, Amazon’s MeshClaw can initiate code deployments, triage emails, and interact with apps such as Slack, according to people familiar with the matter.
Amazon’s internal AI push is reportedly turning into workplace theater
The report claims Amazon recently introduced internal targets encouraging more than 80% of developers to use AI tools weekly. That pressure has reportedly pushed some employees into delegating low-value or completely unnecessary work to AI agents just to climb internal leaderboards and demonstrate adoption metrics.
And honestly, this feels like the most predictable outcome imaginable. The moment companies started tying employee performance and visibility to AI adoption, it was inevitable that some workers would begin optimizing for “looking AI-friendly” rather than actually being productive.
Amazon is hardly alone here either. As reported by Wired, Meta has reportedly been facing internal backlash from employees unhappy about surveillance-heavy AI training practices, including mouse tracking and monitoring systems tied to AI development workflows. Meanwhile, another recent report suggested even Meta’s own staff are struggling to meaningfully integrate AI into daily work despite leadership aggressively pushing it internally.
The funniest part is that AI is becoming more expensive than actual humans
This is where the entire AI gold rush starts looking deeply absurd. Recent reports by Axios have already suggested that, in several cases, enterprise AI systems are becoming more expensive than simply paying human workers, especially once token pricing, infrastructure, and scaling costs are factored in.
And somehow, despite all that, companies are still laying off employees to aggressively chase AI adoption metrics while many AI firms continue selling products at a loss just to capture market share early. That’s the part nobody seems ready to talk about yet. Right now, these tools are relatively “cheap” because the industry is still subsidizing growth. But once businesses become fully dependent on AI workflows and human jobs have already disappeared, those pricing models could change very quickly.
Honestly, this no longer feels like a productivity revolution. It feels like the tech industry is rushing headfirst into another expensive bubble while real jobs quietly disappear in the background.