Employees Are Using Their Jobs’ Super-Expensive AI Tokens for the Most Hilariously Pointless Tasks Imaginable
TL;DR
Futurism frames this as an awkward side effect of corporate AI mandates: employees are told to use AI more and end up burning expensive tokens on low-value tasks. At fintech firm Slash, one employee reportedly spent about $80,000 in AI credits to vibe-code a basic game called „brainrot shooter“. At Accenture, 404 Media reportedly found that much of the token consumption comes from non-engineers using AI for tasks like turning PDFs into PowerPoint decks.
Nauti's Take
This is the next sober phase after the AI hype: not every token bill is innovation, sometimes it is just an expensive screenshot of a broken process. Companies that are serious about AI should stop optimizing for usage and start defining good use cases, cost limits and measurable outcomes.
Otherwise „AI-first“ becomes a permission slip to turn budgets into office theater.
Briefingshow
The issue is not that employees experiment with AI. The issue is that many companies confuse usage with productivity. When budgets, KPIs and leadership signals simply say „more AI“, the result is predictable: expensive automation for tasks that should have been handled with normal tools, clearer processes or no tool at all.
Video
Sources