AI Cost Crisis: "Tokenmaxxing" Hits Microsoft, Meta, Amazon

An AI cost crisis is impacting tech giants like Microsoft, Meta, and Amazon. This comes as employees are using AI tools more, a practice some call "tokenmaxxing." What's happening is that agentic AI can use up to 1,000 times more tokens than standard AI. This increased usage is making AI unexpectedly expensive. For example, one team reportedly spent over $1.3 million in token costs in a single month. Some employees at Amazon have even admitted to using AI for unnecessary tasks to boost internal usage scores. This trend is leading to a corporate pullback on AI usage at these major companies. The bottom line is that the high cost of AI, coupled with limited productivity gains, is making it more expensive than hiring people in some cases.