Tech Giants Admit AI Is Dumb, Expensive, and They Want Humans Back

Microsoft, Uber, and Alibaba are quietly pulling the plug on several internal artificial intelligence projects. It turns out that replacing human brains with silicon chips was a massive, overpriced mistake that actually broke their workflows.

Take IBM as an example. After bragging for years about automating everyone out of a job, they are now frantically hiring junior developers just to patch the digital holes left by their automated "genius" tools. Apparently, someone has to clean up the code salad generated by algorithms that do not even understand what a database is.

Even Nvidia, the ultimate shovel-seller of this gold rush, is starting to spill the beans. Their vice president, Ned Finkle, openly admitted that running these massive AI models is actually more expensive than just paying a normal human salary. When the company making the AI chips tells you your AI project is a money pit, you know the bubble is leaking.

The peak of this corporate comedy happened at Amazon, where managers set a performance metric for employees based on how many AI tokens they consumed. Unsurprisingly, workers spent their days generating endless streams of useless text just to hit their targets, completely wrecking the project's budget.

The grand corporate dream of a fully automated, human-free cash machine has collided head-on with the cold, hard reality of server electricity bills and buggy code. Watching suits crawl back to flesh-and-blood workers they fired just months ago is going to be the most satisfying reality show of the decade.