The narrative of artificial intelligence as a self-sufficient, ever-improving technology is a carefully constructed myth. The dirty secret of the AI industry is that its products are built and maintained by a mountain of stressed, underpaid, and invisible human labor. These are the people who teach the machine, correct its errors, and shield the public from its most toxic outputs.
The journey into this workforce often begins with deception. A technical writer from Texas accepted a job as a “writing analyst,” only to discover her role was to moderate disturbing AI-generated content. She received no warning and was offered no mental health support, a common experience for those in the trenches of AI training. The job quickly led to anxiety and panic attacks, a hidden toll of the AI boom.
Life as an AI rater is a constant race against the clock. The pressure to meet quotas is relentless, with management hounding workers who take too long to complete tasks. One employee was told to “just get the numbers done” and not worry about the quality of what she was “putting out there,” even when dealing with sensitive medical information about cancer treatments. This mindset prioritizes productivity over responsibility, with potentially dangerous consequences.
This system, which one researcher calls a “pyramid scheme of human labor,” treats its human components as expendable. While the AI industry is valued in the trillions, the people doing the essential work of quality control face low wages and frequent layoffs. Their story dismantles the fantasy of AI, revealing it to be a product not of pure code, but of human struggle.
AI’s Dirty Secret: It’s Built on a Mountain of Stressed Human Labor
