The Hidden Cost of AGI: AI Giants Risk Creating a Permanent Underclass

As major tech companies race toward artificial general intelligence (AGI), critics warn that their methods may be cementing a new form of global inequality. The pipeline for training advanced AI systems depends on a vast, largely invisible workforce performing low-wage, repetitive tasks—data labeling, content moderation, and reinforcement learning through human feedback. These workers, often in developing countries, provide the manual labor that refines algorithms, yet they see little of the promised prosperity. While AGI advocates envision a future of universal wealth and productivity, the current trajectory risks locking millions into a permanent underclass, excluded from the very benefits their labor enables. The debate intensifies as companies like Amazon, Google, and Microsoft continue to scale their AI ambitions.