A new pipeline of AI development is emerging from an unexpected source: gig workers in Nigeria and other developing nations training humanoid robots from their homes. Medical student Zeus, working from his Lagos apartment, represents a growing workforce recording themselves performing everyday movements—raising hands, walking, gesturing—which AI companies then use to train robot locomotion and manipulation systems. These workers, compensated at rates far below Silicon Valley standards, form the invisible foundation of robot development that companies publicize as breakthroughs. The arrangement exploits geographic wage disparities, allowing AI firms to access affordable, flexible labor while workers gain supplementary income.