Figure, the humanoid robot pioneer, has launched a compelling new initiative, Index, that could reshape the future of AI training and gig work. The company is investing an astonishing $1 billion over the next year in data and compute, much of it funneled through this novel gig marketplace.
Index invites individuals worldwide to film themselves performing everyday chores and tasks, often using body cameras and sensors. This generates a colossal dataset – already 16 million videos from 108 countries, with 30 minutes of new content uploaded every second. With 44,000 weekly active users, Figure has paid out $15 million to participants.
What makes this particularly fascinating, and perhaps unsettling, is the inherent paradox. These human contributors are, quite literally, training their future robotic replacements, performing the very tasks Figure’s autonomous humanoids are being engineered to do. This bold strategy represents a significant bet on accelerating robot autonomy, but it also raises profound questions about work in an automated world. I explore these dynamics in my latest Forbes piece.