AUTO-UPDATED

Inside the race to train robots on human work

Startups are paying thousands of workers globally to record their daily physical tasks, creating massive datasets intended to train humanoid robots for future labor in homes and factories.

Key Points

  • Companies like Microagi, Mecka AI, and Human Archive use head-mounted cameras and sensor-equipped gloves to capture human movements, force, and physical interactions.
  • Demand for robot training data has surged tenfold in the past year, with robotics firms seeking up to 100,000 hours of annotated footage to improve machine reliability.
  • Global data collection efforts span over 30 countries, with workers performing tasks ranging from cooking and cleaning to repairing jet engines.
  • The AI training dataset market is projected to grow from $11.9 billion this year to nearly $49.8 billion by 2031.
  • Robotics companies raised $33.4 billion in the first half of 2026, driven by investor interest in automating labor and addressing workforce shortages.

Why it Matters

The race to build functional humanoid robots is currently bottlenecked by a lack of real-world training data, turning human labor into a critical commodity for AI development. While this creates immediate income for global workers, it simultaneously accelerates the automation of the very tasks they are being paid to record.
Business Insider Published by rjetha@businessinsider.com (Rya Jetha)
Read original