Gig Workers Around the World Train Humanoid Robots from Home for $15 an Hour
Key Takeaways
- ▸Real-world movement data collected by gig workers is becoming essential for training humanoid robots, as virtual simulations cannot accurately model physics and object manipulation
- ▸Micro1 has built a global workforce of thousands of contract workers across 50+ countries, leveraging tech-savvy populations in lower-cost regions like Nigeria, India, and Argentina
- ▸The robotics industry spent over $100 million annually on training data in 2024, with projections suggesting rapid growth as competition intensifies among humanoid robot developers
Summary
Micro1, a US-based data collection company, has hired thousands of contract workers across more than 50 countries to record themselves performing everyday tasks—folding laundry, washing dishes, cooking—to train humanoid robots. Workers like Zeus, a medical student in Nigeria, mount iPhones on their heads and record their movements, earning $15 per hour, which is competitive pay in their local economies. The real-world movement data collected is becoming essential for training humanoid robots at companies like Tesla, Figure AI, and Agility Robotics, as the robotics industry invests heavily in solving the challenge of teaching machines to manipulate physical objects.
The emergence of at-home data recording as a booming gig economy reflects a paradigm shift in robotics inspired by large language models: just as LLMs learn from vast text datasets, humanoid robots can learn to interact with the world through training on massive amounts of real-world movement data. However, the practice raises significant concerns around privacy and informed consent, with workers often not authorized to discuss their participation. The robotics industry is estimated to be spending over $100 million annually on real-world training data, with investors pouring more than $6 billion into humanoid robots in 2025 alone.
- The gig work raises ethical concerns around privacy, informed consent, and worker autonomy, while providing meaningful income in economically strained regions
Editorial Opinion
The rise of at-home robot training represents both an opportunity and a cautionary tale for the gig economy. While workers like Zeus benefit from competitive wages relative to their local markets, the arrangement exemplifies how AI and robotics companies can outsource labor-intensive data collection to the global workforce while sidestepping some protections and transparency obligations. As billions flow into humanoid robotics, there's an urgent need for clearer consent frameworks and fair compensation practices—especially given that workers' movements are being used to train systems that may eventually displace human workers in factories and homes.
