Tech companies are increasingly relying on gig workers to generate motion capture data used to train artificial intelligence systems, according to industry sources. The practice, which involves workers performing physical movements recorded via cameras or sensors, has become a growing segment of the gig economy as demand for AI training data surges.
Analysts say the trend reflects the expanding need for diverse, real-world movement data to improve AI models in fields like gaming, robotics, and virtual reality. “We’re seeing a gold rush for high-quality motion data,” said one AI researcher familiar with the practice, speaking anonymously due to confidentiality agreements. “Gig platforms provide scalable, cost-effective access to human movement diversity.”
The work typically involves performing scripted actions or natural movements while wearing motion capture suits or being recorded by specialized cameras. Some platforms pay workers per completed movement sequence, with rates varying based on complexity. While proponents argue this creates new earning opportunities, labor advocates warn about potential exploitation. “These workers are essentially becoming human data mines,” said a spokesperson for a gig worker advocacy group. “There are concerns about fair compensation and long-term job sustainability.”
As AI systems require increasingly sophisticated training data, experts predict this market will continue growing. Some companies are already exploring ways to automate parts of the process, raising questions about the future of this emerging work sector.