In a fascinating twist of fate, the bustling world of Indian gig workers is not only delivering food and cleaning homes but also training robots to do the same tasks. Startup Human Archive has tapped into this burgeoning market by equipping workers with special caps that capture their first-person views during everyday activities.

The tech firm’s unique approach involves working with home services, hotel, and restaurant sectors across India. By collecting egocentric data, they aim to fill a critical gap in training AI for real-world tasks, leveraging the vast pool of human workers in the gig economy.

Despite facing rejections from major players like Pronto and Urban Company, Human Archive has found willing partners among smaller startups. They offer customers a choice: pay a discounted price while consenting to data collection or opt for an unrecorded visit. This innovative model not only funds immediate livelihoods but also builds the infrastructure needed for safer and more productive futures.

To further enhance their offerings, Human Archive is developing additional devices such as tactile gloves and full-body motion capture suits. These tools aim to provide a richer dataset, combining video with other sensor data to better fine-tune AI models.