AI companies are paying workers worldwide to record their jobs so robots can learn to do them

This article highlights the emerging ecosystem where human physical labor is being monetized to train robotics AI. Companies like Shift and Encord are paying thousands of workers globally to record everyday tasks, from cooking to cleaning, using specialized hardware to capture first-person data. This initiative addresses a critical bottleneck in the industry: while AI software advances rapidly, robots lack sufficient real-world experience to operate reliably in unstructured environments. The surge in demand for this physical data reflects Silicon Valley’s massive investment in humanoid robots aimed at solving labor shortages and achieving "sustainable abundance." The relevance to open data lies in the commodification of human activity and the opaque nature of this collection. Workers often trade immediate income for the privilege of having their movements recorded, unaware that they are feeding the very systems that might eventually replace them. This practice raises significant ethical concerns regarding consent, wage suppression, and the lack of transparency for contributors. Unlike traditional open data, which often benefits the public good, this "data grab" prioritizes commercial robotics interests, creating a power imbalance where workers are treated as disposable data sources without bargaining power. Ultimately, while the market for AI training data is projected to grow significantly, its long-term viability remains uncertain. Technological shifts, such as improved simulation techniques, could reduce the need for massive physical datasets. However, the current model underscores a troubling trend in AI development: the reliance on exploited labor in developing nations to build advanced technology. This dynamic challenges the open data ethos by privatizing human knowledge and skill, suggesting that the path to autonomous robotics may be built on a foundation of precarious, undercompensated human work rather than truly open and collaborative data sharing.

Source: businessinsider.com
Published on 2026-10-16

AI companies are paying workers worldwide to record their jobs so robots can learn to do them

Related news