The A.I.-Robotics Job Only a Human Can Do

What the report says

The New York Times World reports from India on a growing role in the A.I. economy that still depends on people: workers who wear camera-equipped gear to record how humans move. The footage is then used to train artificial intelligence systems and robotics models to better understand everyday motion.

The piece highlights how this kind of data collection is part of the broader effort to make machines more capable in real-world settings. By capturing human movement from a body-mounted perspective, the work helps developers build systems that can recognize, interpret, and eventually imitate physical actions more accurately.

According to the report, the task illustrates a paradox in automation: even as A.I. and robots are designed to replace or assist human labor, they still require human participation to learn. The story situates India within the global race to gather training data, a process that has become essential to improving machine performance.

More broadly, the example shows how A.I. development often relies on specialized forms of manual labor that are less visible than the technology they support. The report suggests that these jobs may remain important for some time, especially in areas where machine learning systems need high-quality examples of human behavior.