Today, The Verge profiled a new Tesla project aimed at training robots for routine materials movement. The company is hiring workers between 5’7” and 5’11” to move 30-pound packages while wearing sensor arrays and virtual reality goggles. These workers will provide the motion capture data to train Tesla’s Optimus robots, which will eventually begin taking over this work.

Motion capture has typically been used in video games and movies to create realistic human movements in digital characters. By mimicking human motions, Tesla is essentially teaching Optimus to perform tasks with human-like fluidity and precision. If this project and others like it succeed—data bottlenecks are the chief problem—it will mark a significant step toward integrating robots into industries such as manufacturing, logistics, and even the service sector.

Of course, robots are expensive to build—at least for now. But human labor is even more costly, especially under the pressures of tight labor markets as the economy continues to grow. For example, in California’s Central Valley, large farms are already using AI-powered robots with lasers to weed fields. These machines sell for $1.3 million but replace the labor of 30 full-time workers who are increasingly difficult to find. The math works: even these expensive robots pay for themselves within a year and continue to reduce labor costs over time. 

There’s another advantage to using robots in Tesla factories and other industries for some of America’s hard, dirty, and dangerous jobs: they save wear and tear on people. There’s nothing romantic or redeeming about doing repetitive, physically demanding work. Robots may wear out and need replacement, but they don’t suffer from repetitive motion injuries or require opioid prescriptions for chronic pain. Robots are coming for some jobs, and for that, we can be profoundly grateful.