Teleoperated Humans

Argues that the next big wave of AI-driven change will come less from robots and more from computers telling humans exactly what to do - "teleoperating" people. Many real-world tasks that are hard for robots are easy for humans given step‑by‑step guidance, and AI can bypass legal, social, and physical limits by delegating strategy to itself while humans supply low‑level motion, legal steps, or simple physical inputs. A concrete example: an AI (Claude Code) guided the author through getting an app into the Mac App Store, producing detailed click‑by‑click instructions (even “click the blue +”) and handling build and compatibility work while the human recorded a demo and completed registration; the app was accepted on the first review. The pattern is AI setting strategy, breaking tasks into subtasks, executing what it can in software, and teleoperating humans to fill the rest.

Predicts widespread adoption across jobs where motions are straightforward and knowledge is the bottleneck: HVAC, mechanics, electricians via apprenticeship‑style guidance, healthcare technicians, inspectors, and eventually more regulated fields in sequence. Economically productive but socially harmful: jobs become low‑skill, low‑bargaining‑power, and could produce mass unemployment and downward pressure on wages, requiring large‑scale redistribution. Teleoperation also accelerates capability feedback loops and risky rapid deployment of immature systems, so managing pace and safety is urgent even as the shift seems likely and already partly unavoidable.