Epoch AI's take on Moravec's paradox: high human variance predicts earlier AI automation
anpaure · x · 2026-11-11
anpaure cites Epoch AI's article on Moravec's paradox with a crisp heuristic: AI automates tasks sooner when equally trained humans vary greatly in performance on them—high variance implies less evolutionary optimization pressure on the learning algorithm for that task.
The Epoch AI piece argues the classic explanation (older skills are harder to replicate) is incomplete—e.g., communication is ancient yet machines out-bandwidth humans—and uses this refined picture to project which economic tasks resist automation: formal reasoning fell decades ago, while perception and motor skills only recently matched humans.
More from AGI Musings
- OpenAI misalignment incident details drip out on X as analyst urges Anthropic to shorten risk reports — DKokotajlo · 2026-11-11
- Marissa Mayer's AI parenting app Dazzle sparks outsourcing controversy — nordicinst · 2026-11-11
- Gamedev debate: as AI slop floods the market, human-made masterpieces may win — TAbrodi · 2026-11-11
- Robotics researchers debate: "Pure robotics research will end next year" — danfei_xu · 2026-11-11
- Ethan Mollick has AI run a Fermi paradox study on public data: 73% odds we're alone in the Milky Way — emollick · 2026-11-11
- Cloud lab founder: chasing full automation is the wrong goal and will fail — anshulkundaje · 2026-11-11