AI cracking verifiable math is expected — the real test is unverifiable tasks, argues Sigal Samuel
GaryMarcus · x · 2026-11-21
Gary Marcus reshares Sigal Samuel's take: since math is a highly verifiable task, AI breakthroughs on problems like the Riemann hypothesis should be expected, not shocking. The big open question is whether AI's prowess can generalize to unverifiable tasks — that remains unanswered.
More from AGI Musings
- Gary Marcus open letter to OpenAI's Noam Brown: don't imply months-long o1 runs will yield cancer drugs or Riemann proofs — scaling01 · 2026-11-21
- Terence Tao: Riemann Hypothesis needs something beyond current AI, maybe human-AI by 2050 — ChrisGPT · 2026-11-21
- Build for automation that serves humans: staying in control — gdechichi · 2026-11-21
- Multiplayer coding agents: agent attention is unbounded, human attention isn't — manosaie · 2026-11-21
- Konrad Kording doubles down: AI people should hold put options as a hedge — KordingLab · 2026-11-21
- The bitter lesson, interpretability edition: humans steer only the weakest AI in the stack — burny_tech · 2026-11-21