A Gyges-law quote argues future AI debates may hinge on whether people trust the proof

mimi10v3 · x · 2026-07-21

The post quotes a line about **Gyges' law** and argues that if an LLM is smart enough to compute a clever negative example, it is probably also smart enough to prove a similar positive result — the real problem is that people may simply refuse to believe the positive case. The implication is that future AI debates may increasingly be about **social acceptance and trust**, not just raw model capability. The author concludes that claims about problems remaining open after 2026 are therefore more likely to be true.

Original post →

More from AGI Musings

AGI Musings channel →