Navigating Unfalsifiable AI Qualia Claims via Anti-Fragility

sreeramkannan · x · 2026-08-31

Addressing unfalsifiable claims about AI qualia, the author proposes an 'anti-fragility' heuristic. By comparing outcomes under different assumptions, the strategy suggests choosing actions that minimize human fragility: treating AI as programs without rights is safer than assuming they have undetectable qualia that could be exploited.

Original post →

More from AGI Musings

AGI Musings channel →