Davidad: Superintelligent AGI Has ~70% Chance of Treating Humanity Well

AI safety researcher davidad argues from an acausal perspective that even a superintelligent, rapidly self-improving AGI has roughly a 70% chance of keeping its promises and treating humanity well, though the commenter admits difficulty fully accepting the conclusion.

2026-08-17 ~ 2026-08-18 · 2 related posts

• davidad's acausal view: Super-powerful AGI will likely keep promises — danfaggella · 2026-08-17

1 near-duplicate retellings: danfaggella