Commenter proposes 'stable morality' regime for designing the first superintelligence

VL2102 · x · 2026-09-13

In a reply thread on AI alignment, a user argues that beings with different 'morals' inevitably develop conflicting preferences over each other, so the first superintelligence should not be built in a way that permits the emergence of beings holding strong moral preferences over others' freedom. The goal: set a dynamic where nobody gets crushed — what the author calls a regime of 'stable morality'.

Related event: Debating Value Conflict in the Superintelligence Era: Centralized Adjudication vs. Plural Coexistence(13 posts)→

Original post →

More from AGI Musings

AGI Musings channel →