An AI ethics framework rooted in thermodynamics, information, and stability

AryHHAry · x · 2026-07-22

The author argues that ethical AI frameworks should be grounded in physics rather than compliance checklists. They propose pillars such as energy cost, information, Lyapunov stability, and entanglement, and describe a “Golden Triangle” of entropy, symmetry, and information.

Their main claim is that AI should be evaluated against physical law: if a system cannot be assessed that way, it is just expensive noise. They frame the goal as building a scalable governance framework that can extend from classical AI to quantum AI.

Original post →

More from AGI Musings

AGI Musings channel →