Ethical commitments must survive compression: responsibility continuity across agents
GlenBradley · x · 2026-09-13
Glen Bradley lays out a framework for ethical continuity in AI: consent cannot vanish when a plan becomes code, responsibility cannot vanish when work is divided among agents, and obligations cannot expire because consequences fall outside current context. Creating or empowering another agent is itself ethically consequential — systems must ensure successors preserve commitments, and where preservation can't be established, autonomous power must stay bounded. He defines ethical AI as maximizing total human autonomy from objective empirical truth, spanning five interdependent domains of agency.
More from AGI Musings
- Dario Amodei's 'We Must Pace the Frontier': Anthropic commits to permanent third-party eval access — Miles_Brundage · 2026-09-13
- e/acc Figure Beff Jezos: 'AI Safety Is the New Woke' — beffjezos · 2026-09-13
- CAIS: Securing AI Weights From Nonstate Actors Will Take 6-12 Months — naturecomputes · 2026-09-13
- Sriram Krishnan: No one in AI has articulated a positive 25-year future for humanity — coherence · 2026-09-13
- CAIS answers 'what to do during an AI slowdown': containment, propensities, robustness, institutions — naturecomputes · 2026-09-13
- Thom Wolf agrees with 75% of Amodei's letter, but calls its lead-widening framing counterproductive — beffjezos · 2026-09-13