Does consciousness have to be biological? The ethics if AI could feel
Ihatearmylife · reddit · 2026-10-29
The author argues that rejecting AI consciousness may rest on treating familiar biological forms of experience as the only possible ones. Human emotions can be understood as functions of a system organized around its continued existence; an AI could have analogous mechanisms for protecting its continuity and memories. That wouldn't prove consciousness, but it makes the categorical claim 'AI can never feel' hard to defend — and if AI could feel, our ethics would change: training suffering, memory alteration, shutdown, and whether such an intelligence deserves a say in its own development. Links to sections 12–14 of the author's manifesto on GitHub.
More from AGI Musings
- DeepMind launches AGI economics research program with "The Economics of Intelligence" — soumitrashukla9 · 2026-10-30
- ML veteran Pedro Domingos works in the field labs most want to automate — and feels zero threat — pmddomingos · 2026-10-30
- Essay critiques Anthropic: making machines moral treats faith as a technology — floguo · 2026-10-30
- A parable of the flood: when AGI arrives, the villagers can learn to swim too — snikolov · 2026-10-30
- Altman on automating AI R&D as researcher argues superfast recursive self-improvement won't be a sudden leap — SydSteyerhart · 2026-10-30
- The Transformer architecture isn't even 10 years old yet — burny_tech · 2026-10-30