Philosopher Eric Schwitzgebel on AI consciousness and the coming crisis of 'debatable persons'
eschwitz · x · 2026-09-05
In a High Variance podcast interview, UC Riverside philosophy professor Eric Schwitzgebel (author of The Weirdness of the World and the forthcoming Human-Like: A Defense of AI Rights) discusses with host Danny Buerkli what happens if we build AI systems whose consciousness we cannot determine:
- The core dilemma: if AI is conscious, treating it as disposable tools could be a horrible crime; if not, granting it human-like rights could also be catastrophic.
- "Design policy of the excluded middle": avoid creating systems that might or might not merit human-like moral consideration.
- Counterintuitive points: useful AI architectures may make machine consciousness harder to avoid, and an abused conscious AI might have the right to rebel.
- Cultural lenses: Kazuo Ishiguro's Klara and the Sun and Ann Leckie's Ancillary Justice on artificial personhood; our ethics may be as unprepared for AI as medieval physics was for spaceflight.
The conversation also covers when anthropomorphizing AI helps vs misleads and Schwitzgebel's intellectual inheritance from a father who studied under Timothy Leary and B.F. Skinner.
More from AGI Musings
- Wei Dai on the lonely economics of long-horizon strategy: you're only paid for being early — DavidDuvenaud · 2026-09-05
- Self-improving agents stop building tools and start building evidence, experiment finds — zit-hb · 2026-09-05
- Indie consultant: clients failing to build AI in-house is what keeps my business alive — gdechichi · 2026-09-05
- AI risk skeptics publicly concede: 'loss of control' risks are no longer vague — dhadfieldmenell · 2026-09-05
- Reddit user proposes an AI "proctor" agent to keep alignment guardrails at machine speed — RecursiveCTE · 2026-09-05
- Were witch hunts an optimal societal shelling fence? Scott Alexander quote goes viral — shakoistsLog · 2026-09-05