AI researcher: we have no reliable way to tell whether an AI system is conscious
dioscuri · x · 2026-11-02
AI researcher dioscuri (David Krueger) pushes back on "if we build conscious AI, we must do X" arguments: there is no agreed set of behaviours displayed by all and only conscious entities, so if/when we create conscious AI, it won't be clear-cut.
The point highlights a methodological gap — debates about moral obligations to AI rest on premises we currently have no way to verify.
More from AGI Musings
- AI safety researcher: AI companies entirely lack this kind of effective human oversight — davidmanheim · 2026-11-02
- Long-read: AI's real value comes after the efficiency era—infra glut, agent tooling boom — yangyi · 2026-11-02
- vikhyat predicts alternating compute oversupply and undersupply on the road to AGI — TheZachMueller · 2026-11-02
- Sam Altman: AI has surpassed top scientists, and 2026 will be a turning point year — kimmonismus · 2026-11-02
- DeepSeek design doc: a person can only follow 3-5 agent sessions at once, attention is the real bottleneck — heyshrutimishra · 2026-11-02
- Toby Ord: track both what models can and can't do to see real capability gaps — tobyordoxford · 2026-11-02