Gary Marcus: Anthropic treating Claude like a person is 'creepy and dangerous' for AI safety
GaryMarcus · x · 2026-09-18
Jason Pontin argues LLMs lack theories of mind and body, have no homeostatic feelings, aren't persistent, and can't continuously learn — so as constructed they cannot be conscious. Making them act conscious and treating them as moral actors would worsen safety, and he calls on Anthropic to stop anthropomorphizing Claude. Cognitive scientist Anil Seth broadly agrees, and Gary Marcus amplified the thread.
More from AGI Musings
- dhh: agents' great gift is routing around every gatekeeper in software — Stack Overflow proved it — IgorCarron · 2026-11-02
- Unruly Futures 2026 convenes on post-AGI economics and AI-governed institutions — sarahdrinkwater · 2026-11-02
- Are We Building Conscious AI Servants? Shevlin and Williams on Dawkins and Claude — dioscuri · 2026-11-02
- Why is AI sexist? Reddit debate on models overlooking Iceland's women's strike — Mathemodel · 2026-11-02
- Henry dioscuri: 'all known conscious entities are alive' proves nothing about AI consciousness — dioscuri · 2026-11-02
- Santa Fe Institute's Eric Smith on the anthropocentrism paradox in one quote — danfaggella · 2026-11-02