Raphael Millière argues intentional stance usefully describes AI agents without anthropomorphism
sebkrier · x · 2026-09-02
Raphael Millière published a long thread arguing that discussions of AI agents are often forced into a false dichotomy between full anthropomorphism and dogmatic deflationism: intentional-stance descriptions can usefully capture agent behavior without attributing hallmark features of human minds. Melanie Mitchell amplified it, asking whether the intentional stance is equally useful for a fictional Odysseus—whose behavior it predicts and causally explains—so what exactly is the difference?
More from AGI Musings
- Researcher says AI-written papers are flooding literature, urges advisors to review drafts — dianarycai · 2026-09-02
- Cohere's research director: pre-training ROI is slowing, focus shifts to test-time compute — sarahookr · 2026-09-02
- Investors Have Poured $94.3B Into 119 'Neolabs' Betting on Research Before Revenue — WhatTheLJW · 2026-09-02
- Researchers sound alarm on surging AI-written papers, urge advisors to read student drafts — CSProfKGD · 2026-09-02
- Sarah Hooker: The Inference Era Is Breaking a Decade of GPU Design Assumptions — sarahookr · 2026-09-02
- Sarah Hooker: test-time compute and agentic workflows are forcing AI infra to change — sarahookr · 2026-09-02