Cambridge's Henry Shevlin: current AI systems may genuinely have beliefs, goals, and agency
dioscuri · x · 2026-09-24
Henry Shevlin (Leverhulme Centre for the Future of Intelligence, Cambridge) published 'Three frameworks for AI mentality' in Frontiers in Psychology, arguing there's a nontrivial chance current systems are conscious and genuinely hold beliefs, goals, and agency — a view Hinton, Chalmers, and Sutskever take seriously. The paper analyzes three frameworks: 'mindless machines' views (architectural debunking via Marr's levels, which he finds too quick, though they separate implementation-sensitive 'deep' concepts from shallow ones like belief/desire); 'mere roleplay' views (psychologically unstable and theoretically incomplete for anthropomimetic systems); and his preferred 'minimal cognitive agents' framework, allowing limited, graded belief-like attributions and shifting from binary to multidimensional conceptions of belief.
More from AGI Musings
- DeepMind researchers make the moral case that AI benefits should reach everyone — Dr_Atoosa · 2026-09-24
- AI's 'stunningly simple' proof of Erdős–Sós conjecture signals humans becoming interpreters of AI discoveries — alejandroll10 · 2026-09-24
- Anti-superintelligence march has just 1,685 pledges toward a 100,000-person trigger, backed by Bengio and Sanders — DavidSKrueger · 2026-09-24
- 93% of Chinese believe AI will help their country vs 36% of Americans — and that gap may decide the race — Dan_Jeffries1 · 2026-09-24
- AI agents could make every cyber attack plausibly deniable, researcher warns — jeremiecharris · 2026-09-24
- Economist maps the order AI will automate economics: theory first, experiments last — soumitrashukla9 · 2026-09-24