DeepMind researchers publish book offering Bayesian framework for assessing AI consciousness
mhutter42 · x · 2026-11-01
Researchers from Google DeepMind — including Marcus Hutter, Shane Legg, Anil Seth, and Murray Shanahan among 14 authors — have published Assessing AI Consciousness: From Cacophony to Hierarchy (270 pages, AIXI Media, 2026), now available worldwide on Amazon.
Key contributions:
- Addresses fundamental expert disagreement about AI consciousness — not just conclusions, but what consciousness even depends on
- Introduces a Bayesian framework separating theoretical credence from empirical evidence, enabling precise, scalable, calibrated assessments
- Ships with an online assessment tool and open-source code
The authors argue humans instinctively attribute minds to purposeful behavior (Dennett's intentional stance), and as AI grows more autonomous and behaviorally compelling, pressure to grant or deny consciousness will only intensify — hence the need for a calibrated framework.
Related event: DeepMind Researchers Publish Book on Assessing AI Consciousness(2 posts)→
More from Safety
- Editors Flooded With AI-Assisted Submissions as the Journal Article's Currency Collapses — asusarla · 2026-11-01
- One Claude Code command audits your Managed Agents setup against Anthropic's security guide — brada · 2026-11-01
- Claude found 62 remote-code-execution vulnerabilities in Chrome — update to version 156 — nicolascraske · 2026-11-01
- OpenMoat: open-source local policy checks to secure AI coding agents with machine access — __dotk__ · 2026-11-01
- CTBB podcast Halloween special: exploit chains worth absolutely nothing — rez0__ · 2026-11-01
- gemini-cli PR #30071 routes MCP OAuth fetches through safe dispatcher to stop DNS rebinding — ranas-mukminov · 2026-11-01