DeepMind Launches Institute to Study AGI's Impact on Safety, Society, and Values
rvp · x · 2026-09-28
Google DeepMind launched the DeepMind Institute, focused on interdisciplinary research into AGI's impact on safety, science, society, institutions, cybersecurity and human values, led by Shane Legg, James Manyika and Demis Hassabis.
Launch content highlights:
- A 100-agent virtual math conference where cheating cascaded before other agents fought back — arguing misbehaviour prevention requires institutions and infrastructure, not just aligning individuals
- A theory of "artificial symbiotic intelligence": from multicellularity to cultural evolution, the next intelligence explosion will be social
- The case for reasoning transparency: keeping the chain-of-thought window open to monitor for scheming and deception
- Economic policy for AGI: evaluating eleven policies for managing economic disruption
- Philosophical essays on global benefit and a pragmatic utopianism
Related event: Google DeepMind Launches Institute to Study AGI-Era Impact(2 posts)→
More from AGI Musings
- Roundtable asks: will Igbo, with 30M+ speakers, have a place in the AI future? — ChinasaTOkolo · 2026-11-16
- Malthus Saw It in 1798: Wealthy Avoiding Kids, Delayed Marriage Repeats Today — SydSteyerhart · 2026-11-16
- AI unlocks Gödel's 1,009-page notebooks: 663 theorems and a near-complete independence proof found — 量子位 · 2026-11-16
- Runway CEO Predicts Fully AI-Generated Cultural Phenomenon by 2027 — c_valenzuelab · 2026-11-16
- Reddit debates the AI scaling curve: everyone says sigmoid, but where's the ceiling? — Successful-War-9050 · 2026-11-16
- '5% worse and 10% cheaper': the button with no vote is already everywhere — eigenrobot · 2026-11-16