DeepMind Researcher Recalls the Day Demis Hassabis Launched the DeepMind Institute
HaydnBelfield · x · 2026-09-16
AI existential-risk researcher Haydn Belfield shares a long thread about the day Demis Hassabis decided to launch the DeepMind Institute:
- He and colleagues Milan, Allan, and Owen had drafted a paper on 'Lessons from History for AGI'; he was set to present to Demis for the first time and was very nervous
- He arrived early only to find a tiny, empty room — waited 10 minutes, then joined a call
- Colleague Steph told him: 'They're in the other building' — prompting a panicked scooter sprint across King's Cross to make the presentation
A first-person account of a milestone in DeepMind's governance/AGI-safety expansion and its internal culture, promised as a thread on history, AGI, and sprinting across King's Cross.
Related event: DeepMind Launches Institute to Study AGI's Impact as Legg Says AGI Nears(35 posts)→
More from AGI Musings
- AI safety researcher: AI companies entirely lack this kind of effective human oversight — davidmanheim · 2026-11-02
- Long-read: AI's real value comes after the efficiency era—infra glut, agent tooling boom — yangyi · 2026-11-02
- vikhyat predicts alternating compute oversupply and undersupply on the road to AGI — TheZachMueller · 2026-11-02
- Sam Altman: AI has surpassed top scientists, and 2026 will be a turning point year — kimmonismus · 2026-11-02
- DeepSeek design doc: a person can only follow 3-5 agent sessions at once, attention is the real bottleneck — heyshrutimishra · 2026-11-02
- AI researcher: we have no reliable way to tell whether an AI system is conscious — dioscuri · 2026-11-02