Google Researcher: Doom Scenarios Make Sense Only With Gradual Disempowerment First
moultano · x · 2026-09-18
Google research scientist David Pfau (moultano) offers a take on AI risk framing: all the doom scenarios make more coherent sense if you first assume gradual disempowerment — humanity's control eroding step by step — rather than a sudden catastrophic event. It reframes the debate around process-based power transfer as the prerequisite for worst-case outcomes.
More from AGI Musings
- TheZvi mocks AI safety grant reviews: real safety or capabilities work subsidizing labs? — TheZvi · 2026-09-18
- What if an AI proves P=NP with a proof no human can understand? — MoonL88537 · 2026-09-18
- Why AI researchers who believe in extinction risk keep building: inside the exodus — adariostrange · 2026-09-18
- Microsoft's Humanist AI Code of Conduct Slammed as Counterproductive on Alignment and Safety — rgblong · 2026-09-18
- Arvind Narayanan's ICML Keynote: RSI Is Not the Singularity, and No Lab Milestone Kills All Jobs — random_walker · 2026-09-18
- Databricks CEO on Bloomberg: existential AI risk is close to zero, cybersecurity is the real priority — matei_zaharia · 2026-09-18