Boaz Barak puts low odds on AI extinction but offers to debate Scott Alexander anytime in 2035
dhadfieldmenell · x · 2026-09-20
Former OpenAI researcher Boaz Barak says he assigns a low probability to literal human extinction from AI, and offers to publicly debate Scott Alexander on whether AI will kill all humans — "any day in 2035."
However, he rejects the argument that worrying about extinction distracts from other risks: most interventions — pacing, safety and alignment investment, auditing, regulation — mitigate multiple AI risks at once.
While he worries more about non-existential risks such as serious incidents, concentration of power, destabilization, and gradual loss of control, he ends up agreeing with "doomers" on many fronts.
Related event: Boaz Barak Challenges Scott Alexander to Debate AI Extinction Risk(2 posts)→
More from AGI Musings
- Seven shifts in how we use AI: from writing prompts to setting goals — The AI Daily Brief · 2026-09-20
- Nature Says Data-Analysis and Modelling Jobs Are Becoming Obsolete as Data Science Splits in Two — mdancho84 · 2026-09-20
- Nature Report Signals AI Pressure on Data Analysis and Modeling Jobs — mdancho84 · 2026-09-20
- Workshop Recordings Online: Testing for Consciousness in Infants, Animals, and AI — birchlse · 2026-09-20
- A philosophical take: even if AI reaches scientific truths first, humans keep the journey — Fend_st · 2026-09-20
- Musk says AI labs should red-team each other's models — first open-source harness owns the standard — victor_explore · 2026-09-20