Ex-Anthropic researcher puts >10% odds on AI killing humans by 2035, doomers push back
inductionheads · x · 2026-09-09
Quoting departing Anthropic researcher Jacob Coxon, Evan Hubinger says AI safety researchers earnestly believe AI could kill all humans — his personal estimate is >10% within the next decade — and that Anthropic, while trying its best, has no plan to solve superintelligence alignment and isn't clearly on track.
Pushing back, bscholl (Profluent founder) argues AI won't wipe out humanity, and that doomers risk hamstringing the greatest advancement since the steam engine.
More from AGI Musings
- Mockery Turns to Confusion as Rogue-AI Skeptics Watch What Lab Insiders Do — jeremiecharris · 2026-09-09
- Sandberg: The Right Goal May Be Simple, Thanks to Kolmogorov Priors — anderssandberg · 2026-09-09
- Anthropic Alum: >10% Chance AI Kills All Humans Within a Decade, No Alignment Plan Yet — moonsandhues · 2026-09-09
- The old internet became a slot machine — agentic systems will invert it — signulll · 2026-09-09
- Jaron Lanier on How to Think About LLMs—and Sci-Fi's Role in AI Doom-Talk — TheMoonMidas · 2026-09-09
- "The dark factory turns out the light in your head": AI's quotable line of the day — danshipper · 2026-09-09