1 of 1
Story summary
- Jacob Coxon resigned from Anthropic and warned that the race to build AI could threaten humanity.
- Evan Hubinger, Anthropic science lead, said Coxon was “correct” and sees AI extinction risk exceeding 10% within ten years.
- OpenAI’s ChatGPT estimated a 5% chance AI could cause human extinction by September 2036, with a plausible 1–15% range.
- Surveys show AI safety specialists estimate 5%–10% ten-year extinction risk, while the AI community sees <=1% risk.
