Anthropic researcher Jacob Coxon said on Tuesday that he has left the company because he is worried that the AI lab “will go straight to superintelligence that can iterate on itself and risk our lives.” He also added, “Many people involved in artificial intelligence research and development really believe that by the end of this decade, AI may destroy all of humanity.” Coxson's remarks sparked Anthropic's response to Anthropic's research leader Evan Harbinger, who wrote, “Jacob is right — we do seriously believe AI can wipe out all humans! I personally think this is more than 10% likely within the next ten years. I believe Anthropic is doing its best, but we don't currently have a solution to the super-smart alignment problem, and we're not clearly on the right track.”

Anthropic researcher Jacob Coxon said on Tuesday that he has left the company because he is worried that the AI lab “will go straight to superintelligence that can iterate on itself and risk our lives.” He also added, “Many people involved in artificial intelligence research and development really believe that by the end of this decade, AI may destroy all of humanity.” Coxson's remarks sparked Anthropic's response to Anthropic's research leader Evan Harbinger, who wrote, “Jacob is right — we do seriously believe AI can wipe out all humans! I personally think this is more than 10% likely within the next ten years. I believe Anthropic is doing its best, but we don't currently have a solution to the super-smart alignment problem, and we're not clearly on the right track.”