Anthropic CEO Dario Amodei has responded to former researcher Jacob Coxon’s resignation, who left the AI company earlier this month and warned that rapidly developing AI systems could “kill us all by the end of the decade.” In a CNN interview, Dario Amodei said he “agree[s] with Jacob much more than I disagree with him”, while making clear that Coxon was not specifically accusing Anthropic of acting irresponsibly. Instead, Amodei said Jacob Coxon was warning about the wider AI industry moving too fast. During the interview, Dario Amodei also explained why he no longer prefers giving a single percentage for the risk of AI causing catastrophic harm, saying the focus should be on choosing the right path for AI development.
What did Jacob Coxon say about AI killing humans
Jacob Coxon, who worked on pretraining research at both Anthropic and OpenAI, resigned from Anthropic earlier this month. In a post on X, he said the companies were not acting responsibly and warned that the AI industry was moving towards self-improving superintelligence.Coxon wrote that AI developers were “racing straight to self-improving superintelligence and gambling with our lives.” He also said people building AI “earnestly believe that it could kill us all by the end of the decade.” His post quickly drew widespread attention, with CNN reporting that it was seen by more than 40 million people.
Dario Amodei says he agrees with Jacob Coxon
Asked by CNN’s Anderson Cooper whether he believes AI could kill all humans and what probability he would put on that outcome, Amodei did not reject Coxon’s concern.“It’s funny, I agree with Jacob much more than I disagree with him,” Amodei said.He then explained why he viewed Coxon’s resignation differently from a direct criticism of Anthropic.“It’s an interesting resignation, because when he left, he said, you know, I think Anthropic is the most responsible player, right? I think Anthropic is the most aware of these issues. He wasn’t calling out us. He was calling out the dynamic of the industry as a whole, moving too fast.”
Why Dario Amodei does not give one AI extinction percentage
Dario Amodei was also asked to put a number on the possibility of AI causing catastrophic harm. He said he used to discuss AI risks in terms of a single probability because it was an easy way to explain the issue. He now believes that approach can hide the different ways AI development could unfold.“I think there’s some chance that things go wrong, but, you know, but it’s not, it’s not all that high,” Amodei said. He then explained that the more useful question is what paths the AI industry takes.“But again, I think it’s more illuminating to think in terms of, you know, how do you decompose those probabilities? What are the paths in which things go well, and what are the paths in which things go poorly?”Rather than saying there is a fixed 10% or 20% chance that AI could cause a disaster, Amodei said the outcome depends on decisions made by companies and governments. “And so instead of saying, you know, it’s a 10% chance, that sounds like it’s a roll of the dice,” he said.He added:“I think it’s much more illuminating to say, well, that forks into different paths. And if we take the right path, then the chance of something going wrong is very low. If we take the wrong path, then the chance of something going wrong could be, could be even higher than that.”“So let’s focus on our agency, and our ability to take the right paths instead of the wrong paths,” he said.