Jacob Coxon announced his resignation from Anthropic, the AI research company known for its Claude model, on September 8. His departure has raised concerns within the organization, as highlighted by his colleague Evan Hubinger, who supported Coxon’s decision and expressed similar apprehensions in a post on X.
They are racing straight to self-improving superintelligence and gambling with our lives,” adding “Do not underestimate the power of this technology. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing. Coxon continued “A common response is ‘if they truly believe this, why are they still building it? At OpenAI, many have not deeply internalized the civilizational stakes. At Anthropic, the stakes are well-understood, but they are locked in a race to get there first – they believe no one else will act responsibly, so they must do it themselves, despite the risk.”
Coxon wrote “I resigned from Anthropic today.
I personally think it is >10% within the next decade.
He also said “Accepting this race and entering the “endgame” is a hubristic gamble that should not be launched from a private company’s Slack. Attempting to speedrun alignment should require extraordinary confidence that there are no better trajectories available,” adding “I am optimistic about the potential for coordination. Should you put your head down because “it’s happening anyway” – or take this moment to call for different conditions?”. What I am worried about is superintelligence arising from recursive self-improvement, as we have said is happening faster than we thought.
Warning shots like the Hugging Face attack have made pacing agreements between U.S. labs more viable. I don’t feel like we’re on track to prevent a global race, which may require costly actions such as a temporary ban on improving model capabilities. Coxon concluded “If you are a lab researcher, I urge you to consider what the next few years will actually feel like. Do you want to kick off a superintelligent RL run without a rigorous understanding of its mind? Notably, before Anthropic, he worked for OpenAI, the company behind ChatGPT. Hubinger agreed with the part about AI possibly killing all humans. He wrote “Jacob is correct here—we really do earnestly believe AI could kill all humans! I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to. As someone currently working at Anthropic, Hubinger added “To be clear, as we say in our latest Risk Report, I think the risk from present models is low. Hubinger’s previous work experience includes OpenAI, Google, Yelp, and Ripple. He studied at Harvey Mudd College and is a Fellow at the Machine Intelligence Research Institute.
