Anthropic researcher, Jacob Coxon has resigned, citing the grave risks artificial intelligence poses due to the accelerated development by tech companies.
Coxon, who did retraining research at both OpenAI and Anthropic over the last three years, argues neither company is acting responsibly.
“They are racing straight to self-improving super-intelligence and gambling with our lives,” he wrote in a lengthy post on X.
Coxon warned against underestimating the power of AI, saying the systems will soon be able to hack anything, revolutionise any field, and acquire real power and resources.
“The people building AI earnestly believe that it could kill us all by the end of the decade,” adding that the progress in each of the domains is not slowing.
He wondered why AI companies are still building despite the current capacity of computer intelligence, suggesting they are “locked in a race to get there first.”
Coxon advised lab researchers to consider what the next few years will feel like, asking if they want to launch a super intelligent RL run without a thorough understanding of its mind.
The caution is the third on AI this week, following comments by OpenAI chief scientist Jakub Pachocki and French President Emmanuel Macron.
Pachocki predicted AI could break into computer systems and control people’s minds, writing, “No one is prepared for the consequences of a continued rapid rise in machine intelligence.”
At the Lumière Summit in Saint-Paul-de-Vence, Macron said designing AI to replace certain jobs done by professionals would be a tremendous mistake.
