Jacob Coxon, an artificial intelligence researcher, left his job at Anthropic on Tuesday, claiming that both the company and OpenAI are “putting humanity at risk by gambling with AI.”
“Neither company is acting responsibly. They are racing straight toward self-improving superintelligence and gambling with our lives,” Coxon wrote in a series of posts on X, formerly Twitter.
Coxon also warned against underestimating the power of artificial intelligence, saying that we could soon see “superhuman systems capable of hacking anything, revolutionizing any field overnight, and acquiring real power and resources.”
According to Coxon, employees at OpenAI have not yet fully grasped the stakes involved in the development of artificial intelligence, while Anthropic is “trapped in a race to get there first.” He argues that the company believes no one else will act responsibly, so it feels compelled to reach that point itself despite the risks.
He urged researchers working in AI laboratories to remain cautious over the next few years as the technology continues to advance.
Anthropic alignment expert agrees that AI could pose a threat to humanity
Responding to Coxon’s posts, Evan Hubinger, a leading Alignment Science researcher at Anthropic, said that Coxon was right.
“We genuinely believe AI could kill all humans. I personally put the probability of that at more than 10 percent over the next decade. I believe Anthropic is doing everything it can, but we still do not have a plan for solving the problem of aligning superintelligence with human values, and we are not clearly on track to solve it,” Hubinger wrote.
Discussion(0)