Social media debate is growing over an X post from a former employee of Anthropic and OpenAI, who thinks the tech could kill all humans within the decade—and other AI workers agree.
Jacob Coxon, an AI researcher who claims to have worked for both Anthropic and OpenAI over the past three years, went viral for sharing why he resigned from Anthropic. According to Coxon, both OpenAI and Anthropic are acting irresponsibly and “are racing straight to self-improving superintelligence and gambling with our lives.” His post has amassed nearly 79 million views on X.
Coxon wrote that at its current trajectory, AI will soon create “superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources,” all of which it could do to the detriment of the human race.
“The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt,” Coxon continued. “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 that participating in what amounts to an AI arms race is “a hubristic gamble that should not be launched from a private company’s Slack” and called on other AI developers to reflect on their own role in a potential tech-based doomsday.
“If you are a lab researcher, I urge you to consider what the next few years will actually feel like,” Coxon wrote to conclude his thread. “Do you want to kick off a superintelligent RL run without a rigorous understanding of its mind? Should you put your head down because “it’s happening anyway”—or take this moment to call for different conditions?”
Evan Hubinger, an alignment science lead at Anthropic, responded to Coxon’s comments with a viral post of his own—one that affirmed Coxon’s concerns about the future of AI development.
“We really do earnestly believe AI could kill all humans!” Hubinger wrote. “I personally think it is >10% within the next decade. 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.”
Hubinger directed concerned social media users to Anthropic’s latest Risk Report, which he says shows that “the risk from present models is low.” What he is concerned about, he added, “is superintelligence arising from recursive self-improvement, as we have said is happening faster than we thought.”
Another Anthropic employee, Samuel Marks, also replied to Coxon, saying he too sees the potential for an AI-based apocalypse.
“Many AI developer staff desperately want to slow down to figure out how to build AI more safely,” Marks wrote, pointing toward an open letter he signed calling for an industry-wide slowdown. “I work on safety research at Anthropic because I hope my work will reduce the chance of these extinction-level bad outcomes.”
Senator Bernie Sanders also chimed in with a reply, writing that “Coxon is right.”
“The very people building this technology admit that it could threaten the future of humanity,” Sanders wrote. “That is why I will soon be introducing legislation to ban superintelligence and AI development.”
Though Coxon wrote that he has worked at both OpenAI and Anthropic across the last three years (and other Anthropic employees earnestly engaged with his posts), he has almost no digital footprint. Coxon’s X account was created in January of this year, and until his viral thread, he hadn’t made a single post. He also has no presence on LinkedIn, though The Wall Street Journal reported that Coxon moved from OpenAI to Anthropic earlier this year.
Coxon does have an Instagram with the same username as his X account. There, Coxon claims that he’s lost access to his X account since his post went viral.
“My post reached tens of millions of people because I spoke openly about what I witnessed inside the AI labs building the most powerful technology in human history. Hours later, I can no longer access the platform where I said it,” he wrote. “Maybe this is a technical issue. Maybe it isn’t.”
Fast Company reached out to both Anthropic and OpenAI for confirmation that Coxon was an employee. Neither company has responded.