Jacob Coxon, who previously worked as a researcher at both OpenAI and Anthropic, just announced he's left Anthropic. He's calling out both companies for what he describes as reckless behavior, accusing them of fueling a race that puts humanity at risk.

Who said what

Coxon shared that he spent the past three years working on pretraining research at OpenAI and Anthropic. According to him, neither company is acting responsibly—they're "racing" to build self-improving superintelligence, and in his words, "playing with our lives."

The accusations

Coxon says both companies are locked in a dangerous race to develop systems that can upgrade themselves. He argues that this direction is a threat to people and urges them to ditch strategies that, in his view, gamble with human lives.