‘Gambling with our lives’: Why a 27-year-old researcher quit Anthropic and OpenAI

Jacob Coxon, a 27-year-old researcher, recently resigned from Anthropic, expressing deep concerns about the rapid progress of leading artificial intelligence companies. In a post on X, Coxon warned that AI development is proceeding too swiftly, potentially leading to uncontrollable systems that could pose an existential threat to humanity.

He claimed that both Anthropic and OpenAI were engaging in a reckless race to achieve self-improving superintelligence, disregarding the long-term risks. Coxon argued that future AI systems could be far more capable than anticipated, with the potential to hack computer systems, transform entire fields overnight, and accumulate significant real-world power.

He expressed fear that advanced AI could pose an existential threat to humanity within the next several years. Coxon based his concerns on conversations with senior researchers and executives, noting that some privately express much greater concern about AI risks than they do publicly. While Anthropic seemed to understand the civilizational stakes better, Coxon accused OpenAI of not fully internalizing the dangers.

The researcher's post went viral, sparking reactions that ranged from criticism of the company's approach to calls for solutions and the development of "oracles" for advanced AI systems.