AI researchers warn of existential risks as industry faces growing concerns

Employees at Anthropic and OpenAI are raising alarms about the existential risks posed by advanced AI systems. Jacob Coxon, a former Anthropic researcher, resigned after expressing concerns that both companies are racing toward self-improving superintelligence without sufficient safeguards. Coxon claimed that AI leaders privately believe the technology could lead to human extinction by the end of the decade. Geoffrey Hinton, a Nobel laureate, echoed these fears, stating it is 'not unreasonable' to estimate a 10% chance of AI wiping out humanity within a decade. Anthropic's Evan Hubinger and OpenAI's Marcus Williams have also voiced similar concerns, with Williams warning that human extinction is 'very likely' in the next few years without regulation or a coordinated slowdown. OpenAI's Paul Christiano emphasized the need for robust alignment to prevent catastrophic loss of control. Meanwhile, some researchers like Chris Hayduk remain optimistic about AI's potential benefits, while others like Tomek Korbak call for a slowdown in AI development to address alignment challenges.