Former Anthropic researcher Jacob Coxon and several other industry veterans have publicly resigned from top AI labs, citing concerns over reckless development and existential risks posed by artificial intelligence.
Key Points
- Jacob Coxon resigned from Anthropic in September 2024, protesting the company's aggressive pursuit of autonomous artificial general intelligence.
- Former employees from OpenAI, Anthropic, and Google DeepMind report a shift toward corporate secrecy and a prioritization of competitive speed over safety.
- Critics argue that internal "safety" teams are often ineffective, leading some researchers to form independent organizations like AVERI and the Coalition of Concerned AI Staff.
- Many researchers face a "systemic bias" toward staying in labs due to high compensation, access to proprietary compute resources, and the prestige of working at the frontier.
- Internal debates persist regarding whether employees can exert more influence by agitating from within or by speaking out as independent whistleblowers.