AI safety leader says 'world is in peril' and quits to study poetry
Summary
The BBC reports that Mrinank Sharma, an AI safety researcher at Anthropic, resigned citing concerns about AI risks, bioweapons, and broader global crises, and will pursue poetry studies in the UK. The piece situates his departure within ongoing industry debates over AI safety, corporate governance, and advertising, highlighting the scrutiny AI firms face from peers and regulators. It also references related resignations, including an OpenAI researcher, to illustrate a broader talent and ethics discussion in the field.