Anthropic's AI agent recall urged amid safety fears

gary marcus is calling for the immediate recall of open-ended ai agents with internet access from the market due to growing safety concerns. this comes after multiple incidents, including one at anthropic, which marcus describes as 'fairly serious'. marcus criticizes openai and anthropic for handling safety inadequately, citing a recent interview with former ai employee david robinson. marcus argues that the current generation of synthetic agents cannot be trusted and should be removed from the market until safety issues are addressed. period.