LLMs worse at proactive investigation than random heuristicszenodo.org1 point·xylon··0 commentsOpen articleSaveView on HN