LLMs Show Promise and Peril in Personalized Medicine
Large language models demonstrate inconsistent performance in personalized health interventions according to new research. While safety scores remain high, accuracy varies significantly across different medical conditions and patient demographics.