“This randomized clinical trial found that physician use of an…LLM chatbot did not improve diagnostic reasoning on challenging clinical cases, despite the LLM alone significantly outperforming physician participants”

jamanetwork.com/journals/jam...