General-purpose large language models outperform specialized clinical AI toolsnature.com5 points·benwen··0 commentsOpen articleSaveView on HN