There has been work applying AI to medical decision-making, but all used other forms of AI, not LLMs. Now Google trained LLM for diagnoses.
— Ethan Mollick (@emollick) December 4, 2023
In a trial, AI diagnosed 59% of hard cases. 🧑⚕️only got 33% right (🧑⚕️& AI together did well, but not as good as AI) https://t.co/ls3Tu3h8Hp pic.twitter.com/Mv3XzA0zAw
No comments:
Post a Comment