AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Importance of a Large Language Model in the Brain and Internal Medicine
The idea that a large language model would win was surprising. We didn't make any exclusions, even when the patient was asking you to to interpret an image. If it does well, like it did, it gets in general medicine. The source of bias there, I think is, is minimal. This specific, something like just under 200 questions that you asked you there's obviously a large sample that you could have chosen from ask docs. Our target was to have a random sample of 200 question and answer exchanges involved in a physician. But we ended up with 195.