> Based on these observations, I'd bet the house that no, large language models cannot under any circumstance reason about medical questions at present. Did yo

> Based on these observations, I'd bet the house that no, large language models cannot under any circumstance reason about medical questions at present. Did you read the paper? You already lost this bet. And even if this paper didn’t exist, it would be a very naive and short-sighted prediction. We’re just at the start of understanding what LLMs can achieve.