Our research has several limitations and should be interpreted with appropriate caution. Firstly, our evaluation technique likely underestimates the real-world value of human conversations, as the clinicians in our study were limited to an unfamiliar text-chat interface, which permits large-scale LLM–patient interactions but is not representative of usual clinical practice.

> Google AI outperforms medical doctors on diagnostics tasks

> URL: blog.research.google

[X] Doubt

(a) Not using AI at all is costing lives, and (b), using a human at all is costing lives

Given the mediocrity of the PCP I just left (not to mention the attitude and low key insults), I am not surprised.