Agreed, or even the actual paper[^1]. While the methodology seems pretty well-designed, applicability to real-world type situations notwithstanding, they seem to show that LLM's primary advantage lay in the fact that physicians aren't that good at communicating with a lay audience.

I know Epic for sure is/was looking into GPT-4 integration, but probably not in trying to diagnose patients, because that's just too hairy from an accountability/insurance reimbursement standpoint. It's probably more useful in communicating with patients through an interface like MyCharts.

[^1]: https://arxiv.org/pdf/2305.09617.pdf

From the article:

> Corrado said that while he wouldn’t want it to be a part of his own family’s “healthcare journey,” he believes Med-PaLM 2 “takes the places in healthcare where AI can be beneficial and expands them by 10-fold.”

Hah?

The two sentence summary isn't even correct. "In the study, physicians found more inaccuracies and irrelevant information in answers provided by Google’s Med-PaLM and Med-PalM 2 than those of other doctors."

Looking at the poster's submissions, clout building, probably.