“We combine a large language model with evidence-based, peer-reviewed clinical guidelines created and maintained by our physician team.”
As much as I’m a fan of LLMs, I’m not sure how much I’d trust them for supporting clinical decision-making. Their responses, while sounding plausible and authoritative, aren’t anchored to reality. There’s no ‘there’ there.
See this conversation between Sam Harris, Gary Marcus, and Stuart Russell, for a good overview of the concerns with LLMs.