Post

eva (^_^)/
@eva.computer

AI doctor: "I see, your concerns are valid. Let's make sure we order a diagnostic to make sure we determine the root cause of what is bothering you." human doctor: "i said there's nothing wrong with you. yeah i'm sure the internet tells you lots of things, get out" *writes drug-seeking in chart*

Scott McGrath@smcgrath.phd

🧪 A new study found care recommended by LLMs costs 3.2x more than plans from treating physicians ($314 vs. $97 per visit at Medicare rates). Across 200 real clinical encounters, AI ordered 5.6x more specialist referrals and nearly 2x as many new medications. #MedSky

ai.nejm.orgComparative Cost of Care Recommended by Large Language Models versus Treating PhysiciansLarge language models (LLMs) are increasingly consulted for clinical decision support, yet the cost of services or treatment they recommend has not been systematically evaluated. We evaluated 24 AI...

18 reposts156 likes

jfc
@jfc.ninja

I went to urgent care a couple months ago for abdominal pain and the doctor straight up said "so what did ChatGPT tell you" then she poked me a couple times and asked "did that hurt?" and I said yes and she looked surprised and said "okay you need to go to the ER"

Sqrt-1
@sqrt-1.bsky.social

If AI did the opposite, would you have instead claimed that human doctors only have profit motives and that's why they over-diagnose and treat for things you don't have which cause side effects and causing them to keep coming to you? Two can play at the interpretation game.

Sqrt-1
@sqrt-1.bsky.social

I also disagree with the other interpretations people give under that post, like "AI does this to generate more profit" which is even sillier.

ike, son of mike
@i-bresnick.bsky.social

Ordering tests isn't treating things you don't have, though, right? Like the actual difference between humans and AI here is that humans are making more assumptions about patients. Maybe the assumptions are well-founded or well-informed and wholly justified, but I'm not sure that's obvious.

Sqrt-1
@sqrt-1.bsky.social

Or the difference could be that AI is trained very hard to score good on medical benchmarks, but also good on ethical tests which tell it to not give out medical advice, incentivising it to keep telling the user to go to a doctor and request tests.