News for hackers. Stuff that matters.

A comment on GenAI and erroneous medical references. Parent comment

Re:GenAI and erroneous medical references by photochemsyn on Saturday February 24, 2024 @10:58PM (#39497479)
One of the most basic prompts everyone should be using with questions like this is something like "provide detailed arguments both for and against the presented hypothesis."
[ Parent | on HN ]
Re:GenAI and erroneous medical references by kromem on Saturday February 24, 2024 @11:26PM (#39497641)
This works when the person evaluating can distinguish between the subject matter being correct or not.

But the models can generate compelling devil's advocate positions, particularly when they confabulate supporting facts, which might appear convincing to non-experts.

[ Parent | on HN ]