ICYMI: a large language model was better at diagnosis than doctors?
...and letting the doctors use the large language model didn't improve doctor's diagnoses?!
#OpenAccess article in #JAMA: https://doi.org/10.1001/jamanetworkopen.2024.40969
@ByrdNick This is a wild result.
Lets see if it can be replicated.
Hi @UlrikNyman
Adam Rodman (a co-author on the paper) is on other papers that find something similar. And he’s giving talks on not-yet-published stuff — haven’t seen that stuff yet.