OpenAI's o1 correctly diagnosed 67% of ER patients vs. 50-55% by triage doctors

Re:OpenAI's o1 correctly diagnosed 67% of ER patients vs. 50... by fullstop on Sunday May 03, 2026 @08:55PM (#48003361)

Yes, let's pretend that the bias does not exist, that is helpful. It certainly doesn't have to do with the fact that it's currently a 60/40 split in active male vs female physicians. Or that women are more likely to be taken seriously by doctors:

    * https://www.health.harvard.edu/pain/the-dangerous-dismissal-of-womens-pain 
    * https://pmc.ncbi.nlm.nih.gov/articles/PMC10937548/

Are you really unwilling to admit that such a bias exists?

[ Parent | on HN ]

Re:OpenAI's o1 correctly diagnosed 67% of ER patients vs. 50... by heartbreak on Sunday May 03, 2026 @10:23PM (#48003930)

This seems like an especially bad faith interpretation of the comment you were responding to.

[ Parent | on HN ]