Revisited: Who's More Accurate - AI Or Your Doctor?

Revisited: Who's More Accurate - AI Or Your Doctor?

Episode Transcript

Speaker 1 (00:00):
Is AI as smart as your doctor? Is AI as
smart as your doctor? Joel has a very nervous look
kind of face about this one. Yeah, I'm I'm hoping
not I'm hoping the humans went out, but I have
my doubts. Well. The long running Joe since the onset

(00:24):
of the Internet age is that if you search for
what ails you online, all roads will eventually lead to death. Right,
doesn't matter what's going on, you will find a way
in which you will die if you keep going. The
takeaway has been that if you know what ails you well,
you might be able to find additional information about how
to treat your condition or to go about living your

(00:46):
life online. If you don't, you probably won't and you
may be likely to overreact. Now, in the age of AI,
everything is changing once again, and when it comes to
the matter of obtaining a proper diagnosis and or are
coming up with a treatment plan, the question has become
whether AI has an edge over your PA. Recent study

(01:09):
put this to the test. The study entitled Artificial Intelligence
versus Human Clinicians a Comparative Analysis of complex medical query
handling across the USA and Australia. They try to get
to the bottom of this and the stated purpose of
the study. They said, this study has sought to explore
the practical application and effectiveness of AI generated responses in

(01:33):
healthcare and compare these with human clinician responses to complex
medical queries in the USA in Australia. Okay, so what
were the findings. A comparative analysis used seventy one hundred
and sixty five patients to assess AI generated responses versus
human clinicians on accuracy, professionalism, and real time performance using

(01:57):
machine learning algorithms and various tests. The study evaluated AI
in human responses again in both the US and Australia.
What was the conclusive statement what they find They said,
the results showed AI generated responses were generally more accurate

(02:17):
than human responses, suggesting potential benefits like increased efficiency, lower cost,
and enhance patient satisfaction. Sounds pretty good, Joel. I know
you're scoffing at this, but hey, if there's more efficiency,
lower cost, and greater satisfaction, yeah right, it does say. However,

(02:40):
significant concerns such as AI's lack of emotional depth, data bias,
and the risk of displacing human clinicians must be addressed
to fully utilize AI and clinical settings. I take that
to me and kind of like, you know, AIS to
the point. So let's say that you did get like, hey,
you're going to die kind of diagnosis. AI's goes are

(03:00):
going to die right right, rather no bedside matter. And
I was thinking about that because that is the weakness here, right.
They're saying that AI's biggest weakness just the lack of
emotional depth. Okay, I don't know about you, but I've
come across some doctors here and there. Yeah, he doesn't
work in that area too. But anyway, what was the
actual result here? The overall average performance of AI tracked

(03:25):
at about eighty percent and overall quality, which notably is
two percent higher than current patient ratings within the United States.
So you put everything together and in real time in
the study, AI beat your doctor beat your medical service
provider by two percent. So there you go. I don't

(03:48):
want a robot doctor, Brian. You might not want it,
but you might get it.