Would you go to an A.I. doctor?

KERA's Think

KERA

Society & Culture, 071003, Kera, Think, Krysboyd

4.8 • 861 Ratings

🗓️ 13 October 2025

⏱️ 47 minutes

🔗️ Recording | iTunes | RSS

🧾️ Download transcript

Summary

Your doctor might take weeks to diagnose a complicated set of symptoms when A.I. can do it in seconds. Dhruv Khullar is a physician and contributing writer at The New Yorker, and he joins host Krys Boyd to discuss the use of A.I. in medicine, whether doctors will lose the skills to properly diagnose, and how accurate these new computer-aided diagnoses actually are. His article is “If A.I. Can Diagnose Patients, What Are Doctors For?”

Learn about your ad choices: dovetail.prx.org/ad-choices

Transcript

Click on a timestamp to play from that location

0:00.0	Looking for the stories that matter most to Central Texas, I'm Jerry Kihanal, host of the Austin Signal, your daily dive into the news, music, sports, and culture shaping life in Austin.
0:14.1	We bring you reporting from trusted voices at KUT, KUTX, Texas Standard, and more all in one place. Tune in, stay informed, and get connected every day on the Austin Signal. Austin Signal, weekdays at one on KUT News, or on demand at KUT.org, or wherever you get your podcasts.
0:50.9	When somebody I care about is under the weather, before I use my phone to call the doctor,
0:54.5	I often use it to type in their symptoms and see if I can figure out what might be going on. Sometimes the information is straightforward and helpful, but
0:59.5	when it's contradictory or confusing, I move ahead with scheduling an appointment for a real
1:04.6	diagnosis. So I was surprised to learn that there are now a number of developers working on
1:10.4	diagnostic software
1:11.8	meant for doctors to use. From KERA in Dallas, this is Think. I'm Chris Boyd. As a physician
1:19.7	himself, my guest has a deep respect for the almost algorithmic ability of some highly qualified
1:25.6	doctors to reason through patient symptoms and come up
1:29.1	with an accurate assessment of what ails them. It takes some combination of deep knowledge, experience,
1:35.0	and instinct to work through which symptoms really matter and which might be red herrings and
1:39.6	arrive at a conclusion that leads to effective treatment. But if we can upload those intellectual assets
1:45.8	to a machine, could we reduce the amount of time sick people spend waiting for the right
1:50.9	diagnosis? Dr. Dr. Dr. Kular is also a contributing writer at The New Yorker, where he explores
1:56.4	what is happening in this space in an article titled, If AI can diagnose patients, what are doctors for?
2:02.8	Drew, welcome to think.
2:04.8	Thank you for having me.
2:06.1	You start by writing about a guy named Matthew Williams, and like a lot of people every year,
2:11.5	he went to the ER because he knew something was very wrong, although he didn't know what.
2:16.5	Tell us a little bit about his story.
2:19.2	Sure.
	...

Please login to see the full transcript.

Previous episode | Next episode

Disclaimer: The podcast and artwork embedded on this page are from KERA, and are the property of its owner and not affiliated with or endorsed by Tapesearch.

Generated transcripts are the property of KERA and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.