Sue Sheridan's AI story at the WHO and in the New York Times

Also: new KFF study finds AI use is far wider than expected.

Two weeks ago this blog covered how patient safety legend Sue Sheridan suffered a third medical error but achieved a partial rescue using ChatGPT: Sue Sheridan is hit with another medical error. This time, ChatGPT prevented the worst.

The story is starting to get appropriate attention globally. She was already invited to participate in a global consultation on diagnostic errors at the World Health Organization in Geneva, Switzerland, and when the organizers saw the blog post, they switched her topic: they wanted her to tell her AI story.

She did, this Tuesday:

Sue speaking Tuesday at the World Health Organization in Geneva. Photo by Twitter user @PilarAstier.

Note that the facial droop of the Bell’s Palsy she suffered in this incident has not fully resolved; it remains to be seen whether it ever will. (For this condition, the delayed diagnosis and treatment she experienced is a risk factor in the final outcome.)

Then, separately, the New York Times called! I talked with reporter Teddy Rosenbluth for 45 minutes about #PatientsUseAI; at the end she said her next call was with Sue. And when her story “The Chatbot Will See You Now” appeared in today’s Times, the lead was the patient story, as it should be:

Susan Sheridan had heard of ChatGPT but had never used it, so the first question she asked the artificial intelligence chatbot was a bit garbled: “Facial droop, facial pain and dental work.” …

Yes, #PatientsUseAI. KFF finds one in six adults use it for medical advice monthly.

Rosenbluth’s article also links to a study I hadn’t heard of, from KFF, the excellent independent health news source. (I’ve followed them since 2011.) You can read their whole article, but for our purposes I’ll add the key findings below this KFF graphic. Note, this chart is about all AI use, not just healthcare:

Note these key findings:

• A tenth of everyone uses AI several times a day! (Would you have imagined that??)

• That figure is amazingly consistent across all age groups, races, and party lines: 9-14% of every category

• Nearly half of us “old people” use it! (65+; I’m 74.) Only 55% say they’ve never touched it. Who would have predicted that nearly half of Americans 65+ have used AI???

In a future post we’ll get into the health-specific parts of that study - especially the public’s wise skepticism about AI’s health advice. (Hooray!) But for now, notice one key thing:

Until recently, Sue Sheridan and her husband were in the “never touched it” category. No more.

So now a big question faces all of us: how do you use it wisely? We’ll be working on that here, too. Remember rule 1, “Hugo’s Rule”:

I don’t use AI to give me answers.
I use it to help me think.