by Paul Bass
Don’t ask Matthew Facciani whether you should get a flu or Covid shot this fall.
Do ask him whom to ask.
You wouldn’t be alone in asking him for vaccine advice. Lots of people are asking him what shots if any to take. That’s because he makes his living studying the rise of public health misinformation and the decline in trust in traditional medical sources at the expense of AI chatbots, blogs, and viral influencers.
But he won’t give medical advice himself.
“I’m not a medical doctor. I’m a social scientist. So that’s the reason why I want to give medical advice. So I think the best place is still to go to your doctor for health advice, like what vaccines to get.”
Facciani, who’s 37, will answer questions at length about how we got in the mess and how we can get out of it.
He studies those questions as a Yale School of Public Health social scientist. He wrote a book called Misguided: Where Misinformation Starts, How It Spreads, and What to Do About It. He hosts a podcast and writes a newsletter about it. He worked on a study showing that only 51 percent of people surveyed always follow their doctors’ health advice when it conflicts with uncredentialled sources like AI social media.
He spoke about those questions at length Tuesday during a conversation on WNHH FM’s “Dateline New Haven” program.
Those questions have come into focus this vaccination-shot season amid the “zombification” of the federal Centers for Disease Control.
While detailing a decline in trust in reputable sources, Facciani hasn’t lost hope in the possibility of new communications tools having an overall positive impact on public health.
“I’m optimistic,” he said. “Because I see the people doing the work and trying to make a difference. And that brings me hope.”
Hope aside, Facciani offered three suggested ways we can all do a lot to break through the fog of misiniformation to take advantage of new avenues of information distribution.
He suggested that researchers bring their research away from paywalls into public view. That way chatbots can scrape that information in providing advice in summaries to consumers. If they do need to paywall-protect an entire study, at least create a searchable one-page summary that chatbots can crawl. He also suggested that public health institutions create more social media about their research
Secondly, he suggested that we all do more checking when we consult AI sources like Gemini or Claude for health advice. Click on the sources they cite. Then cross-check those sources by seeing what people saying about them.
There’s work to do here to change people’s chatbotting habits, based on a 2024 study Facciani worked on.
“We had people use chatbots, and we gave them different sources. Then we asked them how much they trusted the answer based on the sources we provided. We found that most people don’t look at the sources. They don’t even hover over and see where that source came from.”
Finally, Facciani called on AI companies to make public the models and alogrithms that drive their results so we can better assess them. Transparency, he argued, will build trust and produce better results.
Even the 51 percent figure cited above (of people who take doctors’ rather than chatbots’ advice) doesn’t worry him as much as you might think: He noted that that figure includes many people do follow doctors’ advice for the most part but may occasionally follow a different source.
As for his own decisions — Facciani, a Claude customer, did reveal that he got his fall Covid and flu shots. He asked a doctor first. Not Claude. Then followed the advice.
Click on the video at the top of the story to listen to the full conversation with public-health misinformation researcher Matthew Facciani on WNHH FM’s “Dateline New Haven.” Click here to subscribe or here to listen to other episodes of “Dateline New Haven.” Click here to download the WNHH FM app.