BS Summary: This video contains 34 faulty reasoning types, including Hasty Generalization, Appeal to Authority, and Ambiguity (Equivocation), with Negativity Bias as the most egregious example at 37.2% saturation with 197 hits. Analysis detected 1,776 faulty-reasoning hits from 529 analyzed words, generating a BS Score of 80% and a BS Rank of 87% (2,321 of 16,813 videos). This video is worse (more manipulative) than 86.20% of the video peer group.
Millions of Americans get health information from podcasts.
Now, some medical experts are warning about the harmful impacts of misinformation on these programs.
New data from Podcast Index shows that as of this week, more than 45% of new podcasts are flagged as potentially AI generated.
Experts say some of these podcast episodes are spreading medical sounding advice from fake clinicians with no clinical review.
CBS News medical correspondent Dr. Selene Gounder is here. She's also editor at large for public health at KFFF health news.
So why are AI generated podcasts causing sort of a health literacy problem?
So this week researchers published in BMJ open a paper where they looked at uh five different popular AI chat bots uh on how they performed on various different health and medical questions.
And what they found was that half of the answers were problematic.
AI did worst on open-ended questions.
So the kinds of questions real patients are going to ask.
Now those same AI chat bots ask. Now those same AI chat bots interfaces are also powering this flood of health podcasts that are coming out.
Now one company uh Inception Point AI has some 10,000 shows and released uh uh some 20 different different episodes within a period of just 70 minutes.
And they only have 11 people on staff. So that's numerically impossible to do without the help of AI.
So the topics they were covering ranged everything from uh vaccinations to mental health to addiction.
So these are real clinical subjects where if you're getting bad advice, they could be some truly dangerous implications.
We also spoke to the Inception Point uh CEO who told us that this is entertainment.
It's not medical advice. Uh but that doesn't really change what a listener takes away from it.
And look, I'm not your doctor.
Um, and I provide medical information, but I'm not going to say that's entertainment.
entertainment. I stand behind what I say
and make sure what I say is accurate and reliable for our viewers.
>> How can people know that what they're hearing is real and not AI generated?
>> Well, start with the host. Is this a real person? Do they have credentials?
Uh what institution are they are they affiliated with? And then secondly, uh the content itself.
So, um, a trustworthy podcast is going to have nuance, caveats.
They will say there are limitations, when to talk to your doctor.
Um, if every episode sounds super confident and there's no uncertainty, that is a red flag.
In medicine and science, there is always uncertainty about things.
Um, and then check and see whether the show discloses its AI generated.
Unfortunately, a lot of listeners are not going to dig into the show notes and look for that with every single podcast.
Big picture podcasts are a starting point to learn about things.
I listen to podcasts. I'm sure you do, too.
But that should not be your final source of medical information.
>> I had no idea it was permeating podcast, too.
This is a brave new world. Dr. Ling counter. Thank you.
Analysis
Hover over highlighted words in the article to view the associated bias or fallacy analysis.