A book about how AI distorts the truth contained AI distortions of the truth.

From American Public Media, this is Marketplace Tech.

I'm Megan McCarty Carrino.

File this under the definition of irony.

Last month, the New York Times reported that a high-profile new book,

The Future of Truth, How AI Reshapes Reality,

includes several quotes that appear to be made up or misattributed.

Author Stephen Rosenbaum acknowledged he'd used AI in the writing process.

One of the quotes came allegedly from a book by data journalist and NYU professor Meredith Broussard,

a critique of trusting algorithms to make decisions in medicine.

We have an unexpected connection to this story, so we called her up for a reaction.

I got an email from a New York Times reporter, Ben Mullen, and he said, I am writing about this book, The Future of Truth, and it includes a quote from you, and I wanted to verify that quote with you. And I looked at the quote, and I thought, that does not sound like me. That does not sound like my writing. I searched the text of my most recent book, more than a glitch, confronting race, gender,

and ability, bias, and tech. And no, this quote does not appear. And then I looked in my earlier book, artificial unintelligence, nothing. Then I searched my entire hard drive from the past 11 years.

And no.

And I looked at it more closely and I thought, you know, this does not at all sound like the way I write, but this does sound like the way that I talk.

And I found that I had in fact said this thing in a 2023 interview with you.

Amazing. Well, this is the actual quote that was misattributed to your book.

People will say things like, oh, we should use an algorithm to decide who gets a kidney transplant, because the algorithm will make an objective evaluation of who is most worthy of a donor organ.

I mean, this is just a perfect illustration of the, what I think is the thesis of the book,

you know, which is about the problems that AI may introduce into our epistemics.

I mean, what did you sort of take from this situation?

I mean, I would argue this is an illustration of the fact that AI is not infallible. It does not actually work the way that most people think it

...