Transcript: A Good Name On Bad Books

Interview with Jane Friedman

For podcast release Monday, August 28, 2023

KENNEALLY: “Good name in man and woman,” wrote Shakespeare, “is the immediate jewel of their souls.”

Welcome to CCC’s podcast series. I’m Christopher Kenneally for Velocity of Content.

Jane Friedman, an author who writes on writing, has a very good name in the book business. For 25 years, Jane Friedman has put her name to books and blogs about the publishing industry, including The Hot Sheet, a biweekly newsletter for authors.

Bogus books that trick readers into buying a work because they unsuspectingly recognize a familiar title or an author’s name have long polluted Amazon and e-commerce bookselling sites. It’s an issue Jane Friedman has followed and deplored. But when a Friedman fan emailed her about her new book, she quickly learned that her own good name was attached to some very bad books. Jane Friedman, welcome to Velocity of Content.

FRIEDMAN: Thanks so much, Chris. Great to be here.

KENNEALLY: Well, we’re delighted to have you. This is a very interesting topic and a timely one. The books on sale on Amazon offer the keys to a successful writing career, with titles like A Step-by-Step Guide to Crafting Compelling eBooks and Strategies to Skyrocket Your Book Sales on Amazon. But you didn’t write any of them. In fact, you don’t think anyone wrote these books, but they were manufactured by machines. Did you ever think self-publishing would come to this, Jane Friedman – AI-written garbage books?

FRIEDMAN: Well, actually, yes. (laughter) It was easy to see that this was coming. It wasn’t that long ago, maybe two or three years ago, there was a colleague of mine, Jason Boog, who was experimenting with some of these models. It was probably like a baby version of ChatGPT that he was working with. But I remember him talking about it to a group of writers, and I thought, wow, if what he’s describing gets more advanced, which it will, and then becomes public and free, which I anticipated, we’re going to have an avalanche of material enter the market, and probably a lot of fraudulent work as well.

KENNEALLY: How did you react to discovering you and your own good name made the first major case of such pirated books in the AI era?

FRIEDMAN: Yeah, I did not expect that part. But on reflection, it’s not that surprising, considering that nonfiction is perhaps the easiest thing to maybe publish and profit from on Amazon specifically, because it can be more search-driven, so people are looking for topics that they’re interested in learning about. Recently, The New York Times reported on all of the shoddy travel books that are now making their way into the marketplace through Amazon – although at this point, by people like Mike Steves, rather than Rick Steves. So there’s some slight deception going on there. But so far, I don’t know that many of the fraudsters have gone so far as to appropriate people’s names. It’s more flagrantly illegal and more likely to get taken down. But I do think we’ll likely see more of it.

KENNEALLY: Jane Friedman, who writes these books? Or should I ask, how do you think they are written?

FRIEDMAN: My guess is that they’re using a large language model like ChatGPT. It could be one of the paid versions as well, the types that can generate a lot of content on demand. I guess it’s possible that someone’s actually just kind of scraping up content, or they’re having someone write very cheaply. But I’ve experimented with these tools myself. I see what they’re capable of producing. They can produce sentences – sometimes they’re nice sentences – but the deeper you get into the material, the less useful it is. It tends to be very generic as of today. It’s not particularly meaningful. There’s no voice. There’s a very generic quality to it, unless you’ve trained the model on only just a very narrow set of materials.

So I’m not even sure that the people who generated the books in question that were attributed to me – I’m not even sure that they necessarily trained a model on my work. I think they were just asking ChatGPT about writing and publishing, and it spit out things that could have conceivably been written by me.

KENNEALLY: Amazon and Goodreads seem to bear the responsibility for policing their sites, but as a legal issue, the liability there isn’t clear. Jane Friedman, that leaves it to authors to take up the task, doesn’t it?

FRIEDMAN: Well, it certainly seems that way. Practically, that’s the way it is today. But there are reasons for both Amazon and Goodreads to take this issue seriously. There are state laws about right to publicity. There are certainly issues related to the FTC that come into play related to deceiving consumers, and Amazon and Goodreads probably don’t want to be accused on a mass scale of deceiving consumers. And certainly it is bad in the long term for them if there are a lot of books where people can’t trust that they were written by authors if they claim to be written by authors. It’s one thing to write and publish a book with AI and be transparent about it, and some people are. But the books that were attributed to me and a lot of the other material that you’ll find out there – they’re not tipping their hand that it’s AI-generated, so that is deceiving the consumer. And I think it does become a legal issue over time for Amazon and Goodreads and anyone else who’s trafficking in it.

KENNEALLY: How did Amazon and later Goodreads respond to your own concerns?

FRIEDMAN: (laughter) Not all that seriously. That’s probably unfair. I reached their tier one level of service on this issue. And I’ve used these complaints processes before. They are effective when the case is straightforward. So for instance, if it’s a straightforward copyright infringement case, you can get the material taken down as long as you can point to the infringing work. But of course in this instance, I really couldn’t, because it was a new work, and as of today, AI-generated work is not protected under copyright. There has to be a human hand involved. We don’t know what human hand might have been involved here.

So my case wasn’t straightforward, and they don’t have a system for cases that aren’t like a clear violation of either copyright or trademark. Most authors are not trademarking their names unless you’re – again, to go back to the Rick Steves example, I imagine he does have a trademark, and some other authors may need that. But like most authors, I don’t.

So I do believe that the systems in place today to help authors with these types of books – they need to be created. There needs to be different channels, rather than putting the burden on authors to cite the laws in their favor, which they probably don’t know that well in the first place.

KENNEALLY: Amazon did ask you to provide a trademark registration number associated with your name, Jane Friedman. Does trademark law offer authors comfort from piracy?

FRIEDMAN: Maybe some. There is the Lanham Act in the United States which pertains to trademark law, but you don’t necessarily have to have a registered trademark in order to take advantage of that law, because you can claim false advertising – you know, this is a passing-off situation. And the Authors Guild – in fact, when my situation came to their attention, that’s exactly the grounds on which they demanded Amazon take down these books as a violation of the Lanham Act.

So it can be effective, but I don’t think it’s the ideal vehicle for remedying this problem. Obviously, many authors go by the same name. My name is common enough. So it’s not the best solution. I think there are other things that could be done.

KENNEALLY: ISBNs, the essential identification numbers for the retail book market, are available for free from Amazon’s ebook self-publishing platform. You allege that these free ISBNs are supercharging the distribution of AI-generated materials to the wider market, because the so-called authors are hiding behind Amazon. You also suggest that Amazon should begin charging for ISBNs. How would that move help in the fight against these garbage books and even help self-published authors?

FRIEDMAN: For many years now, Amazon as well as Ingram, Draft2Digital, and some other publishing service providers and distributors – they’ve offered ISBNs to authors in the United States at no cost. It’s a shame that we’re in this position, because in other countries like Canada, New Zealand – those authors get their ISBNs for free. So this has made authors in the US and UK just that much more likely to take the free ISBN. It’s not that I necessarily have the problem with the ISBNs being free. It’s that the way they’re distributed in the US, at least – it does provide this cover for the people who take them, and then that allows the books to be distributed far beyond Amazon and through Ingram’s systems.

So we’re missing some sort of breadcrumb trail or accountability for what is the source of that book and also being able to connect the dots among all the titles that were claimed or purchased or obtained by the publishing or author identity. It’s not necessarily that I think Amazon should start charging, but I do feel like we need stronger identities connected to the ISBNs being used, rather than Amazon having thousands upon thousands that they’re distributing, and the publisher of record is simply Amazon, and they’re the only ones who know where the material originated from. I find that problematic.

KENNEALLY: Well, Jane Friedman, thank you for joining me today on Velocity of Content.

FRIEDMAN: Thank you, Chris.

KENNEALLY: That’s all for now. Our producer is Jeremy Brieske of Burst Marketing. You can subscribe to the program wherever you go for podcasts, and please do follow us on Twitter and on Facebook. You can also find Velocity of Content on YouTube as part of the CCC channel. I’m Christopher Kenneally. Thanks for listening.