At the US Book Show on May 22, the topic of AI arose in conversation across multiple sessions, but it was the focus of only one. Broadly AI was acknowledged as an important technological development with far-reaching implications (and not just for publishing), but few expressed urgency. Mainly, it was spoken of as a business concern like any other, with a good dose of publishing’s usual wait-and-see approach.
During the CEO panel, Simon & Schuster CEO Jonathan Karp didn’t like referring to AI as the elephant in the room and instead characterized it as the cicada in the room, with “lots of buzzing and lots of screwing.” He said that authors’ works are the building blocks of generative AI models, that authors and publishers should be paid, and thus he is very happy about the lawsuits now underway. “Eventually it will work its way through the courts and be resolved to everyone’s satisfaction. It’s not worth going to DEFCON 5 just yet.” (Likely he really meant DEFCON 1.) He acknowledged that AI is a valuable tool that will lead to efficiencies and better information gathering that will allow workers “to do work that’s more interesting and creative for them.” He doesn’t think AI will blow up the world.
Mary McAveney, CEO of Abrams, compared the emergence of generative AI to the birth of the internet. “It’s going to be that big, that insidious in our lives. It’s not going away, the genie is out of the bottle. There’s nothing you can do to put it back in.” On the positive side, she said, “There’s a dearth of capability and time in publishing companies to devote to data analytics, where AI can be enormously helpful,” particularly in relation to direct-to-consumer marketing.
Aman Kochar, CEO of Baker & Taylor—who has a background in computer science—pointed out that generative AI works on the principle of collective wisdom. “It can only look back on the experiences it has in its database. It cannot replace authenticity or replace the laborious work that goes into creating a work of art,” he said. He suggested thinking of AI as IA: an intelligent assistant that can support and enhance your work. He believes the technology is here to stay and that the noisy part we’re all now experiencing “will have a short life.”
Peter Warwick, CEO of Scholastic, discussed how AI has affected music consumption, where personalization quickly leads to a system where people only listen to things that are the same or similar to what they listened to before. However, he doesn’t think that will happen to the book publishing industry, because it enjoys (so far) a bigger ecosystem of discovery that involves diverse retailers, libraries, and so on. He said that’s why it’s so important the industry support bookstores, libraries, and other suppliers who can introduce books to readers that they wouldn’t have necessarily picked up otherwise.
In a panel later that day, on current issues in contract negotiation, AI was front and center as a concern. Jaime Wolf, general counsel for the Association of American Literary Agents, questioned Janet Saines-Cardozo, who is vice president and head of contracts at Big Five publisher Hachette. Hachette has started requiring authors to disclose to their editors the ways in which they’ve used AI to craft their manuscripts. Saines-Cardozo said the purpose of disclosure relates to the many lawsuits now underway against generative AI companies; Hachette must be prepared to deal with any claims of infringement against its published works that have used AI. However, she readily conceded, “Every single day, all of this is changing, and we’re just trying to figure out what’s the best way forward.” And despite asking authors to disclose AI use, Hachette is trying to avoid any sweeping contract language related to AI because they don’t know what’s going to happen with the technology. “AI has not hit the iPhone moment,” Saines-Cardozo said. “All of this AI stuff will standardize, but we’re not anywhere near that yet.”
But is disclosure of AI use really enough? Both Wolf and Saines-Cardozo discussed the problematic nature of asking authors to disclose AI use if publishers lack mechanisms for enforcement or policing. Regardless, Saines-Cardozo said there may be obvious cases of AI use by creators that need to be addressed, like a translator who has typically been able to produce three translations a year who then starts taking on 20 translations in a year.
A focused presentation on AI came from Keith Riegert of Ulysses Press, an independent publisher. He said at the outset, “If you run a publishing company and you don’t have an AI roadmap, it’s time to make one.” He believes it’s a critical time to start using AI and repeated the now common take, “AI is not going to take your job. It’s somebody using AI who is going to take your job.” (People disagree on how much this is true—some are more pessimistic.) He recommended that each person in the audience try using AI for every possible task during the day, and use platforms like ChatGPT, Claude, or Midjourney for at least one hour per day to figure out where it works and where it doesn’t—and to keep a close eye on how quickly this changes. He believes AI will dramatically affect the publishing industry and its jobs.
Riegert then offered a very compelling demo of AI to complete common publishing tasks, such as using ChatGPT to write HR or company policies, using Photoshop’s built-in AI to customize a stock image for a book cover, using Claude to pull out important terms from a book publishing contract—or to draft an addendum to a contract—and using Midjourney to create art from scratch. During nearly every demo, there were audible gasps in the audience at how fast, easy, and straightforward it is to complete tasks using AI that would normally take hours or cost hundreds (even thousands) of dollars. You can view Riegert’s demos yourself at a webpage he created to accompany his presentation.
Bottom line: Riegert described generative AI as “really terrifying and really useful at the same time.” Some in the tech industry describe the current generation of AI tools as a way to have an army of free interns, that is: highly intelligent but very inexperienced assistants. These assistants will get better at publishing tasks the more they are trained and the more questions you answer for them. The most powerful applications clearly come from training your own custom AI models, which requires purchasing versions of generative AI tools, then feeding them your own materials. Not everyone feels comfortable doing that, although Penguin Random House launched its own internal custom ChatGPT for employees back in March. During the CEO panel, McAveney probably spoke for many when she said, “We’ve all recognized it’s a double-edged sword. We want to use it, and we’re afraid of it.”
More coverage of the US Book Show
- Publishers Weekly is the co-host of the US Book Show; here’s their take on the entire day.
- Publishing Trends offers a quick look at various panels.
- Book Riot’s Today in Books offers a more slanted or opinionated take, focusing on big-picture themes such as AI and marketing.
Jane Friedman has spent her entire career working in the publishing industry, with a focus on business reporting and author education. Established in 2015, her newsletter The Bottom Line provides nuanced market intelligence to thousands of authors and industry professionals; in 2023, she was named Publishing Commentator of the Year by Digital Book World.
Jane’s expertise regularly features in major media outlets such as The New York Times, The Atlantic, NPR, The Today Show, Wired, The Guardian, Fox News, and BBC. Her book, The Business of Being a Writer, Second Edition (The University of Chicago Press), is used as a classroom text by many writing and publishing degree programs. She reaches thousands through speaking engagements and workshops at diverse venues worldwide, including NYU’s Advanced Publishing Institute, Frankfurt Book Fair, and numerous MFA programs.
