The most recent issue of Hot Sheet examined how AI is already being used by authors and editors alike as part of the traditional publishing process. However, most traditional publishers draw an ethical and legal dividing line between AI used to write books from scratch versus AI used to support the writing, editing, and publishing process.
But what about translation?
Machine translation has been around for a long time, but advances in generative AI are leading to a new renaissance in translation. Once again, in partnership with the Book Industry Study Group, publishing industry vet Thad McIlroy moderated a panel to discuss how AI is being used right now to translate and to assist human translators. Panelists included Robert Casten Carlberg, the CEO and co-founder of Nuanxed, a translation agency based in Europe, and Len Epp, co-founder of Leanpub.
McIlroy first offered a caveat: Translating books is a broad and insufficient designation. There are many different types of translations: They can be creative, technical, scientific, or corporate. Translations may be word for word, literal, faithful, communicative, semantic, etc. In other words, there is no simple “just go translate that book.”
And, of course, there are differences in how fiction and nonfiction are treated. Fiction translation tends to emphasize style over substance; nonfiction tends toward the opposite. Because generative AI models are “stylists, not fact engines,” McIlroy said, that makes them better suited to translating fiction than nonfiction. (That said, he strongly implied that people involved in literary translation, considered one of the more prestigious areas of translation, are not all that welcoming of AI tools.)
Does AI work better for book translation than humans? No, everyone agrees on that. But it is incredibly cheaper and faster, and for that reason, it has the potential to grow the market for translations. McIlroy suggested AI translation will be used for books where translation has never been economically feasible. Publishing job growth will likely be in the management of translations and quality assurance.
Nuanxed is a European firm that offers project management for translations for traditional publishing houses. Founded in 2021, they work mostly on translating commercial fiction between European languages, using a hybrid process that includes AI tools before, during, and after translation. They pass savings onto the publishers while still paying a good market rate to human translators. Carlberg said, “Most publishers we start working with are very skeptical to the way we are working but realize once they’ve tried it, the quality is good, and the readers really like it.” And the authors also like it, he added.
Translators who work with Nuanxed start from a pure machine translation and work entirely inside the tools that Nuanxed has built (which also connect to the major generative AI models on the market). For translators, it’s a different way of working but, Carlberg said, once they see this way of working, they do like it. He believes AI will bring a lot more jobs to the industry and create more translation projects. “What gets translated today I would argue is the obvious bestsellers … and the passion projects. A lot of authors never get translated. That’s where we want to come in and help out.”
When asked which AI model Nuanxed prefers, Carlberg declined to name a specific one; he said his company tests all of them continuously and also fine-tunes them to do the detail-oriented work required of a good translation (for example: fixing quotation mark punctuation across languages).
One platform that offers authors unedited AI translations is Leanpub. Leanpub is a self-publishing platform and online bookstore that allows authors to write, publish, and sell ebooks from their site. (It’s popular with authors and readers of computer science books.) Leanpub now offers a GlobalAuthor service that will translate any book on the platform into eight or 31 languages. The price for eight translations is $99 for up to 40,000 words, plus a quarter of a cent per word beyond that; that price rises to $249 for 31 languages, plus three-quarters of a cent per word after 40,000. Authors own their translations and can sell them anywhere that allows AI translations to be sold.
But is the quality any good? “Good enough,” said Leanpub’s Len Epp, mainly because these translations are for prescriptive nonfiction books.
Carlberg said of unedited AI translations, “It’s a different market space” and they serve a different purpose. “I don’t think it would hold for a traditional book publisher hoping to get on the New York Times bestseller list without editing, but I mean it’s a different market.” In his experience, only about 20 percent of the sentences in an unedited AI translation don’t need to be touched in some way. “It can look good if you look at a paragraph,” he said, “But for a whole book to be congruent, [AI] is not there yet. Will it get better? Yes. Will we end up in a situation where a big publisher dares to press a button without having anyone look at it before they start to market it? I think we’re very far from there.”
Bottom line: Carlberg’s firm is growing fast, and he’s hearing from more translators who want to work with Nuanxed. He says their big value add is that they pass every translation through the appropriate “cultural lens” and make sure the work is coherent throughout. Meanwhile, regardless of whether an AI-produced translation is carefully edited, the problem of marketing and promoting the translation in another country remains, especially if you don’t know the culture, customs, or language. McIlroy said, “When I think about the value proposition of having 31 languages, clearly it’s not anywhere near what it would be like if I had publishers from 31 other countries approach me and offer to purchase rights. … It would be their commitment to the book and their ability to promote and distribute in those markets. I won’t have that ability.”
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.
