For the first few seconds, the narrator of Kristen Ethridge’s new romance audiobook, Shelter from the Storm, sounds like a human being. The voice is light and carefully enunciated, with the slow pacing of any audiobook narrator, as it begins: “There’s a storm coming, and her name is Hope.”
Then, something about the pacing of the words grates on the ear. It’s a little too regular, even robotic. “I know that sounds a little crazy,” the breathy voice continues, grinding out the words. “That something so destructive could be labeled with such a peaceful name.”
From sentence to sentence, the cadence of the narrator’s voice glides forward, then snags on an artificial syllable. It’s the aural equivalent of watching the gears of a machine rotate under a surface of what looks like human skin.
“Does it sound exactly like a human voice? No,” Ethridge, the novel’s author, told the Guardian. “But I think the quality is great for AI.”
A USA Today bestselling romance novelist from Dallas, Texas, Ethridge is one of the authors recruited a year ago to join a secretive pilot of Apple Books’ recently launched artificial intelligence audiobooks feature. Apple labels the books as “narrated by a digital voice based on a human narrator”.
Google Play also offers its own “auto-narrated audio books” for digital authors, which includes multiple regional accents for books in English, Spanish, French, German and Portuguese.
Even before she knew that Apple was behind the AI narrator pilot she agreed to join, Ethridge was intrigued. “I know the technology is getting better,” she says. “As we listen more to Alexa, telling us what to do, bossing us around, and we get directions from Waze, AI voices are becoming more ubiquitous in our society.
“Did it sound different in my head when I was writing it? Sure,” she says of “Madison”, the artificial voice who narrated her novels. “But the technology’s emerging.”
The market for audiobooks has boomed in recent years, with an estimated $1.6bn in sales in the US in 2021, a 25% increase on the year before. By 2030, the global audiobook market could reach $35bn, one market research firm estimated last year.
Apple and Google’s bids to automate the creation of audiobooks at a massive scale are likely to spark pushback from professional voiceover actors, and has already prompted skepticism from some publishing professionals, who argue that AI is no substitute for the quality of human narration.
But Apple is targeting its “digital narration technology” at small publishers and independently published authors such as Ethridge, who could be interested in making dozens of their titles available as audiobooks but may not be able to afford the cost of hiring voiceover professionals to narrate each novel.
In the fiercely competitive world of digital publishing, which has low barriers for entry and vast quantities of content for sale, many authors struggle to make much of a living from their writing, even in popular genres like romance.
Many of Ethridge’s readers are senior citizens, living on a fixed income, who are voracious romance readers and want to read multiple books a week but are very “price sensitive”, she says. As older readers, they have “accessibility issues” with reading small print, making audiobooks a good option.
But for independently published authors, hiring a voice actor to narrate one of their books is an “expensive proposition”, which may cost $2,000 to $2,500 per finished book, Ethridge says. That might be a reasonable price for hours of an actor’s highly skilled work, she says, but it’s a barrier for an indie author interested in turning dozens of manuscripts into audiobooks.
“My choice was not between a human narrator and a digital narrator,” she says. “My choice was between not doing audio and doing AI.”
The one novel of Ethridge’s that has been turned into an audiobook with a professional actor’s narration is wonderful, but it costs $21.99, she says, a price that is also out of range for many of her readers.
“Human actors provide a full dramatic range,” Ethridge says. “They know when to inflect on a word. They know when to do a longer pause. They know how to pronounce strange words in a science fiction book.”
But cheaper AI narration is likely to be a good option for readers who care about cost and who “may not necessarily need the fully narrated drama experience”, she says. “A lot of people are becoming more used to listening to these voices.” And the quality of the AI speech may not matter as much, she adds, “if you are listening to the book at 1.5 speed, which I do when I walk”.
While some voiceover artists may worry the AI narrators are going to take their jobs, Ethridge, whose entire catalog of independent novels has now been released as Apple AI audiobooks, says she does not believe that AI narration will ever render human voiceover obsolete.
“If you’re expecting AI narration to be exactly the same as someone who has a Sag-Aftra card who’s reading this, you’re probably setting yourself up for disappointment,” she says, referring to the American union that represents professional voiceover actors. She says she expects the book market “will evolve so that there are two different products”: AI narrators and human narrators, just as the publishing industry sells both hardcovers and paperbacks.
A spokesperson for Sag-Aftra did not immediately respond to a request for comment.
