AI-narrated audiobooks: good or bad?

Do you prefer quality over qunatity when it comes to audiobooks?  ( Joyce Busola / Unsplash)
Do you prefer quality over qunatity when it comes to audiobooks? ( Joyce Busola / Unsplash)

Apple has just made publishing an audiobook a whole lot easier, with a little help from artificial intelligence.

The company says it is “empowering indie authors and small publishers” by letting them create audiobooks via its digital narration technology, built with help from linguists, audio engineers, and quality-control specialists. All you need is an English-language ebook (in the fiction or romance categories only, for now) and the rights to it.

This isn’t the end for human narration just yet — Apple says it “remains committed to celebrating and showcasing the magic of human narration” — but it could usher a new age of AI voiceovers.

Is that a good or a bad thing? We have outlined the arguments for and against below.

The case for AI narration

Apple Books AI narration (Apple)
Apple Books AI narration (Apple)

The voices actually sound pretty good

This isn’t your average text-to-speech engine and the results — or at least the results that Apple has highlighted — sound very good indeed.

The company has created digital voices that are “optimised for specific genres” and you can listen to extracts read by Madison, Helena, Jackson, and Mitchell — male and female AI voices specialising in fiction/romance and nonfiction/self-development respectively.

Even if these aren’t entirely representative and cherry-picked for quality, the phrase “read by the author” doesn’t usually inspire confidence when hunting for audiobooks either. So “narrated by Apple Books” may not be too much of a downgrade in some instances — especially if you’re the kind of person who hits the 1.5x speed whenever you start a fresh book.

The cost of entry is much lower

Making an audiobook is an expensive business, and a big part of that cost is getting a professional actor or voice artist to read hours and hours of text coherently. Prices vary on a number of factors, but fees of £1,500 per book and up are pretty standard.

By removing that bar to entry, smaller publishers can take a chance on less popular authors, safe in the knowledge that the risk of making a loss is significantly lower.

This saving could potentially be passed onto consumers who pay for each audiobook individually, rather than purchasing via Audible credits or similar.

More choice for listeners

This lower bar to entry means that creating audiobook versions of texts is suddenly open to a lot more authors.

And that ultimately means more choice for consumers, who will no longer be forced to pick only from titles that publishers deem worthy of investment.

But this, of course, raises its own problems…

The case against AI narration

Voice actors could soon find themselves unemployed as AI encroaches (Karolina Grabowska / Pexels)
Voice actors could soon find themselves unemployed as AI encroaches (Karolina Grabowska / Pexels)

Lower quality control

While the idea of fewer gatekeepers is nice in principle, one of the positives about browsing Audible is that, if a book is available, the chances are it’s there based on merit. Yes, you may not personally like any given author, but if the book was popular enough to warrant inclusion, and was deemed worth the cost of digitising, that’s a solid vote of confidence.

There’s a reason “self-publishing” has a bad rep, and while Apple mentions an “editorial review” of any submissions in its announcement, it clearly won’t be as rigorous a critique as publishers give possible manuscripts when their own finances are on the line.

Even if you’ve written a classic that’s worthy of publication, if lesser titles are greenlit, then “narrated by Apple Books” may unfairly become shorthand for “avoid” in the minds of readers.

The emotional intelligence of a SatNav

While the samples provided by Apple sound pretty good if a little generic, some are sceptical that artificial intelligence can provide a reading with the emotional depth or liveliness that can be delivered by a trained, professional actor.

Crudely, would you rather be read to by your SatNav or Sir Ian McKellen?

“The narrator brings a whole new range of art in creating audiobooks, and we believe that’s a powerful thing,” David Caron, co-producer at a Canadian audiobook publisher, told The Guardian. “They’re creating something that is different from the print book, but that adds value as an art form.

“When you have really great writing and really talented narration, you’re coming up with something special. That’s worth investing in.”

Voice actors need to eat

Speaking of investment, choosing AI narration is a cost saver, as highlighted above, but guess who loses out as a result? Voice artists and actors, that’s who.

“Audiobook narration is a vital source of income for our members, and the fact that Apple has quietly launched a catalogue of books narrated by artificial intelligence is precisely why we launched our campaign to Stop AI Stealing the Show,” Liam Budd, Equity’s industrial official for audio and new media, told The Standard.

“AI presents significant disruption to the creative industries and its workforce, and Equity members have been working hard lobbying their local MPs to update our legal framework so that they are better protected in the ever-evolving digital world.”