Image: Distingué CiDDiQi / Unsplash

AI in audiobooks: another slippery slope

It’s the end of the day, and you’re ready to unwind with your favourite audiobook, lovingly read to you by… an AI-generated voice?

The software company ElevenLabs specialises in AI-generated audio. It has just launched a platform called ElevenReader Publishing, which offers free audiobook production and distribution for authors.

The way it works is that authors upload their books, which are then converted into AI-generated audiobooks. Publishers and authors can then choose the voice of the narrator. When this is finished, customers can listen to the book for free on the ElevenReader app. ElevenReader additionally offers analytics tools so that authors can track listener interactions with their book. Authors completely keep intellectual property rights with no exclusivity restrictions and can update their books at any time. Presently, ElevenReader is targeted at small publishers and self-published authors.

Apple Books introduced AI narrators for their audiobooks in 2023, sparking a debate about whether human narrators will be completely replaced by AI

Jack McDermott, head of mobile growth for ElevenLabs, said that they have plans to introduce subscription models for listeners and a marketplace where publishers and authors can sell audio content. ElevenLabs is not the first company to introduce AI-narrated audiobooks. For example, Apple Books introduced AI narrators for their audiobooks in 2023, sparking a debate about whether human narrators will be completely replaced by AI.

However, many authors who could not afford to pay for a voice actor to narrate their book defend the use of AI-generated voices as a way of making their books into audiobooks. Author of Shelter from the Storm Kristen Ethridge used Apple Book’s AI audiobooks feature, saying that “my choice was not between a human narrator and a digital narrator […] my choice was between not doing audio and doing AI.” In this sense, the use of AI for audiobooks is an effective tool to ensure that more authors have the opportunity to publish their book in audio form. It’s a good way of ensuring that those who cannot read, such as blind people, have more audiobook options, which are cheaper.

Nonetheless, there are issues surrounding the allowance of AI in the publishing industry. First, the fact that some people prefer AI voices over humans could be a cause for concern. For example, Jack Kessler, writing for The Standard, discusses the ElevenLabs audio: “AI narration is surprisingly good, and sometimes better than humans.”

There is something comforting about being read a story by someone who has poured effort into reading it with emotion, so that they can make the experience as enjoyable as possible

It’s impossible to miss the irony that humans, who listen to books written by other humans containing human characters, prefer to hear it being read by an emotionless robot rather than a real human. Of course, there are various reasons as to why this may be. For example, AI voices tend to be smoother, possibly making for a more relaxing experience than a human narrator who may have an annoying voice. Kessler writes: “the tone of voice undulates naturally, the pronunciation is pretty much correct.” It therefore seems that readers are allured by the (near) perfection which AI promises. A human voice, with all its quirks and imperfections, does not always make for the smoothest, most perfect option for narration. The comparison of human narration with AI only serves to highlight such imperfections.

It is easy to say that the narrator is simply a medium by which to enjoy a book. Ethridge argues that people who care about cost “may not necessarily need the fully narrated drama experience”. However, it’s a sad idea that listening to an audiobook with a human narrator is now classed as a luxury. There is something comforting about being read a story by someone who has poured effort into reading it with emotion, so that they can make the experience as enjoyable as possible.

With these developments, there is a very real possibility of a slippery slope which leads to the loss of humanity within this industry

Some believe that AI narrators and human narrators will have the ability to work alongside each other in years to come. Ethridge expects that human narrators will never become obsolete and that the book market “will evolve so that there are two different products”: AI narrators and human ones. Nonetheless, even if they find a way to co-exist in the audiobook market, it is undeniable that in this economy, authors, publishers, and customers will most likely be drawn to the cheapest option made available to them.

By accepting AI narrators into the world of publishing, it is important to ask ourselves where the line lies with the incorporation of AI in the creative industry. For example, what is the difference between an AI-narrated audiobook and an animated show with AI narration? With these developments, there is a very real possibility of a slippery slope which leads to the loss of humanity within this industry.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.