top of page
  • Writer's pictureThe Publishing Post

AI and Audiobooks

By Kathryn Alley, Cameron Phillips and Sarunicka Satkuruparan

With the tremendous growth of the audiobook world, it’s not surprising that the act of narration itself is being challenged by emerging technologies. Recognising that the audiobook production process is extremely time-consuming, many publishing houses are exploring whether artificial intelligence is a worthy supplement to human narration. An AI narrator’s goal is to be indistinguishable from its human counterpart.

Narrators and listeners alike share concerns for the future of audiobooks and the integrity of storytelling. Emily Lawrence, award-winning narrator and co-founder of PANA, argues that the issue is incredibly complicated and nuanced. Not only will narrators be out of a job with the proliferation of AI voices, but the entire editing and production process could be weakened. In addition, ethical concerns arise when actors’ voices are exploited with unfair compensation. Lawrence describes that AI narration threatens the art that she loves and, whether she is out of a job or not, would be devastated to see the art of narration be horribly compromised.

From a publisher’s perspective, AI could serve potentially as the answer to surging demands and increasing production costs, though AI in audio is a very new concept and one that publishers are trialling with the same trepidation as the rest of the industry. Brian Carroll, Rights Manager at Indiana University Press, expresses how the cost of a narrator and audio production results in most titles never becoming audiobooks, particularly at smaller publishers. If AI voices could result in lower costs for listeners and a greater number of titles, including from small indie presses, being made into audiobooks, then it could offer a win-win situation! Especially because this will give people who are unable to visually read, access to a larger range of books.

The catch? Aside from the threats to the livelihood of voice actors, particularly new voice actors or lesser-known ones working on niche titles, the catch seems to always go back to the same factor when it comes to audiobooks – the power of the narrator. Sceptics of AI are not convinced that a computer-generated voice can ever emulate the emotion that is evoked by a quality narrator...but does it need to?

Taylan Kamis, CEO of London-based DeepZen, is one of several AI developers who claim that synthetic narration is not a threat to professional narrators because their technology will be used to make audiobooks that would not have otherwise been recorded. Kamis claims his technology can increase earnings for narrators who allow him to clone their voice because they will receive royalties. It seems, then, that the alternate perspective sees AI and voice actors as serving different purposes entirely. Rather than seeing AI as a replacement for voice actors, it seems more fitting to think about how they can be used alongside each other.

Anthony Goff, VP of Hachette Audio, hopes that publishers will continue to assess where AI will support the audiobook landscape best. Although AI may never fully replace the art of audiobook creation, the technology is a beautiful supplement to accompany a growing world.

We became interested in this topic after the recent news that James Earl Jones, the iconic voice of Darth Vader, signed over the rights to his archival voice recordings to Respeecher, a Ukraine-based startup, that is working with Lucasfilm in order to preserve and recreate the actor's menacing voice using a proprietary AI algorithm. During the time spent researching this trend, one word kept creeping up in the articles: efficiency. For example, Tom Foley, President of the self-publishing company BookBaby, which has recently teamed up with Speechki to work on AI-enhanced audiobooks, said that it takes his company five days to produce a fully, AI-voiced audiobook. According to Foley, “the narration sounds great – pretty much indistinguishable from human narration.”

Being efficient is very important, but if efficiency is the primary mode of growth, then the quality of work will stagnate. Amongst the other things we have argued in this piece, audiobooks are a unique form of literature whose craft needs to be preserved. If AI is a shortcut to a greater volume of audiobooks of a lesser quality, then it is perhaps not the answer.



bottom of page