Woman listening to audiobook

AI-Generated Audiobooks — I’m Amazed!

Posted by

I have mixed feelings when it comes to technology. In 2018, I wrote a song titled Machines Taking Over the World which portrays my disillusionment with some of the negative sides of our technological age. On the other hand, I’m fully aware that technology has enhanced our lives considerably in many ways, and the benefits should not be understated.

My main vocational focus at the moment is making more people aware of the books I have written. Being a self-publishing author requires a lot more work that some might imagine; it seems there’s always more to do as technology advances and distributors make updates to their platforms. One huge recent change for authors has been the proliferation of audiobooks — yet another format for us authors to think about alongside paperbacks, hardcovers, and eBooks.

I regularly listen to a podcast by Joanna Penn, who describes herself as a thriller author and creative entrepreneur. At the beginning of each episode there’s a news segment where she talks about changes occurring in the publishing industry. She has been discussing audiobooks for some time, but recently her focus has been on the world of Artificial Intelligence (AI) in audiobook production.

I have felt very skeptical about the idea of producing audiobook versions of my books using AI. It appealed to me in a sense, because the costs associated with traditional audiobook production are very high. Generally speaking, you need a studio, and you need to either self-narrate or hire a narrator. The costs of producing an audiobook organically are prohibitive for me as they can run into thousands of pounds or dollars. It’s also quite a laborious and time-consuming process recording and editing the audio.

My interest was piqued recently when I noticed on Google Play that the platform is offering AI-generated audiobooks for free while the service is in beta. I was not feeling great about the idea, as I felt the quality of the AI voice would not be good in areas such as intonation and pronunciation. However, today, just as an experiment, I thought I’d have a go at transferring one of my shorter books into an audiobook format using Google’s text-to-speech technology.

I was absolutely shocked at how good the AI is! When you use the tool, you still need to go through every paragraph and correct issues of incorrect pronunciation, but in the book I chose to convert there were not too many changes to make. And the flow of the AI narrator’s voice does in fact very closely match that of a human being, much to my surprise!

I’m anticipating things could get more complicated if I decide to convert one of my books with footnotes into an audiobook using the Google Play platform. I’m not sure how that will work. And also, there were some niggly issues like my surname being pronounced incorrectly, which I had to update a number of times. But seriously, this technology is very impressive, and I feel confident now that I will be able to produce high quality audiobook versions of many of my books for free!

It’s true that nothing beats an audiobook narrated by the author. And perhaps AI-generated audiobooks don’t work too well with certain genres. But for the kind of non-fiction I write, it’s definitely viable.

One issue is how to make the audiobook versions of my books available once I’ve created them. Most book retailers, including the all-powerful Amazon, don’t currently accept AI-generated audiobooks, though this could change in the future. I’ve heard it said that there could be a clear distinction on audiobook retailer platforms so that they are labelled as either AI-generated or human narrated — the latter of which would likely be more expensive.

The process of producing an AI-generated audiobook on Google’s platform still requires quite a lot of work and editing. But the benefits, as far as I can tell, are huge. So I’m hoping to produce audiobook versions for some or maybe even all of my books in the coming weeks (which will keep me busy!).

There is the outstanding issue of how to make my audiobook versions available. I may need to sell direct to customers using a service such as PayHip or Gumroad which allow files to be downloaded in exchange for payment. But then there’s still the issue of making it easy for buyers to get the files onto their device of choice and have a smooth listening experience, without which they probably wouldn’t want to buy the audiobooks. It would be nice if people could seamlessly add my audiobook files to one of their existing apps, but I don’t currently know whether this is possible (leave a comment if you have any knowledge in this regard!).

In any case, I wanted to share my excitement with you over this new opportunity for authors. If you’ve written books yourself, you may wish to investigate this avenue. If you have any questions about the process, you’re welcome to leave a comment below or email me.

2 comments

  1. I have noticed recent adverts about audiobook versions of books using AI.
    The nearest to that method is using text to speech avatars with an online website service
    The quality of speech is impressive and the amount of wording can be up to 2,000 characters

    Liked by 1 person

Comments are closed.