Using AI for Audio-books

Ok peeps, one of the takeaways from the conference was using AI to help authors. Many ways, from actually writing prompts and assists (which I have no intention on using), assisting with administrative stuff (scheduling, billing, etc.) which I may learn in the future, and the latest and greatest in AI generated audio-books. This latter one is the one that intrigues me.

As some of you may know, I’m using AI to translate my books into foreign languages, then hiring editors in those languages to clean it up (and it definitely needs cleaning). I currently have my Spanish versions out, with French, Polish, and German in the works (all of which costs $$$).

Now comes the real fun – creating the audio-books. I’ve got two options for voices – I can either use an AI voice (likely cloned by a narrator – voluntarily – the company doesn’t steal voices) or clone my own voice.

Feedback I’ve had previously is that a young male with a slight southern accent (the military family standard accent, if you will) would be most suitable for the Corps of Discovery. No input yet on the Hayek Chronicles, so willing to listen.

Another option is that I clone my voice (which takes about 3 hours of me actually reading my books so the company can clone it with my inflections, pauses, burps, whatever) and use that.

One of the biggest advantages of using AI for audio-books is the cost savings, which is passed on to you, the reader. To give you an idea, a typical audio-book costs about $200 per finished hour. A finished hour is the hour of audio you hear after all the narrating, proofing, and editing (which is usually twice as much time as what you typically hear). For my books, that would be between 8 and 9.5 hours per book, so roughly $1,600 to $1,900 (that’s why audio-books are so expensive). Using AI, I would be able to create my books for less than $100 (and it would sound like me, or whatever voice you, my loyal readers, select). That means I’ll be able to distribute my audio-books on my site for the same price as the ebook.

Amazon’s ACX doesn’t allow AI generated audio-books at this time, but I’m betting that will change within the year (especially since they’ve recently invested quite heavily in their own AI for audio-book generation). Kobo and other platforms already allow it, and as I’ll be setting up my own sales page (so you can order directly from me), I’ll be allowing it.

Your thoughts on the voice choice?

Also, for those of you who listen to audio-books, how many of you listen to them at a faster speed than normal?