I’m happy to share that Technomancer is live on Amazon, and its first sale happened as well. Additionally, Amazon launched a new beta feature to generate an audio book from a Kindle Direct title using AI voices. I know a lot of people are opposed to AI generated audiobooks, but it is essentially free for self-publishing authors, so I decided to dive in.

AI Narration: Pros & Cons

Pros

• Cost-Effective: AI narration significantly reduces production costs compared to hiring professional voice actors. For a book of Technomancer’s length, a professional narrator through services such as ACX would cost me about $2,500 up-front. While they have royalty-based options, not knowing how many readers are interested, this is a costly proposition.
• Speed & Scalability: AI can generate audiobooks quickly, allowing authors to produce multiple titles in a short time. An ACX narrated book will take about ten weeks to produce and publish.
• Consistency: AI voices maintain a uniform tone and pronunciation throughout the audiobook, avoiding human errors or inconsistencies.
• Accessibility: AI narration makes it easier for independent authors and small publishers to create audiobooks without major financial investment.

Cons

• Lack of Emotional Depth: AI voices, while improving, still struggle to convey the nuanced emotions and dramatic delivery that human narrators bring. This is the biggest drawback, for me. There are elements of the text that would benefit from emphasis. It would be nice to feel the angst in the character’s inner voice through the narration, but that’s not yet possible.
• Limited Adaptability: AI may misinterpret complex phrasing, accents, or dialects, leading to unnatural-sounding narration. I would expand this to include the fine-grained control over the narrative voice for multiple viewpoint characters. In the case of Technomancer, there are two main characters as well as seven others who hold the viewpoint at varies parts of the novel. It would be nice if each viewpoint character had a distinct voice, and any interiority or inner dialog could have inflections to show that context.
• Potential Job Displacement: The rise of AI narration could reduce opportunities for professional voice actors. I feel for voice actors on this one. If my work takes off, I would support a live-narrated edition and explore other channels for audiobooks, but for a quick test of the market, the AI solution is hard to beat.
• Quality Control Issues: AI-generated audiobooks may lack the refinement and artistic interpretation that human narrators provide.

The Amazon Experience

Honestly, using AI narration for a published book could not be much easier. If you want a common narrator across the entire novel, just pick the voice you want from ~47 choices. All voices are English based speakers currently, but I expect Amazon will expand to other languages. It’s also possible it only offered me English voices since the book is not translated currently.

You have options of tailoring the voice by chapter as well. I wish they extended the voice selection to individual scenes, which would have allowed me to vary the narrator based on the point of view character, but it is a beta feature, so I expect that will come in the future. I chose to use this feature and vary the narrator between a male and female voice, based on the viewpoint character at the start of the chapter. You can also add pauses in the narration, and adjust pacing within a manuscript. Finally, you can fix any pronunciation errors simply enough.

Overall, while I would like more fine-grained control of the voice selection to let me create a more dramatized experience based on multiple viewpoint characters, creating the audiobook in its current form took me less than an hour.

Final Thoughts

If you are an Audible subscriber, check out Technomancer. If you have moral objections to AI-based audiobooks, feel free to read via Kindle Direct or order a paperback or hardcover copy today.