When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
Amazon-owned audiobook service Audible invites human narrators to create AI replicas of their voices and earn from it.
Amazon plans to rope in human narrators who will volunteer to create speech models that will be synthesized to create audiobooks faster. Audible noted in a blog that it invited a “small group of narrators” to participate in a beta trial of the program that allows them to monetize their voices to create audiobooks using “AI-generated speech technology.” The trial is currently limited to the US.
Audible will—at least during the beta trial—allow narrators control over the projects where their voices are used. To iron out the robotic effect of the AI-generated voices, narrators will re-record sections to “edit pronunciation and pacing.”
Participation is free for narrators and will involve auditions for using voices in audiobooks and live events. Narrators will be compensated using a royalty-sharing model on a per-title basis. Additionally, Amazon promises not to use these voice replicas to create audiobooks or any content without explicit permission from creators.
The development comes several months after Amazon announced its free (but limited-access) tool to enable self-published authors to create AI-generated audiobooks using their ebooks in the Kindle store. This would allow them a 40% share in sales of these AI-generated audiobooks and their ebook sales. Bloomberg reported that more than 40,000 audiobook titles had already been created with AI “virtual voices” by May this year.
However, the program got complaints about a disquieting discovery experience and concerns about virtual voices replacing human narrators entirely in the not-so-long run.
"Virtual Voices" have made book discovery on Audible.com nearly impossible
byu/MPPreads inaudible
The new program encouraging human voice artists to create audiobooks—with their permission—acts as a band-aid to those concerns. However, it does not eliminate the chances of AI sabotaging the artists’ careers—especially as AI models get better at prosody, as ChatGPT’s upcoming voice mode exemplifies.
Lastly, Amazon still needs to eliminate the possibility of utilizing these AI recording sessions interspersed with human accents to train and refine its models for more realistic virtual voices.
