An Ode to A.I. Audio

I recently came across this post at The New Publishing Standard discussing a new partnership between the book trade distributor Ingram and DeepZen, a UK-based software company that is on the cutting edge of computer-generated audio development. DeepZen offers a synthesized audio narrator that it claims can respond to context clues in the text to deliver several different emotions. Click through to the TNPS post for a link to the demo.

Creating a human-read audiobook is expensive, so when a computer becomes capable of narrating as well as a human, there will certainly be a waiting market. While the DeepZen demo sounds impressive, I don’t think it’s ready to compete meaningfully with a professional human narrator. Being able to mimic four or five emotions is a far cry from what a professional actor reading a story can do.

But I love that work is being done in this field. More than that, I love machine narration. I have a fascination with the voices we hear in the airport and while we’re holding for the next available customer service representative. The idea that words can be spoken by a machine at all fascinates me.

Years back, one of my favorite science fiction authors, Cory Doctorow, shared an audio recording made by one of his readers. This reader was visually impaired, and the recording had been made by the software in use at the time for visually impaired readers. This was in maybe 2007, so you can imagine the level of technology. The “narrator” could never be confused for anything but a computer. Yet I found the narration engaging because it was machine-generated. To me, “she” did have a style all her own. The idea of hearing that airport voice read an entire novel to you might sound painful but I thought it was lovely. Perhaps it is the innate human tendency to anthropomorphize, but I found it easy to imagine “she” was experiencing the story right along with me.

So, no, the days of machine narrators crowding out humans are far away, but I welcome the machines just the same. Perhaps it is because I love science fiction. What better way to read a science fiction story than to have an AI read it to you? And I marvel at the work that has gone into text-speech technology and the way it continues to improve.

I went to DeepZen’s website to see what it would cost to have one of my novels recorded. Alas, they only work with publishing companies. I would need a tax ID and all that good stuff just to get a quote. So the AI audio book of Autumn Leaves will have to wait. One of these days…

Leave a comment

Your email address will not be published. Required fields are marked *