Embracing AI narration
March 6, 2020
by Ivan Raconteur

The Literally Public Relations Blog announced this week that the future of audiobooks is artificial intelligence (AI).

This piqued my interest, because I listen to audiobooks nearly every day.

I started this when days spent in front of a computer screen or editing copy changed reading at the end of the day from the pleasant experience it had always been to a burden, because of the strain on my eyes.

I have always found reading before I go to sleep a good transition that helps me relax.

When age and over-use started to take a toll on my eyes, I wasn’t willing to give up my daily dose of reading for pleasure, so I looked for alternatives.

Audio books were a natural fit.

Listening to an audio book provides much of the same enjoyment of reading, but I can do it while resting my eyes.

Audio books are widely available, and many choices can be checked out digitally through local libraries.

However, not all audiobooks are created equal. I discovered fairly quickly that I enjoyed some readers more than others. In fact, when it comes to my favorite authors, I am very particular about who is doing the reading.

I suppose once I am used to hearing familiar characters read a certain way, it can be disconcerting to hear them read a different way, much like seeing a new actor interpret a role might seem less satisfying if we liked the way it was done the first time we saw a piece performed.

That leads me to question whether or not I will like audiobooks created using AI.

According to the Literally Public Relations Blog, “British company DeepZen (www.deepzen.io) produces and co-publishes audiobooks and other voice content using Artificial Intelligence that replicates the human voice to create a listening experience that is virtually indistinguishable from traditionally-narrated audio.”

That seemed like a rather bold statement, based on some of the digital voices I have heard in the past.

I wanted to find out for myself.

I clicked the links that were thoughtfully provided in the piece, and I have to say, I was impressed.

Although not perfect, the voices I listened to in the samples were quite good. They were easily as good as some of the human readers I have heard in the past, and I had no difficulty in concluding that I would be able to listen to a book read by these voices.

They have a much more natural sound than the digital voices I had heard previously, and instead of being flat, they include the kinds of natural inflections and pauses that a human reader would provide, and, for the most part, the inflections and phrasing seemed to fit the copy in a logical way.

I was skeptical at first, but I will definitely explore the world of AI narration in more depth.

The announcement notes, “DeepZen’s digitally narrated titles are the first to gain broad acceptance and distribution from major re-sellers including Apple Books, GooglePlay, Rakuten Kobo and subscription services Scribd, Nextory, and Storytel. They will also be available on all major library distribution platforms including Overdrive, Bibliotheca, and Baker & Taylor.”

This is good news for those of us who rely on audiobooks. It means there are likely to be more selections available in the future.

I haven’t the faintest idea how these companies produce the AI audiobooks, but whatever they are doing, they seem to be on the right path.

I can also see how this technology could be used in education and in business.

Many of us have likely had the experience of calling a customer service number, only to reach a human being who behaves like a robot, and whose range of ability seems to be working off of a script. A customer service representative created by AI could accomplish at least that much, and it could do so while maintaining a pleasant “attitude,” and speaking in a clear voice that customers can actually understand. In that scenario, AI would be a significant improvement.

Change can be difficult, but I’m cautiously optimistic about where this could lead.

Advertise in over
250+ MN newspapers