A few times in my life, I think I qualified as a bona fide futurist.
Around 1976, while employed in radio and TV, I sat in the station’s lunch area with a TV meteorologist and a few other performers. There I stated that if I had the money, I would open a store that rented movies on tape. My colleagues chuckled and gave the whole idea a thumbs down. The idea seemed a little far-fetched even to me, but I did feel like it would come to pass. Well, the first video rental stores started opening around 1977. They’re gone now, but they were the juggernaut of home entertainment for decades.
In the mid-’80s, at the same radio station, I pulled the Chief Engineer into my office and asked the question; Why couldn’t a broadcast signal be digital? He screwed up his face and said, “It’s all a matter of bandwidth. It will never happen.” Today, digital broadcasts, both over-the-air and online, are standard.
In the ‘90s, I was in the middle of a successful career as a voice-over artist and produced audio tracks for all kinds of projects in advertising and industry. While interacting with fellow voice talents both in-person and online, I predicted that text-to-speech technology would someday threaten our jobs. I don’t recall anyone agreeing with me. But, I persisted in my belief that it was inevitable and said so publicly twenty years ago in my company newsletter in August 2003.
It’s time again for a big ol’ “I told you so!” At the beginning of this month, Apple rolled out a digital narration service for audiobooks. The service uses artificial intelligence-generated voices based on real human readers. The quality of these AI tracks is impressive. They can be heard here.
I’m retired now, but I should disclose that I never saw audiobook narration as a viable business model and decided not to enter that market long ago. I much preferred making $5000 during a three-hour advertising session instead of making the same amount by spending weeks recording, proofing, and editing a 400-page audiobook.
I will make a few more predictions about audiobooks and AI voices:
- This is an easy one – Text-to-speech voices will continue to improve.
- Artificial voices will become so ubiquitous that it will be less and less the goal of developers to match human speech. Instead, humans will take on the patterns of artificial speech. I already hear artificial speech patterns in humans, especially those of the Tik-Tok immersed generation. The odd inflections and tempo of AI speech will become the norm.
- Top-rated human readers will remain in demand for the near future, while mid-tier and mediocre audiobook readers will struggle to find work.
Thought In Memory Of Thomas Brown
“Words mean more than what is set down on paper. It takes the human voice to infuse them with shades of deeper meaning.”