Originally published at: This voice doesn't exist: AI-generated speech that isn't trying to impersonate someone | Boing Boing
…
I wonder if it is fast enough to produce procedurally generated speech for open world video games?
It seems like only yesterday when I first read about this. My how quickly time flies.
The narrative one reading the first paragraph of The Hobbit was good. it wasn’t dissimilar to the audiobook I listened to ages ago. It was close enough, in fact, I wonder if there was any “training” of the AI using the original recording. There’s been a lot said about how this could put a lot of voice actors, people who do not just audiobooks, but commercials, PSAs, foreign language dubbing, etc, out of a job. Hopefully, this company is “ethical” enough to limit availability of its services to authors, publishers who do not have the means to mount a full audiobook production.
James Earl Jones has licensed his voice to Lucasfilm and another AI voice company so that he will be the voice of Darth Vader forever. I wonder if a similar situation is feasible for voice actors who retire or pass on to license their voice through their estate to continue “working” after their demise or retirement. I’m thinking specifically of Roy Dotrice who was the reader on all five of the thus far published Song of Ice and Fire novels. My gut reaction is that I would not immediately opposed to using AI to resurrect his voice for the last two novels in the series, should they ever see the light of day.
This topic was automatically closed after 5 days. New replies are no longer allowed.