Comment by burkaman
1 day ago
Seems like voice actors are safe for now. This is technologically incredible, but the results are really not very good, and usually not particularly close to the prompt. In basically all of these examples some core part of the prompt is completely ignored.
Probably.
I listened to some of the voices. The male voices are believable while all the female voices sound the same and artificial. For some reason, it also reminds me of the voices in Toy Story movies.
Bias in the training data?
> Seems like voice actors are safe for now. This is technologically incredible, but the results are really not very good,
Um, what?
Technologically incredible as in "I cannot believe it's possible for a computer to do this" and not very good as in "these examples are not what was prompted and I can't think of a use case where these would be acceptable".
Imagine someone showing you that they've trained their dog to hold a paintbrush and paint. There would be no contradiction between "this is incredible" and "these paintings suck".
I've used TTS in videos to replace the speech of the person talking so that we could type up whatever we need them to say in editing.
They sounded exactly like that person... but it's like they're angry at me.
Great toys but not production ready.