Voice-to-voice is going to be fantastic, because you still hire Steve Blum, but you have him do six characters who sound nothing alike. But only for the important bits. If you need a thousand lines about mundane background nonsense then you have the devs record themselves and filter it to sound like the character. You’d mimic the serious voice actor’s performance… not their voice.
I just googled, an uprising only has a chance of starting 60-100 years after the first tech, they’ll probably be at least onto stellaris 2 by then so it’s a problem for the modders to solve
This feels like something blown more out of proportion if you actually read Paradox’s reply, and of course trust it. (Reddit warning)
The highlight for me is that it seems more based on being able to add lines in the future, without having to drag the actor back into the studio for 1-3 new lines only (And that’s on the generous side for what gets added per DLC). Paradox claims the original voice artist gets a fair commission from the usage of their voice for new lines, which would be nice if it’s true.
Using the voices of real people who did real work with them, with a real agreement between studio and actor, and paying the actor even when lines are generated, is exactly how it should work.
If it is actual (local, in-engine) Text-to-Speech, I'd see that as more forgivable. Less space taken up by audio files, better for modding/user-generated content.
Though given the mention of AI and a AA/AAA game I highly suspect they aren't going that route.
Read the post, the human still got the job and was paid for it, he was just saved the hassle of going back to read new lines every time they needed a new line while still getting paid.
Complaining about technology eliminating the need for certain labour is glancing over some much deeper issues that result in displaced workers suffering all of the consequences.
Yeah but this is the case for pretty much all technology. Coding getting easier and quicker through new tech = fewer coders necessary. Motion capture = fewer animators, etc etc.
Thats generally not true in arts, the job just changes. Motion cap requires different technicians and artists post capture. Its just the artist drawing the original movement thats gone. No real net loss
This comment implies that no humans were involved with operating the AI. Seems doubtful.
It’s one thing for out of touch executives who blindly replace entire departments with “AI” while fundamentally misunderstanding the role of the department being replaced and the capability of AI, tanking the quality of the product–that’s real self harm for everyone involved; it’s another thing to be advancing the creative processes with more advanced tools and automation, something that we’ve been doing for centuries without much fuss.
The creative part of voice acting isn’t just in moving one’s lips. The creative part of voice acting is just as much, if not more, in feeling and direction–in deciding if a sound sample produces a certain desired emotion, and if that emotion is valuable to the overall experience or not. This is not the territory of generative AI. This is the territory AGI, which does not yet exist. Producing the sound with your lips is just a small part of that. There’s still a human involved in producing the work of art (and if not, then yeah, we are back at that first category, of leadership ignorant of the creative process, and we should bemoan a crappy product lead by executives who have no clue how to retain talent).
They still needed a human to process the files, and another human to provide the voice model.
I mean, the voice files are specifically for a robot, they could have not had a human voice involved at all and eliminated one human from the job pipeline regardless by using FM synthesis or something, which has been around since before the 1990s.
Add comment