135
YouTube Launches Experimental AI Feature With Voice Clones of Major Artists
(www.rollingstone.com)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 16 Nov 2023
135 points (93.0% liked)
Technology
59583 readers
2381 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
....This is a live concert for thousands of people that pay to see a hologram sing and dance. Yes, of course it's in Japan. It's one of their major pop stars.
https://youtu.be/Nfhuj60cJjk?si=HiI_a4YNRKe8weeE
*edit: Holy hell there's a lot of you getting butt hurt over Miku.
Well that kind of proves the actual point I was making, which wasn't attacking "your precious". It's that if AI does it well, tons of people are going to lap it up. They're asking for this.
Dude that's fuckn Miku do you have no idea how much work goes into Vocaloids?
I played around with the vocaloid software a few years ago and that shit is fucking complicated.
It's not just dragging a midi file in and hitting play, you have to hand sculpt every syllable and every fricative throughout the entire performance. It can take literal days when you account for all of the backing tracks and everything else.
Add in synchronizing that to a video and doing all of the mocap and digital effects compositing for that fuck that noise.
I've done some very basic shit with vocaloid 6, UTAU/OpenUTAU, and DeepVocal and they're definitely something above my pay grade in terms of difficulty.
I've never managed to get a single song out because of tuning the voice and getting the lyrics/timing/note pitch just right. It's fun but not for the average person who will just give up after a while, like me.
The thing I've found with AI tech bros is that they're rarely interested or informed about any creative process, and instead are much more interested in typing shit into a web form that spits out some garbage that they can pretend that they created.
It's like arguing with people over NFTs or whatever, we could get down to the technical details and the merits/flaws with the concept and current implementation, but at a certain point it's like trying to tell a flat-earther that the world is round or an anti-vaxxer that vaccines have saved millions of lives over the last century. I really don't want to be an ass to any of them, but after months of trying to get through to them I'm kind of over trying to talk them through it.
Yeah same experience I've had. As a rule of thumb they lack the understanding to even talk about what they are so hyped about. It's sad really. When I'm ready interested in something I want go learn about it. Instead they are interested solely in how it could make them money. They don't care whatsoever about it's actual utility if there even is any.
The songs are still written and composed by humans, only the performance is Virtual, and even then they are mocapped. It's more akin to vtubers rather than AI
You're comparing a collaborative artist project with a soulless Generative Algorithm. Real people made her music, not a computer program owned by some Capitalists who are tired of real artists getting between them and their profit.
Not to mention the artists who created her, animated her, voiced her.
Boo. Boo on this.
Here is an alternative Piped link(s):
https://piped.video/Nfhuj60cJjk?si=HiI_a4YNRKe8weeE
Piped is a privacy-respecting open-source alternative frontend to YouTube.
I'm open-source; check me out at GitHub.