Elevenlabs.io - AI Generate anyone's voice with a minute of audio saying how much they hate trannies

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
I'm sure the fosstards will get us open-source AI tools in due time.
From what i've seen on /g/ its pretty progressing pretty fast, though 11labs is still way ahead of anything open source, people have got the Tortoise (which 11labs might be derivate from) sounding pretty damn good.
Is there any free equivalent or workaround? I don't feel like using a credit card for a toy I will only use for a few days. Nor do I trust them with that information.
Tortoise is worth a look but requires alot more work to get sounding good, if you made a 11labs account before they changed the free plan you can still make custom voices and use them, unless they change anything in about a month your character quota will reset and you can do more stuff with it.
 
So just play it out a speaker and record it on a microphone.
You wouldn't even need it, probably just converting to mp3 or .ogg would already fuck the data up.
There's no 100% fool proof way of certifying that shit, and if there's one, in 5 s there's gonna be someone on GitHub that scrubs any signals clean.
 
You wouldn't even need it, probably just converting to mp3 or .ogg would already fuck the data up.
There's no 100% fool proof way of certifying that shit, and if there's one, in 5 s there's gonna be someone on GitHub that scrubs any signals clean.
Yeah, this kind of verification only really works if the person making it wants it to be known it was a fake in case someone else misuses the file... in which case any of the workarounds work too, and you could probably inject the marker in true files as well making it totally useless.
 
Yeah, this kind of verification only really works if the person making it wants it to be known it was a fake in case someone else misuses the file... in which case any of the workarounds work too, and you could probably inject the marker in true files as well making it totally useless.
I bet some retard news corp or a government uses some crap they got from some scammer company to "certify" videos and Audios and there's gonna be a ton of false positives.

Heck, be prepared to the end of project veritas or true jornalism, every hidden camera, microphone, etc will be dismissed as a.i product unless the big media wants someone fucked over.
You think Taylor Lorenz is bad? Wait until her and her friends start using this shit instead of just putting shit out of context or editing a Interview.
 
You wouldn't even need it, probably just converting to mp3 or .ogg would already fuck the data up.
Yeah but even if you assume some incredible voodoo that would survive such a conversion, even the crude low-tech measure would completely obliterate that, especially if you threw in some background noise. Sort of like those stupid marks printers put in to identify the printer, where just blurring it and photocopying it again obliterates that shit.
 
Here are the ones I did before ElevenLabs cucked out:

James' Quest
View attachment 4441584

Sonic Adventure DX
View attachment 4441588

Edelgard gets woke
View attachment 4441592

Nep-nep alerts Chris Chan
View attachment 4441596

Sonic tries to take it OFF. THE. INTERNET!!
View attachment 4441600

Of course, everything is AI-generated yada yada.
I'm going to be real with you, that bit about Shadow fucking Eggman's wife made me laugh so hard I sprayed a little bit of spit onto my monitor.
 
Posting a classic finally made possible by the magic of AI.
Pretzels.png

 
David Attenborough nature documentary clip on the American Negro in his natural habitat

Hitman's Last Mission (clue: it involves a counterfeit bill and Minneapolis)

The channel that posted these and this /gif/ thread have lots of good ones. The YT guy just had the Biden video striked for Hate Speech so probably wont be around for very long thanks to Susan.
 
Last edited:
This is fantastic.

Man, FNV mods are going to be so sweet from now on.
Just keep in mind that you gotta feed this thing with hours of clean dialogue and you have to learn to parse things properly with the text and edit it a lot in Audacity....just to get it to sound like that, and you can STILL tell it's machine learning.
 
  • Like
Reactions: MrJokerRager
Just keep in mind that you gotta feed this thing with hours of clean dialogue and you have to learn to parse things properly with the text and edit it a lot in Audacity....just to get it to sound like that, and you can STILL tell it's machine learning.
Sure, but I'm certain the tech will continue to improve and 'flat' voices like Ron Perlman and Dagoth Ur seem to come out really well already. At least ending slides will be easy.
 
Back