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

[CervixHammer]

Totally real and not generated, I clipped this from Keffals' stream last night and "accidentally" posted in this thread instead, because it's 150% real:

 

[Super Male Vitality]

Null reading the Warwick Davis copypasta

 

[A-Stump]

I'm going to use Null's voice to invest in a string of alpaca farms then skip town

 

[Epoch]

Dagoth Ur wants TND (Total N'Wah Death)

 

[Cherenkovblue]

View attachment 4387096
Dagoth Ur wants TND (Total N'Wah Death)

Lemme guess... they're already keyword censoring on "niggers", aren't they?

 

[This Is Violence Now]

View attachment 4387096
Dagoth Ur wants TND (Total N'Wah Death)

Too bad the AI can't pronounce N'Wah correctly.

 

[Dan of Steel]

A selection of my personal favorites from the autists on various 4chan boards:

Amy Rose goes on Tucker explaining how she is the TRUE & HONEST wife of Sonic:

Joe Buck and Troy Aikman call the dropping of nukes on Japan:

Joe Rogan & Jordan Sneederson explain the sociological phenomena of janitors:

NO WAY FAG:

RAIDEN PUT YOUR DICK ON THE GAME CONSOLE RIGHT NOW!

I for one, welcome our AI overlords.

 

[Divine right to rule]

The funniest thing about this thread is that the celebrities and vidya characters come out crystal clear, but the AI is completely unable to mimic Josh's low-heartrate mumbling.

 

[EpicBugFacts]

so interesting development, there is a non-zero chance that 11.ai is actually a fork of an existing github project that's open source call "Tortoise TTS"

what this means is, even with all these horrible changes they want to implement, it's entirely possible that you can still continue to generate models if you know a little bit of code and are willing to put in the effort to toy with it.

early warning though, it seems like it currently only works with nvidia, and i'm not so sure how to work around that problem.

[Tortoise-TTS on GitHub]

 

[Trapped Under Rice]

[Tortoise-TTS on GitHub]

Hold up while I dump hours of KingCobraJFS streams into this thing.

I love the diversity statement on the GitHub:
"Tortoise was trained primarily on a dataset consisting of audiobooks. I made no effort to balance diversity in this dataset. For this reason, Tortoise will be particularly poor at generating the voices of minorities or of people who speak with strong accents."

 

[EpicBugFacts]

I love the diversity statement on the GitHub:
"Tortoise was trained primarily on a dataset consisting of audiobooks. I made no effort to balance diversity in this dataset. For this reason, Tortoise will be particularly poor at generating the voices of minorities or of people who speak with strong accents."

tl;dr
i'm pretty sure these "developers" stole an existing project and are forking it to try and make a quick buck.


interesting to note, that is currently a limitation of 11.ai.

what i'm saying very directly is that i'm fairly certain these "developers" didn't actually develop anything; they're running a server with a new-gen Nvidia graphics card and have set T-TTS to run at 33khz instead of the default 24khz.

they may also be running something else to make the end result slightly nicer as well.

 

[AnOminous]

tl;dr
i'm pretty sure these "developers" stole an existing project and are forking it to try and make a quick buck.

That might be why they're concerned about the sudden attention, more than that people are dropping n-bombs into the mouths of celebrities with it.

 

[Sneeds]

Scotty Kilmer finds /o/

 

[Stasi]

Witcher Geralt discovers a gruesome murder of a woman?
View attachment 4374692

That Gerald one is amazing, cracked me up. Almost as good as the Shapiro khazaar milkers one, my sides almost blasted off into orbit listening to that one lmao

 

[Jonah Hill poster]

All I will say is that the moral grandstanding surrounding as to why our “moral betters” will try to get this A.I. technology is going to be tiring and nauseating. Let people have fun for once before trying to make it “empathetic” for people.

That might be why they're concerned about the sudden attention, more than that people are dropping n-bombs into the mouths of celebrities with it.

It’s interesting since before there was a time where celebs were fascinated by the idea of A.I. technology coming in to help change the landscape of what Hollywood has done for years now. Fast forward to today, they’re just afraid that the technology can do their jobs better than them.

 

[AnOminous]

It’s interesting since before there was a time where celebs were fascinated by the idea of A.I. technology coming in to help change the landscape of what Hollywood has done for years now. Fast forward to today, they’re just afraid that the technology can do their jobs better than them.

I bet an AI could write a better novel than Patrick S. Tomlinson, even if you fed it nothing but Patrick S. Tomlinson novels. It would be impossible to get an AI to be as stupid as Pat.

 

[Xarpho's Return]

Tried fiddling around with this and wanted to make Rick Sanchez proclaim how he's devoted his life to Jesus Christ but then when adding samples from a soundboard, the website claimed the files had two datastreams in them instead of one, and I wasn't able to do it.

 

[Epoch]

Tried fiddling around with this and wanted to make Rick Sanchez proclaim how he's devoted his life to Jesus Christ but then when adding samples from a soundboard, the website claimed the files had two datastreams in them instead of one, and I wasn't able to do it.

your audio is in stereo retard, mix it down to mono

 

[JackoVerde]

this isn't the same one but it's interesting none the less, it uses a different approach to voice AI
it changes your voice to match another persons. the thing is, it's not so good as of yet.
this type of approach is harder to sensor, since it cant just filter out words as easily.

 

[Lieutenant Rasczak]

Saw this and thought it was fucking gold.