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 made clips of the great Kevin Conroy (RIP) saying King Novik's intro and outro to Doom Eternal. I had to splice together multiple clips to create each one because there really isn't a good way to get Tortoise to emphasize specific parts of sentences, and because it keeps cutting paragraphs off too early. I also added the reverb to match the game's effect.


 
1678319741839.png
Totalbiscuit's wife is butthurt that people are using his voice lmao
 
View attachment 4731862
Totalbiscuit's wife is butthurt that people are using his voice lmao

In the span of, what? Two years? In the span of two years, we've seen AIvoice, drawing, and writing come fully into their own. You could literally make a cartoon show of professional quality for nothing. It reminded me of the heady days back in 99 2000 where the internet put culture and knowledge at man kinds fingertips. Art, philosophy and science where once you had to go to a university or major urban library, costing a fortune could be enjoyed by the entire human race via modem.

Then boof. the DMCA

Now we're back here again. This tech is all censored to hell.

And why? Because they can't have you make your own Simpsons or South Park. Because that is one of the foremost levers of power in society. Media. I'm dreading the new DMCA, and what other draconian, abusive powers it will give figures. Just look at the damage to fair use DMCA strikes has caused.
 
Then again, it did this horrifying shit when I fed it Ronald Reagan's voice:
View attachment 4708166

So at this point I have no fucking idea wtf I'm doing. How much sample data are you feeding it?
I encountered the same problem when playing around with that old thing from github on a 2gb Nvidia card. It was kind of like this and it was fun to play around with even if it was unintended.
 
Then again, it did this horrifying shit when I fed it Ronald Reagan's voice:
Pardon the creepypasta spergout, but this would fit really well for the possessed Ronald Reagan tape. I bet people have already used the imperfect text to speech for horror purposes at some point.

1678396214782.png
 
I'm dreading the new DMCA, and what other draconian, abusive powers it will give figures. Just look at the damage to fair use DMCA strikes has caused.
No don't you get it! The heckkin valid and wholesome twitter artists/VAs would be saved knowing that their jobs aren't at risk of the AI overlords and instead are at the helms of Corporats paying them pennies to nickels for their work at the cost of freedom! Don't you understand you ai nazi chud, the poor starving artists will be saved by harsher copyright and DMCA actions to know they are barely scraping by in life. This is so wholesome 💯
967.jpg
 
Then again, it did this horrifying shit when I fed it Ronald Reagan's voice:
View attachment 4708166

So at this point I have no fucking idea wtf I'm doing. How much sample data are you feeding it?
I've gotten the same thing at times and I have like an hour of Kevin Conroy audio. Just a quirk of the training, I suppose.
 
I caught a 'tism and made a thing. It's not done yet, but I have all of the audio generated. Just need to do editing and regenerate the shit parts. I started getting lazy after 5 days after work of playing with this, but here's the first hour. I'll probably not finish it or just throw up the unfinished parts as a file.

Intro

2 - THE PSYCHOLOGY OF MODERN LEFTISM

3 - FEELINGS OF INFERIORITY

4 - OVERSOCIALIZATION

5 - THE POWER PROCESS

6 - SURROGATE ACTIVITIES

7 - AUTONOMY

8 - SOURCES OF SOCIAL PROBLEMS

9 - DISRUPTION OF THE POWER PROCESS IN MODERN SOCIETY

10 - HOW SOME PEOPLE ADJUST

11 - THE MOTIVES OF SCIENTISTS

It's not perfect, but I'd say fair for a first go. Learned a lot about how this stuff likes to be handled. I left in some misspelling because they were used multiple times and might have been intentional. There was a shitton of editing and certain words will get cut off at the end of a generation no matter what you do.
 
Then again, it did this horrifying shit when I fed it Ronald Reagan's voice:
That's horrifying. It actually reminds me of this old SCP featuring Reagan:
FUCK! I just noticed you actually tried to generate SCP-1981.

Please tell me this is an elaborate troll.
 
That's horrifying. It actually reminds me of this old SCP featuring Reagan:
FUCK! I just noticed you actually tried to generate SCP-1981.

Please tell me this is an elaborate troll.
Easy ad revenue and clicks. Way I figured, if I could get decent results, I’d post a link to the full AI videos on the article’s comments.

Use that initial boost of popularity to game the YT algorithm, posting enough videos to at least get some form of monetization to cover my monthly cloud GPU fees.

Then I found out the SCP jannies are afraid of AI content and did a blanket ban on the stuff, so I shelved that plan.
 
Pardon the creepypasta spergout, but this would fit really well for the possessed Ronald Reagan tape. I bet people have already used the imperfect text to speech for horror purposes at some point.
The text about abominations is actually from that. It's the first transcript. I only noticed because I went to look up the number and ended up re-reading the thing. I just found it incredibly creepy that it did that thing for that particular text and voice.
 
So the creator of Tortoise has dropped support for the Colab. His claim on the github is that it's google's fault, however this week he went on an autistic rant about how the tech was being used to scam people, and mocking the idea that anyone should be able to use it.

Looks like some other people are trying to maintain Tortoise without him, but we'll see how that goes.

Interesting to note is he seems to believe that Eleven Labs is using a further trained version of Tortoise to run their service. Not sure whether he's onto something or if he's just sneeding.
 
Back