The question is, what does this do for blackmail? They're all too far in bed with the globalists to back out now, there won't be any quarter on the freedom side of the line.
Stuff like this seems like it'll be a boon for RPG video games. Voice acting really ruined them in my opinion, as having to record dialog meant they had to 'streamline' the number of interactions to save on disk space and budget. Now a seemingly infinite amount of responses can be put through an AI to create good enough voiced interactions. And the AI is only going to get more and more convincing.
Sky's the limit, now. You could have a Dwarf Fortress-style engine feeding prompts into a ChatGPT-style response generator and then have VoiceAI speak it into being, and to give it a more human feel, let StableDiffusion draw up a little portrait.
All that's missing is an inflection/tone/mood prompt, really.
You can get a voice that sounds similar with like, a minute of audio. I don't know what it takes to get it exact but you can definitely trigger neuron activation with very little.
This is pretty cool, I been hoping something like this would be accessible to people like me so I can copy paste transcripts of people on the right and use voices from the left and have someone I know with TDS listen and see his response. Like take excerpts from Tucker's show and have Trevor Noah's voice read it
The point is that this thing is few sound engineers away from tinkering with modulation until it sounds convincing enough to deploy very convincing deepfakes.
I'd imagine some of it can be even masked by ambient noise in the background or stock wind blows.
It's not like news media would bother anyway. Remember Biden's green screen bullshit?
All of these have about the same cadence, but I wonder if through the use of punctuation and/or direct commands (e.g. wait:duration instructions) that might be a simple problem to solve.
They'll mix the deepfakes with the real ones.
The question is, what does this do for blackmail? They're all too far in bed with the globalists to back out now, there won't be any quarter on the freedom side of the line.
The Rogan clip about 4chan jannies is fucking hilarious. Toppest of keks.
Stuff like this seems like it'll be a boon for RPG video games. Voice acting really ruined them in my opinion, as having to record dialog meant they had to 'streamline' the number of interactions to save on disk space and budget. Now a seemingly infinite amount of responses can be put through an AI to create good enough voiced interactions. And the AI is only going to get more and more convincing.
Sky's the limit, now. You could have a Dwarf Fortress-style engine feeding prompts into a ChatGPT-style response generator and then have VoiceAI speak it into being, and to give it a more human feel, let StableDiffusion draw up a little portrait.
All that's missing is an inflection/tone/mood prompt, really.
The Cenk one is spot on.
It doesn't sound exactly right, but it's pretty close. Trump sounds like an impersonator doing it. Shapiro is a bit off, too.
Rogan is spot on, probably because they have thousands of hours of him talking to train on, and he sounds the exact same in all of it.
You can get a voice that sounds similar with like, a minute of audio. I don't know what it takes to get it exact but you can definitely trigger neuron activation with very little.
Now I want to make videos.
Trudeau is brilliant. Entirely believable script too!
"You don't wanna be a nigga." -- Martin Luther King, Jr. on his 2023 standup comedy tour
This isn't nearly as good as the Boondocks episode the words came from.
This is pretty cool, I been hoping something like this would be accessible to people like me so I can copy paste transcripts of people on the right and use voices from the left and have someone I know with TDS listen and see his response. Like take excerpts from Tucker's show and have Trevor Noah's voice read it
This sounds nothing like them. It sounds like people doing impressions of them.
You're right but that is not the point.
The point is that this thing is few sound engineers away from tinkering with modulation until it sounds convincing enough to deploy very convincing deepfakes.
I'd imagine some of it can be even masked by ambient noise in the background or stock wind blows.
It's not like news media would bother anyway. Remember Biden's green screen bullshit?
Geralt of Rivia performs an autopsy on a new type of horror: https://vocaroo.com/1gR2pw70Zt3x
All of these have about the same cadence, but I wonder if through the use of punctuation and/or direct commands (e.g. wait:duration instructions) that might be a simple problem to solve.
HNNNNG I'M COOMING
Tim Pool's is a bit off.
Had to click the Abby Shapiro one.
Wasn't disappointed.