Tucker Carlson's last warning - Upgoat

Tucker Carlson's last warning
submitted by shitface9000 to AI 1.3 yearsFeb 14, 2023 17:41:22 ago (+38/-3) (files.catbox.moe)
https://files.catbox.moe/3tdcjf.mp3

29 comments block

sort by: hot new top old low

[ - ] deleted 8 points 1.3 yearsFeb 14, 2023 18:09:55 ago (+8/-0)

deleted

[ + ] Monica

[ - ] Monica 6 points 1.3 yearsFeb 14, 2023 23:06:35 ago (+6/-0)

It's gonna be weird when AI is a squirrel that walks up to you in the park and tells you this.

[ + ] oldblo

[ - ] oldblo 1 point 1.3 yearsFeb 15, 2023 00:40:25 ago (+1/-0)

Wait til they can scan things and create realistic replicas. Shit is gonna get weird when you need to ask yourself if youre being followed by a real bird or not.

[ + ] Rowdybme

[ - ] Rowdybme 0 points 1.3 yearsFeb 15, 2023 22:51:51 ago (+0/-0)

More like our lives are going to be a bitch because idiots use it to make dumb shit like this. That isn't even believable and wouldn't be surprised if it's illegal.

[ - ] Monica 4 points 1.3 yearsFeb 14, 2023 23:12:07 ago (+4/-0)

Is it possible to take Adolf Hitlers voice and have him speaking in English? If so please tell me how.

[ + ] oldblo

[ - ] oldblo 1 point 1.3 yearsFeb 15, 2023 00:43:10 ago (+1/-0)

Until just recent mimicking voices was not a thing we could do. But I imagine we are not far off from that. Giving the voice some emotion will likely require all sorts of specialized settings. Paywalled settings no doubt.

[ + ] Cantaloupe

[ - ] Cantaloupe 1 point 1.3 yearsFeb 15, 2023 01:36:19 ago (+1/-0)

It should work, it'd should just use phonemes

[ + ] x0x7

[ - ] x0x7 1 point 1.3 yearsFeb 15, 2023 02:28:23 ago (+1/-0)*

Yeah. One way is to take some speeches and process them through a neural network that generates the same audio, but with a significant bottle neck in the neural network. That's called an identity transform when you generate the same input and output. When it has that bottleneck in it it's called an auto-encoder. Essentially what you created on the down stream side of the bottleneck is a Hitler voice generator. Now you can likely even keep the same encoder side and it will likely work. If it still sounds too much like you try speaking more like him or loop it through twice.

Now add two levels of complexity to it to make it better. Upgrade it to a variational autoencoder that pushes the bottlenecked layer to be more likely to have a normal distribution among themselves by adding its deviation from a normal distribution to the loss function. Then take it one step further and upgrade it to an untangled variational autoencoder which reduces anomalies. The last one basically clamps the model down to always sound like Hitler no matter what.

https://en.wikipedia.org/wiki/Variational_autoencoder

Now because you are dealing with audio the portions of the neural net to either side of the bottleneck will have to be recurrent neural networks or convolutional because its dealing with time series data. Basically means graphics card on fire lmao.

[ + ] Spaceman84

[ - ] Spaceman84 0 points 1.3 yearsFeb 16, 2023 16:29:50 ago (+0/-0)

That would be much more difficult. Different languages use different phonemes. There are some common sounds between languages but without a recording to work from much more synthesizing would be necessary.

[ - ] HeyJames 4 points 1.3 yearsFeb 14, 2023 20:43:03 ago (+5/-1)

Imagine hacking into the emergency broadcast system and putting this out nationwide

[ - ] The_Reunto 3 points 1.3 yearsFeb 14, 2023 21:46:25 ago (+3/-0)

The quality of the deep fakes gets more and more sophisticated, and many of the technologies are moving to convincing real-time deep fakes for voice and visual.

I think it is fortunate that many of the critical verification functions in society don't rely on voice and visual alone. It's the reason we still use ID cards and other forms of authentication.

But man, this opens up an entire arena of psychological warfare where a person could deep fake your spouse cheating on you, a close friend abusing your child, etc. in order to grind away trust between you and the people close to you.

[ - ] lord_nougat 3 points 1.3 yearsFeb 14, 2023 17:53:20 ago (+3/-0)

Damn. I should watch him again. I didn't realize he had gotten this good.
Well, watch his episodes leading up to this one, I mean.

[ - ] Not_C 2 points 1.3 yearsFeb 14, 2023 19:27:25 ago (+2/-0)

At first I thought I'd be able to tell it was fake because when people are talking, they're breathing, and an AI can't fake that.
But that AI did actually simulate the breathing.

You could pick up on the fact that the breathing wasn't perfect though. You could tell it wasn't natural if you listened closely. Especially when going from one sentence to the next.

But throw in some background noise to prevent people from listening too closely, and it'll be even more convincing.

[ + ] Spaceman84

[ - ] Spaceman84 1 point 1.3 yearsFeb 15, 2023 00:33:21 ago (+1/-0)*

The constant cadence and lack of emphasis for emotional effect makes it sound flat. The software will need to improve a lot. AI doesn’t understand what it is reading and so can’t assign an emotion or importance to any particular word. A human will be required to do this manually to the script the AI is reading.

[ + ] GetFuckedCunt

[ - ] GetFuckedCunt -1 points 1.3 yearsFeb 15, 2023 02:12:54 ago (+0/-1)

The software doesn't improve itself, which would be AI. Instead humans are going to be editing the coding and the scripts in order to fix these quirks.

Does anybody here actually understand what the theory of AI actually is?

[ + ] Monica

[ - ] Monica 0 points 1.3 yearsFeb 14, 2023 23:05:37 ago (+0/-0)

AI can simulate environments .. thats easy

[ + ] GetFuckedCunt

[ - ] GetFuckedCunt -1 points 1.3 yearsFeb 15, 2023 02:11:58 ago (+0/-1)

How is this AI?

These are just lines of code and script being read by the machine, coded by humans, not even close to being AI.

[ - ] RMGoetbbels 2 points 1.3 yearsFeb 14, 2023 19:01:55 ago (+2/-0)

I'd actually like tuck the cuck if this were him.

[ - ] voatersaredumbasses 1 point 1.3 yearsFeb 15, 2023 04:33:07 ago (+1/-0)

You guys are sealing your own fate by using this tech. You just can't wait for your shiny new digital id huh? Nice going asshole.

[ + ] Rowdybme

[ - ] Rowdybme 0 points 1.3 yearsFeb 15, 2023 22:50:35 ago (+0/-0)