AsiaCel
Pray For The Extermination of XMAF
★★★★★
- Joined
- Nov 24, 2017
- Posts
- 30,279
- Online time
- 1d 7h
I fucking hate Generative AI. I'm not an artist, so I will not talk about how it replaces creativity or whatever.
Instead I will focus my angle as the average person.
While I dislike obvious AI slop, like the disturbing food pictures and cringe human cat videos, they are largely easily spotted and harmless.
Even the AI knowledge shorts videos are not that bad. "Have you ever wondered" scripts at least provide some value by having the AI do the research and provide facts that you can easily check yourself.
The reason I trust the AI slop videos is because they can be easily verified, and AI is actually surprisingly good at macro facts and aggregated trends and bad at micro details.
"Why people like Big Mac?" - AI will give you an accurate summary
"What is the chemical XYZ in Big Mac, what is the percentage of it" - AI will hallucinate there
I hate generative AI, because it is mostly used to create hyper realistic fake videos. You can no longer trust videos on the internet; how bad is it? Bad enough that it is literally creeping onto completely mundane videos, such as pet videos. I'm not talking about the pet videos where a dog does impossible tricks, but completely mundane ones like a dog hugging a owner — why do these have to be faked too?
In addition, footages from non-famous/established channels have to be constantly scrunitized because you never know whether something is real or not.
I'm not being against the small YouTubers, but they have to be established in their categories/niches, like for example, for WW2 aviation, Greg's Airplanes and Automobiles is an example of a non-famous (outside of niche) and yet established channel.
It has gotten so bad that I have to use "before:2023" tag when I want to watch authentic content, ranging from porn (this field literally spawned deep fakes) to camera footages.
Instead I will focus my angle as the average person.
While I dislike obvious AI slop, like the disturbing food pictures and cringe human cat videos, they are largely easily spotted and harmless.
Even the AI knowledge shorts videos are not that bad. "Have you ever wondered" scripts at least provide some value by having the AI do the research and provide facts that you can easily check yourself.
The reason I trust the AI slop videos is because they can be easily verified, and AI is actually surprisingly good at macro facts and aggregated trends and bad at micro details.
"Why people like Big Mac?" - AI will give you an accurate summary
"What is the chemical XYZ in Big Mac, what is the percentage of it" - AI will hallucinate there
I hate generative AI, because it is mostly used to create hyper realistic fake videos. You can no longer trust videos on the internet; how bad is it? Bad enough that it is literally creeping onto completely mundane videos, such as pet videos. I'm not talking about the pet videos where a dog does impossible tricks, but completely mundane ones like a dog hugging a owner — why do these have to be faked too?
In addition, footages from non-famous/established channels have to be constantly scrunitized because you never know whether something is real or not.
I'm not being against the small YouTubers, but they have to be established in their categories/niches, like for example, for WW2 aviation, Greg's Airplanes and Automobiles is an example of a non-famous (outside of niche) and yet established channel.
It has gotten so bad that I have to use "before:2023" tag when I want to watch authentic content, ranging from porn (this field literally spawned deep fakes) to camera footages.
Last edited:





