Yes. The answer is Yes. And Hank Green brings receipts.
I remember watching this when it first came out and I was honestly as upset as him. But don’t worry, it’s for the benefit of everyone so they can make AI as best as they can or some shit. But really it’s just stealing with extra steps.
Remember that every single piece of AI art, music, and video you have ever seen was made entirely out of stuff found online and most likely taken without people’s consent.
But really it’s just stealing with extra steps.
Accurate.
Shouldn’t it be obvious?
I think the screenshot with “YES” in the background helped ;)
Half of YouTube videos are written and read by AI these days. That will be a wonderful feedback loop if they train it on those videos.
None of these companies are stupid enough to train on unvetted channels full of random videos. They are selectively taking them on a channel by channel basis.
They’re all in in the insane race for dominance against each other. I doubt they rate quality that high right now. They want data fast to get money first.
You watched the video and saw there is a list of channels. Right?
Obviously they are since i got a youtube premium feature the other day that gave me a button to skip a sponsored segment and it’s most likely an ai that said the segment starts here and ends there from learning the sponsorship patterns.
Oh you know they are. Did you have to ask?
Oh good. So Google’s AI is going to be just as bad at proselytizing, skepticism, and logic as Christians are.
Can’t wait to debate a computer that “thinks” fine-tuning and the Kalam cosmological argument are the best ideas nobody could ever debunk.
Raise your hand if you’ve never trained on a YouTube video
I’ve trained on a lot of YouTube videos, I love Antranik’s yoga and related vids (website if interested).
He addresses this.
If you are trying to make the argument learning from media is the same as taking that media and performing calculations on it.
🙋🏻♂️
Whatcha u gonna about it, pussy?
Call the police lol