It is still an open legal question whether or not training AIs on something is any less legal than training humans on something. And unless it’s a trademark or so identical as to be difficult to distinguish, it’s unlikely it will be found to be illegal.
Like YT creators are really original. Most of you just rehash other people's videos or use information that can easily be found elsewhere. Especially you wimpy 'Shorts' creators. Cry. Me. A. River.
I expect most of you have heard of a case where a Court of Appeals ruled this kind of scraping is legal and the Supreme Court of the United States declined to review that decision. The courts held that scraping and transforming content to serve a new purpose is covered by the doctrine of Fair Use. You might not remember the parties, but you'll appreciate the irony. It was... Authors Guild v. Google, and of course, Google operates YouTube.
Apple definitely has its flaws but I gotta agree with them there. The internet is a free place to put stuff out, whether you’re a company or a creator once you put something out there it’s no longer yours. You made it no doubt about that but people can and should do whatever they want with it, that’s how we get innovation. Would you be happier if Alexander graham bell’s family still owned the concept of a phone and it had to be the same as it was?
Every time I see some shit like this I feel even more supportive of piracy lol big companies steal other people's work but don't want the same happening to them
Who would have guessed that a company wouldn’t have your best interests at heart.
Seriously though, are we really surprised about this? AI companies are notorious for this sort of thing. Are we really surprised that a AI company is doing what a AI company does.
Current "AI" is theft, plain and simple. Stolen writings, stolen pictures, and now stolen videos with extra goodies on the side. That's the whole business.
Because when people find out where data is being used to train AI, they deliberately try to make it offensive or misleading, to mess with the AIs head.
Remember that regulations only ever gets written when companies do something unacceptable and everyone agrees that it's unacceptable and we write rules making it illegal. Until it's explicitly proven to break regulation they will continue to try and get away with bs like that.
They probably could get away with taking data from small, little known websites. I know several got their data from Twitter, even. But Google was certainly a high mark.
They stole five of my videos and I don't mean clown on myself here but who the f*ck am I? Seriously, I have 4k subs. I am no one. If I got robbed, thousands of people did.
It's not like AI is trying to duplicate specific people, it's just learning from public information. Would a person have to pay a youtuber if they watched a tutorial or how to video? Of course not.
You know, news companies put videos of events onto youtube
If AI was trained on youtube, someone could make a video about the president doing or saying illegal stuff...
On the plus side, if that happens, the government would make _so many_ laws against AI.
Wait a second first of all content creators don’t even own their content because it’s not their platform lets just think think for a moment that YouTube goes down can you do anything about your videos no because you don’t own them and Google has been trained AI for years with all google photos maps youtube and everything they have why do you think because they just can and Apple it’s not less guilty than Google or Meta and last one Apple didn’t scrape anything they use a company that has data for research only but I can’t protect any company is this kind of stuff
Apple saying it wasn't their fault that that work ended up in their hands is the same as saying it's justifiable to keep using Lightning: a complete f--king JOKE
Scrapping companies should inform the persons involved with the work that they are using their data and if the action is deemed unfavorable, allow an easy way to request the infomation to not be used for training.
I can download a transcript on my paid YouTube account and use it for my means. I can create new content or store the information and study it. This is no different. I don't see companies going after re-posters of the exact content without linking it to the original creator. I believe this needs to change. We need to give original creators more control of their creations.
I've used YouTube videos to train myself how to work on cars and computers and I've made a decent amount of money because of it. I don't give them credit unless I'm asked where I learned the stuff which is something the ai can do. There really isn't an issue.
Like marques said Apple and a lot of companies are getting it from one company who just gave them the data it’s not like Apple is scrapping the data
Edit:
ah I am sorry I paused the video because normally I forgot the comment also Marques said in his vid he pays money so people don’t need to use auto captions but this makes auto captions better
コメント
1. Not accurate
2. Based on stolen work
It's why I don't use or trust AI for anything except looking up silly trivia or getting a laugh at a generated pic.
Seriously though, are we really surprised about this? AI companies are notorious for this sort of thing. Are we really surprised that a AI company is doing what a AI company does.
The only way to scrape data is secretly.
2024: DIBS ON UR DATA!1!1!1
-apple
This is so fucked up.
must be something new (cries in artist)
If AI was trained on youtube, someone could make a video about the president doing or saying illegal stuff...
On the plus side, if that happens, the government would make _so many_ laws against AI.
And we are all forced to accept in order to use the software we pay money for.
Lmfao
classic
Edit:
ah I am sorry I paused the video because normally I forgot the comment also Marques said in his vid he pays money so people don’t need to use auto captions but this makes auto captions better