The only way to have Sota model is to use copyrighted data so this ultimately benefits big AI in the end. They get a monopoly because of the cost that’s added to making the model. That cost is passed on to consumers who are now forced into a subscription model. At least Hollywood and YouTube get paid though.
They analyze statistical patterns, they don’t cross reference the training data.
It’s picking up the text as generated because it’s mostly guess work and constantly spits out false positives, especially with non native speakers.
Good on him to actually give the source of the joke. Awesome!
The output of the training, the model (LLM) itself, is transformative. I’m not talking about the output of the llm nor was the person I was responding to.
The pirating issue has nothing to do with the current lawsuit (the New York Times one mentioned in the article) nor did the government mention it in the current context.
That being said, I’m a pirate in my personal life, and I also think it’s kind of a necessary evil when it comes to building SOTA models. I wish we had a copy left solution to it, so you can use pirated data but need to open source it, so open source has a fighting chance and big AI can buy it if they want it. I’m okay with openAI and anthropic getting brought into court about it I guess, but the legislation ends up being market capture in their favor in the end.
I understand the issues though, very “rule for thee but not for me” currently.
Call in sick bro, you deserve it.
Per the article, he’s talking about doctors.
No one will believe you once they pull out the guillotines for one
From what I’ve seen happen with short story and poem competition, I bet any address you find online to send your script to is already getting flooded with automated junk.
I also don’t trust Hollywood to not steal it.
That being said, I don’t think we are very far off from being able to generate proper tv. I’m expecting to see stuff pop up very soon because of the new minimax model. Needs a lot of know how currently to make something that isn’t garbage though.
It might seem like a lot but when you calculate the amount of overhead and numbers of employees, that’s basically only 15 billion in profits.
Ya it doesn’t make sense. I think it’s probably facial recognition and since she’s in the training dataset, her face pops up when the model gets pulled in that direction.
They wouldn’t have access to the dataset, I don’t see how they could know for the Arts explanation. Maybe it learned the hash and reproduces it but that doesn’t seem likely. Even the part about adding hashs to track the images doesn’t make much sense. That would imply they are distributing it somehow.
I don’t get why X isn’t running age detection on the output though.
You can make a pokemon ripoff with Photoshop, should Photoshop be illegal? Is it the tools fault or the user’s fault?
Precendent was already there. Copyright has never been enforceable when it’s transformative. Collage is legal, so was training neural nets before the copyright industry saw money in it.
None of what you said is wrong but I think it’s a bit simplistic to compare it to a resolution jump or to bring up Indy graphics vs AAAA as a sign that nobody cares anymore.
FPS is indeed the real killer though, you can’t really run this without high end cards and the difference is simply not worth the price for most people.
If you have copyright juggernauts gatekeeping the data through legislation, then you can’t have open-source models.
I’m surprised how many want to shoot consumers in the face just to protect mafia like companies like Universal Studios and YouTube.
Training a model with copyright material = fine
Creating copyright material and distributing it = not okay
I can infringe copyrights with Photoshop, or just copy pasting the image itself.
We don’t need to knee cap tools and force everything behind subscription services with censorship just to satisfy the copyright industries craving for blood.
Bring an expert that says the contrary instead of throwing vague shade. It’s the consensus in most fire arm circles from what I understand.
Lots of other things don’t add up as well (like paving over the crime scene a week later).
Can’t wait for some doc to get declassified and we learn the CIA has been doing it for decades.
There’s also selling it to another party that isn’t a vampiric parasite as an option. I doubt they were the only potential buyer.
There’s a version of this world where they didn’t die out and are now kept as industrial animals to dig subway tunnels and also sometimes as pets.
Only difference between a pumpkin is that he smashes teenagers instead of the other way around.