Datasets:
PSA: False DMCA claims
DMCA is intended for copyright violations.
Fanfiction is using other people's copyrighted works for enjoyment purposes.
You cannot copyright someone else's work, because you do not own 'Naruto', 'One Piece', 'Supernatural', or any other copyrighted material which your derivative work is actually owned by.
If an original work that you hold the copyright of, NOT a derivative work, is present, that is the only circumstance you can legally file a valid DMCA claim here.
Making false claims leaves you open to legal action.
me when i'm wrong
me when i'm wrong
Feel free to assert copyright over works you don't have any legal basis to claim. Of course, I doubt any lawyer would take you up on it. If they do, they just want your money.
Your assertion that we cannot file a DMCA claim for our work is incorrect and demonstrates a misunderstanding of copyright law. While our work is derivative of existing copyrighted material, the specific expression of ideas, plot, dialogue, and descriptions within our works are original creations and are protected by copyright.
The DMCA is precisely the mechanism by which copyright holders, including those who hold copyright in original elements of derivative works, can address unauthorized reproduction of their work online. The scraping and republication of our work without explicit consent constitutes copyright infringement of our original contributions.
Your claim that DMCA is only for "original works that you hold the copyright of, NOT a derivative work" is a misinterpretation of the law. Copyright protection extends to the original aspects of derivative works.
Please cease and desist from any further unauthorized use of our work.
Miren al gringuito que no sabe ni cómo funcionan las leyes de su propio país, LOL.
Your assertion that we cannot file a DMCA claim for our work is incorrect and demonstrates a misunderstanding of copyright law. While our work is derivative of existing copyrighted material, the specific expression of ideas, plot, dialogue, and descriptions within our works are original creations and are protected by copyright.
The DMCA is precisely the mechanism by which copyright holders, including those who hold copyright in original elements of derivative works, can address unauthorized reproduction of their work online. The scraping and republication of our work without explicit consent constitutes copyright infringement of our original contributions.
Your claim that DMCA is only for "original works that you hold the copyright of, NOT a derivative work" is a misinterpretation of the law. Copyright protection extends to the original aspects of derivative works.
Please cease and desist from any further unauthorized use of our work.
Copyright only extends to 'original contributions' within a work, which, considering the entire work and characters are both based on infringing content, and are thereafter contextualised by your violation of copyright, is not likely.
Stop using copyright holders work without consent.
Did you even ask them before writing those infringing works? Do you care?
At best, you're a hypocrite—wanting to violate other people's copyright whilst trying to assert your own falsely and leading a campaign of harassment.
The many violent threats and hate comments on your brigading reddit thread have been noted.
Hello, sock puppet.
You clearly have zero idea how the DMCA works, or how copyright works. The DMCA has SUCCESSFULLY defended fanworks in the past.
Fanworks do not violate copyright because they're transformative, and because we do not make a profit from them. AI datasets like nyuuzyou's, however, are used to train AI, that is in turn used for profit--meaning the AI violates copyright, and so does the dataset. Because the AI is being trained on a specific dataset, that means the AI will also generate content that is extremely similar, which does not meet the requirement for "transformative"--another violation.
@FrostKitten can you explain who is making money from this dataset?
Let's look at OpenAI for example, since so many people love ChatGPT.
ChatGPT is free, right? Like. You can enter text into the box and get an answer without paying for it, right? So who's making money?
OpenAI does. Because when you enter text into that box, it takes the information you put in and adds it to the dataset used to train the AI. ChatGPT is ALSO USED FOR BUSINESS PURPOSES, and those businesses pay money to use that same AI. The company profits.
For art, how about Midjourney AI Art, which is also popular. You can generate images via text prompts. That AI was trained by scraping image galleries without permission, including many commercial works and fanworks. If you put in a prompt, you can generate images until you get the one you're looking for. Of course, at the free level, it's only five generations...and then you start paying money if you want more than that. The company profits.
It's become a recent trend to generate art or stories with these AIs and SELL THEM. That's what people get these datasets for. So now it's not just the company that's profiting, but the individual. Off of fanworks taken and entered into the AI without permission.
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
It is STILL theft. It is STILL illegal.
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
It is STILL theft. It is STILL illegal.
model training is fair use
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
You do know what an example is, right?
But alright, fine. Hugging Face makes a profit, if someone decides to deploy a model from here. The more use the model gets, the more is charged to the uploader...hourly. That type of thing wouldn't really be worth it unless the uploader was ALSO making a profit somewhere to offset it. So.
Actually, EDIT: nyuuzyou takes donations, which means they're directly making money off datasets, including this one if it's re-enabled.
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
You do know what an example is, right?
Usually, examples are supposed to be relevant to whatever is going on, which yours is not, because this isn't the action of some huge AI company building something for profit. You'd probably find that the majority of users here don't even like OpenAI that much. I rarely hear any other researchers say anything but complaints about them, with the occasional backhanded compliment.
But alright, fine. Hugging Face makes a profit, if someone decides to deploy a model from here. The more use the model gets, the more is charged to the uploader...hourly. That type of thing wouldn't really be worth it unless the uploader was ALSO making a profit somewhere to offset it. So.
You literally do not know how this site works at all, or what people generally do here. For the vast majority of us, this is a hobby which we LOSE money on. No, Huggingface doesn't offer kickbacks for people using your models or datasets. Why on earth would you assume that they're doing that, especially without even trying to check? They might make some money from hosting (I don't seem to have a limit on public dataset storage for my account though, so as far as I can tell I would pay nothing to upload this as long as I have it set to public), but that doesn't make it make any more sense to complain about Huggingface than it does to complain about hard drive manufacturers profiting from this data hoarding. People uploading datasets like this aren't working on behalf of any giant tech company, they're posting these because toying with AI models and datasets like this is their idea of fun. Big AI companies have almost certainly already scraped this data for themselves long ago so I have no idea why you would think that taking this down would in any way be getting back at them. There is no conspiracy, there are no sinister ulterior motives here, but apparently some people think it's still worth sending death threats over.
You do not care about the people who's work was taken without consent. Understood.
@FrostKitten you do realize that openAI is pulling/curating data themselves, right? they're not using random datasets off of huggingface.
the most likely thing that this dataset would be used for is training some random gooner model finetune, which would then be uploaded here for free.
You do know what an example is, right?
Usually, examples are supposed to be relevant to whatever is going on, which yours is not, because this isn't the action of some huge AI company building something for profit. You'd probably find that the majority of users here don't even like OpenAI that much. I rarely hear any other researchers say anything but complaints about them, with the occasional backhanded compliment.
But alright, fine. Hugging Face makes a profit, if someone decides to deploy a model from here. The more use the model gets, the more is charged to the uploader...hourly. That type of thing wouldn't really be worth it unless the uploader was ALSO making a profit somewhere to offset it. So.
You literally do not know how this site works at all, or what people generally do here. For the vast majority of us, this is a hobby which we LOSE money on. No, Huggingface doesn't offer kickbacks for people using your models or datasets. Why on earth would you assume that they're doing that, especially without even trying to check? They might make some money from hosting (I don't seem to have a limit on public dataset storage for my account though, so as far as I can tell I would pay nothing to upload this as long as I have it set to public), but that doesn't make it make any more sense to complain about Huggingface than it does to complain about hard drive manufacturers profiting from this data hoarding. People uploading datasets like this aren't working on behalf of any giant tech company, they're posting these because toying with AI models and datasets like this is their idea of fun. Big AI companies have almost certainly already scraped this data for themselves long ago so I have no idea why you would think that taking this down would in any way be getting back at them. There is no conspiracy, there are no sinister ulterior motives here, but apparently some people think it's still worth sending death threats over.
Look at you, completely ignoring that nyuuzyou takes donations for this shit. Doesn't matter if they're losing money overall if they're taking money for shit they don't own to begin with, especially when the ultimate goal of taking donations like that is to eventually get it to the point YOU PROFIT FROM IT. And Hugging Face still makes a profit; I'm well within my rights to complain about their continuing to host this dataset when they have more than enough proof the data doesn't belong to nyuuzyou. You wanted a more specific example of who profits, I gave you two.
Go kneel on rice for a couple hours.
Look at you, completely ignoring that nyuuzyou takes donations for this shit. Doesn't matter if they're losing money overall if they're taking money for shit they don't own to begin with, especially when the ultimate goal of taking donations like that is to eventually get it to the point YOU PROFIT FROM IT.
Yes, I am sure that the person living in a heavily sanctioned country who is only able to take donations via cryptocurrency is definitely hoping to strike it big one day! Realistically, no, that isn't the goal; the goal is to at least partially cover the costs of doing this.
Go kneel on rice for a couple hours.
I hope that you can one day realize that nobody here is actually doing anything that hurts you.
model training is fair use
from the TOS... "You represent and warrant that you have ownership, control, and responsibility for the Content you post or otherwise make available on our Website, or otherwise have the right to do so."
He does not have ownership of the content. Full stop. It is against the site's TOS.
model training is fair use
from the TOS... "You represent and warrant that you have ownership, control, and responsibility for the Content you post or otherwise make available on our Website, or otherwise have the right to do so."
He does not have ownership of the content. Full stop. It is against the site's TOS.
or otherwise have the right to do so
like through... fair use? presently what is being done with this data is no different from what the Internet Archive does on a regular basis. it's archival.
They just don't care. At all.
I'm sure, if it was their writing--their creativity, their characters, their stories--they would.
BUT
They don't use their brains for that. Instead they steal.
And they don't give a damn about who they harm. And nothing will change their minds.