Hey, can you upload the archiveofourown data set somewhere else like on a discord, i'd like to have a backup before people try to flag your whole account

#223
by SpacehogSutra - opened

I think it would be nice to have it still around, information deserves to be free :)

Maybe on a site like zenodo?

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

This comment has been hidden (marked as Spam)

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

@enby-ghost And now we know that they definitely used methods to circumvent people purposefully locking/restricting their fics from public view/guests. So it's not just copy/pasting, it's specifically being a creep to steal what they couldn't come up with themselves, because they can't make them work without words that aren't theirs. If it was about just getting what was "available," why not use works from Project Gutenberg or other works in public domain?

@enby-ghost And now we know that they definitely used methods to circumvent people purposefully locking/restricting their fics from public view/guests. So it's not just copy/pasting, it's specifically being a creep to steal what they couldn't come up with themselves, because they can't make them work without words that aren't theirs. If it was about just getting what was "available," why not use works from Project Gutenberg or other works in public domain?

this, 100000%

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

ehh, i dont really see the point

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

If you didn't want your work to be scraped and used for non-commercial AI training, why did you upload it to a site that allows scraping and use in non-commercial AI training?

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

???

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

If you didn't want your work to be scraped and used for non-commercial AI training, why did you upload it to a site that allows scraping and use in non-commercial AI training?

They don't
The supposed proof you sited was under the section of what you CAN'T DO
Screenshot_20250430_083145_Chrome.jpg

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

man you're obsessive

because I care about people having the right to keep their works out of the hands of ai

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

If you didn't want your work to be scraped and used for non-commercial AI training, why did you upload it to a site that allows scraping and use in non-commercial AI training?

They don't
The supposed proof you sited was under the section of what you CAN'T DO
Screenshot_20250430_083145_Chrome.jpg

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes.

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

man you're obsessive

because I care about people having the right to keep their works out of the hands of ai

Who are you responding to?

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

If you didn't want your work to be scraped and used for non-commercial AI training, why did you upload it to a site that allows scraping and use in non-commercial AI training?

They don't
The supposed proof you sited was under the section of what you CAN'T DO
Screenshot_20250430_083145_Chrome.jpg

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes.

that doesn't make it ok

@enby-ghost man (gn), I've never encountered pubic lice, but I can only imagine the experience is similar to dealing with the people on this site: parasitic, invasive, vaguely nauseating, and no one knows what they actually contribute to their environment. wild.

anyway, FYI, the lousy are trying to meeble about AO3 allowing scraping, but it's worth noting that IIRC the data they allow to be pulled is for stats purposes, not the actual prose stored on the site. so they're still wrong, fwiw. ¯_(ツ)_/¯ best of luck.

you know where you can find all these fics in a not scummy way?
Go on AO3 and read them the way all these authors intended!
You don't need a dataset to read these stories.
Don't support this blatant example of theft

Wasn't this the same thing everyone laughed at the NFT people for a couple years back? Copying and pasting isn't theft lol

this isn't just "copy and pasting"
this is someone taking things people worked on and feeding it into ai without the people's consent.
It'd be one thing if these fics were simply stolen and reuploaded somewhere, but they are actively being used to train ai, which many of us didn't consnet to.
Not just fan works, but original works and art as well.

https://archiveofourown.org/tos

I.F. What you can't do
You agree not to use AO3 (as well as the email addresses and URLs of OTW sites):
...
5. to conduct any commercial activity, whether for direct or indirect commercial advantage, including (without limitation) making available any advertising, spam, or other solicitation, or scraping Content in order to commercialize it;

Scraping content is allowed by the TOS of the site as long as the scraped data isn't used commercially. You did consent to scraping occurring when you signed up to use the site.

no one consented for their works to be USED TO TRAIN AI
It's not just the fact someone scraped our works, but they are using them to train something that many of us DO NOT WANT in our creative spaces
There are people who allow their works to be used to train ai
That doesn't gove someone a free pass to do that with EVERYONE'S works

If you didn't want your work to be scraped and used for non-commercial AI training, why did you upload it to a site that allows scraping and use in non-commercial AI training?

They don't
The supposed proof you sited was under the section of what you CAN'T DO
Screenshot_20250430_083145_Chrome.jpg

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes.

that doesn't make it ok

Sounds more like suggestions to me 👍

@enby-ghost man (gn), I've never encountered pubic lice, but I can only imagine the experience is similar to dealing with the people on this site: parasitic, invasive, vaguely nauseating, and no one knows what they actually contribute to their environment. wild.

anyway, FYI, the lousy are trying to meeble about AO3 allowing scraping, but it's worth noting that IIRC the data they allow to be pulled is for stats purposes, not the actual prose stored on the site. so they're still wrong, fwiw. ¯_(ツ)_/¯ best of luck.

you seem bitter

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes

that doesn't make it ok

Is that not the whole point though? That using it commercially isn't okay? If the use of it to train something commercial isn't what you're objecting to, then is it just that it's part of a bulk download? If that's the problem, then we had best start targeting stuff like the archive.org backup of the site. When you follow this line of thought you can conclude that you ought to be objecting to your work being downloaded at all, even when AO3 natively allows you to download works in multiple formats. The entirety of the DataHoarder subreddit could be seen as doing some sort of moral infringement for saving and sharing websites where people have previously shared creative works by that logic.

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes.
that doesn't make it ok
Is that not the whole point though? That using it commercially isn't okay? If the use of it to train something commercial isn't what you're objecting to, then is it just that it's part of a bulk download? If that's the problem, then we had best start targeting stuff like the archive.org backup of the site. When you follow this line of thought you can conclude that you ought to be objecting to your work being downloaded at all, even when AO3 natively allows you to download works in multiple formats.

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes

that doesn't make it ok

Is that not the whole point though? That using it commercially isn't okay? If the use of it to train something commercial isn't what you're objecting to, then is it just that it's part of a bulk download? If that's the problem, then we had best start targeting stuff like the archive.org backup of the site. When you follow this line of thought you can conclude that you ought to be objecting to your work being downloaded at all, even when AO3 natively allows you to download works in multiple formats. The entirety of the DataHoarder subreddit could be seen as doing some sort of moral infringement for saving and sharing websites where people have previously shared creative works by that logic.

I was able to sell AI generated books on this dataset in batches and you are all sitting here wasting time man... Gotta have a different mindset

Yeah, that's correct. What you can't do: scrape for commercial purposes. What this repo did: scrape for non-commercial purposes

that doesn't make it ok

Is that not the whole point though? That using it commercially isn't okay? If the use of it to train something commercial isn't what you're objecting to, then is it just that it's part of a bulk download? If that's the problem, then we had best start targeting stuff like the archive.org backup of the site. When you follow this line of thought you can conclude that you ought to be objecting to your work being downloaded at all, even when AO3 natively allows you to download works in multiple formats. The entirety of the DataHoarder subreddit could be seen as doing some sort of moral infringement for saving and sharing websites where people have previously shared creative works by that logic.

I was able to sell AI generated books on this dataset in batches and you are all sitting here wasting time man... Gotta have a different mindset

I find that hard to believe. If there are models trained this fast on this dataset (big if), it would take you some time to properly edit and get everything in place to even make a book. Professional writiers use proprietary models to help with creative writing, yet you somehow managed to get some random 8B model, stuffed on this data to overcook and write cohesive pieces. I don't think so.

Your need to confirm your account before you can post a new comment.

Sign up or log in to comment