5 ms·

Hi all. I would like your help with an ethics problem. I run a site called The Nose, a safe haven for AI training data. It operates overseas in a region out of

Hi all. I would like your help with an ethics problem. I run a site called The Nose, a safe haven for AI training data. It operates overseas in a region out of reach of DMCAs. (Past info: https://news.ycombinator.com/item?id=37512147 https://news.ycombinator.com/item?id=37512147) This was necessary because I felt it was unacceptable for entire datasets to be forced offline by one lawyer. The ethical problem is that I'm sympathetic with people who want to remove their content from AI training data. I received an email from the Danish Rights Alliance about Books3: https://pastebin.com/6qw3yMWZ https://pastebin.com/6qw3yMWZ They point out that this is illegal in Denmark and elsewhere, and threaten to ban thenose.cc from Denmark. Obviously, the threats are meaningless. But I'm interested in your views on whether we should comply with the request by removing the specific titles they list. I was thinking of saying "If you say 'please', I will remove the listed titles." There are 109 entries, so it wouldn't be too much hassle to just remove those from the tarball, and it would be amusing to force a lawyer to ask nicely. For now, I asked for a complete list of the full filenames they want to be removed, along with proof that they represent the listed rightsholders. I'm more interested in how you feel. It seems reasonable to let people opt out of training. We could formalize this process by setting up a way to do this. We could also just ignore takedown demands. What do you think?

19h 3y ago
Make those files torrents and I’ll make sure your data will transcend into the unenforceable infinities of the internet.
nostril 3y ago
One other question: Is there legal basis in Denmark for asking for proof that they control the copyright for the listed works? DMCAs operate on "good-faith belief, under penalty of perjury" but in international situations it becomes trickier. If anyone knows of a Danish lawyer I could consult with, or someone versed in international affairs, please let me know. (Or if you care to contribute funding. Hosting costs around $140/mo right now, which isn't free, but paying for consultation is costlier.)
- rmbyrro 3y ago
  If you're worried about DMCA, you should comply immediately. If you're not obligated to comply with DMCA, why do bother with such question? If there's no process set by law, set yours yourself. By "overseas" and "outside the reach of DMCA", be careful how you draw the lines. Did you incorporate overseas? How are you separating you personally from your corporation? If you are based in a country that obligates you to follow DMCA and if your corporation is nothing but paper and you're the only person involved, a judge might disconsider the corporation as a mere way for you to escape your local jurisdictional obligations.
  - nostril 3y ago
    We're not worried. Our operation is anonymous, and as long as we don't slip up, we'll be fine. Though saying "don't slip up" is very "draw the rest of the owl"; it's most of the work: https://news.ycombinator.com/item?id=37346620 https://news.ycombinator.com/item?id=37346620 But we'd like to do the right thing ethically, which is hard to figure out. Hypothetically, if you were going to set up a process for yourself outside of the law, what criteria would you use?
    - rmbyrro 3y ago
      Disconsidering the law, I'd use the books to train AI. I could use it to train myself, right? What's wrong with training a machine But if you're distributing the contents of these books, that's another story. You're pirating, not training AIs. It didn't end up well for the guys behind The Pirate Bay, unfortunately. They can find you. If they can't bust you for copyright infringement, they'll just make stuff up until they put you in jail. Especially if you offend their personalities. Be careful!..
- yuliyp 3y ago
  I don't think (read: about 99% sure) that DMCA safe harbor applies to someone serving a repository that they themselves have compiled so there's no sense in a rights holder using that process. They can ask with varying levels of niceness and/or sue.
rmbyrro 3y ago
What is the nature of the data in question? Is it related to people, businesses, facts, creative works? How was this data created and how did you have access to it?
- nostril 3y ago
  As far as I can tell, Books3 seems to be training data for language models. I'm not sure how it was created, but it contains a lot of books. I got it by torrenting The Pile after it was forced offline by the Danes. They're asking to remove 109 books from the dataset, which I can do. But I'm not sure whether to. Once you set aside the question of law, it becomes a matter of ethics, and these questions aren't so easy.
  - rmbyrro 3y ago
    I wouldn't set the law aside. Do philosophy over the ethics if you want, but only after you are sure to have the legal side covered, because this one can ruin your finances and your life in general. Unless you're based and incorporated in Iran, Iraq or North Korea, your country has signed the Berne Convention and has implemented in law some level of copyright protection that almost certainly makes the distribution of those books illegal. If you're not taking very careful technical and legal measures to remain anonymous, you can get in serious legal trouble for breaking the law. What is the upside for you? Companies like Uber, Google, etc break the law all the time. But they profit billions from that and then pay millions in fines and lawyers. What's your game? Are you profiting enough to make sense - financially-wise - to break the law? Last but not least, I wouldn't play with lawyers' personalities trying to make them "please" you. Respect them, otherwise, they'll do whatever they can to make you regret it. And believe me, they can do a lot against you. These people are evil. Don't cross their paths.
    - nostril 3y ago
      > What's your game? Are you profiting enough to make sense - financially-wise - to break the law? Not at all. Hosting costs $130/mo, and I feel the sting each month. I'm not sure we'll even get enough donations to cover that, let alone have some kind of profit motive. But we wouldn't want to profit off the works anyway, or else we'd be no better than the corporations. My game is to help people like you be able to train your own models. If I don't help you, who will? Companies will have the final say in what you're allowed to do on your own hardware, because they control the data. No data, no training. The hard part is to balance this with doing the right thing. I'd like to figure out the right thing from first principles and by asking thoughtful people like you, rather than from fear of consequences. As for consequences, we're being careful enough that it seems worth the risk. (You can read more about our precautions at https://news.ycombinator.com/item?id=37346620 https://news.ycombinator.com/item?id=37346620.) But I agree that staying out of jail is preferable to being in one.
jstarfish 3y ago
Just do the right thing. Put yourself in the other party's shoes and consider how you'd feel about approaching this from their side. Textbook publishers aren't improving their offerings with each iteration. They re-release the same shit with a different cover and charge schools (and taxpayers) a premium for this "service." In some cases, the content they republish was already paid for with taxpayer money. Their business model is exploitative on every level. Fuck them. A fiction author puts effort into a work of art. They're not forcing sales or doing anything shady; they're just someone trying to make a living selling copies of their art. Respect that and don't play games with them, unless they can't be civil.
- nostril 3y ago
  This is an excellent point. Thanks. Do you mind if I quote your comment in our official policies? It's interesting because libgen also provides most fiction titles, but everyone is rooting for them. For example, one of the books they want taken offline is from 1954, republished in 2008. So in this case they operate closer to the textbook model than the author model.
  - jstarfish 3y ago
    Thanks; go for it :) I can't speak to libgen's current fiction policy. Just be cognizant of the human element. Your last point is good; I meant to add something about dead authors too. Fuck estates for that very reason. Lazy-ass kids should write their own damn novel.
- refurb 3y ago
  This is ridiculous take. Textbooks aren't publishing what "was already paid for with taxpayer money". By that same logic if I write a book that summarizes all the scientific research in a certain area, then I don't deserve copyright. That makes no sense. Writing a textbook is no different than writing a piece of fiction. It takes actual work to do, and it's original content. And if you don't want to buy the latest edition for your class blame the professor. Most of them are too lazy to actually use older editions and save student hundreds of dollars.
potsandpans 3y ago
> What do you think? Suspect this will be unpopular, but.. Information wants (and deserves) to be free. People make sophisticated and convincing arguments for incentivizing creation, they resonate with me but ultimately I just do not agree with them. I think projects like yours are on the right side of history, but it will take a long while before we collectively agree. Your ethical dilemma hinges on whether or not you agree with the above.
- nostril 3y ago
  Could you elaborate a little more on how you came to this belief? I'm interested in the process of deciding whether to agree or disagree. A good way to get better at that is to get perspectives from thoughtful people.
  - droopyEyelids 3y ago
    People naturally share useful information and this collaboration is the basis of all human achievement and the mechanism by which human society evolves. Putting artificial obstacles in the way of sharing useful information is an act against progress and society itself. For my part im not an absolutist in this (not all information) but i enthusiastically support zlib and libgen because keeping books and papers from those who cant afford it (half the people on the planet!) is, in my view, extremely antisocial.
  - CamperBob2 3y ago
    My take: copyright as a general concept was and is vital to the well-being of a creative society. But: (a) Copyright law is so badly thought-out that I don't feel bad about breaking it; and (b) What's happening in ML is nothing less than the next stage in human intellectual evolution, after thousands of years of relative stasis. It will prove far more important than copyright in the long run, and if a choice is forced the path is clear. I don't have much use for the Roko's Basilisk argument, but I'm loath to take any action that might either hold back progress in this field, or that might make it possible for the technology to be captured and owned by powerful commercial interests. It will be humans, and not machines, who curse us in the future for allowing archaic values and corrupt copyright laws to slow progress down... or for allowing Facebook and Microsoft to control it. TL,DR: party on.
flir 3y ago
Just an opinion: "I think this data set is valuable, and I want to keep it available for use in free countries. After some thought, I think the best solution is the one you propose: go ahead and ban me in Denmark." Shorter version: "Your move, asshole." But that's just me, and it's easy to talk big when it's not your neck on the line. So I reckon you should go with what you think; you're the one in the firing line if they figure out how to come after you.

19h 3y ago

Make those files torrents and I’ll make sure your data will transcend into the unenforceable infinities of the internet.

nostril 3y ago

One other question: Is there legal basis in Denmark for asking for proof that they control the copyright for the listed works? DMCAs operate on "good-faith belief, under penalty of perjury" but in international situations it becomes trickier. If anyone knows of a Danish lawyer I could consult with, or someone versed in international affairs, please let me know. (Or if you care to contribute funding. Hosting costs around $140/mo right now, which isn't free, but paying for consultation is costlier.)

rmbyrro 3y ago
If you're worried about DMCA, you should comply immediately. If you're not obligated to comply with DMCA, why do bother with such question? If there's no process set by law, set yours yourself. By "overseas" and "outside the reach of DMCA", be careful how you draw the lines. Did you incorporate overseas? How are you separating you personally from your corporation? If you are based in a country that obligates you to follow DMCA and if your corporation is nothing but paper and you're the only person involved, a judge might disconsider the corporation as a mere way for you to escape your local jurisdictional obligations.
- nostril 3y ago
  We're not worried. Our operation is anonymous, and as long as we don't slip up, we'll be fine. Though saying "don't slip up" is very "draw the rest of the owl"; it's most of the work: https://news.ycombinator.com/item?id=37346620 https://news.ycombinator.com/item?id=37346620 But we'd like to do the right thing ethically, which is hard to figure out. Hypothetically, if you were going to set up a process for yourself outside of the law, what criteria would you use?
  - rmbyrro 3y ago
    Disconsidering the law, I'd use the books to train AI. I could use it to train myself, right? What's wrong with training a machine But if you're distributing the contents of these books, that's another story. You're pirating, not training AIs. It didn't end up well for the guys behind The Pirate Bay, unfortunately. They can find you. If they can't bust you for copyright infringement, they'll just make stuff up until they put you in jail. Especially if you offend their personalities. Be careful!..
yuliyp 3y ago
I don't think (read: about 99% sure) that DMCA safe harbor applies to someone serving a repository that they themselves have compiled so there's no sense in a rights holder using that process. They can ask with varying levels of niceness and/or sue.

rmbyrro 3y ago

What is the nature of the data in question? Is it related to people, businesses, facts, creative works? How was this data created and how did you have access to it?

nostril 3y ago
As far as I can tell, Books3 seems to be training data for language models. I'm not sure how it was created, but it contains a lot of books. I got it by torrenting The Pile after it was forced offline by the Danes. They're asking to remove 109 books from the dataset, which I can do. But I'm not sure whether to. Once you set aside the question of law, it becomes a matter of ethics, and these questions aren't so easy.
- rmbyrro 3y ago
  I wouldn't set the law aside. Do philosophy over the ethics if you want, but only after you are sure to have the legal side covered, because this one can ruin your finances and your life in general. Unless you're based and incorporated in Iran, Iraq or North Korea, your country has signed the Berne Convention and has implemented in law some level of copyright protection that almost certainly makes the distribution of those books illegal. If you're not taking very careful technical and legal measures to remain anonymous, you can get in serious legal trouble for breaking the law. What is the upside for you? Companies like Uber, Google, etc break the law all the time. But they profit billions from that and then pay millions in fines and lawyers. What's your game? Are you profiting enough to make sense - financially-wise - to break the law? Last but not least, I wouldn't play with lawyers' personalities trying to make them "please" you. Respect them, otherwise, they'll do whatever they can to make you regret it. And believe me, they can do a lot against you. These people are evil. Don't cross their paths.
  - nostril 3y ago
    > What's your game? Are you profiting enough to make sense - financially-wise - to break the law? Not at all. Hosting costs $130/mo, and I feel the sting each month. I'm not sure we'll even get enough donations to cover that, let alone have some kind of profit motive. But we wouldn't want to profit off the works anyway, or else we'd be no better than the corporations. My game is to help people like you be able to train your own models. If I don't help you, who will? Companies will have the final say in what you're allowed to do on your own hardware, because they control the data. No data, no training. The hard part is to balance this with doing the right thing. I'd like to figure out the right thing from first principles and by asking thoughtful people like you, rather than from fear of consequences. As for consequences, we're being careful enough that it seems worth the risk. (You can read more about our precautions at https://news.ycombinator.com/item?id=37346620 https://news.ycombinator.com/item?id=37346620.) But I agree that staying out of jail is preferable to being in one.

jstarfish 3y ago

Just do the right thing. Put yourself in the other party's shoes and consider how you'd feel about approaching this from their side. Textbook publishers aren't improving their offerings with each iteration. They re-release the same shit with a different cover and charge schools (and taxpayers) a premium for this "service." In some cases, the content they republish was already paid for with taxpayer money. Their business model is exploitative on every level. Fuck them. A fiction author puts effort into a work of art. They're not forcing sales or doing anything shady; they're just someone trying to make a living selling copies of their art. Respect that and don't play games with them, unless they can't be civil.

nostril 3y ago
This is an excellent point. Thanks. Do you mind if I quote your comment in our official policies? It's interesting because libgen also provides most fiction titles, but everyone is rooting for them. For example, one of the books they want taken offline is from 1954, republished in 2008. So in this case they operate closer to the textbook model than the author model.
- jstarfish 3y ago
  Thanks; go for it :) I can't speak to libgen's current fiction policy. Just be cognizant of the human element. Your last point is good; I meant to add something about dead authors too. Fuck estates for that very reason. Lazy-ass kids should write their own damn novel.
refurb 3y ago
This is ridiculous take. Textbooks aren't publishing what "was already paid for with taxpayer money". By that same logic if I write a book that summarizes all the scientific research in a certain area, then I don't deserve copyright. That makes no sense. Writing a textbook is no different than writing a piece of fiction. It takes actual work to do, and it's original content. And if you don't want to buy the latest edition for your class blame the professor. Most of them are too lazy to actually use older editions and save student hundreds of dollars.

potsandpans 3y ago

> What do you think? Suspect this will be unpopular, but.. Information wants (and deserves) to be free. People make sophisticated and convincing arguments for incentivizing creation, they resonate with me but ultimately I just do not agree with them. I think projects like yours are on the right side of history, but it will take a long while before we collectively agree. Your ethical dilemma hinges on whether or not you agree with the above.

nostril 3y ago
Could you elaborate a little more on how you came to this belief? I'm interested in the process of deciding whether to agree or disagree. A good way to get better at that is to get perspectives from thoughtful people.
- droopyEyelids 3y ago
  People naturally share useful information and this collaboration is the basis of all human achievement and the mechanism by which human society evolves. Putting artificial obstacles in the way of sharing useful information is an act against progress and society itself. For my part im not an absolutist in this (not all information) but i enthusiastically support zlib and libgen because keeping books and papers from those who cant afford it (half the people on the planet!) is, in my view, extremely antisocial.
- CamperBob2 3y ago
  My take: copyright as a general concept was and is vital to the well-being of a creative society. But: (a) Copyright law is so badly thought-out that I don't feel bad about breaking it; and (b) What's happening in ML is nothing less than the next stage in human intellectual evolution, after thousands of years of relative stasis. It will prove far more important than copyright in the long run, and if a choice is forced the path is clear. I don't have much use for the Roko's Basilisk argument, but I'm loath to take any action that might either hold back progress in this field, or that might make it possible for the technology to be captured and owned by powerful commercial interests. It will be humans, and not machines, who curse us in the future for allowing archaic values and corrupt copyright laws to slow progress down... or for allowing Facebook and Microsoft to control it. TL,DR: party on.

flir 3y ago

Just an opinion: "I think this data set is valuable, and I want to keep it available for use in free countries. After some thought, I think the best solution is the one you propose: go ahead and ban me in Denmark." Shorter version: "Your move, asshole." But that's just me, and it's easy to talk big when it's not your neck on the line. So I reckon you should go with what you think; you're the one in the firing line if they figure out how to come after you.