5 ms·

All large-scale public machine learning stuff is depending on being exempt from copyright restrictions, under fair use doctrine. Look at my responses to all of

All large-scale public machine learning stuff is depending on being exempt from copyright restrictions, under fair use doctrine. Look at my responses to all of the threads about Copilot + GPL for more info about that application of it: https://hn.algolia.com/?query=chrismorgan+copilot+gpl&type=comment https://hn.algolia.com/?query=chrismorgan+copilot+gpl&type=c.... When that is finally tried in court, if it fails to any meaningful extent at all (including going all the way up to Supreme Courts as it doubtless will), then Copilot is dead, DALL·E is dead, GPT-3 is dead, all of these things will be immediately discontinued in at least the affected jurisdictions, at least until such a time as they get the laws changed or judgements overturned.

eslaught 4y ago
To me this feels like the argument that we should allow Uber and Airbnb because they're sufficiently "transformative" use cases. When clearly they are playing fast and loose by the rules and have taken advantage of being early enough to do so. As soon as the rulemakers caught up, it became obvious that they didn't have a license to operate differently from everyone else, just because they're new and popular. Personally, I agree that a strike against AI fair use would kill these current generation of tools. But I don't see why that would be the end of it. What it would do is to create a market for open source data sets with liberal licenses. We'd lose something by not being able to train models on every piece of media that has ever been on the internet anywhere, but it's not obvious to me that was ever really reasonable in the first place. If the only way to make AI that can produce good writing is to train it on every piece of writing ever produced in the history of the human race... aren't we missing something? Surely if AI has a future, it'll have to overcome this at some point.
- musicale 4y ago
  > As soon as the rulemakers caught up, it became obvious that they didn't have a license to operate differently from everyone else, just because they're new and popular. This is well said. One of the primary advantages of these businesses was evading the regulation and taxation that their competitors were subject to.
  - jokethrowaway 4y ago
    Indeed, but this is the proof the market didn't want the regulation and taxation and wanted better technology. Shouldn't we listen to what people want over bureaucrats?
    - ale42 4y ago
      People want a lot of things, but people can be egoistic. What people want is not necessarily good for the society as a whole. That's why laws exist.
- vivegi 4y ago
  > What it would do is to create a market for open source data sets with liberal licenses. This is exactly right. Open Datasets is the way to go. I would also say that in the spirit of the Open Access movement for journals and publications, it might be useful to set up an Open Access protocol for training data sets, methods (these are just the algorithms; publishing them openly might be the way to go) and computed models. This will ensure that models are evaluated for risk by a large set of people and any risks/shortcomings could be addressed soon. Quite similar to how cryptographic algorithms are designed / analyzed in public. Obfuscation might look like it helps, but it doesn't in the long run and just creates more headaches.
- kayodelycaon 4y ago
  There is no "fair use" when it comes to laws and regulations.
egypturnash 4y ago
My feeling is that while these things may be technically falling under fair use, I really feel like they are running roughshod over a lot of ethical and moral lines and that perhaps "fair use" needs to be redefined to explicitly exclude this kind of processing. And if it kills these things, oh well. "Being an artist" is a precarious enough existence in this world as is, I'd be delighted to stop worrying about having to compete with an endless sea of algorithmically-generated barely-good-enough spam.
- int_19h 4y ago
  You will still be competing against it. It'll just be outsourced and presented as manual work.
  - omnimus 4y ago
    Exactly this. It will just become tool used by those who can afford it in secret. No matter what copyright it breaks.
  - egypturnash 4y ago
    True. But there wouldn’t be a growing number of easily accessible websites happily letting everyone type whatever prompt they desire, and there wouldn’t be further development being done in this area. “Zero people trying to disrupt my job with art generators” would be ideal, but “a lot fewer people trying to disrupt my job with illegal art generators they can get in trouble for using” is still better than “hey we released this great copyright infringement machine for anyone to use for free”.
    - int_19h 4y ago
      There will be. They'll just be outside of your country's jurisdiction.
brd529 4y ago
Great points but scary. If training ML models on copyrighted data becomes illegal in the US but remains legal in say China or Russia then the US will quickly fall Behind on ML capabilities - major national security implications at the very least. I suspect if the decision went the way you suggest congress would have to change the law to allow training.
- asdff 4y ago
  Isn't that true for all technology? In the U.S. we have the specwriter system which leads to inefficiency to get around copyright. In China or Russia they just copy the code and iterate.
  - thesz 4y ago
    As a Russian programmer, worked in companies big and small, I cannot say this is even remotely right. All companies I worked with really cared about cleanness of the origin of the code.
XorNot 4y ago
It's not going to fail: the US courts are big company biased, and all the big companies are going to show out in force and money to ensure they get the result they want. But even extending that: knocking copyright'd images out isn't going to stop these systems. We know they work now, so if you have to be careful about licensing then that's just going to be done. The idea that any of these platforms will "die" if copyright fair use doesn't automatically apply is magical thinking. Most art is worthless - companies hoovering up huge corpuses with the correct rights assignment for machine learning is going to be the new business. A company like Disney will drop every piece of output from their staff into a dataset for Disney, then license it under terms to other companies - the tech works, so "invent me a disney character looking like..." would have value internally, just to Disney, for idea generation and refinement - arguably a lot more then to anyone else because they would still retain the artist resources to capitalize on it. Right now, a bunch of people who told themselves that despite the pay, they weren't going to be replaced by AI are shrieking that it's turned out not to be the case (it was obvious for a few years something like this was coming though). They're reaching for every legal tool that they hope will kill these things, forgetting that it's never worked out like that. Copyright being a problem when it happens to you as an individual, is different to when it happens to MegaCorp Inc. which is constantly being sued, has limited liability, and puts payouts down as a line-item expense.
- chrismorgan 4y ago
  For my part, I think you’re right, and that it’s unlikely to fail in at least the USA, and, after consideration of what you wrote, that if it did fail in any meaningful way it would push things in the direction of copyright pools (like patent pools); but at the very least, it would be a massive disruption which would take some time to be sorted out and require a certain degree of starting from scratch in data sets; and all up it’d probably favour big business even more heavily than the current informal consensus. I think there’s also a fair chance in such a situation that European countries with their different approach to copyright philosophy would act as a balancing force, striking down overly-general copyright-assignment-equivalent clauses in terms of service and the likes, which would be the only real way of sucking up as much everything as these models need to work well, especially for providing retroactive relicensing (to avoid a big hole in their sources).
anigbrowl 4y ago
Nah, you could zap the training sets tomorrow and start over with public domain material and it would be fine. In fact I think you could easily get paid to generate more content for it.
- chrismorgan 4y ago
  For text (the GPT-3 case), that’d work to train a model that had no knowledge of the last century of popular culture or idiom, and was significantly biased to more formal and traditional writing styles. The effects of this would be really quite interesting, but I think it would significantly limit the places it could be usefully applied. For DALL·E and Copilot, I’m confident that you couldn’t find anywhere near enough material to produce results anywhere near as good as what there is now. I strongly suspect the results would be too poor to be useful in most places where they may be useful now.
  - gnopgnip 4y ago
    There is tons of english text and images with permissive licensing. All of stack overflow, wikipedia is creative commons. Anything created by the US government or many other governments is public domain
    - chrismorgan 4y ago
      The terms of the CC-BY-SA licenses that Stack Overflow and Wikipedia largely use cannot practically be satisfied in a data model. By design, all outputs derive from all sources to some extent, and the licensing requires that they generally be specifically identified, so you can’t just say “from Wikipedia” or “from Stack Overflow” but “from such-and-such a page, by so-and-so”. “Permissive” is not enough. You need no-strings-attached, and attribution is a string. Hence mostly talking about public domain materials, which make up the vast majority of suitable materials. I suspect that covered works of the USA federal government would be quite a large fraction of the public domain material (as reckoned by the USA) from the last 70 years. I don’t believe it’d be enough to be particularly useful, certainly not for pop culture knowledge or colloquial idiom.
TaylorAlexander 4y ago
I have been saying for a long time. These big companies with an investment in getting these systems working could collaborate with twitter and instagram to include image license options. They could give users the option to opt-in to data collection. This would help the public feel like they’re giving consent, and it would also lead to the creation of massive, properly licensed open data sets. There’s already some big open image datasets to start with. And of course, they could always work more on sample efficiency.
- chrismorgan 4y ago
  I’m not particularly familiar with open image data sets, but I doubt that most of them are suitable. Open Images, for example, uses CC-BY images (https://storage.googleapis.com/openimages/web/factsfigures.html https://storage.googleapis.com/openimages/web/factsfigures.h...). Without the fair use exemption, this would suggest that if you used a model trained on that, you would have to comply with the license of every image, which would mean providing attribution for every single item in the set, which is somewhere between infeasible and impractical. The only types of licenses suitable are ones that require nothing like attribution. This is why you’d mostly be limited to public domain materials (though if it went down this way, you’d find terms of service popping up that included a license grant for model training and selling either your data for model training or trained models without any sort of attribution or remuneration).
  - TaylorAlexander 4y ago
    Actually, I think they could comply by providing a list of every author in the dataset, though this is following the letter rather than the spirit. They could also do research on ways to get the model to return the top ten influential works for some output, and make a legal argument that this is a best effort given technical challenges with tracing every source. For the first point, I get that by reading the license at this image: https://commons.m.wikimedia.org/wiki/File:3H8A7368.jpg https://commons.m.wikimedia.org/wiki/File:3H8A7368.jpg “attribution – You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. share alike – If you remix, transform, or build upon the material, you must distribute your contributions under the same or compatible license as the original.” So, output has to have the same license and worst case the image is accompanied by a link to a list of every author in the dataset. This is a far cry from the “this research would be useless if copyright was enforced” as some people suggest. Here’s an open dataset I found which does not require attribution: https://www.pexels.com/creative-commons-images/ https://www.pexels.com/creative-commons-images/ And Wikimedia commons, some of which require attribution: https://commons.m.wikimedia.org/wiki/Category:Images https://commons.m.wikimedia.org/wiki/Category:Images And it is easy to go take a 4k video camera and start collecting tens of thousands of frames of your own images. My point is that people are throwing up their hands saying oh, respecting the copyright of artists is impossible. But it feels very unfair that these huge companies are walking all over the copyright of small artists, but if we took their code to re use their lawyers would sink us overnight. This upsets a lot of people and it’s a bad look. I don’t actually like copyright but if everyone else has to follow the rules I don’t like giving them a free pass.
saurik 4y ago
I personally feel like that lawsuit happens the moment someone builds the version of this that works on music; in my experience arguing before the copyright office at the library of congress, the people who tend to be the most omnipresent is the RIAA, and when someone releases an AI-generated piece of music that sort of sounds like some recent Taylor Swift song but using that infamous sample from Under Pressure / Ice Ice Baby, the lawsuit will be filed within days.
throwaway290 4y ago
They will be discontinued, but of course the profits made during all this time-- with everyone including those companies knowing how it is basically laundering intellectual property-- will stick.
remram 4y ago
As a private citizen, what can I do to hasten this trial in court? I have a feeling that the sooner the better, before this becomes a big industry that the US government might not want to hurt.

eslaught 4y ago

To me this feels like the argument that we should allow Uber and Airbnb because they're sufficiently "transformative" use cases. When clearly they are playing fast and loose by the rules and have taken advantage of being early enough to do so. As soon as the rulemakers caught up, it became obvious that they didn't have a license to operate differently from everyone else, just because they're new and popular. Personally, I agree that a strike against AI fair use would kill these current generation of tools. But I don't see why that would be the end of it. What it would do is to create a market for open source data sets with liberal licenses. We'd lose something by not being able to train models on every piece of media that has ever been on the internet anywhere, but it's not obvious to me that was ever really reasonable in the first place. If the only way to make AI that can produce good writing is to train it on every piece of writing ever produced in the history of the human race... aren't we missing something? Surely if AI has a future, it'll have to overcome this at some point.

musicale 4y ago
> As soon as the rulemakers caught up, it became obvious that they didn't have a license to operate differently from everyone else, just because they're new and popular. This is well said. One of the primary advantages of these businesses was evading the regulation and taxation that their competitors were subject to.
- jokethrowaway 4y ago
  Indeed, but this is the proof the market didn't want the regulation and taxation and wanted better technology. Shouldn't we listen to what people want over bureaucrats?
  - ale42 4y ago
    People want a lot of things, but people can be egoistic. What people want is not necessarily good for the society as a whole. That's why laws exist.
vivegi 4y ago
> What it would do is to create a market for open source data sets with liberal licenses. This is exactly right. Open Datasets is the way to go. I would also say that in the spirit of the Open Access movement for journals and publications, it might be useful to set up an Open Access protocol for training data sets, methods (these are just the algorithms; publishing them openly might be the way to go) and computed models. This will ensure that models are evaluated for risk by a large set of people and any risks/shortcomings could be addressed soon. Quite similar to how cryptographic algorithms are designed / analyzed in public. Obfuscation might look like it helps, but it doesn't in the long run and just creates more headaches.
kayodelycaon 4y ago
There is no "fair use" when it comes to laws and regulations.

egypturnash 4y ago

My feeling is that while these things may be technically falling under fair use, I really feel like they are running roughshod over a lot of ethical and moral lines and that perhaps "fair use" needs to be redefined to explicitly exclude this kind of processing. And if it kills these things, oh well. "Being an artist" is a precarious enough existence in this world as is, I'd be delighted to stop worrying about having to compete with an endless sea of algorithmically-generated barely-good-enough spam.

int_19h 4y ago
You will still be competing against it. It'll just be outsourced and presented as manual work.
- omnimus 4y ago
  Exactly this. It will just become tool used by those who can afford it in secret. No matter what copyright it breaks.
- egypturnash 4y ago
  True. But there wouldn’t be a growing number of easily accessible websites happily letting everyone type whatever prompt they desire, and there wouldn’t be further development being done in this area. “Zero people trying to disrupt my job with art generators” would be ideal, but “a lot fewer people trying to disrupt my job with illegal art generators they can get in trouble for using” is still better than “hey we released this great copyright infringement machine for anyone to use for free”.
  - int_19h 4y ago
    There will be. They'll just be outside of your country's jurisdiction.

brd529 4y ago

Great points but scary. If training ML models on copyrighted data becomes illegal in the US but remains legal in say China or Russia then the US will quickly fall Behind on ML capabilities - major national security implications at the very least. I suspect if the decision went the way you suggest congress would have to change the law to allow training.

asdff 4y ago
Isn't that true for all technology? In the U.S. we have the specwriter system which leads to inefficiency to get around copyright. In China or Russia they just copy the code and iterate.
- thesz 4y ago
  As a Russian programmer, worked in companies big and small, I cannot say this is even remotely right. All companies I worked with really cared about cleanness of the origin of the code.

XorNot 4y ago

It's not going to fail: the US courts are big company biased, and all the big companies are going to show out in force and money to ensure they get the result they want. But even extending that: knocking copyright'd images out isn't going to stop these systems. We know they work now, so if you have to be careful about licensing then that's just going to be done. The idea that any of these platforms will "die" if copyright fair use doesn't automatically apply is magical thinking. Most art is worthless - companies hoovering up huge corpuses with the correct rights assignment for machine learning is going to be the new business. A company like Disney will drop every piece of output from their staff into a dataset for Disney, then license it under terms to other companies - the tech works, so "invent me a disney character looking like..." would have value internally, just to Disney, for idea generation and refinement - arguably a lot more then to anyone else because they would still retain the artist resources to capitalize on it. Right now, a bunch of people who told themselves that despite the pay, they weren't going to be replaced by AI are shrieking that it's turned out not to be the case (it was obvious for a few years something like this was coming though). They're reaching for every legal tool that they hope will kill these things, forgetting that it's never worked out like that. Copyright being a problem when it happens to you as an individual, is different to when it happens to MegaCorp Inc. which is constantly being sued, has limited liability, and puts payouts down as a line-item expense.

chrismorgan 4y ago
For my part, I think you’re right, and that it’s unlikely to fail in at least the USA, and, after consideration of what you wrote, that if it did fail in any meaningful way it would push things in the direction of copyright pools (like patent pools); but at the very least, it would be a massive disruption which would take some time to be sorted out and require a certain degree of starting from scratch in data sets; and all up it’d probably favour big business even more heavily than the current informal consensus. I think there’s also a fair chance in such a situation that European countries with their different approach to copyright philosophy would act as a balancing force, striking down overly-general copyright-assignment-equivalent clauses in terms of service and the likes, which would be the only real way of sucking up as much everything as these models need to work well, especially for providing retroactive relicensing (to avoid a big hole in their sources).

anigbrowl 4y ago

Nah, you could zap the training sets tomorrow and start over with public domain material and it would be fine. In fact I think you could easily get paid to generate more content for it.

chrismorgan 4y ago
For text (the GPT-3 case), that’d work to train a model that had no knowledge of the last century of popular culture or idiom, and was significantly biased to more formal and traditional writing styles. The effects of this would be really quite interesting, but I think it would significantly limit the places it could be usefully applied. For DALL·E and Copilot, I’m confident that you couldn’t find anywhere near enough material to produce results anywhere near as good as what there is now. I strongly suspect the results would be too poor to be useful in most places where they may be useful now.
- gnopgnip 4y ago
  There is tons of english text and images with permissive licensing. All of stack overflow, wikipedia is creative commons. Anything created by the US government or many other governments is public domain
  - chrismorgan 4y ago
    The terms of the CC-BY-SA licenses that Stack Overflow and Wikipedia largely use cannot practically be satisfied in a data model. By design, all outputs derive from all sources to some extent, and the licensing requires that they generally be specifically identified, so you can’t just say “from Wikipedia” or “from Stack Overflow” but “from such-and-such a page, by so-and-so”. “Permissive” is not enough. You need no-strings-attached, and attribution is a string. Hence mostly talking about public domain materials, which make up the vast majority of suitable materials. I suspect that covered works of the USA federal government would be quite a large fraction of the public domain material (as reckoned by the USA) from the last 70 years. I don’t believe it’d be enough to be particularly useful, certainly not for pop culture knowledge or colloquial idiom.

TaylorAlexander 4y ago

I have been saying for a long time. These big companies with an investment in getting these systems working could collaborate with twitter and instagram to include image license options. They could give users the option to opt-in to data collection. This would help the public feel like they’re giving consent, and it would also lead to the creation of massive, properly licensed open data sets. There’s already some big open image datasets to start with. And of course, they could always work more on sample efficiency.

chrismorgan 4y ago
I’m not particularly familiar with open image data sets, but I doubt that most of them are suitable. Open Images, for example, uses CC-BY images (https://storage.googleapis.com/openimages/web/factsfigures.html https://storage.googleapis.com/openimages/web/factsfigures.h...). Without the fair use exemption, this would suggest that if you used a model trained on that, you would have to comply with the license of every image, which would mean providing attribution for every single item in the set, which is somewhere between infeasible and impractical. The only types of licenses suitable are ones that require nothing like attribution. This is why you’d mostly be limited to public domain materials (though if it went down this way, you’d find terms of service popping up that included a license grant for model training and selling either your data for model training or trained models without any sort of attribution or remuneration).
- TaylorAlexander 4y ago
  Actually, I think they could comply by providing a list of every author in the dataset, though this is following the letter rather than the spirit. They could also do research on ways to get the model to return the top ten influential works for some output, and make a legal argument that this is a best effort given technical challenges with tracing every source. For the first point, I get that by reading the license at this image: https://commons.m.wikimedia.org/wiki/File:3H8A7368.jpg https://commons.m.wikimedia.org/wiki/File:3H8A7368.jpg “attribution – You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. share alike – If you remix, transform, or build upon the material, you must distribute your contributions under the same or compatible license as the original.” So, output has to have the same license and worst case the image is accompanied by a link to a list of every author in the dataset. This is a far cry from the “this research would be useless if copyright was enforced” as some people suggest. Here’s an open dataset I found which does not require attribution: https://www.pexels.com/creative-commons-images/ https://www.pexels.com/creative-commons-images/ And Wikimedia commons, some of which require attribution: https://commons.m.wikimedia.org/wiki/Category:Images https://commons.m.wikimedia.org/wiki/Category:Images And it is easy to go take a 4k video camera and start collecting tens of thousands of frames of your own images. My point is that people are throwing up their hands saying oh, respecting the copyright of artists is impossible. But it feels very unfair that these huge companies are walking all over the copyright of small artists, but if we took their code to re use their lawyers would sink us overnight. This upsets a lot of people and it’s a bad look. I don’t actually like copyright but if everyone else has to follow the rules I don’t like giving them a free pass.

saurik 4y ago

I personally feel like that lawsuit happens the moment someone builds the version of this that works on music; in my experience arguing before the copyright office at the library of congress, the people who tend to be the most omnipresent is the RIAA, and when someone releases an AI-generated piece of music that sort of sounds like some recent Taylor Swift song but using that infamous sample from Under Pressure / Ice Ice Baby, the lawsuit will be filed within days.

throwaway290 4y ago

They will be discontinued, but of course the profits made during all this time-- with everyone including those companies knowing how it is basically laundering intellectual property-- will stick.

remram 4y ago

As a private citizen, what can I do to hasten this trial in court? I have a feeling that the sooner the better, before this becomes a big industry that the US government might not want to hurt.