6 ms·

> When it comes to open source code I suspect it will be found to be similar, if the code is freely visible on the web by a human without an active agreement to

> When it comes to open source code I suspect it will be found to be similar, if the code is freely visible on the web by a human without an active agreement to view it, then it will be possible to "scrape" it. I don't think the license the code is under will apply if that is the case. I don't see the scraping case applying here -- the idea that all human-readable code accessible on the public internet can be ingested into such a system without regard for its license would effectively mean that anything posted to the public internet is entered into the public domain, which seems ridiculous on its face (especially when the author(s) is including an explicit license alongside that code, which the scraper can theoretically also read and account for).

visarga 4y ago
I think we're facing a copyright extinction event. The whole concept is out of touch with the new reality - when you can generate 100 variations for your text, code or image with the click of a button, what does it even mean to hold copyright over the original? "In the style of" killed copyright in 2022.
- UncleMeat 4y ago
  How many people want to read AI generated text in the style of Lord of the Rings vs how many people want to read Lord of the Rings?
  - visarga 4y ago
    Maybe it's not interesting to see what other people generate with it. I want to have chatGPT answer my questions, and reference materials when doing that. For example I could paste an article and start asking probing questions, debating it, asking for summary, ELI5, etc. It would be a research assistant and tutor.
  - throwaway4aday 4y ago
    If an AI can generate an infinite number of stories that take place in the LotR world and do it well and faithfully in the style of JRR Tolkien then I would happily read it. You can only read the trilogy and The Hobbit so many times.
    - UncleMeat 4y ago
      You'd happily read it, but that's an answer to a different question. You've already read LotR. An AI generated novel in a similar style doesn't take money from Tolkien's estate. Some people will want to read the AI generated knockoff. That number of people will usually be much smaller than the number of people who want to read the original and many of those people interested in the AI generated knockoff will also read the original.
      - int_19h 4y ago
        The proportion really depends on 1) the quality of what the AI generates, and 2) expectations imposed by the society - in particular, how consumption of AI-generated stuff is perceived as a matter signaling of one's social class. Both are going to evolve rapidly, so it's really hard to tell where we're going to be in 10 years.
  - PeterisP 4y ago
    With sufficient quality of the model I can certainly imagine people preferring to read AI generated text in the style of Lord of the Rings instead of the original, as the AI generation can adjust the content and style towards the interests of the particular reader, customizing it to deliver what they want. For example, one obvious observation from looking at fan fiction creations is that 'shipping' is popular, that certain people would strongly prefer certain characters to have romantic relations - and AI generated Lord of the Rings lookalike could have the story with the particular relationships that particular reader would prefer, based on an AI analysis of their earlier reactions to other books; this is not something that we have working today (as far as I know) but we're not that far from it becoming real.
- londons_explore 4y ago
  As someone who would vote to repeal copyright entirely, because I think its downsides outweigh its benefits, this is a good thing.
  - satvikpendem 4y ago
    Agreed. I don't understand how people on threads like these turn into copyright hawks just because it seems to undermine their livelihoods, same as in the AI art threads. We should be abolishing copyright, not extending it.
  - londons_explore 4y ago
    I'd actually cut the term of all existing copyrights/patents by a factor of 10 (ie. 70 years goes down to 7, 1 additional year if you aren't dead yet.). I'd then make a new combined copyright/patent system. You would get 1 year on anything creative. You can double the remaining time if you publish all the info necessary to easily recreate what you did. For software, that would be the source code. For music, the score and source recordings. For paintings, the source material and types of paint used, etc. For toys, the 3d design files, etc. 1/2 years of head start is plenty of head start to get your business going. I might make say a 10 year attribution requirement. I'd leave trademark law mostly as-is.
- AlexandrB 4y ago
  This is a pipe dream. There's too much money behind strictly enforcing copyright protections on commercial products. If anything gets killed, it's going to be automatic copyright protection for "little guys". Microsoft will be able to copy your publicly shared code/art/images willy-nilly but will still send their compliance officers to check that your company has a valid Office 365 license if they notice you writing a private letter in Word. Edit: If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better?
  - visarga 4y ago
    I am sure they have an in-house model trained on all their code. It would be immensely more useful than a generic model.
  - sebzim4500 4y ago
    >If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better? Presumably because there might be trade secrets in there that they don't want to leak. That seems entirely separate to copywrite to me.
  - moyix 4y ago
    > If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better? I hear this argument a lot but I think the answer is actually pretty mundane: the model behind Copilot was trained by OpenAI, not Microsoft. Microsoft has a large investment in OpenAI, but they don't own the company, and AFAICT OpenAI did all of the scraping for Copilot on their own, without any special access to MS code.
- marginalia_nu 4y ago
  Haven't stepped on the right toes yet. Create code that will generate art "in the style of Disney/Nintendo" and hey would you look at that, copyright is back and alive again!
SkyBelow 4y ago
>the idea that all human-readable code accessible on the public internet can be ingested into such a system without regard for its license would effectively mean that anything posted to the public internet is entered into the public domain It clearly isn't in public domain. But if we define something new, public knowledge domain. This being material it is legal for a human to look at and learn from. They gain no rights to copyright or IP, but they can learn from it and use it per the existing limits of copyright and IP. Saying that anything posted online gets entered into public knowledge domain seem agreeable. Few things wouldn't be allowed here, generally material that is agreed upon to be illegal world wide (and some countries may have tighter limits, like a theocracy banning someone from learning anything from material deemed blasphemous). Then it is a question of if an AI can also learn off of such material as long as it doesn't produce works that violate existing copyright or IP laws, same as a human. This doesn't seem, on its face, inherently ridiculous. There are still corner cases and potential for abuse, but those also exist with copyright law yet we don't throw the whole system away and just ban all forms of copying or selling the right to copy.
rockemsockem 4y ago
I see lots of folks equate "trained on" to "available verbatim" and that simply isn't the case for the vast majority of training data. It becomes hard to have a productive discussion when there is such focus on the examples that are regurgitated verbatim (often by people with explicit knowledge of the expected output, so they would *know* that they are going to infringe if they republished it) to the exclusion of talking about an in-general system that is trained on data and outputs unique data. IMO that second case is a FAR more interesting question.
- int_19h 4y ago
  The reason why people focus on snippets regurgitated verbatim is because even one such snippet, if sufficiently long and non-trivial, could be sufficient to claim the model itself as a derived work.
8note 4y ago
I'm not clear how it's particularly different from ingesting it into a browser, and then rendering it as part of the html into pixels

visarga 4y ago

I think we're facing a copyright extinction event. The whole concept is out of touch with the new reality - when you can generate 100 variations for your text, code or image with the click of a button, what does it even mean to hold copyright over the original? "In the style of" killed copyright in 2022.

UncleMeat 4y ago
How many people want to read AI generated text in the style of Lord of the Rings vs how many people want to read Lord of the Rings?
- visarga 4y ago
  Maybe it's not interesting to see what other people generate with it. I want to have chatGPT answer my questions, and reference materials when doing that. For example I could paste an article and start asking probing questions, debating it, asking for summary, ELI5, etc. It would be a research assistant and tutor.
- throwaway4aday 4y ago
  If an AI can generate an infinite number of stories that take place in the LotR world and do it well and faithfully in the style of JRR Tolkien then I would happily read it. You can only read the trilogy and The Hobbit so many times.
  - UncleMeat 4y ago
    You'd happily read it, but that's an answer to a different question. You've already read LotR. An AI generated novel in a similar style doesn't take money from Tolkien's estate. Some people will want to read the AI generated knockoff. That number of people will usually be much smaller than the number of people who want to read the original and many of those people interested in the AI generated knockoff will also read the original.
    - int_19h 4y ago
      The proportion really depends on 1) the quality of what the AI generates, and 2) expectations imposed by the society - in particular, how consumption of AI-generated stuff is perceived as a matter signaling of one's social class. Both are going to evolve rapidly, so it's really hard to tell where we're going to be in 10 years.
- PeterisP 4y ago
  With sufficient quality of the model I can certainly imagine people preferring to read AI generated text in the style of Lord of the Rings instead of the original, as the AI generation can adjust the content and style towards the interests of the particular reader, customizing it to deliver what they want. For example, one obvious observation from looking at fan fiction creations is that 'shipping' is popular, that certain people would strongly prefer certain characters to have romantic relations - and AI generated Lord of the Rings lookalike could have the story with the particular relationships that particular reader would prefer, based on an AI analysis of their earlier reactions to other books; this is not something that we have working today (as far as I know) but we're not that far from it becoming real.
londons_explore 4y ago
As someone who would vote to repeal copyright entirely, because I think its downsides outweigh its benefits, this is a good thing.
- satvikpendem 4y ago
  Agreed. I don't understand how people on threads like these turn into copyright hawks just because it seems to undermine their livelihoods, same as in the AI art threads. We should be abolishing copyright, not extending it.
- londons_explore 4y ago
  I'd actually cut the term of all existing copyrights/patents by a factor of 10 (ie. 70 years goes down to 7, 1 additional year if you aren't dead yet.). I'd then make a new combined copyright/patent system. You would get 1 year on anything creative. You can double the remaining time if you publish all the info necessary to easily recreate what you did. For software, that would be the source code. For music, the score and source recordings. For paintings, the source material and types of paint used, etc. For toys, the 3d design files, etc. 1/2 years of head start is plenty of head start to get your business going. I might make say a 10 year attribution requirement. I'd leave trademark law mostly as-is.
AlexandrB 4y ago
This is a pipe dream. There's too much money behind strictly enforcing copyright protections on commercial products. If anything gets killed, it's going to be automatic copyright protection for "little guys". Microsoft will be able to copy your publicly shared code/art/images willy-nilly but will still send their compliance officers to check that your company has a valid Office 365 license if they notice you writing a private letter in Word. Edit: If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better?
- visarga 4y ago
  I am sure they have an in-house model trained on all their code. It would be immensely more useful than a generic model.
- sebzim4500 4y ago
  >If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better? Presumably because there might be trade secrets in there that they don't want to leak. That seems entirely separate to copywrite to me.
- moyix 4y ago
  > If you doubt this, notice that co-pilot was trained on public, open-source code on Github. Not on Microsoft's/Github's own proprietary code. If co-pilot is truly so transformative that copyright doesn't apply, why not feed it all of Microsoft's code to train it better? I hear this argument a lot but I think the answer is actually pretty mundane: the model behind Copilot was trained by OpenAI, not Microsoft. Microsoft has a large investment in OpenAI, but they don't own the company, and AFAICT OpenAI did all of the scraping for Copilot on their own, without any special access to MS code.
marginalia_nu 4y ago
Haven't stepped on the right toes yet. Create code that will generate art "in the style of Disney/Nintendo" and hey would you look at that, copyright is back and alive again!

SkyBelow 4y ago

>the idea that all human-readable code accessible on the public internet can be ingested into such a system without regard for its license would effectively mean that anything posted to the public internet is entered into the public domain It clearly isn't in public domain. But if we define something new, public knowledge domain. This being material it is legal for a human to look at and learn from. They gain no rights to copyright or IP, but they can learn from it and use it per the existing limits of copyright and IP. Saying that anything posted online gets entered into public knowledge domain seem agreeable. Few things wouldn't be allowed here, generally material that is agreed upon to be illegal world wide (and some countries may have tighter limits, like a theocracy banning someone from learning anything from material deemed blasphemous). Then it is a question of if an AI can also learn off of such material as long as it doesn't produce works that violate existing copyright or IP laws, same as a human. This doesn't seem, on its face, inherently ridiculous. There are still corner cases and potential for abuse, but those also exist with copyright law yet we don't throw the whole system away and just ban all forms of copying or selling the right to copy.

rockemsockem 4y ago

I see lots of folks equate "trained on" to "available verbatim" and that simply isn't the case for the vast majority of training data. It becomes hard to have a productive discussion when there is such focus on the examples that are regurgitated verbatim (often by people with explicit knowledge of the expected output, so they would *know* that they are going to infringe if they republished it) to the exclusion of talking about an in-general system that is trained on data and outputs unique data. IMO that second case is a FAR more interesting question.

int_19h 4y ago
The reason why people focus on snippets regurgitated verbatim is because even one such snippet, if sufficiently long and non-trivial, could be sufficient to claim the model itself as a derived work.

8note 4y ago

I'm not clear how it's particularly different from ingesting it into a browser, and then rendering it as part of the html into pixels