4 ms·

I for one am quite happy that AI folks are basically treating copyright as not existing. I strongly hope that the courts find that LLM weights, and the datasets

I for one am quite happy that AI folks are basically treating copyright as not existing. I strongly hope that the courts find that LLM weights, and the datasets are "fair-use" or whatever other silly legal justification. Aaron Swartz was a saint.

suction 3y ago
[dead]
slashdev 3y ago
Agreed, copyright has gone too far. I hope the advent of AI serves to weaken it.
axblount 3y ago
Swartz distributed information for everyone to use freely. These companies are processing it privately to develop their for-profit products. Big difference, IMO.
- Der_Einzige 3y ago
  I am sad about closed source LLMs like ChatGPT, but Llama is in that grey area where it's freely available if you choose to ignore their silly license stuff, which of course pirates and AI developers are all too keen to ignore. Even if they win the lawsuit, LLM development will simply go underground, and as we see from what the coomers at civitai and in the stable diffusion world have done, that may in fact ironically speed up development in AI.
colechristensen 3y ago
While I’ma proponent of free information and loosening copyright, allowing billion dollar companies to package up the sum of human creation and resell statistical models that mimic the content and style of everyone… is a bit far. Fair use is for humans.
- brnaftr361 3y ago
  Yeah, but hypothetically should open source projects be offered special protections? I feel like they should, and with certain caveats where, say, a company like Meta is allowed to claim fair use if and only if they free up the entire ecosystem as they deploy it. But yeah, having private ostensibly profitable models based on other people's work without giving them free access to it is not fair. Give some get some.
  - colechristensen 3y ago
    Maybe… maybe restrictions freed up for resources in the public domain… that is no patent, copyright, license or ownership rights of any kind. As in you could distribute the thing unmolested but that’s it.
pessimizer 3y ago
Copyrights shouldn't exist, but > LLM weights, and the datasets are "fair-use" or whatever other silly legal justification. Would just be a carve out for the wealthy. If these laws don't mean anything, everyone who got harassed, threatened, extorted, fined, arrested, tried, or jailed for internet piracy are owed reparations. Let Meta pay them.
JoshTriplett 3y ago
> I for one am quite happy that AI folks are basically treating copyright as not existing. I strongly hope that the courts find that LLM weights, and the datasets are "fair-use" or whatever other silly legal justification. I would be very happy if either a court or lawmakers decided that copyright itself was unconscionable. That isn't what's going to happen, though. And I think it's incredibly unacceptable if a court or lawmakers instead decide that AI training in particular gets a special exception to violate other people's copyrights on a massive scale when nobody else gets to do that. As far as a fair use argument in particular, fair use in the US is a fourfold test: > the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; The purpose and character is absolutely heavily commercial and makes a great deal of money for the companies building the AIs. A primary use is to create other works of commercial value competing with the original works. > the nature of the copyrighted work; There's nothing about the works used for AI training that makes them any less entitled to copyright protections or more permissive of fair use than anything else. They're not unpublished, they're not merely collections of facts or ideas, etc. > the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and AI training uses entire works, not excerpts. > the effect of the use upon the potential market for or value of the copyrighted work. AI models are having a massive effect on the market for and value of the works they train on, as is being widely discussed in multiple industries. Art, writing, voice acting, code; in any market AI can generate content for, the value of such content goes down. (This argument does not require AI to be as good as humans. Even if the high end of every market produces work substantially better than AI, flooding a market with unlimited amounts of cheap/free low-end content still has a massive effect.)
- lubujackson 3y ago
  It is the equivalent of making a 3D map of a museum and getting sued by one artist of one painting in the museum. Ant individual work in an AI dataset is nearly worthless - only in aggregate does it have value. If that doesn't count as a "transformative work" I don't know what does.
  - thebears5454 3y ago
8organicbits 3y ago
There's a difference between "information wants to be free" and "Facebook can produce works minimally derived from your greatest creative work at a scale you can't match". LLMs seem to aggregate that value to whoever builds the model, which they can then sell access to, or sell the output it produces. Five years from now, will OpenAI actually be open, or will it be a rent seeking org chasing the next quarterly gains? I expect the latter.
- PostOnce 3y ago
  "Will OpenAI actually be open" That ship sailed, friend. OpenAI is no longer a charity in any meaningful sense of the word anymore, it's now an adversarial organization working against the public good with the sole aim of making a few rich men richer. After privatization, they sent their PR people to lobby congress to make it impossible for anyone to compete with them (important note: not out of any interest in actually "protecting" the public from the very AI they're building), and perhaps worst of all, they're no longer being open with the scientific theories and data behind their new models.
  - red-iron-pine 3y ago
    we are rapidly achieving the cyberpunk future, and it's much worse than we thought.
  - two_in_one 3y ago
    > an adversarial organization working against the public good with the sole aim of making a few rich men richer. I wonder what is _not_ in the list?

suction 3y ago

[dead]

slashdev 3y ago

Agreed, copyright has gone too far. I hope the advent of AI serves to weaken it.

axblount 3y ago

Swartz distributed information for everyone to use freely. These companies are processing it privately to develop their for-profit products. Big difference, IMO.

Der_Einzige 3y ago
I am sad about closed source LLMs like ChatGPT, but Llama is in that grey area where it's freely available if you choose to ignore their silly license stuff, which of course pirates and AI developers are all too keen to ignore. Even if they win the lawsuit, LLM development will simply go underground, and as we see from what the coomers at civitai and in the stable diffusion world have done, that may in fact ironically speed up development in AI.

colechristensen 3y ago

While I’ma proponent of free information and loosening copyright, allowing billion dollar companies to package up the sum of human creation and resell statistical models that mimic the content and style of everyone… is a bit far. Fair use is for humans.

brnaftr361 3y ago
Yeah, but hypothetically should open source projects be offered special protections? I feel like they should, and with certain caveats where, say, a company like Meta is allowed to claim fair use if and only if they free up the entire ecosystem as they deploy it. But yeah, having private ostensibly profitable models based on other people's work without giving them free access to it is not fair. Give some get some.
- colechristensen 3y ago
  Maybe… maybe restrictions freed up for resources in the public domain… that is no patent, copyright, license or ownership rights of any kind. As in you could distribute the thing unmolested but that’s it.

pessimizer 3y ago

Copyrights shouldn't exist, but > LLM weights, and the datasets are "fair-use" or whatever other silly legal justification. Would just be a carve out for the wealthy. If these laws don't mean anything, everyone who got harassed, threatened, extorted, fined, arrested, tried, or jailed for internet piracy are owed reparations. Let Meta pay them.

JoshTriplett 3y ago

> I for one am quite happy that AI folks are basically treating copyright as not existing. I strongly hope that the courts find that LLM weights, and the datasets are "fair-use" or whatever other silly legal justification. I would be very happy if either a court or lawmakers decided that copyright itself was unconscionable. That isn't what's going to happen, though. And I think it's incredibly unacceptable if a court or lawmakers instead decide that AI training in particular gets a special exception to violate other people's copyrights on a massive scale when nobody else gets to do that. As far as a fair use argument in particular, fair use in the US is a fourfold test: > the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; The purpose and character is absolutely heavily commercial and makes a great deal of money for the companies building the AIs. A primary use is to create other works of commercial value competing with the original works. > the nature of the copyrighted work; There's nothing about the works used for AI training that makes them any less entitled to copyright protections or more permissive of fair use than anything else. They're not unpublished, they're not merely collections of facts or ideas, etc. > the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and AI training uses entire works, not excerpts. > the effect of the use upon the potential market for or value of the copyrighted work. AI models are having a massive effect on the market for and value of the works they train on, as is being widely discussed in multiple industries. Art, writing, voice acting, code; in any market AI can generate content for, the value of such content goes down. (This argument does not require AI to be as good as humans. Even if the high end of every market produces work substantially better than AI, flooding a market with unlimited amounts of cheap/free low-end content still has a massive effect.)

lubujackson 3y ago
It is the equivalent of making a 3D map of a museum and getting sued by one artist of one painting in the museum. Ant individual work in an AI dataset is nearly worthless - only in aggregate does it have value. If that doesn't count as a "transformative work" I don't know what does.
- thebears5454 3y ago

8organicbits 3y ago

There's a difference between "information wants to be free" and "Facebook can produce works minimally derived from your greatest creative work at a scale you can't match". LLMs seem to aggregate that value to whoever builds the model, which they can then sell access to, or sell the output it produces. Five years from now, will OpenAI actually be open, or will it be a rent seeking org chasing the next quarterly gains? I expect the latter.

PostOnce 3y ago
"Will OpenAI actually be open" That ship sailed, friend. OpenAI is no longer a charity in any meaningful sense of the word anymore, it's now an adversarial organization working against the public good with the sole aim of making a few rich men richer. After privatization, they sent their PR people to lobby congress to make it impossible for anyone to compete with them (important note: not out of any interest in actually "protecting" the public from the very AI they're building), and perhaps worst of all, they're no longer being open with the scientific theories and data behind their new models.
- red-iron-pine 3y ago
  we are rapidly achieving the cyberpunk future, and it's much worse than we thought.
- two_in_one 3y ago
  > an adversarial organization working against the public good with the sole aim of making a few rich men richer. I wonder what is _not_ in the list?