Hi, I'm Mark Pilgrim. You may remember me from such classics as "Dive Into Python" and "Universal Character Encoding Detector." I am the original author of chardet. First off, I would like to thank...
If the llm they used was trained on the original code, the result was not legally rewritten. To change licensing without buy in from all original authors, the new code must be fully original from spec. Ignoring the legal definitions for convenience opens the door for corporations to steal open source and copyleft materials and strip away the licensing requirements.
That’s a wild claim you’re making. So far, it looks like the code is completely new, and for this case, it doesn’t really matter where it comes from. New code - new license.
If you write new code looking at the old code in another editor window, that’s likely derivative work. If you’ve never seen the original code and are looking only at the API, that’s likely not derivative work. Determining whether the code is ‘new’ is insufficient.
okay, you have to be able to prove the LLM didn’t learn off of the original source material. Because if it is, its dertivitve work, making it subject to LGPL.
That’s valid in a debate, but not quite how courts work?
I’m not a lawyer, just someone petty enough to read laws.
The discovery requests in the law suit will require yo turn over all training data. From there, it will be up to the AI makers to prove that it wasn’t used, if it was fed into training data. Which if it was open source, almost certainly was.
That as side.
Your making an equal claim that it wasn’t. With an equal amount of proof. So what your sating bears as much weight as the other person.
I have not made any claims, and I am not affiliated with this project in any way. I don’t know how this could be dealt with in court, or whether anyone will even bother with it.
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: [email protected]
This is the official technology community of Lemmy.ml for all news related to creation and use of technology, and to facilitate civil, meaningful discussion around it.
Ask in DM before posting product reviews or ads. All such posts otherwise are subject to removal.
Rules:
1: All Lemmy rules apply
2: Do not post low effort posts
3: NEVER post naziped*gore stuff
4: Always post article URLs or their archived version URLs as sources, NOT screenshots. Help the blind users.
5: personal rants of Big Tech CEOs like Elon Musk are unwelcome (does not include posts about their companies affecting wide range of people)
6: no advertisement posts unless verified as legitimate and non-exploitative/non-consumerist
7: crypto related posts, unless essential, are disallowed
If the llm they used was trained on the original code, the result was not legally rewritten. To change licensing without buy in from all original authors, the new code must be fully original from spec. Ignoring the legal definitions for convenience opens the door for corporations to steal open source and copyleft materials and strip away the licensing requirements.
That’s a wild claim you’re making. So far, it looks like the code is completely new, and for this case, it doesn’t really matter where it comes from. New code - new license.
If you write new code looking at the old code in another editor window, that’s likely derivative work. If you’ve never seen the original code and are looking only at the API, that’s likely not derivative work. Determining whether the code is ‘new’ is insufficient.
If the LLM training data is based on / has used GPL code, this might set an interesting legal precedent.
okay, you have to be able to prove the LLM didn’t learn off of the original source material. Because if it is, its dertivitve work, making it subject to LGPL.
LLM is not the copyright owner, it’s a developer of the LGPL package… IMHO, it’s an obvious violation of the original developer rights.
Well, I do not have to, the burden of proof lies on the person making the claim.
That’s valid in a debate, but not quite how courts work?
I’m not a lawyer, just someone petty enough to read laws.
The discovery requests in the law suit will require yo turn over all training data. From there, it will be up to the AI makers to prove that it wasn’t used, if it was fed into training data. Which if it was open source, almost certainly was.
That as side.
Your making an equal claim that it wasn’t. With an equal amount of proof. So what your sating bears as much weight as the other person.
I have not made any claims, and I am not affiliated with this project in any way. I don’t know how this could be dealt with in court, or whether anyone will even bother with it.
You claim
https://vger.to/lemmy.ml/comment/24346212
That its completlt rewritten, with the implication that its not using the project as input.
So yes, you do should back that up
That’s not my claim, this is what the chardet maintainers say.
okay, repeating an unverified claim is better?
https://en.wikipedia.org/wiki/Clean-room_design