Codeberg seems to ban vibecoded Projects; reason might be german copyright law
>It looks like Codeberg want only copyrighted material in their service, so it is reliable in the future that e.g. licenses must be followed (e.g. GPL), and copyright doesn’t suddenly get declared as being of the model owner, and it isn’t a copy of something else.
That is a cautious reasonable position - in early days of LLM coding (3 years ago!) indemnity from model companies was a major issue globally because of the lack of clarity of the law around this. The US specifically has settled on it being (effectively?) public domain. But I don’t think that is fully settled, and it certainly isn’t settled in international copyright law.
The goal of the vague “mostly” in the Codeberg change is to ensure there is enough human input to the code they host, to be reasonably sure under German copyright law it is copyright of the person sharing it.
edit: link to poll that caused it (might be down due to high traffic) https://codeberg.org/Codeberg/org/pulls/1253#issuecomment-19820434
edit 1: i dont defend/oppose this move, i just find it interesting
You must log in or register to comment.
Please update with the [AIT] tag in the title.
It means if the guy running Codeberg doesn’t like it. This is his fiefdom, not a public square. Make your own fiefdom for AI coding. You could call it Vibeberg.
LMAO! That’s hilarious and true. Look, until you own your own platform, with your own rules for said platform, or are a significant investor in said platform, then you are subject to the whims of that platform.
Those comments on HN are hilarious 😅 That site really works as a parody of bad tech bro takes if you ignore that most of them are sadly dead serious.
Huh. I mostly read people moralizing the use of AI and wanting to slap a scarlet AI on anyone who uses it while derisively using the terms “slop” and “vibe” to poison the well.
Huh. I mostly read the use of “slop” and “vibe” to be fair depictions
Looks like the bad HN commenter takes are leaking 😑
Same here. I think Lemmy is the den of vipers when it comes to hate and prejudice of AI.
Oh you’ll smoke a turd in hell for your insolence.
The US specifically has settled on it being (effectively?) public domain.
The US Copyright Office has declared that it would not grant copyright protection to AI-generated material due to a lack of human authorship, but it is very far from being codified by law, nor does it say anything about the ownership of the material.
This is still a problem for OSS licensing, since in order to license a work you need to be the copyright holder.
This might mean that OSS projects find it even more difficult to assert their licenses against violators in court, since they would need to prove that the part of the code that was copied was not AI output.
Banning the output of AI tools in OSS codebases is a sensible move for any project that actually cares about enforcing the license that their code is released under.
At the very least it’s a mistake to say the law is “settled” at all. It’s barely been tested.
Indeed, this ruling says nothing about something being a derivative work, all it says is that AI generated works can not create new copyright claims, which is totally sensible and non-controversial.
The bigger problem is that AI tools have ingested a lot of copyrighted and FOSS code and totally will use that code in your projects. So even if AI output can’t be copyrighted, you are still on the hook if your AI copies copyrighted code into your own project. People will notice if your project just changed the variable names
Yes, that is what the term “derivative work” referrs to.
Please add the [AIT] tag - thanks!
This is just to let their prejudice flag fly. There’s no way to actually tell AI was used if the planning and assisting meta-data files aren’t committed to the repo. Nothing has truly changed. They might as well ban anyone who thinks about the color pink more then twice per day - it would be just as easy to enforce.
You might not be able to tell, but the signs something is LLM generated are pretty obvious to more experienced programmers.
Yeah, just like more experienced artists can see the AI generated pixels.
Are you trying to make some point? Because, yes they can.
Yes, I am. I have a global pre-processor directive for all of my prompts, so it always includes “match existing code style, formatting and conventions.” The result is indistinguishable from hand written code. Thinking you can magically discern what was written by AI is delusional.
Sorry but you are clearly delusional. There is no magic involved, just being able to look at the structure of the code and the pattern of the commits. It’s fairly easy in fact.
Oh, I see. You assume agentic AI that can perform it’s own commits. That would make it easy to tell, but you’d have to be really stupid to give AI such unfettered access to your repository.
No need to assume that. Just by using LLMs to assist with code writing, the human in the loop changes their commit pattern.
That, plus, it’s driving away vibe coders just because they don’t want to be there. The rule doesn’t have to be enforceable 100% in order to be effective.
If copyright status is the reason, they should ban any, even a single-line AI contributions.
And it seems a short-sighted decision. Now AI is slop, but what about in 10? 20 years? They would rewrite TOS again?
And what about people they want to migrate from GitHub?
There are plenty of other public git forges you can migrate your LLM assisted projects to 🤷
They would rewrite TOS again?
Yes, that is how that works.
And what about people they want to migrate from GitHub?
Slop to Microslop, code to Codeberg. That simple.
