in part:
Collaboration at danger
It is not purely the digital and civic infrastructures that are impacted by the use of LLMs. The free/libre software ecosystem, of which we consider Codeberg an important part of, is a social phenomenon centered on collaboration. Working in this way is only possible thanks to free sharing and mutual learning. This includes even very small tools that are shared and re-used and around which collaboration can start out. In contrast, by adopting LLMs people tend to code single-use software from scratch. While this leads to an increase in 'shared' code, it is mostly code that not only has not been 'written' by anyone but is also not maintained by anyone.
Losing trust in each other
The widespread use of LLMs in FLOSS is instead becoming a multidimensional attack on the trust between contributors and the very idea of convivial collaboration itself. Maintainers are under an increased work-load due to people submitting (often well-meaning) low-effort, LLM-generated contributions that require substantial amounts of time to review. At the same time it is becoming increasingly less clear which projects are maintained by experienced developers and which ones are LLM-generated without any meaningful human oversight and input. In the case of copyleft projects, LLMs additionally also lead to 'license laundering', where copyleft code is stripped of its reciprocity requirements by 'generating' it out of the training data.
We observe an increasing trend of mistrusting each other, up to the point where people who put in actual effort to analyze issues or share their suggestions are being accused of having used LLMs when they did not. At the same time, others instruct their LLMs to hide their traces and actively avoid common patterns, prompting others in reviewing contributions and communication more carefully for signs of machine generation.
Entering a vicious cycle
Together, these forces make collaboration not only harder but also less rewarding: With the transaction cost of collaboration increasing, people are becoming less likely to contribute to creating high-quality software projects and more likely to 'vibecode' a one-off software that is specific to your need, and won't evolve beyond. We get a vicious cycle where collaboration is becoming less and less rewarding, while the amount of single-use software that's unmaintained and never sees any improvements is going up.
Although often well intentioned, sharing the result of an prompt and calling it "libre software" does not make the world a better place. Codeberg is not and does not want to be a place to dump such generated single-use software that no one else will ever look at. We are a place for people to collaborate and improve software together. Within this context, the recent votes can be understood as a reconfirmation of those principles: As we want to center on human collaboration, we will not actively support or engage in the creation of LLMs and will not put our limited resources to use for storing single-use software that would pollute our FLOSS commons.> ## Collaboration at danger
It is not purely the digital and civic infrastructures that are impacted by the use of LLMs. The free/libre software ecosystem, of which we consider Codeberg an important part of, is a social phenomenon centered on collaboration. Working in this way is only possible thanks to free sharing and mutual learning. This includes even very small tools that are shared and re-used and around which collaboration can start out. In contrast, by adopting LLMs people tend to code single-use software from scratch. While this leads to an increase in 'shared' code, it is mostly code that not only has not been 'written' by anyone but is also not maintained by anyone.
Losing trust in each other
The widespread use of LLMs in FLOSS is instead becoming a multidimensional attack on the trust between contributors and the very idea of convivial collaboration itself. Maintainers are under an increased work-load due to people submitting (often well-meaning) low-effort, LLM-generated contributions that require substantial amounts of time to review. At the same time it is becoming increasingly less clear which projects are maintained by experienced developers and which ones are LLM-generated without any meaningful human oversight and input. In the case of copyleft projects, LLMs additionally also lead to 'license laundering', where copyleft code is stripped of its reciprocity requirements by 'generating' it out of the training data.
We observe an increasing trend of mistrusting each other, up to the point where people who put in actual effort to analyze issues or share their suggestions are being accused of having used LLMs when they did not. At the same time, others instruct their LLMs to hide their traces and actively avoid common patterns, prompting others in reviewing contributions and communication more carefully for signs of machine generation.
Entering a vicious cycle
Together, these forces make collaboration not only harder but also less rewarding: With the transaction cost of collaboration increasing, people are becoming less likely to contribute to creating high-quality software projects and more likely to 'vibecode' a one-off software that is specific to your need, and won't evolve beyond. We get a vicious cycle where collaboration is becoming less and less rewarding, while the amount of single-use software that's unmaintained and never sees any improvements is going up.
Although often well intentioned, sharing the result of an prompt and calling it "libre software" does not make the world a better place. Codeberg is not and does not want to be a place to dump such generated single-use software that no one else will ever look at. We are a place for people to collaborate and improve software together. Within this context, the recent votes can be understood as a reconfirmation of those principles: As we want to center on human collaboration, we will not actively support or engage in the creation of LLMs and will not put our limited resources to use for storing single-use software that would pollute our FLOSS commons.
Lemmygraders on their way to call codeberg a reactionary luddite organisation
Lemmygraders on their way to point out for the umpteenth time how qwen only needs a $500 GPU and a power supply and missing the point entirely.
There is no published or accessible LLM, they are all for-profit products, including the Chinese ones.
Not sure why you're singling out lemmygrad. LLM use in software development is a contentious topic everywhere. My observation is that most of us agree that AI in general is a useful tool, but disagree with how the west and its corporate oligarchy approach its development, implementation, and primarily focus on word prediction chatbots.
Not a grader, but the issue with AI is how the capitalists own it and the people do not have control over it, right?
And also on how content is scraped, error rates and datacenters are not made while taking care to reduce pollution and avoid utility price increases.
Markov chains are useful, especially for central planning. We've also had them since the 20s and have understood the systems that modern LLM architectures are built on for about as long.
The issue isn't with the existence or use of Markov chains, it's with the fact that the specific form of LLM is utterly useless for anything beyond propaganda and advertising (but I repeat myself).