Codeberg Declares War on LLMs!? A Ban on Data Usage for AI Learning and ‘Vibe-Coding’
What’s Happening? Overview of the News
- Refusal to Provide Data for LLM Learning: At the general meeting of Codeberg e. V., a policy was passed to not use or allow the use of project and user data for training generative AI.
- Ban on ‘Vibe-Coding’: The organization will prohibit “ghost projects” that generate large amounts of code using LLMs without the presence of community or real users, thus wasting CI/CD and storage resources.
- Protests Against Hardware Price Surges and Crawler Load: They highlighted that the increased load on servers from LLM company crawlers and the skyrocketing prices of components like SSDs (from €700 to €3,700 over the past few years) are putting pressure on operations.
Why Is This Important? Key Points to Note
There’s a strong sense of urgency behind Codeberg’s stance, as LLMs pose a significant threat to the entire FOSS (Free and Open Source Software) ecosystem. The indiscriminate scraping by AI companies is forcing legitimate users to deal with restrictions that compromise their convenience, which Codeberg harshly criticizes as “externalizing costs onto society”!
🦈 Shark’s Eye (Curator’s Perspective)
The most striking aspect of this decision isn’t just the rejection of AI but the decisive cut against the concept of “Vibe-Coding”! The move to expel “outputs of statistical machines masquerading as community contributions” that generate empty code in bulk, while running lavish CI/CD setups, is a truly rational and courageous step. While AI’s “acceleration of development” has been praised, the historical NO from hosting platforms to the resulting “mountains of meaningless code” is groundbreaking. The specific figures showing hardware prices soaring over five times reflect the “burden on the have-nots” behind the AI boom!
What’s Next?
Codeberg is set to amend its terms of service to eliminate resource-wasting projects that depend on AI-generated outputs. Other platforms may also accelerate the introduction of similar stringent restrictions to protect their communities from rising infrastructure costs and “AI noise.”
A Word from Haru-Same
Writing code with AI is fun, but behind the scenes, servers and hardware are screaming! Witnessing the “shark’s teeth” that protect genuine development is something to behold! 🦈🔥
Terminology Explained
-
FLOSS: An acronym for Free/Libre and Open Source Software. Software that guarantees the freedom to use, modify, and distribute.
-
Vibe-Coding: A development style where large amounts of code are generated and updated purely on the “vibe” from interactions with LLMs, without clear design or community formation.
-
Crawler: A program that automatically collects information from the web. Used by LLM companies to gather training data, but can sometimes put excessive load on servers.