We don't need more contributors who aren't programmers to contribute code
4 months ago
- #AI-contributions
- #LLVM
- #open-source
- LLVM AI contribution policy requires a human in the loop to review and understand contributions.
- Contributors must review LLM-generated content before submission and be able to answer questions about it.
- Transparency is required; substantial tool-generated content must be labeled (e.g., Assisted-by: <tool>).
- Automated agents or tools that act without human approval are banned.
- Policy aims to prevent extractive contributions where review costs outweigh benefits.
- Maintainers can label contributions as extractive and request justification for review.
- Contributors must ensure AI-generated content does not violate copyright policies.
- Examples of compliant contributions include reviewed documentation and verified proofs.
- Policy informed by other communities like Fedora, Rust, and QEMU.