LLVM: Concerns about low-quality PRs being merged into main (discourse.llvm.org)

The LLVM project has recently raised concerns regarding the increasing submission of low-quality pull requests (PRs), which may be linked to the rise of AI-generated code contributions. Reviewers are expressing fatigue as they navigate through numerous PRs that require extensive feedback and multiple iterations, often resulting in patches that are later reverted. This trend highlights a potential risk where less experienced contributors, bolstered by AI tools, may submit superficial changes without a deep understanding of the codebase, leading to burnout among seasoned reviewers. The implications for the AI/ML community are significant, as they underscore the challenges of maintaining quality in open-source contributions. As AI tools make it easier to generate plausible-looking code, LLVM advocates for reevaluating review processes and guidelines for PR submissions. Suggestions include rejecting PRs that utilize markdown excessively in commit messages, which can indicate AI involvement but may also create confusion among genuine contributors. To address these issues, LLVM is considering ways to balance welcoming new contributors while ensuring the integrity and reliability of the codebase, emphasizing the need for proactive strategies in managing AI's influence on software development practices.