Linus Torvalds: The AI Slop Issue Is *Not* Going to Be Solved with Documentation (www.phoronix.com)

Linus Torvalds has weighed in on the debate surrounding AI-assisted patches for the Linux kernel, dismissing the notion that the "AI slop issue" can be resolved through improved documentation. He argues that equating large language models (LLMs) with traditional tools underestimates the complexity and challenges they introduce in kernel development. Torvalds stresses that those who contribute poorly crafted AI-generated code are unlikely to document their contributions, making the idea of documentation ineffective for addressing the quality concerns raised by AI-generated outputs. This statement is significant for the AI/ML community as it highlights the ongoing tension between the adoption of AI tools and maintaining high standards in software development. Torvalds’ insistence on viewing AI as merely another tool reflects concerns about the potential decline in code quality and the implications for kernel stability. By rejecting the notion that documentation can serve as a panacea for AI-related issues, he calls for a responsible use of AI in development, emphasizing a need for developers to personally uphold coding standards rather than relying on external guiding documents.