Linus Torvalds rebukes anti-AI stance in Linux kernel

The open-source world is currently grappling with an emerging dilemma: the infiltration of AI-generated code, often dubbed “slop code,” into projects like Gentoo Linux, Curl, and Ghostty. This flood of machine-created contributions has sparked debates about quality control, authorship, and the future of collaborative development.

The resistance to these contributions has been vocal, leading some to call for outright bans on LLM-created work. Yet, amidst this friction, a subtle shift is occurring. Figures in the open-source community are beginning to recognize that as AI models improve and users become more discerning, the landscape may evolve from one of prohibition to one of careful integration.

At the center of this evolving discussion is Linus Torvalds, the creator of Linux. In a long comment on the kernel mailing list, Torvalds offered a clear perspective on his view of artificial intelligence in software development: “I realize that some people really dislike AI, but this is an area where I’m willing to absolutely put my foot down […] Linux is not one of those anti-AI projects, and if somebody has issues with that, they can do the open-source thing and fork it. Or just walk away.”

This stance reflects a broader philosophy: technology, and AI in particular, should be viewed as tools rather than existential threats. Torvalds emphasized that Linux is fundamentally about improving technology, not engaging in a “social warrior” debate. He asserted that AI is simply another tool, useful for improving code, regardless of how it was generated.

The discussion often centers around specific tools designed to assist development, such as Sashiko, a multi-stage code review instrument for kernel patches. Sashiko claims to identify up to 53.6% of bugs in proposed patches, suggesting its analysis is already superior to human assessment on certain metrics. However, the debate quickly pivoted away from mere metrics and into the philosophy of oversight.

One developer suggested that before any AI-generated suggestions are applied, they must be rigorously triaged, especially when dealing with sensitive contributions. This focus on human judgment over automated output highlights a core tension in the development process.

When pressed on these automated findings, the consensus moved toward practicality. Torvalds pointed out that the value of tools lies in their ability to find “embarrassing bugs,” and insisted that developers should learn to disregard those who resist beneficial technological evolution.

Ultimately, the narrative surrounding AI code is less about a battle against technology itself and more about how we manage our relationship with it. It underscores a poignant point: while AI tools may not possess natural intelligence, they serve a crucial function in improving software quality. The key lies in recognizing that these systems are powerful aids, requiring human oversight to ensure that the resulting code remains robust, effective, and ethically sound.