Linus Torvalds says AI slop issue won't be solved with kernel documentation

Linus Torvalds Rejects AI Documentation in Linux Development

Linus Torvalds has ignited fresh debate by declaring there is "zero point in talking about AI slop" in kernel documentation1

1

. The Linux creator's blunt dismissal came in response to Oracle kernel developer Lorenzo Stoakes, who argued that LLMs require unique documentation and flagging in Linux kernel development. Torvalds insisted that people submitting AI-generated code won't document their patches as such, making any documentation effort futile. "The AI slop people aren't going to document their patches as such. That's such an obvious truism that I don't understand why anybody even brings up AI slop," he wrote1

1

.

The Debate Over AI Tools in Software Engineering

The controversy emerged as a team works on guidelines for LLM-assisted contributions to the kernel, as reported in November2

2

. Lorenzo Stoakes had responded to Dave Hansen, arguing that AI tools aren't "just another tool" and need special consideration. "We're noticing a lot more LLM slop than we used to. It is becoming more and more of an issue," Stoakes noted. He warned that thinking LLMs are just another tool suggests the kernel is immune from problems, which he called a silly position. Stoakes expressed concern that tech press reports stating "the kernel accepts AI patches now" could trigger an influx of AI-generated content.

Torvalds Wants Kernel Documentation to Stay Neutral

Torvalds emphasized he doesn't want kernel development documentation to take any stance on AI. "We have enough people on both sides of the 'sky is falling' and 'it's going to revolutionize software engineering', I don't want some kernel development docs to take either stance," he stated1

1

. He insisted documentation should treat AI as "just a tool" and is meant for good actors, calling anything else "pointless posturing"2

2

. This position reflects a pragmatic view that banning AI tools would simply drive their use underground, with users denying their involvement in AI-assisted coding.

Real-World Impact of LLM-Assisted Coding

People are already using coding assistants for Linux work. Wikimedia developer Dmitry Brant blogged about using Claude Code to modernize a 25-year-old kernel driver2

2

. Torvalds previously said 90 percent of AI marketing is hype and commented he was OK with vibe coding as long as it's not used for anything that matters2

2

. The reality is that LLM coding assistants are popular and currently cheap, making them attractive despite concerns about low-quality patches and potential malicious intent.

Future Outlook for Generative AI in Development

The Register suggests a third AI winter may be imminent, following previous cycles around 1984 and the end of Japan's Fifth Generation Computer Systems project2

2

. The generative AI industry currently spends vast amounts subsidizing models in hopes of future trillion-dollar profitability. Some large model prices are already increasing sharply, and if the industry collapses, plain human brainpower may prove much cheaper than hundreds of billions of dollars' worth of datacentres2

2

. This could make the AI slop issue self-correcting, though for now the debate continues to grow despite Torvalds's assertion there's zero point in discussing it.