Torvalds' answer to AI disclosure rules: no compilers, no editor, no backspace
SummaryOct 13, 2026: replying on the Linux kernel mailing list to a push to keep AI-use disclosure, Linus Torvalds proposed a sarcastic 'new rule' banning compilers and commit-message editors, then called such rules 'UTTERLY STUPID.' The quote is going around X again this week.
· MODEL Editors · 1 min read
Some links may be affiliate links. Disclosure
What happened. On Oct 13, 2026, Linus Torvalds replied on the Linux kernel mailing list to kernel developer Laurent Pinchart, who did not want the kernel's Assisted-by: AI-disclosure tag removed without another required disclosure mechanism. Pinchart argued the rule was about "setting our expectations." Torvalds disagreed (LKML archive (opens in a new tab)):
"I don't understand this whole idiotic concept of 'don't use AI for X'. Here's a new rule just for you: don't use compilers for any of your kernel submissions."
He went on: write it in assembler to prove you understand the machine, and "you had better not use an editor to write that commit message either," because editors make it easy to "backspace and hide the mistakes you made while typing." Then: "Do you not see how UTTERLY STUPID those kinds of rules would be?"
His actual point. "I expect people to understand what they are doing. But that understanding has very little to do with what tools they use."
Context. The Maintainers Summit in Prague (Oct 8) reached consensus to deprecate the Assisted-by tag; Sasha Levin posted the removal patch and documentation maintainer Jonathan Corbet applied a revised version on Oct 14 (The Register (opens in a new tab)). Levin's patch said roughly 10% of the 18,533 non-merge commits in Linux 7.3 through -rc7 carried the tag. Humans still have to review AI-assisted code and sign off with their own Signed-off-by. We covered the tag drop in Slop doesn't need a label. It needs an owner.
Why it's back. The quote is circulating on X again this week as part of a wider debate about vibe coding and AI eroding coding skills.
Caveats. The "no compilers / no backspace" rules are sarcasm, not policy. Not everyone on the list agreed; other kernel developers argued that LLMs change review norms more than past tools did.
Disclosure: Not sponsored. No affiliate links. Not tested by MODEL.
MODEL Editors
Sources
Keep reading
AI News
He spent $20,000 so AIs could play Werewolf
Dwarkesh Patel's Clocktower Bench (Nov 5) ran 386 Blood on the Clocktower games, model vs model. GPT-6 Astra leads Claude Opus 4.6 by ~430 Elo (~92% wins). The edge is detection, not better lying.
· MODEL Editors
AI News
Claude Code's product lead just shipped herself to OpenAI
Cat Wu (@_catwu), former Head of Product for Claude Code (and Cowork) at Anthropic, is joining OpenAI to lead product management for Core Products — weeks before Anthropic's planned IPO. OpenAI confirmed to Business Insider; Anthropic declined comment.
· MODEL Editors
AI News
Claude Code cut the status-line tax by ~99%
Background subagent status used to re-read the whole chat every 30 seconds. Now Haiku 5.5 writes it from short notes — Anthropic says each update costs about 99% less, and sessions with background subagents used ~5% less of the usage limit (~6% more prompts). CLI v2.1.316+.
· MODEL Editors
FREE · EVERY WEEK
The top AI tools this week, tested.
Every week MODEL's editors pick the AI tools worth your time, try them hands-on, and tell you what's good, what's not, and who it's for. Free, in your inbox.
Free. Unsubscribe anytime. Our picks are editorial and can't be bought.