Linux legt das Gesetz für KI-generierten Code fest, sagt „Ja“ zu Copilot, „Nein“ zu KI-Schwachsinn, und die Menschen nehmen Fehler in Kauf – nach Monaten heftiger Debatte kommen Torvalds und Betreuer zu einer Einigung

    https://www.tomshardware.com/software/linux/linux-lays-down-the-law-on-ai-generated-code-yes-to-copilot-no-to-ai-slop-and-humans-take-the-fall-for-mistakes-after-months-of-fierce-debate-torvalds-and-maintainers-come-to-an-agreement

    Share.

    22 Kommentare

    1. IcetistOfficialz on

      The Linux kernel will accept AI-assisted code but not AI-generated slop. Meanwhile startups accept AI-generated slop but not AI-assisted thinking, funny

    2. I mean, it makes sense to me. Especially that the author has to take the responsibility for it

    3. TheMericanIdiot on

      AI code needs to have a human sponsor. Without it, it should be rejected

    4. AbeFromanEast on

      >*humans take the fall for mistakes* 

      The Linux maintainers are ahead of the wider culture in this. rn businesses absolutely love being able to blame ‚buggy AI,‘ mistakes. (throws up hands) „*Nothing we could do to prevent this.*“

    5. haecceity123 on

      >The new guidelines mandate that AI agents cannot use the legally binding „Signed-off-by“ tag, requiring instead a new „Assisted-by“ tag for transparency.

      >Late last year, NVIDIA engineer and kernel maintainer Sasha Levin faced massive community backlash after it was revealed he submitted a patch to kernel 6.15 entirely written by an LLM without disclosing it, including the changelog. While the code was functional, it include a performance regression despite being reviewed and tested. The community pushed back hard against the idea of developers slapping their names on complex code they didn’t actually write, and even Torvalds admitted the patch was not properly reviewed, partially because it was not labeled as AI-generated.

      I have no idea how the „new“ situation is different from the old. Before, the stance was „we have no way to control your use of LLMs, so please don’t be lazy about it“. The new stance is … the same?

      Or did I miss the part of the article where they describe how they plan to reliably compel transparency from someone with a motivation to just *not*?

    6. itsprobablytrue on

      I’m glad the democrats are finally standing up to the conservatives and AI

    7. At work we made the following rule a while back: „We don’t care how code is written, we do care that it passes PR requirements. Whoever opens the PR is responsible for the code“.

    8. AvailableReporter484 on

      This is how it should be everywhere. AI is just a tool. If someone pays you to build a house a hammer isn’t going to do it on its own.

      Use Bob, co-pilot, whatthefuckever to help you ideate or pseudo code and then you’d better review the fuck out of it and make sure you understand it before moving forward.

    9. How did they do it pre-AI when people just copy pasted code from stackoverflow they didn’t understand?

      Like this shouldn’t be about AI or not AI. it shoudl be about code you understand and would write like that yourself or not.

    10. Fuzilumpkinz on

      Honestly this is the way it should be every where. You have to hold people accountable. Use AI, it’s great and can do amazing things. But you have to hold that person accountable. If the person does their due diligence and proper set up along with code review it’s going to be fine. But when they don’t and no one holds them accountable or they just point at Claude, that’s where you get slop.

    11. Like it should be. Linux been people working together since distros started coming out. It doesn’t need terrible A.I. code in it

    12. It ain’t rocket science. You put that code out in the wild, you hold responsibility for what it goes on to do.

    Leave A Reply