@werdahias How will you be able to do this, if the kernel itself has AI contributions as Linus just seemed to have allowed. Or is this to explicitly raise the pressure on the kernel developers?
@werdahias thanks for putting this together and up for a vote!
> While enforcement could be a challenge, this is a statement of intent by the Debian community, and we trust this community to adhere to it in good faith.
Nice, I still expect you to get a lot of "how will you enforce this" gotcha style questions from people who don't read past the title
@jych @werdahias "It does not include:
Upstream projects using LLMs for development
AI-related software
Upstream patches/security fixes etc.
"
> Licensing and Attribution:If any pre-existing copyrighted materials (including pre-existing code licensed as free software) authored or owned by third parties are included in the AI tool’s output, prior to contributing such output to the project, the contributor should verify they have the right to submit it under the relevant open source license.
I suspect this part (from the second proposal) would be even harder to achieve than enforcing the first proposal :D
Like, maybe there will be a couple of hours saved by using an LLM, but then you need to hire a whole team for a year to figure out where the output came from, and if any of it is licensed under conflicting terms 👻
https://linuxiac.com/debian-developers-debate-ban-on-ai-assisted-contributions/
@werdahias eh. Proud to have had the chance to AM you in the project!
@werdahias it will be interesting to see how many people who tried to block systemd because it was too dependent on contributions from a single vendor, or who won't ship ZFS kernels for copyright reasons, argue in favour of making Debian dependent on non-copyrightable code with no provenance and entirely dependent on a couple of company toolchains.