CONTRIBUTING.md: added guidelines about the disclosure of AI/LLM assisstence in submissions#8662
CONTRIBUTING.md: added guidelines about the disclosure of AI/LLM assisstence in submissions#8662firewave wants to merge 1 commit into
Conversation
|
This is just a basic guideline on how to handle the disclosures without any given restrictions. I think this a reasonable and should be non-controversial. But I would still get some feedback/approval from several recent/regular contributors on this but merging it. CC'ing @danmar @chrchr-github @pfultz2 @ludviggunne @swasti16 @wjakobsson @francois-berder @olabetskyi @glankk @gruenich @orbitcowboy @mptre @IOBYTE |
|
Disclosure on my side - I have a rigid anti-AI stance (and voiced it several times in the past already - that should not be up for discussion here though) but I am not going to enforce this on this project or other people and I would just like to see some reasonable disclosure of it. |
|
Sounds about right to. You are not (yet) restricting, just asking for transparincy. |
|
|
||
| If you choose to use AI/LLM to assist with your work please disclose this in the description of the PR (so it will be part of the commit message of the merged changes) by stating the name and the version of the model in question. | ||
|
|
||
| Also please attribute any such changes with `Co-authored-by:` for the respective model (if available). |
There was a problem hiding this comment.
I don't think this is useful, there is some consensus that AI should not be regarded as an author: https://academia.stackexchange.com/a/225428
There was a problem hiding this comment.
That is unfortunate as that would have greatly helped to easily identify (one click) commits which were authored with the help of AI.
I do have seen people attributing it to GitHub accounts for these models.
No description provided.