Handling AI assisted contributions & improved general guidance for new contributors

[edwardchalstrey1]

@tturocy @rahulsavani @StephenPasteris @d-kad With our recent experience related to GSoC of a high volume of contributions, including some that are likely AI-assisted in nature (or potentially even AI agents themselves), it might be a good time to improve our contribution guidance further; opening this discussion thread as a place to voice opinions on the topic.

I noticed a LinkedIn post today sharing this list of open source project's AI contribution policies. e.g. from matplotlib:

Restrictions on Generative AI Usage

We expect authentic engagement in our community.

Do not post output from Large Language Models or similar generative AI as comments on GitHub or our discourse server, as such comments tend to be formulaic and low content.

If you use generative AI tools as an aid in developing code or documentation changes, ensure that you fully understand the proposed changes and can explain why they are the correct approach.

Make sure you have added value based on your personal competency to your contributions. Just taking some input, feeding it to an AI and posting the result is not of value to the project. To preserve precious core developer capacity, we reserve the right to rigorously reject seemingly AI generated low-value contributions.

In particular, it is also strictly forbidden to post AI generated content to issues or PRs via automated tooling such as bots or agents. We may ban such users and/or report them to GitHub.

Also recommended: https://llvm.org/docs/AIToolPolicy.html

What we have currently

Here is the current contributors guide which focusses on how to contribute from a practical perspective, without much in the way of a code of conduct etc - I also just made a PR #771 which increases the find-ability of this page from the documentation homepage.

We could also update the pull request template I added with a "How, if at all, did AI assist you in these changes" and say "Pull requests which do not disclose how AI was used cannot be accepted"... or similar.

[edwardchalstrey1]

Pull requests created by AI must have been fully verified with human use. AI must not create hypothetically correct code that hasn't been tested. Importantly, you must not allow AI to write code for platforms or environments you don't have access to manually test on.

From polars well worded