Hi, On Thu, Jul 4, 2024 at 4:11 PM <rgosw...@quansight.com> wrote: > > Personally, I wouldn't (as a maintainer) take a decision to reject code based > on if I feel it is generated by AI. It is much easier to rule on the quality > of the contribution itself, and as noted, at least so far the AI only > contributions are very probably not going to clear the barrier of being > useful in their own right. > > We are a large, visible project, but I think it is weirder to have a policy > and fail to enforce it (e.g. don't use NumPy for military purposes, never us > an AI tool etc). than to not have a specific policy at all.
It is not a meaningless requirement to ask contributors not to use AI, even if it might be hard to detect. Yes, sure, if we ask people not to use it, sometimes they will because they don't care that we've asked them not to, sometimes they won't realize that they've used it. But especially in the former case, we can, and I think should, reject that code unless they can assure us that the code is free from any incompatible copyright. Which case they could make, if they wanted to. Cheers, Matthew _______________________________________________ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
