How are #opensource maintainers dealing with #AI contributions? I have seen some PRs from new contributors that felt AI generated and lacks substance.

E.g. the PR would change code which is out-of-scope of the issue, or it will apply "best-practices" without sound reasoning.