[openstreetmap/openstreetmap-website] How should we deal with zero-effort AI PRs? (Issue #6954)

[Andy Allan]

gravitystorm created an issue (openstreetmap/openstreetmap-website#6954)

I'm finding the number of zero-effort AI PRs to be getting a bit tiresome. 

I'm not against people using AI to help them write code (I use claude myself, and find it useful) and I'm not against new contributors joining our team (as I'm sure you all know from my years of working on DX and helping new developers with their code). 

But the zero-effort AI PRs are wasting an increasing amount of time and energy. The zero effort ones are easy to spot - they are usually fixing a recently-created issue, have PR descriptions and/or commit messages written entirely by AI, and often have tell-tale signs that they haven't been proof-read or even been tried by the developer (e.g. outright syntax errors). 

I'm not sure what the best approach is for these zero-effort PRs? In an ideal world, the developer will read our feedback, update their PR, maybe go through a few review cycles and we get some useful code merged at the end along with a newly-upskilled developer who can help further. But is trying to review this zero-effort PRs even worth the hassle? Or should we concentrate our efforts on developers who are genuinely trying to be helpful, and have made at least some demonstrable effort first?

I'm open to suggestions.

-- 
Reply to this email directly or view it on GitHub:
https://github.com/openstreetmap/openstreetmap-website/issues/6954
You are receiving this because you are subscribed to this thread.

Message ID: <openstreetmap/openstreetmap-website/issues/6954 at github.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/rails-dev/attachments/20260401/3b6f961b/attachment-0001.htm>