Because to decide if it's crap, you still have to read it.And because AI respect coding guidelines, you have to actually understand what the code does to detect crap. Also the sheer number is unmanageable.

In the Monkey Selfie case - https://en.wikipedia.org/wiki/Monkey_selfie_copyright_disput... - courts decided that copyright requires a human author and a human merely setting the conditions for a copyrighted work to appear is not enough.

This reasonably means AI contributions where a human has guided the AI are not subject to copyright, and thus can't be supported by a project's license.

From a less self-centered viewpoint there are plenty of reasons to be critical of LLMs and their use.