I commend to you jonny's thread on the tests:

https://neuromatch.social/@jonny/116666900898570791

It keeps turning out that when you look at the AI output, it's shit.

On one of the BlueSky threads going over over the test code, one of the things they uncovered was some stuff running as root which in no world should be necessary. He may not have just prompted Claude to "convert test suite to python", but there's a lot there that seem like clear red flags in terms of AI slop code.

Which is no surprise, really, given that properly proof-reading AI code is often much more labour intensive than just writing the code oneself. It's easy for things like this to slip through the cracks, even if you are trying to check the AI output

It's a perfect example of how "using LLMs for test coverage" can also be harmful. He expected the tests to to prevent introduction of said regressions, probably based on a combination of the quantity of tests and their style (they look like what decent human written tests look like). But the tests are AI slop, and so they give a lot less value per line of code than he expects, hence a significant regression.

It is literally useful to call these tests AI slop, and the problem is in part caused by not calling them AI slop, and having consequent inflated expectations. LLMs are not any better at writing tests than at writing other code! It is merely that the bar for tests can, legitimately, be a lot lower (in projects where there would otherwise be no tests at all). Making an exception to calling AI generated tests "slop" is thus counter productive, because it leads people to act as if LLMs are actually better at writing tests than at writing other code, and not just because the bar for tests is frequently very low.

edit: actually scratch that I looked at the PR and those tests even look like dogshit and worse than the tests I seen claude write at a workplace that was into vibecoding (which i since quit).