So some people think that even if you write all the actual program code and the tests for it by hand, it's fine to use #slop machine to create some missing tests for old code. You know, the stuff that's been on the backburner since forever, nice-to-haves that nobody has time to work on.
Let's say you don't care about ethics. Let's say you don't care about society. Let's say that you're making sure that the slop machine only adds new tests and doesn't touch existing code, that you double check that it didn't incidentally introduce any exploits in the process. What could possibly go wrong?
For a start, the added tests could be meaningless. They could be increasing test coverage without actually testing anything. They could be giving you false confidence without actually checking for regressions. Of course, many software developers can do that manually as well, but now you've streamlined the process. Not to mention you're wasting time and energy on not doing anything meaningful.
Worse, the tests could be actually slightly wrong. Yes, they aren't causing any breakage right now. But when the tested code changes, they may suddenly start failing, and give you or your contributors the wrong idea of what the problem is. Did you never get a misleading test result, and fix the wrong thing? And now it isn't even your test, it's just slop.
#NoAI