28
Took me 6 hours to track down a memory leak in a small Python script I wrote for work - thought AI coding assistants would catch that stuff by now
I kept getting mad at the tool for not flagging it, but then I realized my prompt was too vague and the leak was from a variable I declared wrong outside the loop, so now I'm wondering if anyone else has had to spend way more time debugging AI-generated code than just writing it yourself?
2 comments
Log in to join the discussion
Log In2 Comments
hannaht291mo ago
Felt that. Spent a whole afternoon chasing a bug AI wrote wrong.
5
kim_nelson27d ago
Ugh, the "AI wrote it so it must be right" trap. My trick now is to manually walk through every line the AI generates before I even run it, catches like 90% of the nonsense.
4