Maybe it's just me but the few times I ask a language model to do some repetitive code for me, it fails. Sometimes it gets a few variables mixed up but you wouldn't notice if you didn't understand your code.
If this is being used in mass, I'm very concerned for the future of software.