A big enough quantitative change is a qualitative change. There is a big difference between a bad programmer who banged on that code for three days before finally getting it to do what they wanted it to do, and never went back to try to minimize it and clean it up, and that same programmer pushing a prompt into X-GPT and getting that code in five minutes, then moving on to do it again and again and again dozens of times faster than before.

Like everyone else here of any experience I too have waded through gooey code that was impossible to discern any purpose or design in, because there really wasn't any, after everyone was done hacking on it. But the hacking was still bounded by human speeds.

Our only two options for a code base produced that way would be 1. discard it and start over or 2. hope that the next-generation AIs that aren't just LLMs are able to clean it up, since "automatically cleaning up LLM-generated code bases" is going to be a rather lucrative field. LLMs, no matter how much you hypetrophy them, aren't suitable for coding at scale, and they can't be. Their architecture is just wrong. But that's a claim I only make about LLMs, not AI in general.