For me programming is about reducing complexity, in one way or another. By reducing needless CPU cycles or RAM, or by reducing the complexity of the code when I read it, making it "more logical", which is of course in tension with what would be the absolute most optimal (e.g. perfectly optimized assembly for everything). You make low-level things to turn into higher-level functions to do even higher level things with, and how joyful that is to navigate and use has a lot, if not everything, to do with how things are named.

LLM are great at creating code, obviously, but what is your (or anyone's) experience with them making code "better" in that way? Finding patterns in messy things and making it more "tidy", renaming things to be less verbose yet still descriptive, or even more descriptive, that sort of thing? Where you can "apply" agents to a codebase and have it come out more accessible to LLM and humans, because it's better code and less of it, if you will. I know rewriting things is more enticing and often more practically useful, but are there efforts/experiment in that direction? With actual code I can look at, comparing before and after?