Code written for humans to work with has to fit into a human brain's limited context window.

Which requires layers of abstraction; leverage for thinking.

A massive accomplishment and benefit... but also leaky, hard to reason about cleanly.

LLMs can maintain a quite large context window to patiently sift through.

Code written for LLMs will have less abstraction, less leverage.

Code written by LLMs will still use abstraction, because it will mimic code it has seen: code written for humans.

But as more code is created that humans don't have to understand, that can be more detailed and less abstracted and still work, the LLMs will learn to code in that way too, slowly.