How do we make AI-generated code readable ?

Some believe that we invented programming languages so humans wouldn’t have to read machine code. This is a very narrow perspective.

We invented higher-level programming languages to make coding more abstract, expressive, portable, and accessible, allowing humans to communicate and share algorithmic logic so that we can effectively solve complex problems with the assistance of computers. If you lose the purpose, you lose the meaning; if you lose the meaning, you lose the direction.

And if we allow AI-generated code to become incomprehensible to humans, we risk losing not just readability, but the very layer through which human intent is expressed, communicated, verified, and preserved.