The trick is to split the code into smaller parts.

This is how I code using ChatGPT:

1. Have it analyze how to structure the program and then give me the code for the outline with not yet implemented methods and functions.
2. Have it implement the methods and functions one by one with tests for each one.
3. I copy the code and test for each method and function before moving on to the next one So that I always have working code.
4. Despair because my code is working and I have no idea how it works and I have become a machine that just copies code without an original thought of my own.

This works pretty well for me as long as I don't work with obscure frameworks or in large codebases.

I just use it for snippets - "here's my function, how would I go about changing x?" Or, "here's my block of code, I'm getting this error, what am I missing?" (I know, I'm fine to share my code but not company code)