These GPT programs literally don't know what the code does (or the meaning of their output in general). They are Large Language Models. They just know (roughly) how to make responses in a particular language that make sense (english, C++, wingdings). They do not understand the "sense that is made" however.
This requires subject matter experts, like yourself, to use and implement.
This requires subject matter experts, like yourself, to use and implement.
These LLM are tools. They are not sentient.