LLMs manifestly cannot code.
They actually can't do it.
I feel like a lot of this debate is kinda like people making reference to a famous book, but no-one has actually read it.
LLMs are a weird bunch of statistics applied to the ordering of symbols.
An LLM is always waiting for a prompt from a human, and will frequently give answers with simple logical errors.
LLMs do not understand what you are saying. They are not alive or conscious or sentient.
They are just another kind of program.
