> It understands the framework I've built, what functions to use when and writes code looking and working as if it was my own.
Yes, that's what LLMs do. They build statistical models based on their context and training data. They then give the most likely statistical output.
There's nothing creative about it, it's all statistics based on the inputs and they can at best extrapolate but they cannot fundamentally move outside of their inputs.
Humans uniquely can.
