Putting a human in the loop with an LLM makes the result better.

But the human runs many, many orders of magnitude slower than the computer can.

Every bit of feedback takes what feels like a millenia to the computer.