Comment by alansaber

This has always been the premise. They can't fix the fundamental problems with LLMs but they can continue to optimise them for IE parsing large volumes of data quickly