Or you could take the view that, in fact, this is one of the things LLMs are very good at, ie making sense of complexity.
I'd love to hear that LLMs can be used to trim and simplify complexity, but I don't believe it. They generate content, not reduce it.