It seems the authors conflate a problem bring easy to state with little processing power being needed to solve it. This obviously isn't true, very complex mathe

It seems the authors conflate a problem bring easy to state with little processing power being needed to solve it. This obviously isn't true, very complex mathematical problems can be stated in very few bits. Human interactions often can be extremely complex, even though they are relatively slow. Reading a text isn't about matching symbols to words. It is about taking these words and putting them into a social context, potentially even doubting their content or imagining the inner world of the author. Obviously that is what the "inner" brain (which existence seems very dubious to me) has to do. I see absolutely no paradox at all.