Expert systems are in facts 'pattern recognition' systems. Input are collected, re-arranged into patterns then matched with 'known patterns' (in a data-base or knowledge-base). So, expert systems work like:

"IF" input pattern-X matches with stored pattern-Y "THEN" output a stored or known solution-Z.

Expert systems are great for "common" (normally distributed) repetitive problems ("repeating histories"). But expert systems cannot output "NEW SOLUTIONS" (innovations).

A question: Are our kids being trained in school to become "experts" in sciences, literature, ...? ;-)