This reminds me of the statement Norvig makes at the beginning of SICP that all AI programs seem like magic until you understand how they work, when they just s

This reminds me of the statement Norvig makes at the beginning of SICP that all AI programs seem like magic until you understand how they work, when they just seem quite unimpressive. What we're learning is not so much that fish are wonderfully intelligent as that the things we take as intelligence are actually not-so-complex patterns.