Unfortunately, there is not a drop of numerics in that book. It's a good book for learning about symbolic AI. There is, to a solid first order approximation, zero symbolic AI in a system like, say, Google speech recognition.
Any particular reason why? I feel that the problem of mapping sensory inputs like sound etc onto internal concepts for reasoning is an important part of AI.
My personal reason is that I don't consider the output of speech recognition to be "internal concepts for reasoning". In a Spoken Dialog System, this task is typically performed by a subsequent component that does natural language understanding (rather than recognition).
One of the best books on AI and Programming ever.