This question on the Theoretical Computer Science Stack Exchange generated some good answers (via Nicolas Seriot): So “why” exactly do we study deterministic and non-deterministic finite automata (DFA/NFAs)?