[jfz]

I'll admit that finding the right fitness function is hard, but I find that multi-dimensional fitness functions are well-suited to finding a set of solutions that a human can choose from. If you are finding paths for a fleet of delivery trucks, you can optimize separately for time, distance, and cost. Then, with the solutions that are better at all three than every other solution, pick 10 different ones and let a human make a decision.

I agree that when there's a fast, perfect solution, it doesn't make sense to use genetic algorithms. But when finding solutions to a non-general problem (optimize CNC tooling to produce a list of orders, each of which has a series of operations that require a certain amount of time, on certain machines, and require being moved from machine to machine, such that you produce the most on-time orders for high-priority clients), genetic algorithms can work very well.