Now we know how action-oriented AI can be built into any game. We have seen simple methods to convey the illusion of intelligent behavior, from chasing a spaceship to choosing an attack move in a fighting game. However, these behaviors can hardly qualify as "intelligent thinking." They more closely resemble the behavior of a robot. They are simple sequential tests, which, despite looking intelligent as a whole, are not very far from conventional programming.
Now we will move one step higher on the reasoning scale as we advance into the realm of tactical AI. In this chapter, we won't bother with the behavior itself, but instead will worry more about analyzing and making the right decisions in complex scenarios. We will learn to create intelligent paths through game worlds, to sense and analyze combat strategies, and to synthesize general solutions to complex problems. We will learn to make informed decisions or, even better, to think.