AI has been making a splash lately in the game-playing arena. DeepMind, the Google research group that created AlphaGo, have been proponents of utilizing reinforcement learning methods for game-playing applications. More and more video game companies are using deep learning methods for their AI players. So, where are we coming from, and where are we going with AI in the gaming space?
Traditionally, game-playing systems have been made up of a combination of hardcoded rules that have covered the range of behaviors that the AI is supposed to cover. Have you ever played an older adventure, first-person shooter, or strategy game where the AI players were clearly operating off a hardcoded strategy? More often than not, these AIs used a pathfinding algorithm such as Dijkstra's Algorithm to find the shortest distance between two points. IBM's Deep Blue, which won...
