Present video games need artificial intelligence (AI) to create fun and challenging experience. However, game AI usually needs some level of “stupidity” in order to be defeatable by humans. Therefore, game AI is limited to elementary AI techniques such as finite state machines, rule-based systems, and basic search algorithms. This has created a huge gap between computer science AI and game AI. Since early 2000s, developers and researchers have started to realize that recent advances in academic AI has a lot to offer to the gaming industry, including automated game testing, procedural content generation, and player modelling. This talk gives a gentle introduction to game AI research and its most important concepts. We will also demonstrate the state of the art by highlighting some of the most recent successful applications of game AI research. This talk can be useful for anyone who has a basic understanding of AI concepts and is wondering what is happening in the game AI community.