Computer games have been classified as the ``Human-level AI's Killer Application''. State-of-the-art computer games recreate real-life environments with a surprising level of detail.
However, even though there have been enormous advances in computer graphics, animation and audio for games, most games contain very basic artificial intelligence AI techniques. The purpose of the talk will be two fold. First I will present the domain of computer games, and in particular or Real-Time Strategy (RTS) games and explain why are they interesting from an AI point of view. Second, I will introduce the Case-Based Planning (CBP) paradigm, and explain how case-based planning is a promising technique in the context of RTS games. Finally, I will present the Darmok system, a CBP system capable of learning how to play RTS games.