Image courtesy of
We developed a mastermind player which played very well using a completely general definition of curiosity

Most computer games involve a human playing against an intelligent computer opponent. Artificial intelligence (AI) rivals the best human players in many traditional games including checkers, backgammon, chess, and poker. It also plays a major part in modern computer games such as Halo and The Sims.

Computer opponents in more traditional games, such as board games and card games, are created using a variety of techniques. Some are borrowed from game theory (a crossover field between mathematics and AI), while others are based on core AI techniques, such as reinforcement learning or neural networks.

Computer opponents for many modern computer games, such as first person shooters, are usually driven by simpler rule based approaches. An agent has a set of states it can be in, a set of actions it can take, and a set of observations it can make. It then has rules such as “If the state is ‘in a battle’, find the nearest opponent and shoot at him.”

If you are developing a new game or trying to improve an existing one, and you need better AI to drive the game, we would be glad to talk to you about it, so please contact us today!