Skip to Content

Agents' fun

Mary Lou Maher asked: can agents have fun? Her research suggests that agents can be curious, and perhaps as we model "fun" we may soon be able to say that agents that model human fun, know what fun is, and are thus also capable of feeling the sense of having fun. Burelli, Togelius, and Yannakakis work on systems that learn "fun" in the context of certain games in order to create levels that are perceived by humans as being more fun. Can we say their system feels the sensation of fun?