A procedural critique of deontological reasoning

This paper describes a prototype game that learns its rules from the actions and commands of the player. This game can be seen as an implementation and procedural critique of Kant's categorical imperative, suggesting to the player that (1) the maxim of an action is in general underdetermined by the action and its context, so that an external observer will more often than not get the underlying maxim wrong, and that (2) most ingame actions are morally "wrong" in the sense that they do not contribute to wellbalanced game design. But it can also be seen as an embryo for an authoring tool for game designers, where they can easily and fluidly prototype new game mechanics.