As per the emotion theory, the following simple examples are based on situations Steve desires to have come about, or to have not come about.
Goal: I will engage the user Goal: The user will retain task knowledge (adapted from Herman the Bug) Goal: The user will be cautions. Goal: The user will perform increasingly well over a previously assessed baseline Goal: I will give explanations about the subject domain, and discuss interesting details about the domain with the user.
By caring about how well the student performs, Steve shows the student that he is interested in the user's progress. When combined with a friendship relationship, Steve also can exhibit fortunes-of-others emotions which is a rudimentary form of ``caring'' about the student himself.
By exhibiting happiness when he has a chance to share information about the subject domain, Steve manifests a rudimentary form of enthusiasm about the subject.
The bi-valenced goal for caution can be implemented, for example, as fear and hope over a prospective future event. For a caution-sensitive Steve, there might be both a recurring hope, and a recurring fear with respect to an accident happening to a user, subsequent to training with Steve. Each instance of user caution raises the threshold which controls whether or not the rumination occurrence (implemented as a cyclic, self-generating, event) actually leads to an instance of fear. Similarly each instance of sloppiness, leading to a simulation accident (or potential accident), leads to a reduced threshold. With identical structure, but in a contrasting manner, ruminations leading to hope that an accident will not occur would take the same form: the more the user exhibits caution, the more hope will tend to arise; the more the user exhibits sloppiness the less hope will tend to arise.