As the title implies, this week, we discussed the interdependence, interaction, and relationships in robotics. This was also the third background-reading paper that we had. We talked about the theory of the mind, which helps a robot reason the state of its partner.
We discussed that behavior is shaped by three different things: propoerties of a person, features of the situation, and social environment. Interaction is also defined in a similar way; it is conceptualized in terms of need, thoughts, and motives in relation to one another, when dealing with two or more people. When trying to represent interactions, it is best to use outcome matrices and transition lists. That way, we can consider all future implications of any current decisions that a robot must make. One type of interaction is called cyclic interaction. There are two primary features of cyclic interaction: time is important and the fact that you are in the current state becuase of something that happened before. All of this interaction and decision making leads to dealing with interdependence space.
Interdependence space consists of three main parts: control, correspondence, and interdependence. All of these are nicely related in a diagram, for any given situation. While trying to figure out different courses of actions, people also create diagnostic situations. A diagnostic situation is one in which the person that is in the situation (un)consciously puts themselves in that situation in order to figure out the other persons’ character. This is one step closer to figuring out how to program trust into robots and helping robots figure out characters of the human beings they will be working with. Diagnostic situations are also used to figure out what situation is best and selection of situations.
On the other hand, relationship-specific motives are often very helpful in guiding robot behaviors. Relationship-specific motiations are inclinations that help robots respond to particular situations in a specific manner with a specific partner.
That was the basic discussion and summary of the paper that we read. We are also reading another paper, “A survey of socially interaction robots”, for two weeks from now. It is a review of “socially interactive robots”, where human-robot interaction is very important. There will be more to discuss about this next time.