Abstract

Phoebe Sengers: Closing the Affective Gap

In order to enrich human experiences with computers, many researchers are advocating affective computing, in which computers automatically sense, process, and respond to human emotions.  While recognizing that emotions are an important part of human experience, this research also raises difficult ethical, technical, and design issues.  Substantial privacy concerns arise when computers can become aware of and reason about human emotional states without the active participation, awareness, and consent of human users. In addition, it is extremely challenging to accurately sense emotions, since experienced emotions are complex and elusive, influenced by
many physiological, individual, and cultural factors.  Finally, the need to simplify emotions in order to computationally model them often leads to interfaces that simplify, formalize, and flatten our emotional experiences.  In short, there is a gap between the rich, situated, personal ways in which people experience emotions and the formalized, objective ways in which computers model them.

In this talk, I will show how to close this affective gap by using computing, not to acquire and formally reason about users' emotional states, but rather to provide opportunities for users themselves to experience and interpret their own emotions.  These systems reflect perceived affect in open-ended ways that trigger user reflection by demanding active interpretation.  By shifting the center of interpretation and reflection from the system to the user, we can minimize intrusive sensing, address emotions that computational systems alone cannot truly understand, and focus our design efforts not on formalizations of affect but on rich, complex, idiosyncratic, and enigmatic emotional experiences.  In this talk, I will describe ongoing projects that explore how to computationally support affective experience without reducing it to its formalizations.