Ghosts in the Machine: Personalities for Socially Adroit Software Agents

Abstract

Recent research indicates that people respond socially to computers and perceive them as having personalities. Software agents especially are human-machine interaction artifacts that embody those qualities most likely to elicit social responses: contingent behavior, fulfilling a social role, and using language. People may perceive agents as one or many discreet social entities, each with a personality that will shape people's interactions with the machine. An industrial team's research on and experience with the design of software agent personalities is described, in the context of related social science and computer science research. The conclusions from this experience are summarized as guidelines for future agent developers.