Emotional Intelligence for Intuitive Agents

Currently, there are no machines with emotions that influence their reasoning, perception and decision-making abilities to the degree that emotions affect human behaviour in these areas. This could be for two reasons. Firstly, emotions have traditionally been broadly defined and no discrete categorization had been formulated, and secondly, emotions have been viewed as opposing logic, the very basis for computational machines, and as a disruption to rational reasoning and function. It is the very contrasting evidence in recent research that has seen a renewed enthusiasm into emotional research. The role of emotion in rational human behaviour may have a larger impact on cognitive processes than first thought. In this paper, we define emotions and discuss the importance that they will have on artificial intelligences of the future. Dawkins, in "The Selfish Gene" (1976) wrote, "A duck is a robot vehicle for the propagation of duck genes." This sentence could be construed somewhat controversial when viewed in context for AI researchers. When he wrote this, Dawkins was commenting on evolutionary and survival instincts in line with Charles Darwin's view that emotions being prevalent in both humans and other animals are crucial to the survival of the species. If AI researches were to build an artificial robot duck, the least use they would have in mind for it would be the parenting of baby robot ducks. This broaches a fundamental question in the AI domain. Do emotions have any significant contributions to make to the area? Our research seeks to answer this question and concentrates on the emotional assessment of atomic elements within the virtual world of an artificial agent. We envisages a meshing of several psychological theories on emotion generation with agent technology, in order to produce an intuitively rational reasoning and decision making artificial being. To these ends, our continuing research endeavours to expand our understanding of cognitive emotion theories and their influence on affective reasoning in humans and the translation of these models into an affective agent architecture.