Engineering empathy

Abstract

Abstract Human intelligence has a built-in locus of concern, the self, to drive and structure cognition and action. The creativity and adaptability (as well as characteristic biases and defects) of human thinking are products of an architecture that embeds emotional processing and self representation at the core of cognition. These themes are raised and addressed in different ways by work in artificial intelligence that connects artificial emotions, general intelligence and self awareness. Although much of that work is conducted in ‘classical’ AI paradigms the themes are equally to forms of generative AI that define the current landscape. The aim of this chapter is to show how these issues arise for AI and to suggest that current debates about AI, reasoning, consciousness, morality, and autonomy lack a key ingredient: the attention to the nature of the self for whom these activities are existentially vital. There are as yet no AI systems that co-ordinate systemic function based on a predictive model of the consequences of action for a self on whose behalf action is performed.