Criteria for consciousness in artificial intelligent agents

Abstract

Accurately testing for consciousness is still an unsolved problem when applied to humans and other mammals. The inherent subjective nature of conscious experience makes it virtually unreachable to classic empirical approaches. Therefore, alternative strategies based on behavior analysis and neurobiological studies are being developed in order to determine the level of consciousness of biological organisms. However, these methods cannot be directly applied to artificial systems. In this paper we propose both a taxonomy and some functional criteria that can be used to assess the level of consciousness of an artificial intelligent agent. Furthermore, a list of measurable levels of artificial consciousness, ConsScale, is defined as a tool to determine the potential level of consciousness of an agent. Both the mapping of consciousness to AI and the role of consciousness in cognition are controversial and unsolved questions, in this paper we aim to approach these issues with the notions of I-Consciousness and embodied intelligence.