Control and Consciousness in the Animal and the Machine

Control and consciousness are two very different areas, but recent developments in the theory of consciousness show that they may be linked through the role played in both by internal self-models. This paper first reviews the likely nature and structure of self-models and world models used for decision and control in natural and artificial agents, and makes parallels with some aspects of Metzinger's self-modelling theory of consciousness. It then describes some work undertaken within the CRONOS project, an attempt to create the infrastructure for machine consciousness by building a complex humanoid robot equipped with an internal model of itself and an internal model of the world. In the realms of decision and control, the robot can manipulate and use these models to discover and select actions for achieving particular goals in the world, although its competence is currently limited by technical factors. The paper concludes with a discussion of how, even at the present stage of development, the system can be used to investigate aspects of machine consciousness.