Robotics Inspired in the Organism

摘要

The major technical and conceptual advances in cognitive science have always been connected directly or indirectly to artificial intelligence and, over the last 20 years, specifically to the field of autonomous robotics. It is thanks to the elegance and practical value of work in various regions of robotics that certain fundamental questions in cognitive science can be revisited. This is particularly the case for the question of meaning. What makes an engagement with the world meaningful for a robot ? In this paper I argue for an approach that addresses this question based on a life-mind continuity perspective supported as much by notions embodiment, complex adaptive systems and enaction as well as phenomenologically and existentially informed theorising. I concern myself with the question of what constitutes the identity of a cognitive system and answer this question in terms of autonomy, defined as operational closure. Then I examine how autonomy lays down a possibility for normative regulation of environmental interactions (agency) in terms of sense-making, the hallmark of cognition. I discuss several questions that are opened by this systemic approach : like whether habits may be considered as higher forms of autonomous identities beyond the organismic self. Some direct implications for robotics are explored by revisiting and extending models of homeostatic adaptation to distortions of the visual field.