Artiicial Life and Natural Stories

This paper speculates about`stories' as potentially useful units to construct historically grounded artii-cial life forms. We draw our motivation from the structure and organisation of biological life forms and the organisation of human memory and sociality. A fresh approach towards historical artiicial life creatures is to develop story-telling systems supported on situation-ally grounded, discrete dynamical systems. This involves creating autobiographic agents which, by telling through remembering, by constructing stories, mightùnderstand' the world. We address these issues using methods from the area of Algebraic Engineering, growing out of the algebraic theory of automata developed by Krohn and Rhodes. Our approach aaords mathematical methods for expressing events during the existence of an agent in an associative multiplica-tive structure that adequately captures the notion of irreversibility of events in life histories. Furthermore, story-telling and autobiographic reconstruction are supported within this framework by appropriate historical expansions (which are themselves amenable to algebraic manipulation). Upcoming implementations of this framework in simulation and robotic agents have to demonstrate the validity of this approach. The basic Artiicial Life (AL) endeavour concerns the study of life-as-it-could-be in physical and other media rather than life-as-it-is just on the single example of planet Earth. AL's methods typically involve the study of emergence of complexity, deening primitive units, e.g. cellular automata, and deening (often non-linear) local interactions between these units 10]. Putting the system in the appropriate context, e.g. executing these rules, can lead to a phenotype with properties (behaviour or appearance) which might not be predictable in advance from knowing the set of transformation rules and the initial state. Thus, emergence of something \new" which requires a new level of description depends crucially on the choice and the representation of the units and the transformations. So, what are the appropriate units for designing or evolving physical artifacts, robots, supposed to behave in a lifelike way, animal-like, or humanoid, i.e., a bit like us? Each natural living system can only be understood and its phenotype can only be interpreted by taking into account its historical context, its evolutionary context, phylogeny, and individual developmental context, ontogeny. Each creature has its own history. Such an`history' encompasses its morphology, behaviour, and cognition over time. In this paper we will present a particular approach and philosophy of thinking about historical creatures, story-telling, and socially embedded creatures. 2 Lifelike Stories The past decade has shown a number of successful a-historical artiicial life implementations, Brooks's in-sectoid walking …