We highlight the limitations of formal methods by exhibiting two results in recursive function theory: that there is no effective means of finding a program that satisfies a given formal specification; or checking that a program meets a specification. We exhibit a simple MAS which has all the power of a Turing machine. We argue that any pure design methodology will face insurmountable difficulties in today's open and complex MAS. We recommend instead a methodology based on experimental method -- scientific foundations for MAS construction and control.

The overall behavior and nature of complex natural systems is in large part determined by the number and variety of the mechanisms involved – and the complexity of their interactions. Embodied natural communication belongs to this class of systems, encompassing many cognitive mechanisms that interact in highly complex ways, both within and between communicating individuals, constituting a heterogeneous, large, densely-coupled system (HeLD). HeLDs call for finer model granularity than other types of systems, lest we risk them to be not only incomplete but likely incorrect. Consequently, models of communication must encompass a large subset of the functions and couplings that make up the real system, calling for a powerful methodology for integrating information from multiple fields and for producing runnable models. In this paper I propose such an approach, abstract module hierarchies, that leverages the benefits of modular construction without forcing modularity on the phenomena being modeled.

Abstract — This paper presents an imitation learning system capable of learning tasks in a complex dynamic real-time environment. In this paper we argue that social learning should be thought of as a special case of general skill learning, and that the biases it presents to the skill learning problem radically simplify learning for species with sufficient innate predisposition to harness this power. We decompose skill learning into four sub-problems, then show how a modification of Roy’s CELL system [1] can address all these problems simultaneously. Our system is demonstrated working in the domain of a real-time VR game, Unreal Tournament. Index Terms — Imitation, Social Learning, Memetics, Language.

Robot-Cub undertakes to (1) develop an open platform for developmental cognitive robotics and (2), using that platform, to perform experimental work in line with theories of early childhood development. The platform (in the shape of a 2-year old child) will be available to the research community at large. Work in the first year proceeded essentially as foreseen in the 18 months implementation plan. At this stage the reviewers did not notice any major flaws or shortcomings that might jeopardise the success of the project. By and large they appear to be supportive of both the project objectives and the approach taken to achieve these objectives. They were critical though of a number of details and pointed out potential pitfalls. They agreed on a set of recommendations that should be taken into account in revising the implementation plan for phase two (months 13-30). Resources match results.

We highlight the limitations of formal methods by exhibiting two results in recursive function theory: that there is no effective means of finding a program that satisfies a given formal specification; or checking that a program meets a specification. We exhibit a simple MAS which has all the power of a Turing machine. We argue that any pure design methodology will face insurmountable difficulties in today's open and complex MAS. We recommend instead a methodology based on experimental method -- scientific foundations for MAS construction and control.

Abstract. If we aren’t sure what consciousness is, how can we be sure we haven’t already built it? In this article I speak from the perspective of someone who routinely builds small-scale machine intelligence. I begin by discussing the difficulty in finding the functional utility for a convincing analog of consciousness when considering the capabilities of modern computational systems. I then move to considering several animal models for consciousness, or at least for behaviours humans report as conscious. I use these to propose a clean and simple definition of consciousness, and use this to suggest which existing artificial intelligent systems we might call conscious. I then contrast my theory with related literature before concluding. 1