The concept of an agent has become important in both Artificial Intelligence (AI) and mainstream computer science. Our aim in this paper is to point the reader at what we perceive to be the most important theoretical and practical issues associated with the design and construction of intelligent agents. For convenience, we divide these issues into three areas (though as the reader will see, the divisions are at times somewhat arbitrary). Agent theory is concerned with the question of what an agent is, and the use of mathematical formalisms for representing and reasoning about the properties of agents. Agent architectures can be thought of as software engineering models of agents; researchers in this area are primarily concerned with the problem of designing software or hardware systems that will satisfy the prop-erties specified by agent theorists. Finally, agent languages are software systems for programming and experimenting with agents; these languages may embody principles proposed by theorists. The paper is not intended to serve as a tutorial introduction to all the issues mentioned; we hope instead simply to identify the most important issues, and point to work that elaborates on them. The article includes a short review of current and potential applications of agent technology.

We report on an experiment to implement an autonomous creature situated in a two-dimensional world, that shows various learning and problem-solving capabilities, within the Society of Mind framework. This goal is approached from a developmental perspective, where phases in the experiment correspond broadly to cognitive stages in the development of an infant. This paper describes the first stage, the creature being a newborn whose behavior is strongly driven by motivational states---impulses to action based on bodily needs---and basic emotions---peripheral and cognitive responsestriggered by the recognition of a significant event. Physiological parameters are used to model both concepts, which are seen by analogy with control systems. Motivations drive behavior selection and organization based on the notions of arousal and satiation, and the exploitation principle. Emotions exert further control by sending "hormones" that may affect the intensity of the selected behavior, enable it, or p...

... This paper makes two claims: first, psychologism is true, and thus a natural behaviorist analysis of intelligence that is incompatible with psychologism is false. Second, the standard arguments against behaviorism are inadequate to defeat this natural behaviorist analysis of intelligence or to establish psychologism. While psychologism is of course anathema to behaviorists, it also seems wrong-headed to many philosophers who would not classify themselves as behaviorists. For example, Michael Dummett says: If a Martian could learn to speak a human language, or a robot be devised to behave in just the ways that are essential to a language speaker, an implicit knowledge of the correct theory of meaning for the language could be attributed to the Martian or the robot with as much right as to a human speaker, even though their internal mechanisms were entirely different. (Dummett, 1976) Dummett's view seems to be that what is relevant to the possession of a certain mental state is a matter of actual and potential behavior, and that internal processing is not relevant except to the extent that internal processing affects actual and potential behavior. I think that this Dummettian claim contains an important grain of truth, a grain that many philosophers wrongly take to be incompatible with psychologism. This grain of truth can be elucidated as follows. Suppose we meet Martians, and find them to be behaviorally indistinguishable from humans. We learn their languages and they learn ours, and we develop deep commercial and cultural relations with them. We contribute to their journals and enjoy their movies, and vice versa. Then Martian and human psychologists compare notes, only to find that in underlying psychological mechanisms the Martians are very different from us. The M...

this paper we investigate the claim that complex causal interactions cause trouble for the notion of inner representational vehicles. We review some of the cases supposed to put pressure on a representational-vehicle based understanding and conclude that the threat, even in these ongoing, interactive cases, is more apparent than real. The main contribution of the present paper, however, is to go beyond this negative thesis

Can theories of mental imagery, conscious mental contents, developed within cognitive science throw light on the obscure (but culturally very significant) concept of imagination? Three extant views of mental imagery are considered: quasi-pictorial, description, and perceptual activity theories. The first two face serious theoretical and empirical difficulties. The third is (for historically contingent reasons) little known, theoretically underdeveloped, and empirically untried, but has real explanatory potential. It rejects the “traditional ” symbolic computational view of mental contents, but is compatible with recent situated cognition and active vision approaches in robotics. This theory is developed and elucidated. Three related key aspects of imagination (non-discursiveness, creativity, and seeing as) raise difficulties for the other theories. Perceptual activity theory presents imagery as non-discursive and relates it closely to seeing as. It is thus well placed to be the basis for a general theory of imagination and its role in creative thought.

this paper I'm indebted to three anonymous referees (whose insights were especially helpful), John Searle (whose seminal discussion of zombies in his The Rediscovery of the Mind provides the first round of my ammunition), Daniel Dennett, Stevan Harnad, David Rosenthal, Robert Van Gulick (who offered particularly insightful comments on the remote ancestor presented at the 1994 Eastern APA Meeting), Peter Smith, Kieron O'Hara, Michael Zenzen, Jim Fahey, Marvin Minsky, Larry Hauser and Pat Hayes. David Chalmers provided helpful analysis of a previous draft, and I profited from reading his The Conscious Mind, wherein zombies are taken to be logically possible. Conversations with Ned Block and Bill Rapaport also proved to be valuable.

Abstract not found

This paper will appear in The Turing Test and Turing Machines, Clark and Millican (Eds.), Oxford University Press, and was originally presented at the Turing90 Colloquium, University of Sussex.) 1 The intentional stance and intentional systems

Many elements of culture made their first appearance in the Upper Paleolithic. Previous hypotheses put forth to explain this unprecedented burst of creativity are found wanting. Examination of the psychological basis of creativity leads to the suggestion that it resulted from the onset of contextual focus: the capacity to focus or defocus attention in response to the situation, thereby shifting between analytic and associative modes of thought. New ideas germinate in a defocused state in which one is receptive to the possible relevance of many dimensions of a situation. They are refined in a focused state, conducive to filtering out irrelevant dimensions and condensing relevant ones. Introduction: A Cultural Revolution

Cognitive science is, more than anything else, a pursuit of cognitive mechanisms. To make headway towards a mechanistic account of any particular cognitive phenomenon, a researcher must choose among the many architectures available to guide and constrain the account. It is thus fitting that this volume on contemporary debates in cognitive science includes two issues of architecture, each articulated in the 1980s but still unresolved: • Just how modular is the mind? (section 1) – a debate initially pitting encapsulated mechanisms (Fodorian modules that feed their ultimate outputs to a nonmodular central cognition) against highly interactive ones (e.g., connectionist networks that continuously feed streams of output to one another). • Does the mind process language-like representations according to formal rules? (this section) – a debate initially pitting symbolic architectures (such as Chomsky’s generative grammar or Fodor’s language of thought) against less language-like architectures (such as connectionist or dynamical ones). Our project here is to consider the second issue within the broader context of where cognitive science has been and where it is headed. The notion that cognition in general—not just language