with the real world – design principles for

It is appealing to consider hybrids of neural-network learning algorithms with evolutionary search procedures, simply because Nature has so successfully done so. In fact, computational models of learning and evolution offer theoretical biology new tools for addressing questions about Nature that have dogged that field since Darwin [Belew, 1990]. The concern of this paper, however, is strictly artificial: Can hybrids of connectionist learning algorithms and genetic algorithms produce more efficient and effective algorithms than either technique applied in isolation? The paper begins with a survey of recent work (by us and others) that combines Holland's Genetic Algorithm (GA) with connectionist techniques and delineates some of the basic design problems these hybrids share. This analysis suggests the dangers of overly literal representations of the network on the genome (e.g., encoding each weight explicitly). A preliminary set of experiments that use the GA to find unusual but successf...

The paper addresses the question how a group of physically embodied robotic agents may originate meaning and language through adaptive language games. The main principles underlying the approach are sketched as well as the steps needed to implement these principles on physical agents. Some experimental results based on this implementation are presented. 1 Introduction In the past five years, a large number of robotic agents, i.e. physical systems capable of sensori-motor control, have been built in order to investigate a bottom-up approach to artificial intelligence (see the overview in [8]). Important results have been achieved, particularly by using behavior-oriented architectures [14] and learning methods based on neural networks [6] or genetic algorithms [3]. Nevertheless, it is still largely an open question how these robots may reach sufficient complexity in order to qualify as cognitive agents. Most of the experiments have focused on `low level' tasks like obstacle avoidance o...

We describe a project to capitalize on newly available levels of computational resources in order to understand human cognition. We are building an integrated physical system including vision, sound input and output, and dextrous manipulation, all controlled by a continuously operating large scale parallel MIMD computer. The resulting system will learn to "think " by building on its bodily experiences to accomplish progressively more abstract tasks. Past experience suggests that in attempting to build such an integrated system we will have to fundamentally change the way artificial intelligence, cognitive science, linguistics, and philosophy think about the organization of intelligence. We expect to be able to better reconcile the theories that will be developed with current work in neuroscience.

Research on potential interactions between connectionist learning systems, i.e., artificial neural networks (ANNs), and evolutionary search procedures, like genetic algorithms (GAs), has attracted a lot of attention recently. Evolutionary ANNs (EANNs) can be considered as the combination of ANNs and evolutionary search procedures. This paper first distinguishes among three kinds of evolution in EANNs, i.e., the evolution of connection weights, of architectures and of learning rules. Then it reviews each kind of evolution in detail and analyses critical issues related to different evolutions. The review shows that although a lot of work has been done on the evolution of connection weights and of architectures, few attempts have been made to understand the evolution of learning rules. Interactions among different evolutions are seldom mentioned in current research. However, the evolution of learning rules and its interactions with other kinds of evolution play a vital role in EANNs. As t...

temporal relations requires the joint evaluation of responses from more than one neuron, only experiments that permit simultaneous measurements of responses 60528 Frankfurt from multiple units are considered. These include multi-Federal Republic of Germany electrode recordings from multiple individual cells, but also measurements of local field potentials (LFPs) and electroencephalographic (EEG) or magnetoencephalo-Most of our knowledge about the functional organization of neuronal systems is based on the analysis of the firing patterns of individual neurons that have been recorded one by one in succession. This approach permits as-sessment of event-related variations in discharge rate, but it precludes detection of any covariations in the amplitude or timing of distributed responses if these graphic (MEG) recordings. The signals of these latter

Quartz, S. & Sejnowski, T.J. (1997). The neural basis of cognitive development: A constructivist manifesto.

The paper proposes a set of principles and a general architecture that may explain how language and meaning may originate and complexify in a group of physically grounded distributed agents. An experimental setup is introduced for concretising and validating specific mechanisms based on these principles. The setup consists of two robotic heads that watch a scene in which a robot moves around in its ecosystem. The first results from experiments showing the emergence of distinctions, of a lexicon, and of primitive syntactic structures are reported. 1 Introduction Artificial Intelligence research has made remarkable progress the last decades by showing how operations over symbolic models may explain various aspects of intelligent behavior, such as planning, problem solving, natural language processing, etc. However, the problem of the origin of these symbolic models has so far not been adequately addressed. Most of the time it is the programmer who designs formalisms and datastructures, ...

Behavior-oriented AI is a scientific discipline that studies how behavior of agents emerges and becomes intelligent and adaptive. Success of the field is defined in terms of success in building physical agents that are capable of maximising their own self-preservation in interaction with a dynamically changing environment. The paper addresses this artificial life route towards artificial intelligence and reviews some of the results obtained so far. 1 Official reference: Steels, L. (1994) The artificial life roots of artificial intelligence. Artificial Life Journal, Vol 1,1. MIT Press, Cambridge. 1 Introduction For several decades, the field of Artificial Intelligence has been pursuing the study of intelligent behavior using the methodology of the artificial [104]. But the focus of this field, and hence the successes, have mostly been on higher order cognitive activities such as expert problem solving. The inspiration for AI theories has mostly come from logic and the cognitive...

The Nature-Nurture controversy has been with us since it was first outlined by Plato and Aristotle. Nobody likes it anymore. All reasonable scholars today agree that genes and environment interact to determine complex cognitive outcomes. So why does the controversy persist? First, it persists because it has practical implications that cannot be postponed (i.e., what can we do to avoid bad outcomes and insure better ones?), a state of emergency that sometimes tempts scholars to stake out claims they cannot defend. Second, the controversy persists because we lack a precise, testable theory of the process by which genes and environment interact. In the absence of a better theory, innateness is often confused with (1) domain specificity (Outcome X is so peculiar that it must be innate), (2) species specificity (we are the only species who do X, so X must lie in the human genome), (3) localization (Outcome X is mediated by a particular part of the brain, so X must be innate), and (4) learnability (we cannot figure out how X could be learned, so X must be innate). We believe that an explicit and plausible theory of interaction is now around the corner, and that many of the classic maneuvers to defend or attack innateness will soon disappear. In the interim, some serious errors can be avoided if we keep these confounded issues apart. That is the major goal of this paper, i.e., not to attack innateness but to clarify what