Narrow Results

We are fascinated by the idea of giving life to the inanimate. The fields of Artificial Life and Artificial Intelligence (AI) attempt to use a scientific approach to pursue this desire. The first steps on this approach hark back to Turing and his suggestion of an imitation game as an alternative answer to the question "can machines think?".¹ To test his hypothesis, Turing formulated the Turing test¹ to detect human behavior in computers. But how do humans pass such a test? What would you say if you would learn that they do not pass it well? What would it mean for our understanding of human behavior? What would it mean for our design of tests of the success of artificial life? We report below an experiment in which men consistently failed the Turing test.

This paper presents a new approach to solving N-queen problems, which involves a model of distributed autonomous agents with artificial life (ALIFE) and a method of representing N-queen constraints in an agent environment. The distributed agents locally interact with their living environment, i.e. a chessboard, and execute their reactive behaviors by applying their behavioral rules for randomized motion, least-conflict position searching, and cooperating with other agents, etc. The agent-based N-queen problem solving system evolves through selection and contest, in which some agents will die or be eaten if their moving strategies are less effective than others. The experimental results have shown that this system is capable of solving large-scale N-queen problems. This paper also provides a model of ALIFE agents for solving general CSPs.

This paper presents a novel approach to image texture classification, which involves a model of artificial organisms i.e. Artificial Crawlers (ACrawlers) and a series of evolution curves representing the features of the texture. The distributed ACrawlers locally interact with their living environment, i.e. textured regions, and each ACrawler acts according to a set of homogenous rules for isotropic motion, energy absorption and colony formation, etc. The ACrawlers evolve through natural selection, which produces the specific curves of agent evolution, habitant settlement and colony formation as well as the scale distribution of all colonies. The feasibility and effectiveness of the proposed method have been demonstrated by experiments.

Deriving from the artificial life theory, this paper proposes an artificial co-evolving tribes model and applies it to solve the image segmentation problem. During the evolution process, the individuals in this model making up the tribes effect communication cooperatively from one agent to the other in order to increase the homogeneity of the ensemble of the image regions they represent. Two remarkable properties, that is, the monotone contraction and the conservation of the system are proved. Stability and scale control of the proposed method are carefully analyzed. Experimental results are presented and compared with two latest segmentation methods, both quantitatively and visually. We also discuss the results matching with human visual perception.

In this paper we analyze the replication properties of additive or parity Cellular Automata (CA). The objective of the paper is twofold. Firstly, to review and extend the existing results in the cases of one- and two-dimensional CA. Secondly, to report a general result that states the replication properties of an n-dimensional CA, whose evolution law depends on an arbitrarym number of neighbor cells.

The Bee System (an artificial bee swarm) is introduced in this paper. The proposed approach is applied to the Traveling Salesman Problem. The obtained results are very promising. The potential applications of the developed Bee System in the field of transportation engineering are discussed. The Bee System represents the new, successful application of emergent techniques based on natural metaphors, such as simulated annealing, genetic algorithms, and neural networks, to the complex engineering and management problems.

This paper explores the application of evolutionary techniques to the design of novel sounds and their characteristics during performance. It is based on the "selective breeding" paradigm and as such dispensing with the need for detailed knowledge of the Sound Synthesis Techniques involved, in order to design sounds that are novel and of musical interest. This approach has been used successfully on several SSTs therefore validating it as an Adaptive Sound Meta-synthesis Technique. Additionally, mappings between the control and the parametric space are evolved as part of the sound setup. These mappings are used during performance.

Foraging behavior of bees in finding food resource is one of the useful patterns to develop algorithms for solving complex problems. This article by simulation of such behavior and consider a memory for them proposed a method in discrete spaces. The proposed method is applied to Travel Salesman Problem (TSP) and successfully solved it. Simulation results have been proved the performance of our algorithm compared to similar strategies.

The embryologists found the realistic human organ models and animations of development necessary for their studies. The main idea of this paper is a methodology producing a realistic animation of development by combining the L-system growth model with a physical model. The skeleton of a digestive system is a line skeleton with a tree structure. Therefore, its growth in length can be simulated by an algebraic L-system which controls the growth of skeleton segments. The global deformations of the skeleton due to the gravity and the lack of space in abdominal cavity are simulated by a dynamics of skeleton segments. The movements that have no physical reasons such as looping are implemented by external forces applied on links controlling the organ movement in space. The convolution surfaces generated by skeletons define the final shape for growth animation. The entire system consists of two steps: First, the actual number of skeleton segments and the length of each skeleton segment is calculated from growth functions, second, the skeleton deformation in space is updated based on dynamics.

Despite spectacular progress in biophysics, molecular biology and biochemistry our ability to predict the dynamic behavior of multicellular systems under different conditions is very limited. An important reason for this is that still not enough is known about how cells change their physical and biological properties by genetic or metabolic regulation, and which of these changes affect the cell behavior. For this reason, it is difficult to predict the system behavior of multicellular systems in case the cell behavior changes, for example, as a consequence of regulation or differentiation. The rules that underlie the regulation processes have been determined on the time scale of evolution, by selection on the phenotypic level of cells or cell populations. We illustrate by detailed computer simulations in a multi-scale approach how cell behavior controlled by regulatory networks may emerge as a consequence of an evolutionary process, if either the cells, or populations of cells are subject to selection on particular features. We consider two examples, migration strategies of single cells searching a signal source, or aggregation of two or more cells within minimal multiscale models of biological evolution. Both can be found for example in the life cycle of the slime mold Dictyostelium discoideum. However, phenotypic changes that can lead to completely different modes of migration have also been observed in cells of multi-cellular organisms, for example, as a consequence of a specialization in stem cells or the de-differentiation in tumor cells. The regulatory networks are represented by Boolean networks and encoded by binary strings. The latter may be considered as encoding the genetic information (the genotype) and are subject to mutations and crossovers. The cell behavior reflects the phenotype. We find that cells adopt naturally observed migration strategies, controlled by networks that show robustness and redundancy. The model simplicity allow us to unambiguously analyze the regulatory networks and the resulting phenotypes by different measures and by knockouts of regulatory elements. We illustrate that in order to maintain a cells' phenotype in case of a knockout, the cell may have to be able to deal with contradictory information. In summary, both the cell phenotype as well as the emerged regulatory network behave as their biological counterparts observed in nature.

We recently formulated an approach to representing structures in cellular automata (CA) spaces, and the rules that govern cell state changes, that is amenable to manipulation by genetic programming (GP). Using this approach, it is possible to efficiently generate self-replicating configurations for fairly arbitrary initial structures. Here, we investigate the properties of self-replicating systems produced using GP in this fashion as the initial configuration's size, shape, symmetry, allowable states, and other factors are systematically varied. We find that the number of GP generations, computation time, and number of resulting rules required by an arbitrary structure to self-replicate are positively and jointly correlated with the number of components, configuration shape, and allowable states in the initial configuration, but inversely correlated with the presence of repeated components, repeated sub-structures, and/or symmetric sub-structures. We conclude that GP can be used as a "replicator factory" to produce a wide range of self-replicating CA configurations, and that the properties of the resulting replicators can be predicted in part a priori. The rules controlling self-replication that are created by GP generally differ from those created manually in past CA studies.

The understanding of the evolutionary transitions is a major area of research in artificial life and in biology. We follow an artificial life approach to investigate these phenomena, using a system inspired by Anabaena cyanobacteria (which exhibit rudimentary multicellular differentiation and cooperation) in order to look for evidence of emerging differentiation and multicellular cooperation in colonies of individual cells.

We first evolve single free-living cells with the help of a Genetic Algorithm (GA). These cells are controlled with genetic regulatory networks. The single cells are evolved to each perform both of two tasks: an abstraction of house-keeping metabolism and a reproductive cycle. Once such a cell was evolved with the GA, the cell is used to seed the growth of a multicellular filamentous colony, whose constituent cells continue to reproduce and evolve. Two types of colonies generated from the seed cell are studied: one with intercellular communication ability and one without.

We introduce and apply new measures for assessing the impact of multicellular interaction on individual reproduction and on life span.

The conclusion of these studies shows that the colony with the ability to communicate shows, with the help of our new measures, behaviors that hint at the emergence of early cooperation.

After half a century of research and development, our requirements as far as robots are concerned now seem clear. People want robots to be mobile, intelligent, autonomous, able to cooperate with man and other machines, and to act within the world according to our wishes. Of course, that aim has still to be reached, and the means to reaching it cannot arise ex nihilo from the researcher's brain. However, one possible solution could be to take a look at the world around us and to ask ourselves the following question: are there, in the world, any entities that possess the characteristics we are looking for in robots? Indeed, there are billions of these entities. They are living beings in general, and animals and man in particular. Although bio-inspiration for robot design is not a new issue, it would appear that very few works have addressed the basic issues that might enable us to exploit this source of inspiration to the full. This paper aims to set out some basic premises which could then help designers to find adequate solutions. We should point out that, although the ideas presented in this paper come from observation and knowledge about living beings, the aim is neither to make living robots, nor to make an artificial real animal, and even less to make an artificial human. The goal consists in identifying the most important features of living beings that could be used to meet our requirements in the matter of robot design. It is obvious, moreover, that these needs or demands are not static, but may change according to the pace of new technological and scientific achievements. On the other hand, needs as far as robots are concerned may be formulated in terms of human (or animal) appearance, of mental and physical concerns or behavior, and not at all in terms of human (or animal) reality. Finally, the basic issue comes down to this: what "parts" of humans can be transferred to machines, and do these "parts" correspond to what is required nowadays of robots?

Perhaps the best known example of user-guided evolution is furnished by evolving expressions, an image generation technique first introduced by Sims. In this version of artificial evolution, images are evolved for aesthetic purposes, hence any fitness measure used must be based on aesthetics. We consider the problem of guiding image evolution autonomously on the basis of computational, as opposed to user-assigned, aesthetic fitness. Due to the difficulty of formulating an absolute criterion for aesthetic fitness, we adopt a coevolutionary approach, relying on hosts and parasites to establish relative criteria for aesthetic fitness. To sustain the coevolutionary arms race, we allow coevolution to proceed in stages. This permits appropriate fitness levels to be maintained within the parasite populations we use to infect host image populations. Using staged coevolution produces two beneficial results: (1) longer survival times for subpopulations of host images, and (2) stable phenotypic lineages for host images.

Machine (artificial) consciousness can be interpreted in both strong and weak forms, as an instantiation or as a simulation. Here, I argue in favor of weak artificial consciousness, proposing that synthetic models of neural mechanisms potentially underlying consciousness can shed new light on how these mechanisms give rise to the phenomena they do. The approach I advocate involves using synthetic models to develop "explanatory correlates" that can causally account for deep, structural properties of conscious experience. In contrast, the project of strong artificial consciousness — while not impossible in principle — has yet to be credibly illustrated, and is in any case less likely to deliver advances in our understanding of the biological basis of consciousness. This is because of the inherent circularity involved in using models both as instantiations and as cognitive prostheses for exposing general principles, and because treating models as instantiations can indefinitely postpone comparisons with empirical data.

A simplified model of Darwinian evolution at the molecular level is studied by applying the methods of artificial chemistry. A chemical reactor (chemostat) is composed of molecules that are represented by strings of tokens and these strings are autoreplicated with a probability proportional to their fitness. Moreover, the process of autoreplication is not fully correct, there may appear sporadic mutations that produce new offspring strings that are slightly different from their parental templates. The dynamics of such an autoreplicating system is described by Eigen's differential equations. These equations have a unique asymptotically stable state, which corresponds to those strings that have the highest rate constants (fitness). A generalized version of a rugged fitness landscape realized by a Kauffman KN function is used for an evaluation of strings. Recently, Newman and Engelhardt have demonstrated that this simple type of fitness surface simulates in fact almost all basic results about molecular Darwinian evolution achieved by Schuster with his associates. Schuster et al. used a physical model of RNA molecules with fitness specified by their ability to be folded into a secondary structure. The presented model with Kauffman rugged function induces a detailed look at mechanisms of molecular Darwinian evolution, in particular to ameaning and importance of neutral mutations.

Tierra is a digital simulation of evolution for which the stated goal was the development of open-ended complexity and a digital “Cambrian Explosion.” However, Tierra failed to produce such a result. A closer inspection “ Tierran evolution's adaptations show very few instances of adaptation through the production of new information. Instead, most changes result from removing or rearranging the existing pieces within a Tierra program. The open-ended development of complexity depends on the ability to generate new information, but this is precisely what Tierra struggles to do. The character of Tierran adaptation does not allow for open-ended complexity but is similar to the character of adaptations found in the biological world.