Chroniques d'un scybernéthicien [sotoBlogue]

"What is the fundamental organization of autonomous systems ?
Can we model it ?"

"As early as 1974 Maturana, Varela and Uribe's published a paper that included the first simulation model of an autopoietic system. Latter development of the concept of autonomy (Varela 1979) and its convergence with related concepts such as Rosen's closure to efficient causation, Kauffman's autocatalytic networks and Ganti's chemoton model (to mention but a few) has put autonomy at the center of research interest those researchers interested on the origins of biological and cognitive organization."

MODELLING AUTONOMY WORKSHOP

“Enaction is a term coined in psychology (J. Bruner) and used in a particular biological approach (F. Varela), according to which cognition is fundamentally a feature of living organisms in a dynamic adaptive relationship with their environment. Only recently the term has gained widespread currency in domains such as human-computer interaction. In the same kind of view, the physicist Jean Petitot proposes the idea of a "Phenophysics", based on the theory of "Morphogenesis" of the mathematician R. Thom, according to which the necessary condition for categorization to occur, is the presence of specific singularities in the dynamics of physical sensorial events, on which categorization can get a grip. These theories are used in signal processing, shape recognition and can be used in extraction of emotive patterns in signals. Other related concepts are presently developed also in relationship with artificial intelligence and robotics. The proper domain is that of the so-called embodied cognition, which gives much importance to action and perception in the definition and simulation of intelligent behaviors, by focusing the attention on parallel, distributed architectures, on adaptive behavior of different kinds (not only high order symbolic capacities) and on the possibilities offered by the dynamic systems modeling of behavior.”

http://acroe.imag.fr/enactive07/enaction.php

N'y a-t-il pas un "malentendu"... Cette approche est, à mon sens, trop tendue par l'opérationnel externaliste. Je pense qu'il faudrait beaucoup plus mettre l'accent sur l'humain et l'apprentissage que sur la machine. La pensée énactive est une compréhention alternative à l'exclusivité représentationnelle mais aussi, c'est corollaire, à l'adaptationisme. La cognition incorporée, ce n'est pas que pour les "autres".

Un site d'amateurs (avec vidéos) dédié à des rencontres sur le thème "Arts, Sciences et Techniques" : le Café des Arts des Sciences et des Techniques. A voir, les arabesques présentées par Jean-Marc Castera (rare).