Constraint and pathology in spontaneously activated systems

Neural network modeling is approaching a strange attractor. Mathematical demonstrations of the theoretical behavior of non-Lipschitzian dynamical systems claim to show how such self-developing computational networks may be fully deterministic as to their microstructure yet without any prescription of their final state in advance. Justification for these simulations stems from applications which demand more biological realism in the design of neurobiotics. Realization of such modeling in actual hardware does not exist. While the best of these models are surely heuristic for neuroscientists who want to know how brains work, their attempted realization in control systems for planetary high technology seems confounded by several intractable problems, which, a few years ago, would have interested only neurophilosophers in retirement.