The Potential Variability of Neural Networks in the Context of L. V. Krushinsky's Ideas

We discuss the concepts contained in the article by A. L. Krushinsky. We specifically focus on the idea that each informational act within a period of nervous activity leads to an increase in the entropy of the neural network, which requires an initial excessive orderliness. We present arguments in favor of the hypothesis proposed by L. V. Krushinsky, that initial redundancy is likely implemented in the brain as an originally large number of nerve cells, as yet undifferentiated by their connections. However, it is suggested that this redundancy has its normative values, the increase of which may lead to patho- logical phenomena.