ABSTRACT
The Hopfield model is in fact an academic exercise—instructive from a theoretical point of view, but not very useful for practical applications. As a model of the brain, it is based on unrealistic simplifications, such as:
full and symmetric connectivity, based on the Hebb rule; and
the choice of uncorrelated, random references.
As a technological recognition and memory device, it is limited by:
a capacity in the number of references which is a small fraction of the number of automata (0.14N if we tolerate a small error rate of the order of 1%); and
the dramatic deterioration of performance as soon as we attempt to exceed the maximum capacity.
