We do not see that any good can come of experiments which pay no attention to limiting factors that will assert themselves as soon as the small model… - Marvin Minsky

Most theories of learning have been based on ideas of "reinforcement" of success. But all these theories postulate a single, centralized reward mechanism. I doubt this could suffice for human learning because the recognition of which events should be considered memorable cannot be a single, uniform process. It requires too much "intelligence." Instead I think such recognitions must be made, for each division of the mind, by some other agency that has engaged the present one for a purpose.

This book... too, is a society — of many small ideas. Each by itself is only common sense, yet when we join enough of them we explain the strangest mysteries of mind.

More concretely, we would call the student's atten­tion to the following considerations: 1. Multilayer machines with loops clearly open all the questions of the general theory of automata. 2. A system with no loops but with an order restriction at each layer can compute only predicates of finite order. 3. On the other hand, if there is no restriction except for the absence of loops, the monster of vacuous generality once more raises its head. The perceptron has shown itself worthy of study despite (and even because of!) its severe limitations. It has many features to attract attention: its linearity; its intriguing learning theorem; its clear paradigmatic simplicity as a kind of parallel computation. There is no reason to suppose that any of these virtues carry over to the many-layered version. Nevertheless, we consider it to be an important research problem to elucidate (or reject) our intuitive judgment that the extension is sterile. Per­haps some powerful convergence theorem will be discovered, or some profound reason for the failure to produce an interesting “learning theorem” for the multilayered machine will be found.