This book feels antiquated but it is interesting to read what a top scientist
thought of the emerging field of automatic computation in 1950
(when "programming" was called "taping"!)
Wiener begins by paying tribute to Gibbs (and only marginally to Boltzmann) for turning Physics into a stochastic science. Gibbs recognized that any physical description is limited by a degree of "incomplete determinism" and therefore we de facto live in a probabilistic world. Wiener even makes a connection to St Augustine, who defined "evil" as "incomplete love" (in order to resolve the contradiction that an infinitely good god would have created evil). Gibbs naturally understood entropy in terms of probability, and the probability of chaos tends to increase over the probability of order. Life, however, is a process that defies that tendency (as Schroedinger had already discussed in "What is Life") and creates pockets of the universe in which the second law of thermodynamics is violated, pockets of increased organization.
Wiener points out that the theory of messages is probabilistic in the same way that Willard Gibbs' Physics is. Messages control machines and societies, and therefore the theory of control is a theory of communication. Communication and control are one and the same: a message is a form of pattern and organization. The information carried by a message is a measure of organization. Information is therefore the opposite of entropy (the negative logarithm of its probability) Confusingly, another founder of the theory of information, Claude Shannon, at the same time, defined information as proportional to entropy. Both machines and living beings control entropy through feedback, feedback that is processed by a regulatory mechanism. This way a machine or a living organism counters nature's tendency toward disorder.
There is a chapter in which Wiener speculates that computers would be a bit cheaper if they were built in series of "tens of twenties"...
Wiener, writing in 1950, was worried that this new machines, capable of automating jobs, would cause massive unemployment. It turned out they would create many more jobs than they destroyed.
TM, ®, Copyright © 2014 Piero Scaruffi