Cybernetics or Control and Communication in the Animal and the Machine

by Norbert Wiener

A classic and influential work that laid the theoretical foundations for information theory and a timely text for contemporary informations theorists and practitioners.

With the influential book Cybernetics, first published in 1948, Norbert Wiener laid the theoretical foundations for the multidisciplinary field of cybernetics, the study of controlling the flow of information in systems with feedback loops, be they biological, mechanical, cognitive, or social. At the core of Wiener's theory is the message (information), sent and responded to (feedback); the functionality of a machine, organism, or society depends on the quality of messages. Information corrupted by noise prevents homeostasis, or equilibrium. And yet Cybernetics is as philosophical as it is technical, with the first chapter devoted to Newtonian and Bergsonian time and the philosophical mixed with the technical throughout. This book brings the 1961 second edition back into print, with new forewords by Doug Hill and Sanjoy Mitter.

Contemporary readers of Cybernetics will marvel at Wiener's prescience—his warnings against “noise,” his disdain for “hucksters” and “gadget worshipers,” and his view of the mass media as the single greatest anti-homeostatic force in society. This edition of Cybernetics gives a new generation access to a classic text.

Genres: Science, Philosophy, Nonfiction, Technology, Mathematics, Psychology, Artificial Intelligence

352 pages, Kindle Edition

First published January 1, 1948

About the author

Norbert Wiener was an American mathematician and philosopher. He was Professor of Mathematics at MIT. Wiener is considered the father of cybernetics, a formalization of the notion of feedback, with implications for engineering, systems control, computer science, biology, philosophy, and the organization of society.

Reviews

[Ben · Rating 3 out of 5]

Most of my reaction to this book is incidental to its content: First, it's funny what once passed for pop science. The reviews on the back from the Saturday Evening Post and others hail its readability for the layman, yet at least 10% of the pages are devoted to difficult equations and proofs, and I had to skip a couple of chapters because the math was way, way over my head.

Second, much of the science Wiener pioneered has been obviated by the development of more powerful tools in neurology and computing, among others. One on hand, his ruminations on the potential of the machine to perform this or that task feel lost in a hastening genealogy of technological developments.

On the other hand, Wiener was both philosopher and scientist. As a scientist he was evidently peerless at the time; as a philosopher he reads as ... quirky. But at least he's trying. A dyed-in-the-wool materialist, his assertion that the body is a machine - a wonderfully complex machine, but a machine nevertheless - apparently had not been so internalized by his intended audience (again, a mathematically literate lay audience) that it was unnecessary to make the point. But this hardly exhausts his argument. Wiener was clearly an instrumentalist, and explicit in his theory of the human body is the idea that the workings of neurons, synapses, and brain waves could be translated to computing technologies for many applications, good or bad, which he discusses at length in other books like "The Human Use of Human Beings." It was prophetic of Wiener to predict that computing power would scarcely be limited, except by the efficiency of the vessel in which the computer operated, including considerations like energy conservation.

The last and most pressing reason that the book is interesting is because Wiener, probably the highest authority in the world on the science of intelligence at the time the book was written, was clearly committed to a program of ethical research and development. He warned of the danger of developing dangerous computing applications, and dismissed the idea that we can always "turn off" machines that we don't like, since it isn't always clear that the danger exists until after the damage is done. We have to be careful lest we find ourselves cobras fighting mongooses (mongeese?)

[Bernardo Moreira · Rating 3 out of 5]

Disclaimer: o livro é bom, muito interessante. Tô dando 3 estrelas por dois motivos: 1. muita matemática (principalmente na primeira parte do livro), o que eu reconheço como importante para o desenvolvimento do argumento mas nossa senhora sofri muito (obrigado JG, você me salvou 30 mil vezes e agora eu "entendo" movimento browniano e funções integrais); 2. essa minha distância fez com que eu aproveitasse menos a experiência da leitura (o que explica quase um mês para ler um livro de 240 páginas - apesar de estar lendo mil coisas simultaneamente, é um período longo).
O livro já começa (no prefácio) com uma introdução sobre sistemas não lineares e teoria da predição para estudar máquinas que operam com funções de movimento browniano (inputs randômicos) e os mecanismos de feedback dessas máquinas. Temos um momento mais calmo com a Introdução, que é um grande resumo das pesquisas do Wiener e de outros colegas sobre cibernética, o que ajuda a ilustrar os problemas que estavam em pauta na época (o sistema de mísseis antiaéreos por exemplo, que exige uma predição do movimento de um avião para atingí-lo com sucesso - tendo que considerar o caráter não-linear desse movimento e o aprendizado necessário da máquina para considerar as possibilidades de trajetória).
Tempo newtoniano e bergsoniano é interessante, expondo as questões da irreversibilidade do tempo da evolução e biologia em Bergson, do problema de estabelecer unidades e das soluções estatísticas para prever o imprevisível. Há uma reflexão filosófica interessante, e a mistura de filosofia e técnica é bem envolvente.
Mas aí então começa A Grande Pica. Grupos e Mecânica Estatística é quase frustrante de tão complicado. Sistemas binários, grupos de transformação, derivação e integral, teoria ergódica, invariantes linares de grupo, parâmetros de distribuição estatística, termodinâmica, teoria do caos e movimento browniano, demônio de Maxwell - tudo isso pensado pelo sistema binário para a teoria da predição. Soei bastante.
Mas Séries Temporais, Informação e Comunicação consegue ser ainda mais tenso, matematicamente. Apesar de muitas noções já estarem mais consolidadas ao decorrer do livro, as equações são bastante desafiadoras. À esse ponto, eu já recomendo que cê leia ou com ajuda de algum amigo de exatas, ou depois de ter estudado muita matemática ou simplesmente acreditando que o Wiener tá provando o que ele tá falando. Aqui discutiremos registro, filtragem e transmissão da informação: mensagem e ruído em um matematiquês sinistríssimo para pensar os temas do capítulo anterior nas séries temporais para a predição na teoria das mensagens dependentes linearmente do movimento browniano. Teoria multidimensional, mecânica quântica, teoria da entropia e equilíbrio estatístico se juntam a cena pra complicar tudo mais ainda.
Feedback e Oscilação é um pouco menos assustador. Um dos conceitos principais (senão o principal) da cibernética - o feedback - foi devidamente entendido em suas definições e implicações (funções temporárias, inputs e outputs, periodicidade, antecipação e compensação, estabilidade e informação, processos homeostáticos etc) o que complica um pouco são as suas aplicações técnicas.
Computadores e o Sistema Nervoso é bem mais tranquilo. Discussão interessantíssima sobre memória, cálculo, circuitos, repetição, limitações da lógica, registro etc. A questão do tônus afetivo já nos dá um prelúdio dos reflexos nos mecanismos de aprendizagem, mobilizando Locke e Pavlov para pensar a experiência e o condicionamento. Outro ponto interessante é sobre a economia energética das máquinas, que levanta questões calorosas sobre matéria e informação.
Gestalt e Universais continua a discussão pelos mecanismos de associação e seus feedbacks, perspectiva, mecanismo visual e imagem. Cibernética e Psicopatologia apresenta uma diferença curiosa entre perturbações funcionais e orgânicas, demonstrando a importância do feedback para o funcionamento dos mecanismos psicológicos mais básicos - grande complexidade do tráfego no sistema nervoso. Aponta uma relação interessante entre a psicanálise e a cibernética, sobre a questão do condicionamento e dos ciclos da memória.
Informação, Linguagem e Sociedade pensa a transmissão de informação comunal e apresenta uma crítica importante à suposta homeostase da livre competição no mercado: não há homeostase social. Aponta limitações da teoria dos jogos por seu caráter de ênfase na "escolha racional", ressaltando os meios de comunicação na sociedade como baseados no favorecimento de meios lucrativos, na limitação da posse dos meios em poucos homens abastados e no contexto de poder de tal Monopólio. Desta forma, o processo homeostático está nas mãos dos principais elementos antihomeostáticos da comunidade. Wiener afirma entretanto que não há lugar para um excessivo otimismo de pensar uma possível homeostase social, tendo em vista os abismos entre as diferentes ciências: a cibernética que considera partículas e tem "controle" sobre o experimento já trabalha com procedimentos de anulação da diferença, de acoplamentos frouxos para garantir o mínimo de exatidão em um fenômeno - algo inconcebível para a ciência social, que não pode possuir mecanismos estatísticos como os da física atômica. Wiener aponta então que a teoria dos jogos é muito mais adequada para algo como uma máquina enxadrista. E em Sobre Máquinas de Aprender e Autorreprodutoras, é exatamente isso que está em jogo: aprendizado ontogenético e filogenético, predição e memória para avaliar padrões de comportamento e hábitos, especulações sombrias sobre os perigos de atribuição de objetivos para mecanismos de aprendizagem ("aprenda a vencer uma guerra nuclear" - atribuição de valores de cada lance, elemento, vitória e derrota), transdução linear e não linear, deslocamentos de inputs e outputs.
Ondas cerebrais e Sistemas Auto-organizadoras é o último capítulo e temos a volta do matematiquês sinistro para discutir os temas das máquinas de aprendizagem e autoreprodução aplicadas às questões neurológicas de eletroencefalografia. Uso de computadores para análise harmônica de espectros cerebrais, feedbacks não-lineares, médias de curvas, densidade espectral, autocorrelação, método heteródino, a construção de modelos estatísticos para o espectro de ondas cerebrais com base na integração no espaço das funções, a questão da "porta lógica" e as atrações de frequência em sistemas auto-organizadores.
Admito que a experiência de leitura desse livro não é de "quero mais", mas de "quero menos". O livro é bom e interessante, mas o jargão técnico e matemático cansa um pouco. De forma alguma digo que é ruim por ser difícil, é um livro realmente bom. Mas não é um passeio no parque.

[Dan Rera · Rating 4 out of 5]

This was a very interesting book. It had some idiosyncrasies that prevented me from giving it a higher rating, but other than that, the subject matter and breath were fascinating enough for me to certainly recommend it to anyone with any interest in cybernetics as a broad concept.

Let me first say that there were a few chapters in the beginning and end that were needlessly technical and mathematical. It isn’t so much that I object to the existence of mathematical proofs in their proper context, but it seemed unnecessarily detailed for the overall purpose and thrust of the book. After working my way through a few of the demonstrations, I eventually gave up an took the author’s word on the soundness of his conclusions. I think this is fair enough to do, so much so, in fact, that I started to wonder why he was making me flip past pages upon pages of dense calculus only to arrive at a summary paragraph that would elucidate the meaning of his findings. At one point he attempts to justify this technique by saying something to the effect of: it would take me much longer to put these formulae into common English, so read them for their condensed shorthand value. I, as the reader, would have been perfectly willing to let him dispose of the rigor for the sake of cleaner text. But whatever; maybe people really wanted to delve into that level of demonstration.

In addition, for as direct and focused Wiener seems to be as a mathematician, his thoughts, and even his prose, seem disjointed and meandering. He quickly moves from one large concept to the next, sometimes leaving the reader reeling trying to catch up. He goes on digressions that seemed opinionated and lengthy, and, when returning to the original thread, makes no real indication that he has returned. Other times these digressions will be nearly freeform transitions between concepts, similarly lacking indication that he has departed one concept and started addressing another. That he does both of these leaves the reader wondering if the text is moving forward or looping back. I would like to think that the author was making some larger point about the nature of cybernetics though this ambiguity, but this seems doubtful to me.

These criticisms are small, however, taken in relation to the positive aspects of this book. The conclusions being reached by Wiener might seem banal to a contemporary reader, but this only lends credibility to their influence. Weiner not only anticipates a great deal of the future of computing, he also strongly develops a theory of the animal (and human) as, essentially, an organic computing machine; not just the brain, but the whole organism. Early in the work, he distinguishes this position from simplistic Cartesian materialism (i.e. with respect to Descartes’ conception of animals as sophisticated machines), and, instead, argues for a vitalism that explains the soul as a material concept. It is non-symmetrical feedback, as unfolding through time, that makes the system seem vital as opposed to mechanical, and it is this level of complexity that makes the behavior of animals and humans seem so radically different than the motions of planets or pendulums.

The remainder of the work goes on to apply the cybernetic concept of feedback to a whole range of biological phenomena and computational questions in a way that demonstrates the power of the theory and the broadness of its application. So much of what Wiener says is taken as understood in modern times that it is easy to lose sight of how striking his claims really are.

[X]

“Those of us who have contributed to the new science of cybernetics thus stand in a moral position which is, to say the least, not very comfortable. We have contributed to the initiation of a new science which, as I have said, embraces technical developments with great possibilities for good and for evil. We can only hand it over into the world that exists about us, and this is the world of Belsen and Hiroshima.

We do not even have the choice of suppressing these new technical developments. They belong to the age, and the most any of us can do by suppression is to put the development of the subject into the hands of the most irresponsible and most venal of our engineers. The best we can do is to see that a large public understands the trend and the bearing of the present work, and to confine our personal efforts to those fields, such as physiology and psychology, most remote from war and exploitation.

As we have seen, there are those who hope that the good of a better understanding of man and society which is offered by this new field of work may anticipate and outweigh the incidental contribution we are making to the concentration of power (which is always concentrated, by its very conditions of existence, in the hands of the most unscrupulous). I write in 1947, and I am compelled to say that it is a very slight hope.”

What a way to end an Introduction!


Cybernetics, or the Control and Communication in the Animal and the Machine, was published first in 1948 and then again in 1961 with two additional chapters - this book is both broader, and much narrower, than Wiener’s The Human Use of Human Beings: Cybernetics and Society, the more readable companion piece. If you’re going to read one, I definitely recommend you read that one, but I still found Cybernetics interesting as a piece of science history.

The Introduction largely focuses on Wiener and the men (yes, only men) he worked with. His most frequent collaborator, interestingly, was Dr. Arturo Rosenblueth, a Mexican physiologist who organized cross-disciplinary discussions which seem to have been the origin of much of “cybernetics,” at least as Wiener was interested in the subject. Some of The Human Use, and much of this book, focuses on current (bah dum ching) research about how electricity works in the body (brain waves, neurons, etc.), and on analogizing between how biological organisms work, and how machines can be made to work. (Wiener and Rosenblueth worked together regularly for years, to the extent that when Rosenblueth left Harvard Med to run a lab at the Instituto Nacional de Cadiología in Mexico City, they had to work out a joint research plan with their institutions so Wiener could regularly make longterm visits to the Instituto to allow them to continue their research.)

The Introduction is full of these details about not just *what* Wiener and other researchers did, but *how* they did it - it’s full of names with interesting backstories attached, and descriptions of travel. (In 1947, for example, Wiener went to Nancy, France for a conference on harmonic analysis. On the way back, he spent three weeks in England to meet with computer scientists there, “and above all to talk over the fundamental ideas of cybernetics with Mr. Turing at Teddington.”)

It also has an interesting (read: depressing) section on “computing machines” and their effect on the labor market. Wiener refers to computers as “mechanical slaves,” and says that with the computing “revolution,” “the average human being of mediocre attainment or less has nothing to sell that it is worth anyone’s money to buy.” He has an extremely clear solution: “The answer, of course, is to have a society based on human values other than buying or selling.” But will that happen? He says that “we need a good deal of planning and a good deal of struggle” to get there, but doesn’t seem optimistic. Apparently he reached out to people at the C.I.O. about this and was essentially told that he was right, and that they didn’t think at the union was prepared to handle the problem... and the conversations ended there!

The rest of the book is made up of fairly technical chapters on specific topics - below are the quotes I thought were particularly interesting:

*

“Even before the existence of any adequate dynamical theory, even as far back as the Babylonians, it was realized that eclipses occurred in regular predictable cycles, extending backward and forward over time. It was realized that time itself could be better measured by the motion of the stars in their courses than in any other way.”

Chapter I, “Newtonian and Bergsonian Time,” has some interesting material on how we measure time and how much easier it is to measure anything, in fact, in space, this big empty space with just a few things moving around, than on Earth, in our atmosphere.

*

“The very fact that we see a star means that its thermodynamics is like our own.” And “Within any world with which we can communicate, the direction of time is uniform.” And “The individual is an arrow pointed through time one way.”

*

“In general, there is no set of observations conceivable which can give us enough information about the past of a system to give us complete information as to its future.”

*

“the brain, in its course of nature, never even approximately clears out its past records. Thus the brain, under normal circumstances, is not the complete analogue of the computing machine but rather the analogue of a single run on such a machine.”

*

And of interest for people reading about the environmental impact of AI, cryptocurrency, etc.: “No other computing machine approaches the economy of energy of the brain.” And “living mechanisms tend to have a much smaller space scale than the mechanisms best suited to the techniques of human artificers, although, on the other hand, the use of electrical techniques gives the artificial mechanism an enormous advantage in speed over the living organism.”

*

In his chapter on “Cybernetics and Psychopathology,” he makes the dry observation that although lobotomies have recently been in “a certain vogue” for making care of patients easier afterwards, “Let me remark in passing that killing them makes their custodial care still easier.”

*

This quote just cracked me up since we’ve definitely all been there: “I do not care to pronounce an opinion on this matter; I leave it as an interesting speculation.”

*

And I found interesting his observation that “All the great successes in precise science have been made in fields where there is a certain high degree of isolation of the phenomenon from the observer.” It reminded me of Mill’s argument in The Subjection of Women that it is always easier (or maybe, only possible) for an outsider to overthrow unequal power dynamics. There’s always that push and pull between being able to look at something objectively, and knowing enough about what you’re looking at that your thoughts are of value.

*

One of the additional chapters, “On Learning and Self-Reproducing Machines,” drew on stories of magic as good analogies for dealing with computers - The Sorcerer’s Apprentice, the monkey’s paw, etc. - which was fun. “In all these stories the point is that the agencies of magic are literal-minded; and that if we ask for a boon from them, we must ask for what we really want and not for what we think we want. The new and real agencies of the learning machine are also literal-minded. If we program a machine for winning a war, we must think well what we mean by winning.”

[Ushan · Rating 3 out of 5]

This is a disorganized book that mixes super-brief mathematical treatment of statistical mechanics, ergodic theory, control theory and stochastic processes with an explanation of the difference between digital and analog computers, and speculations about learning machines and the human nervous system. It finds the same patterns of control and feedback occurring throughout the world, from the nervous system through society, and notes that while automatic control is not new (centrifugal governors on steam engines were invented by James Watt in the 18th century, and analyzed by James Clerk Maxwell in the 19th, as mentioned in this book), ever more powerful electronic digital computers will carry it to a new level. The book warns that humanity needs to be super-careful not to become a sorcerer's apprentice from the fairy tale, and have its power taken away by mindless computers. In the next 65 years, computers did become far more powerful, but they did not usurp power from humanity because they operate on data humans give to them, and if humans have an incentive to lie, garbage will both flow into the computers and come out of them. The Soviet economy was a classic example. If a factory manager was given a production plan, his compensation depended upon the fulfillment of the plan, but his suppliers let him down, he had an incentive to lie and to bribe any inspectors, and so on down the supply chain through the entire economy. As a child I heard adults say that the Soviet State Planning Committee had an American VAX computer acquired in circumvention of CoCom restrictions; a Cray would add and multiply fake numbers much faster, but the sums and products would be no less fake. A more modern example is the Enron bankruptcy scandal. If Enron executives reported fake profits, no stock-trading computer could understand that they are fake and make them stop; only human prosecutors and judges could.

In the Soviet Union in the early 1950s there was an ideological campaign against cybernetics, "a reactionary pseudoscience", "an ideological weapon of the imperialist reaction." Having read this book, I still don't understand the reason for the campaign. The idea that there are some common control and feedback patterns between mechanisms, animals, the human mind and human society can be construed as contradicting Marxism-Leninism, which proclaims the primacy of matter over immaterial information; at around the same time, genetics was also persecuted, and the idea that the development of an organism follows from the information in its DNA was denounced as a bourgeois pseudoscience. But then, every branch of non-Communist human thought can be construed as contradicting Marxism-Leninism.

[Nils · Rating 5 out of 5]

A truly remarkable book, which shows that the real vision of computer scientists at the dawn of the computer age was literally to build an artificial of synthetic brain — albeit rooted in the era’s highly reductive and behavioristic understanding of the brain. The brain is simply a feedback mechanism based on the binary firing of synapses — and this is exactly what a computer is in material form.

But Wiener was also a visionary about what the wider impact that the proliferation of artificial brains would entail. Wiener foresaw that computing would “devalue the human brain,” and result in the elimination of vast swathes of intellectually oriented jobs; he observed further that if these systems were to be deployed on the basis of what he called “the open market,” it would generate a disaster. “We have contributed to the initiation of a new science,” which, he wrote, “embraces technical developments with great possibilities for good and evil. We can only hand it over into the world that exists about us, and this is the world of Belsen and Hiroshima.” He suggests that we need to establish a system of economic organization based on values other than profit, that workers will never figure out how to do this on their own, and that this process of revaluation is the responsibility of philosophers and statesmen.

Seventy years on, his prediction seems right on the money (as it were), and the work he called for remains to be done.

[Oleg Dats · Rating 4 out of 5]

Think of this book as a part of history. Must read for those who is interested in the origins of the AI field.
Wiener described foundation of a new interdisciplinary subject: Cybernetics.

1. Newtonian approach will not work. World systems are statistical in nature and impossible to get time reversersibility.
2. We should use statistical mechanics to model the world. Modern subject is Ergodic theory
3. How to predict the future. Information theory. Modern subject is statistical time series.
4. Feedback is crucial. Modern subject Optimal control.
5. Binary operations, energy and computation,… basic blocks for software and hardware. Modern subject: Electrical engineering
6. Gestalt as idea of some object. Generalization. We can perceive (vision) a cat from different angles, light conditions, shadow,… but we still see a cat (cat == gestalt). Ideas about transformation groups (affine transformatio…). Modern subject is computer vision.
7. How math can help us to understand diseases of mind and vice versa. Modern subject is Neuroscience
8. Economics and Game theory
9. What is learning and how to teach computer to play chess. Reinforcement learning.
10. Brainwaves and basic insights. Even now we have just some typization of brainwaves without deep understanding.

As you might see, most ideas live is separate fields now. Creating a new science was not successful.
Study modern Optimal Control to get more insights.

[Ami Iida · Rating 5 out of 5]

everything is for science.

[Vygandas Ostrauskis · Rating 5 out of 5]

Viena iš labai svarbių knygų mano gyvenime, nulėmusi mano inžinerinį išsilavinimą, kai negalėjau studijuoti literatūros (trys metai Sovietinėje armijoje man rodės per didelė auka, kad mokyčiausi VU humanitarinių dalykų). Daug įdomaus šioje knygoje gali rasti ir eilinis skaitytojas (nors ji rekomenduojama inžinieriams ir mokslininkams) – tai knyga, pirmoji padėjusi pagrindus labai plačiai mokslo šakai – kibernetikai.
Turiu 1968 metų rusišką laidą (pirmąją toje srityje, kai buvo suvokta, kiek atsilikta skaičiavimo technikoje nuo Vakarų šalių ir ypač nuo JAV). Knyga, patekusi į Lietuvą, ėjo iš rankų į rankas tarp jaunųjų mokslininkų ir elektronikos inžinierių). Įkelsiu į šį portalą tą knygą ir pabandysiu parengti apžvalgą.

[Anthony Bello · Rating 4 out of 5]

I was surprised to learn that Wiener actually cautioned against applying game theory and its models too generally. Also, I didn't realize that the math of a book this influential, with applications to a diverse range of subjects, would require such an advanced understanding of calculus. Suffice it to say, I was unable to appreciate the finer points of Wiener's mathematical scope. Nevertheless, this book will only get more influential with time, and the easy parts are still worth reading.

[Jona · Rating 5 out of 5]

Wiener ist die Lichtgestalt der Kybernetik, ihr Gründervater.
Und sein der Wissenschaft ihren Namen gebendes Buch ist nach wie vor hoch interessant, vor allem, wenn man es aus einer historischen Perspektive liest.
1947 veröffentlicht passt es wie die Faust aufs Auge auf eine Zeit, die nach den Umwälzungen der Physik in den ersten Dekaden des Jahrhunderts im Zuge der Weltkriege und speziell des Mannhattan Projekts wesentliche technologische Fortschritte erlebt.
Unweigerlich bahnt sich in der Nachkriegszeit der nahezu nahtlose Übergang in das digitale Zeitalter an und so ist es kein Wunder, dass andere wissenschaftliche Superstars dieser Ära (Hallo John von Neumann) eine prominente Rolle in Wieners privatem wie akademischen Freundeskreis gespielt zu haben scheinen.

Als Kybernetik-Student habe ich mich immer gefragt, warum der Begriff "Kybernetik" heute so vergessen zu sein scheint, obwohl an allen Ecken und Enden auf die fundamentalen Konzepte, die Wiener hier präsentiert, verwiesen wird.
Häufiger habe ich im Voraus gehört, Wiener versuche, eine Art Universaltheorie aufzustellen, die sich letztlich also gar nicht so universal herausgestellt hat.
Im Rückblick würde ich dem relativ entschieden widersprechen.

Aber der Reihe nach:
Wiener startet mit einer Art Chronik der Entstehung des Gebiets als Resultat vielfältiger Kollaborationen, die vor allem auf dem Interesse fußen, interdisziplinär zu forschen ("It is these boundary regions of science which offer the richest oppportunities to the qualified investigator.").
Das Resultat dieses Abschnitts ist die Namensgebung des jungen Felds und ein grober Abriss dessen, was im weiteren Verlauf des Buches folgt.
Dabei bemüht sich Wiener direkt, um eine Einordnung der Chancen und Risiken, die sich aus den entwickelten Methoden und ihren technologischen Realisierungen ergeben.
Sich daran nicht zu beteiligen, helfe nicht, die Gefahren einzugrenzen - im Gegenteil muss man die Verantwortung an- und die Entwicklung in die Hand nehmen ("We do not even have the choice of suppressing these new technical developments.").
Auch in dieser Hinsicht ähnelt er von Neumann ("For progress there is no cure."), sieht aber gleichzeitig voraus, dass die kybernetische, also digitale, ähnlich wie die industrielle Revolution knapp 150 Jahre zuvor, ihre Opfer haben und damit zu gesellschaftlicher Veränderung führen wird.
Gewissermaßen offenbaren sich in Zeiten von GPT erst heute einige der Potentiale, die Wiener und viele seiner Zeitgenossen bereits vor mehr als 75 Jahren vorhergesehen haben.

Nach dieser eher allgemeinen Einleitung wird schnell offensichtlich, dass Wiener ein unumstößlicher Materialist ist, auch wenn er sich bemüht, seinen vitalistischen Materialismus in die moderne Physik, insbesondere im Kontext statistischer Mechanik, einzuordnen bzw. in dieser zu begründen.
Dazu holt er ein wenig aus und wird für den Beweis des Zusammenhangs zwischen räumlichen und zeitlichen Mittelwerten in Gibbs' Mechanik mit Lebesgue Integralen, affinen Gruppen und einem Schwenk zur Ergodentheorie (Hallo Bernard Koopman) fast schon unnötig kompliziert und mathematisch.

Es folgt die Einführung in einen der zentralen Begriffe des Buchs, den der Information.
Über die Äquivalenz zu negativer Entropie und dem Maxwellschen Dämom kommt Wiener zu Zeitreihen, Filtern, Vorhersageproblemen und damit einem weiteren eher technischen Abschnitt.
Dabei erwähnt er quasi in einem Nebensatz, dass die vorgestellte Theorie auf Kenntnis der Beobachtungsdaten der kompletten Vergangenheit beruht und daher um gute Samplingstrategien erweitert werden muss (Hallo Optimal Design).

Im Anschluss dreht es sich dann endlich um den zweiten zentralen Begriff der Kybernetik, den Holy Grail der Regelungstechnik: negative Rückkopplung ("Feedback") und ihr Potential zur Stabilisierung von dynamischen Systemen.
Hier zeigt sich Wieners Arbeit mit Rosenblueth in Mexiko, sind die einführenden Beispiele doch alle physiolgisch motiviert (Stichwort Ataxie), bevor wiederum eher mathematisch die komplexe Übertragungsfunktion hergeleitet und das Schwingungsverhalten des geschlossenen Regelkreises untersucht wird.
Von diesem eher einfachen Basisfall ausgehend beschreibt Wiener einige fortgeschrittene Regelungsstrategien, die er "anticipatory feedback" (Hallo MPC) und "control by informative feedback" (Hallo Iterative Learning Control und Reinforcement Learning) nennt.
Aber auch Ideen, die sich in der adaptiven und robusten Regelung wiederfinden, tauchen hier auf.

Da Feedback letztlich auch nur Information ist, ist es dann auch nur ein Katzensprung von der mathematischen Beschreibung physiologischer Prozesse und ihrer Regelung zum engen Zusammenhang der Informationsverabeitung im Gehirn und der Entwicklung moderner Computer.
An dieser Stelle war ich dann doch ein bisschen überrascht, wie viel des grundsätzlichen Designs der digitalen Berechnungsmaschinen von der Struktur und Arbeitsweise des Hirns übernommen ist oder zumindest dem damaligen Stand des Wissens entsprach (Optimalität des Binärsystems, Informationsspeicherung).
Auch die Frage, ob Maschinen lernen können, diskutiert Wiener kurz und eher prophetisch, bevor er mit dem Lösen von partiellen Differentialgleichungen ein schon damals relevantes Anwendungsproblems anspricht. Arg viel weiter sind wir also in 75 Jahren nicht gekommen.

Die letzten Kapitel befassen sich mit weiteren Anwedungsgebieten, in denen die Konzepte Information und Feedback und damit die Kybernetik als Brückenwissenschaft zu neuen Einsichten gerreichen kann.
Dabei ist es durchaus interessant zu sehen, dass Themen wie Computer Vision, Pattern Recognition und Reinforcement Learning in ihrer Essenz und philosophischer Dimension (zumindest eines Teils derselben) beschrieben werden.
Noch interessanter wird es beim Thema Psychopathologie, das laut Wiener explizit kein materialistisches Feld ist. Im Gegenteil können die betrachteten Fehlfunktionen des Gehirns durchaus rein funktional sein, im Grunde aus fehlerhafter Informationsverarbeitung resultieren.
Seine Schlussfolgerung ist, dass es sich im Kern also um kybernetische Probleme handelt und dass insofern andere informationsverarbeitende Organismen, zum Beispiel Computer, als Blaupause für die Behandlung der pathologischen Zustände dienen können.
Hier führt er dann auch einige Beispiele auf, erklärt, wie die Psychoanalyse in das geschaffene Bild passt, wieso der Mensch anfälliger für psychische Störungen als andere Tiere ist und dass das menschliche Gehirn potentiell die Komplexitätsgrenze bereits überschritten hat, die für einen langen Fortbestand der Art eigentlich eine natürliche Barriere bildet - alles im Kontext von Kommunikation und Zirkularität.
Insofern präsentiert sich sein neu intrepretierter Informations-Materialismus hier im schönsten Gewand und ich kann gut verstehen, wieso viele der systemtheoretischen und kybernetischen Köpfe der Folge-Jahrzehnte von diesem Geist beseelt wurden.

Ganz zum Abschluss kommen dann auch noch sozialwissenschaftliche Themen auf den Tisch und ich war hier vor allem deshalb sehr gespannt, weil ich vielen Aspekten in anderen Kontexten schon begegnet bin (Hallo Future Histories), Wieners, als die "ursprüngliche" kybernetische Perspektive aber noch nie gelesen habe.
Neben den zu diesem Zeitpunkt des Buchs schon wenig überraschenden Vorausblicken auf aktuelle Entwicklungen (Unsupervised Machine und Imitation Learning; Google als Online-Bücherei zur Organisation der gigantischen verfügbaren Informationsmengen) trifft man hier vor allem auf harsche Kritik am Glauben an den freien Markt und seine stabilisierende Wirkung. Die Antwort ist von Neumanns Spieltheorie ("Unfortunately, the evidence, such as it is, is against this simple-minded theory. The market is a game.[...] There is no homeostasis whatsoever."), wobei gleichzeitig deren Limitierungen zumindest erwähnt werden.
Wieners meines Erachtens nach größte Einsicht ist dabei, dass in ausreichend großen Gemeinschaften, das Individuum nicht von der Menge an verfügbarer Information profitiert und intelligenter wird. Im Gegenteil bietet die Kontrolle der Kommunikationsmittel hier besonders große Potentiale ökonomische und damit letztlich politische Macht auszuüben.
Die Instrumentalisierung von sozialen Netzwerken für politische Zwecke und Marketing sind heute so allgegenwärtig, dass dieser Abschnitt fast schon gruselig wirkt, vor allem wenn man die damit einhergehende Machtkonzentration, die Wiener ebenfalls voraussieht, berücksichtigt.

“Thus on all sides we have a triple constriction of the means of communication: the elimination of the less profitable means in favor of the more profitable means; the fact that these means are in the hands of the very limited class of wealthy men, and thus naturally express the opinions of that class; and the further fact that, as one of the chief avenues to political and personal power, they attract above all those ambitious for such power.“

Ihm zufolge muss die Antwort darauf eine freie, demokratische Gesellschaft sein, in der mehr Wert auf kleine, lokale Gemeinschaften gelegt wird, da nur hier Information in gesunden Maß geteilt und stabilisierende Praktiken etabliert werden können.
Und während er das schreibt, betont er gleichzeitig explizit, dass er nicht (!) an die soziale Dimension der Kybernetik im gestalterischen Sinne glaubt, einfach deshalb, weil die Übertragung von Methoden aus den Natur- auf die Sozialwissenschaften nicht so einfach möglich ist. Die Kopplung zwischen Beobachter und Beobachtem ist zu eng, als dass die Beobachtung das zugrundeliegende System nicht zu stark beeinflussen könnte.
Und so lautet seine passende Schlussfolgerung:

"There is much which we must leave, whether we like it or not, to the un-'scientific', narrative method of the professional historian."

Edit: Nachdem ich Ben Recht im InControl Podcast dazu gehört habe, sei noch ergänzt, dass man bei Wiener durchaus rausliest, dass er die Probleme, die er im Laufe des Buchs betrachtet, schon in seinem Vorwort zur zweiten Edition weitestgehend als gelöst ansieht. Für ihn hat sich das kybernetische Denken durchgesetzt und wenn man auf die elektro- und informationstechnischen Umwälzungen seit den 1950ern schaut, lässt sich das wohl auch nur schwer bestreiten. Als Anwendungsdisziplin hat die Kybernetik, aufgefächert in viele Teilbereiche, gesiegt, hinter ihrem biologischen Anspruch blieb sie wohl auch aus Berechnungsgründen zurück und so verwundert es eigentlich nicht, dass im aktuellen KI- bzw. AGI-Hype auch eine Renaissance der Disziplin schlummert.

[Marco Sán Sán]

Hito. La parte algorítmica me vuela la cabeza, no mentiré, pero la teoría de la comunicación sustentada en impulso neurofisiológicos es una pasada. Entre la maquina y el animal, entre lo que conmuta y lo que rumia.

[Grace · Rating 2 out of 5]

The book was interesting to read. Cybernetics had everything combined into one book. It covers the biological factors as well as the technological and the historical. It gave me a new perspective on how technology is similar to the human body. An example from the text: “The synapse is a coincidence-recorder, and the outgoing fibre is only stimulated if the number of incoming impulses… ” (29).

Overall the book was fascinating to learn about technology in a new perspective, but because this was for a class, I would not have chosen it to read on my free-time unless I wanted to fall asleep.

[Sultan Daniels · Rating 5 out of 5]

I finally understand what convolution is. That single-handedly made it a five star. This book definitely made me more comfortable with Fourier Transforms too. I should definitely reread the section on prediction after taking a probability course. And I should also reread the control theory section because that shit was confusing. A lot of cool anecdotes in here too, like how right-handedness/dominant sides of the brain decrease wire delay since the human brain is such a behemoth.

[Alexander · Rating 4 out of 5]

This is technically a very great book, much of the logic and mathematical proofs are lost on me. Quite a special polymath, the 203 pages are incredibly dense and the author covers a massive amount of sciences, theories, and topics; all relating of course to cybernetics. If you’re looking for an introduction, this isn’t it, although it seems it was written to be one. This requires the knowledge of a great variety of disciplines to appreciate.

[Barack Liu · Rating 5 out of 5]

544-Cybernetics-Norbert Wiener-Engineering-1948

Barack
2024/11/24


Cybernetics was first published in 1948. This is the first time the word "cybernetics" was used publicly to refer to self-regulating mechanisms. The book provides a comprehensive guide to servomechanisms (whether electrical, mechanical, or hydraulic), automatic navigation, analog computing, artificial intelligence, and other fields. Intelligence, neuroscience, and reliable communication lay the theoretical foundations. The second edition, with slight revisions and two additional chapters, was published in 1961.

Norbert Wiener was born in Columbia, Missouri, US in 1894 and died in 1964. He studied at Tufts College (BA), Cornell University (MA), and Harvard University (PhD). He was an American computer scientist, mathematician, and philosopher. He later became a professor of mathematics at the Massachusetts Institute of Technology (MIT). Wiener was a child prodigy who became an early researcher in random and mathematical noise processes, contributing to electronic engineering, electronic communications, and control systems. Wiener is considered the founder of cybernetics, a science of biological and machine-related communications that has implications for engineering, system control, computer science, biology, neuroscience, philosophy, and social organization. His work had a significant impact on computer pioneer John von Neumann, information theorist Claude Shannon, and anthropologists Margaret Mead, and Gregory Bateson, among others. Wiener is considered one of the first people to propose the theory that all intelligent behavior is the result of feedback mechanisms that could be simulated by machines, an important early step in the development of modern artificial intelligence.

Table of Contents
I.Newtonian and Bergsonian Time
II.Groups and Statistical Mechanics
III.Time Series, Information, and Communication
IV.Feedback and Oscillation
V.Computing Machines and the Nervous System
VI. Gestalt and Universals
VII.Cybernetics and Psychopathology
VIII.Information, Language, and Society
IX.On Learning and Self-Reproducing Machines
X.Brain Waves and Self-Organizing Systems

The author mentioned at the beginning that he had participated in an informal group discussion when he was at Harvard Medical School. This reminds me of a phenomenon: when people talk about their past, where they choose to start often reflects which events they think have the most direct impact on their current behavior and the topic of discussion. In other words, the author may feel that this discussion has a greater impact on him in the future, so it is the beginning of his topic. In the academic environment, I have seen two different styles of mentors. One mentor will regularly convene members of the research group for group meetings, and each time a student will share his research topic. Although other students have different research directions, the process of listening sharing, and asking questions can inspire many unexpected inspirations. Another mentor tends to avoid group discussions, believing that this form is a waste of time, and prefers to meet with each student individually. I agree more with the former approach. Because I now believe that the importance of discussion cannot be ignored. Whether it is academic discussion or daily debate, their value is reflected in communication and collision. Looking back, my neglect of discussion stems from an arrogant mentality always felt that I couldn't learn anything from others and think that discussing with others is a waste of time. This arrogance may not have obvious consequences in the short term, but in the long run, it is likely to have far-reaching negative consequences. The core theme of this book is "system theory". As the name suggests, it looks at everything from a system perspective and emphasizes the causal relationship between input and output. Any system can be regarded as a feedback system. If we can understand its operating mechanism, we have the opportunity to make the output develop in the direction we expect. This is exactly the purpose of cybernetics: to achieve goals by understanding the system and controlling the feedback.

In the development of physics, there is one person who cannot be avoided: Newton. So, what did Newton do? In the final analysis, the phenomenon he studied is a question that countless people around the world have thought about for thousands of years - what are the rules of this world? Religious believers attribute everything to God's will but driven by the strong curiosity of human beings, there are always people who try to explore and reveal the laws. The greatness of Newton lies in that he revealed the most laws in the simplest form based on the existing research results at that time. His explanation method is highly general and universal, which injected new vitality into scientific thinking at that time. From this perspective, Newton's work is not only the result of scientific research but also reflects a style of the entire history of human civilization - the pursuit of revealing complex truths with simple logic. The core of this scientific exploration is a unique thinking ability: seeing the essence through the phenomenon. Only by seeing the essence can we extract infinite laws from limited experience and not be limited by appearances. This ability is not only the foundation of scientific research but also the key to human understanding and transformation of the world.

The book mentions two scientists, one from the United States and the other from France. Although the two may not know each other's research, they inadvertently form a kind of complementarity - one person's research results can answer the questions raised by the other. This reminds me that such a phenomenon exists not only in academic research but also in daily life. Sometimes we are troubled by certain problems, but we may find that these problems have been clearly answered by others, and it is not worth worrying about at all. Therefore, the academic community encourages communication and inspiration among peers. This idea of "learning from others' stones can polish jade" has long been deeply rooted in people's hearts. Many breakthroughs in scientific progress are based on such mutual reference and cooperation. The worldview of Newtonian mechanics is a deterministic concept: as long as we know the initial position, speed, and direction of an object and various parameters in the environment (such as resistance), we can predict the future state of the object. This deterministic concept emphasizes that among infinite possibilities, the probability of realizing a specific event is almost zero, but countless possibilities close to zero are accumulated to form a certain event with a probability of 1. It's like when we shoot an arrow, the probability of the arrow landing at any specific point is close to zero, but the probability of the arrow finally landing at a certain position is 1. This logic involves the concept of "infinity" - when countless values approaching zero are added together, is the result 0, 1, or infinity? The most magical thing in nature is precisely this "infinite" leap. It is like a deep fault that makes it difficult for us to perceive and understand through intuitive real experience. "Infinity" is like a gap that appears out of thin air on the land.

Information is a continuous existence, which unfolds with time as the axis. The information we pay attention to is meaningful content, while those we ignore are called noise. How to extract rules from this time series information and use these rules to predict future development is an important issue that we have always been concerned about. Today's technology, from the earliest telegraph and telephone to today's Internet communication, the basis of these information transmission methods is built on our understanding of "information". With the continuous deepening and precision of understanding, our ability to process information is also increasing. From thousands of years ago when information was transmitted by fireworks during the day and firelight at night to today's digital communication protocols that are accurate to bits, the efficiency and accuracy of human information transmission have reached a new height. I once saw an interesting statement describing how to count the information on the entire earth: Assuming that we can convert all information into binary form, for example, the result is "1010", then we only need to mark a point on a spacecraft, and the position of this point can be marked as "0.1010" with a scale. In this way, this point carries the essence of all information. In other words, information can be compressed to an extremely simple form but still retain all its connotations. From this perspective, information seems to be a conserved existence. It does not appear or disappear out of thin air but only transforms into different forms. All the information contained in the colorful universe we see today may have been contained at one point at the beginning of the Big Bang. It was difficult for me to understand this point at first, but if I imagine this point as the point marked "0.1010" on the spacecraft, this concept seems to become more intuitive.

In order to solve real-life problems theoretically, humans invented language and mathematics; then, in order to solve these mathematical problems more efficiently, we invented machines for calculation. From calculators used by elementary school students to today's world-leading supercomputers, the essence of all is "calculation". There are two main methods of calculation: one is a continuous analog method that uses analog signals to process data; and the other is a digital method based on state switching, which uses discrete 0s and 1s to express information. Take music as an example. Although electronic music or music on digital streaming media sounds continuous, it is actually composed of discrete digital signals. Humans rely on the instinct of visual persistence or auditory persistence to "complete" these discrete signals and understand them as continuous sounds. As long as the discrete intervals are small enough, the human ear cannot detect the difference. Interestingly, some people prefer music on vinyl records because they believe that vinyl records provide continuous signals and the sound quality is closer to "perfection". In this pursuit of "perfection", people jump back and forth between discreteness and continuity. This made me think, does the real world itself also have this discreteness? The world we see with our naked eyes seems to be continuous, but when we go deeper into the microscopic level, we find that things are made up of molecules. There are gaps between molecules, and molecules are made up of atoms, and atoms have smaller particles such as nuclei and quarks. So, is there a possibility that the so-called "continuity" in the physical world we live in is actually just an illusion? Perhaps, the "continuity" we perceive is only because the discrete units that make up the world are small enough to exceed the recognition ability of our naked eyes. This reminds me of the pictures in video games, such as shadows. Although it looks smooth, if you zoom in and take a closer look, you will find that its edges are covered with "burrs". These burrs are small enough that they cannot be detected by the human eye, so we think it is continuous. Perhaps, the " discontinuity" of the real world can also be used as evidence that the world is virtual.

[David · Rating 3 out of 5]

For the widest category of reader, I suggest skipping most chapters due to the swaths of trivial mathematics and focusing on the following ones: preface; introduction; “Newtonian and Bergsonian Time;” “Information, Language, and Society;” and the 1961 “On Learning and Self-Reproducing Machines.” These are sufficient to provide you with a general yet sufficiently granular understanding of the namesake field and its implications. The genius of the work is found in the fact that, today, the contents of the book are so ubiquitous as to render them banal—pervasive to point of finding itself in engineering, psychology, biology, sociology, and economics to name a few. The book is an intellectual memoir of the field of cybernetics, concerning not only its inception but related ideas throughout its evolution.

The field was born out of a practical need to reduce the redundancies of scientific research, opting for an interdisciplinary scientific program wherein scientists themselves are knowledgeable about adjacent fields. This was also the period in which close-system approaches, like Newtonian mechanics, were encountering their limitations in being able to deal with dynamic systems involving a great deal of complexity and randomness. Open feedback systems were much better suited for modern scientific research. Ultimately, cybernetics became a discipline of statistics, communication, and control. This framework worked so well that it appeared applicable to every other field, be it the study of atoms, the nervous system, or the economy. So well, in fact, that Wiener understood the field to be era-defining, and what I argue is the beginning of the all too modern ‘human-as-a-computer’ metaphor.

Precisely because he understood the importance of his field, Wiener was nervous about its implications on society and the labor economy. On this topic, you can see his Socialist influences. In our capitalist economy, efficient and cost-saving machines would undoubtedly execute acts of physical and low-level mental labor, leaving the average person with nothing to offer of significant value to the market. However, the abstract study of this issue and the practical everyday doings of labor are so far removed that any solution seems hopeless. Labor conditions are hardly ever optimal when competing with machinery. Similarly, the concentration of wealth and power that is enabled by our political-economic systems means that communication and control fall in the hands of the few that manipulate the majority to attain their financial and political goals. Here, Wiener recognizes the value of the social sciences, as inexact as they may be, as distinct from the hard sciences that have little to do with the ordinary conditions of human life.

[John Doe · Rating 5 out of 5]

I am more into the first half of the book, which basically covers physics mathematics and computer science. I went completely dumbfound for the later parts which covers biology and physiology stuff. I have no idea what he was talking about. But its fine, i get the basic idea, humans are like machines, could be mathematically modeled right?

So, modern science is based on mathematics, more specifically, Newtonian physics and later quantum mechanics both were developed hand in hand with intro level algebras to calculus & probability theory and all the way to group theory(abstract algebra), differential equations, linear algebra and stochastic processes.

the philosophical implication of those fundamental thoughts is what Dr. Wiener was trying to convey. My understanding is that, first we perceive everything as independent and self contained objects with simple shapes and simple relationships with one another, and that's how we developed basic number system and Rene Descartes' Cartesian system in analyzing algebraic relationships between objects. on the physics side, the Newtonian physics kicks in. at that time, space and time are indisputably and inseparably definitive, we takes everything for granted, we recognize the world we live in as a giant mechanical world as if all the motions and all the future could be predicted, since everything obeys the Newton's basic laws.

But as mathematics and science develops, the original perception of space and time alters, so does our mathematics system, everything falls into the hand of probability theory, especially in the world of atoms, which means the physics world also agrees, that our mechanical world, after all is not that mechanical and definitive, instead it is stochastic and intricate. time and space intertwined, the speed and position of a particle could not be decided at the same time without interference, we have to model things as a whole, as an intricate system.

Luckily, scientists and mathematicians developed more methods to model the uncertain and imperfect non-Newtonian world, we had information theory, we had computers, with the superpower of statistics(now maybe at the courtesy of machine learning and artificial intelligence), uncertainty gets measured too, we can develop machines with modern tools that mimic the natural organism, call it astrophysics simulation, weather man's weather simulation, biologist's organism simulation or whatever. But the point is, we can and we will use mathematical models to abstract all natural phenomenons and figure out the mechanisms of all the "machines".

[Attay Kremer · Rating 3 out of 5]

This book is rather unique, and perfect for a specific kind of reader. Unfortunately, that reader is not me. We start off with a chapter on science and time, dedicating itself to the difference between Newtonian and Bergsonian time. From a philosophical perspective, this chapter is on that rare cusp between mind-numbing boredom and usefulness. For a more science and science fiction oriented person this could be a pleasant and interesting read, but I found it to be mostly busy belabouring obvious points and being all around cumbersome in its style.
This strange interplay between a boring style of writing and a blend scientifically oriented manner of thought is carried along through the book, even at its most explosive and fascinating points. To add to this mess of a read, mathematical explorations are sprinkled across the book, mostly for scientific completeness. By scientific completeness, I mean that they add very little, and do not help to further construct or expose the main thesis of the book.
All this being said, the book is revolutionary in its thesis: to bridge the gap between vitalist and uniquely humanist philosophies of the world (Wiener mostly identifies Bergson with this view) and materialistic and machinic philosophies of the world. This gap has been aching to be bridged since its unbecoming formation with Cartesian philosophy. This bridge is constructed in two layers: the obvious and surface level mechanization of control and feedback loops, and the deeper level of finding in humanity essentially machinic components. Statistics, feedback and parallel computation form qualitative multiplicities and dureé-tional time, thus forming the "uniquely" human from that which is unmistakably machinic and alien.

The book is perfect for the scientifically literate sci-fi enthusiast. The philosophically inclined should take heed but still read the book, for it bears, from beyond the mask of humanity, the underlying universal machinic nature of everything. The universe is not united Newtonically, by isolated and simple collision and causal relations, but by a machinic substructure of feedback and parallel action.

[Sicofonia · Rating 3 out of 5]

Cybernetics is an introductory text to one of the founding figures of the field, Norbert Wiener. Defining Cybernetics as the field of control and communication theory, whether in the machine or the animal, this work goes into explaining the different areas of study concerning this field. Examples of those would be feedback and oscillation, information theory, language, communication, computing machines and nervous system or learning to name a few.

We have to bear in mind the calibre of the author of this book, so Wiener being an excellent mathematician and scholar, resorted to demonstrate certain points on this book by using mathematical formulas. For the reader who is not familiar or up-to-scratch with the mathematical concepts this could be all meaningless (as was my case). However, it is still possible to read the entire book and gain an understanding of what cybernetics is about, why Wiener considered important and the role it could play in furthering human progress or knowledge.

One of the highlights for me was that Wiener didn't shy away to state his opinions on a variety of subjects. Like for example, the role of mass media in influencing the general public opinion in the way the owners of the media want it to be. From a cybernetic point of view, it is relatively easy to explain this influence in the feedback mechanisms present in this form of communication. And also about the possible future, now a present reality, of a society where humans would be find themselves expendable in the labor market due to the advancement of machines.

We have progressed leaps and bounds in fields like computing or artificial intelligence since this book was published. I came across another review in here that speaks about the value of this book as a historical document, and just because of it this is worth reading. That, and to also realize that some of the challenges we face nowadays in this highly mechanized society were already anticipated by Wiener in this book.

Only 3 stars because, even it it was good, it is a bit out of date and I couldn't cope with the maths. But totally recommended all the same.

[Gulliver's Bad Trip · Rating 4 out of 5]

What Octavio Paz were (with his poetry and essays acting as positive, exemplary or negative, untrustworthy cardinal points in the mazy 20th century) for writers such as Roberto Bolaño amongst others latin american poets and novelists, even a few european ones, Norbert Wiener was this (with his essays on cybernetics, post-quantum physics, the sciences in general and contemporary society whether flirting with a inadvertent apology of it all or with the most staunch denunciation of a predictable catastrophe) for american writers (New Wave science fiction at first) and filmmakers (at least since Kubrick) and for a few other european artists and a couple of philosophers as well, even becoming a beacon for some foreign, latin american cosmopolitan imigrants and no less than a eccentric asian filmmaker, believe it or not. Finding them or not.

There's no more doubts that one of the most known writers influenced by Wiener recently was David Foster Wallace, belatedly becoming famous for his wacky non-fiction or self helppish essays for his turn, who wrote in his relatively far-famed novel Infinite Jest the ignominious character James Incandenza. There's no need for more than a superficial knowledge of the cybernetician's work and life in order to understand from whom primarily came the inspiration for a all-too-much successful, catastrophic scientist but such a clearly disastrous family man and exposed failed artist ('entertainer' as the novel puts it) to the point of self-destruction and worldwide tragedy. Since I barely have any college level knowledge of mathematics and the like, only a couple of essay books, two autobiographic books (there's also at least one recent biography as well), a inacessible enigmatic novel called The Tempter and some presumably rare, magazine-published, short science fiction stories are left to be exhumed.

[Fred · Rating 3 out of 5]

I really wanted to like this book more. I really did. Granted, it didn't get off to a good start, and that really soured the taste for me throughout.

More specifically, I am talking about Chapter 3. Chapter 3 felt so unnecessary and was kind of a disaster. I am a trained mathematician, and a big fan of the "Motivation => Intuition => Formalization pipeline" in mathematics education, but that chapter failed in a huge way on those first two points -- the most important ones. The material felt completely unmotivated. It's like yes, I know how to drive (follow the mathematics), but I don't know where I am going or even where I came from.

If Chapter 3 was removed from the book (honestly, it felt like an "I need to hit 200 pages"-type chapter) then my review would be much more gracious. There is a lot of useful insight to gain from this book -- even as an adventurous layperson -- as long as you skip that chapter.

I would caution delving into this book for most people, even for people who have a strong background in mathematics and are interested in the intersection of Mathematics, Computer Science, and Cognitive Science.

If you haven't already, go read Gödel, Escher, Bach. It does Motivation, Intuition, and Formalization a lot better, even if it's not on exactly the same topic.

Side Note: Frankly, I think the review from The Saturday Review of Literature on the inside cover saying

"It is a beautifully written book, lucid, direct, and despite its complexity, as readable by the layman as the trained scientist..."

is offensive. I know that the education standards in the US have fallen since the 60s but...

[to'c · Rating 4 out of 5]

A deeply challenging, insightful, and thoughtful book about how we may work. How we think and how we function. Brilliant at its time and maybe even brilliant now. Wiener was an undoubted genius who saw things his contemporaries did not. And perhaps things we do not yet see even today.

Some of the challenge is the mathematics. Wiener was a mathematical genius and saw the world as deeply mathematical. I admit that I didn't grok all the math. But I was able to follow along, especially since much of it gets explained in not-quite-so-plain English. But don't give up! The math isn't quite *that* dense and most readers probably won't find it as challenging as my old brain far removed from my college days did.

But deeper than the maths are the thoughts and insights that drive it. And you don't need to fully understand the math to gain those insights so don't be frightened away by it!

Of course, the book was written in 1947 so there's lots of out-of-date stuff here. Bear that in mind and perhaps look up some more modern descriptions of how things work. One of my favorite passages is in the final chapter. Wiener spends several pages deriving the math and analog hardware necessary to analyzing brain waves. Then he says he got access to a digital computer and pretty much everything he just detailed was no longer needed. Welcome to the modern world, Norbert! ;0)

[Maksym Shcherban · Rating 4 out of 5]

Somewhat mixed feelings. I usually enjoy these retro scientific and popular-scientific books much more. I love to see how some of their predictions came fully to life, while some quirky things were either completely wrong, or simply didn't come to fruition.

In this case, unfortunately, I've studied Cybernetics and Robotics for 5 years in a university, and it must have sucked out the fun out of the topic for me 😂 Also, for some reason, the author has decided to add some heavier mathematical derivations, but not to every chapter. And because of that, some chapter read lightly, like a popular science book, and some would be more suitable for students who already had several courses in calculus.

It was somewhat amusing to read Wiener's thoughts regarding possible applications of control theory in future nuclear warfare, or discussion of lobotomy as "a viable method of fighting mental distress". But most of author's ideas regarding feedback control, machine learning etc. turned out to be correct. One thing is for sure, I now understand why they call Norbert Wiener the "father of cybernetics".

[Robert Mason · Rating 3 out of 5]

3/5 after a casual reading - it may require more than this to take everything in.
I picked this up as I believe it is an important work in the history of communication, but it spends little time on this and is much more broad - from AI and the nature of life to anti aircraft guns that track planes.
The book is extremely well written and some areas of the book feel like they were actually written for a general audience, however it can often get bogged down with maths without giving you the bigger picture which often left me wondering why a particular calculation was being done.
As it is a book whose first edition was released in 1949 (with two additional chapters added in the 1960s) some of the information is dated, but it is interesting to see what people thought of e.g. chess playing computers long before deep blue and some of the ideas are so foundational to a subject that it is still somewhat relevant to think about. However, from a modern perspective, some of the information seems unimportant for an overview of the subject because it's age.

[Stephen · Rating 3 out of 5]

Not gonna rate this on the scientific merits, cause obviously this book lays the groundwork for a number of things which have been massively influential to a point that can't be understaded. But those ideas were developed further and more completely in academic publications (rightly so), while this book touches on some of them in chapters that are heavy in math detail, while other chapter are largely expository. So after reading this, i'm just kind of curious, who was this book for? Did it succeed on communicating what "Cybernetics" is and it how it encompasses feedback control, information theory, pathopsychology, and a whole slew of other topics? Just seems unlikely. As a scientist, Wiener is unsurpassed. As a communicator of Big Ideas and how they fit together, maybe could've used another draft...

[Eddie · Rating 5 out of 5]

May spend some time soon properly reviewing this. Stay tuned.

Immediate thoughts: Extremely dense, very thoroughly argued. Advances theoretical basis for modern analytics discipline. Relevant concepts for data management and other information technology disciplines explained as well. Written from a seemingly optimistic overall perspective on analytics technology. Should be read and considered with sufficient knowledge of its historical context. Requires a fair degree of technical understanding. Potentially quite relevant for computer scientists, technologists, economists, and the like today.

[Felix Delong · Rating 5 out of 5]

Norbert Wiener is a character with strong moral values, well worthy of being read and his example to be followed. I'm so glad I finally got to read this absolute gem.
This book is a basis for information theory of life, which I find fascinating. It basically unifies life, physics and technology under the banner of exchanging signals. I do not claim to understand all the mathematics herein, but otherwise the message in this book is crystal clear, even without the math.
I would recommendt it to anyone - it should be a canon for everyone striving to count himself as a well educated person.

[Yates Buckley · Rating 4 out of 5]

This is a very important text for anyone interested in the history and philosophy of computation. Weiner was extremely influential in his ideas post war and thus shaped much of the cultural context of early computer science up to the late ‘60s. The ideas and theoretical work are still fundamental and echo behind many technological systems today. A brilliant mind with perhaps too much influence on contemporary ideas giving rise to several unintended consequences.

[Steve Kelly · Rating 5 out of 5]

The mathematical derivations are archaic, but also intriguing because they are wholly sui generis. The mathematics can be difficult to follow, but he usually explains the concept at after a long chain of integrals. The psychological and sociological discussions also are very apt any make up the first third and last third of the book. Almost all his predictions about controls and computation are now reality.