Cybernetics A to Z (1974) | Hacker News Reader

Cybernetics A to Z (1974)

archive.org

81 pointsthe-mitr17 comments

Cybernetics needs a comeback. It was the most important and underexplored attempt of creating an unified attempt of interpreting reality having at its core ubiquitous concepts such as systems, information and complexity. Unfortunately we've been stuck with a reductionist paradigm for a long time, and most people don't seem to care much about the obvious presence of systems all around (and even inside) them.

Couldn't agree more.

for me, systems theory is a potent heir of cybernetics, and thriving. at least in my field the management sciences. especially with new incarnations incorporating new materialism, or the actor network theory.

Isn't the historical progression cybernetics turns into control theory (via Norbert Weiner) and control theory turns into reinforcement learning. So just had a few rebrandings and never went away.

"Control theory" isn't a term I'm particularly familiar with, but I can at least say that this isn't correct. Cybernetics didn't exist in any unified fashion prior to Wiener writing the book on it in 1948, and it didn't last more than a couple decades afterwards before being laundered piecemeal into mathematics, technical subfields, philosophy/anthropology, and a couple other things of varying interest (Stafford Beer's "Management Cybernetics" is probably the most notable example).

There was a particular schelling point in the '40s surrounding Wiener (and Shannon, Von Neumann, etc) and the Macy Conferences who were searching for an interdisciplinary umbrella, and found it in Cybernetics. It began as something decidedly "para-academic" and never really grew beyond that. The scope of Cybernetics was — and probably is — too broad to fit into a rigid departmental system. Instead of thinking of it as "going" away, it is probably more accurate to say that it never really arrived.

You could argue that the resurgence of neural networks being the dominant paradigm in "AI"/ML is a partial comeback of cybernetics (dynamics of simple systems leading to complex behavior). I'd like to understand the history of cybernetics better, but my current impression is that much of cybernetics, or at least ideas in the spirit of cybernetics, just became rebranded after the original practitioners died.

I agree with your general take though. The history of aviation safety is a good example of taking this systemic way of thinking to heart, but most other fields don't seem to take such a wholistic approach in their analysis. I.e. part of the reason planes are so safe nowadays is that the people in the field worked really hard to understand the systems from their basic mechanisms to the psychological and physiological effects on the pilots/crew (granted wanting to not kill people is a great incentive to figure this out). I'm sure there are other examples of intensive analysis throughout the system, but that's the only good example I can think about.

I agree, it's unfortunate how it ended. My first real contact with cybernetics while being younger was a book "Cybernetics and character" by Marian Mazur (printed in 1976). Later chapters bored me but the beginning of book had impact on me as showed me a bit different mode of thinking which I believe improved my understanding of complex systems.

In book he tried to model different psychological behaviors, like conflicts, using cybernetics language and try to draw conclusion and what such simple models can tell us.

Also in introduction Mazur argued that science should not have artificial bounds as "subjects" and closed walls - there are many problems that arise in many different fields, with many different point of views. That also changed me in a way that I started to actively engage with totally different fields of science. At all, cybernetics is about abstract ideas that are quite universal.

> Cybernetics needs a comeback.

I'm inclined to agree, but it'll need a name change...

Why do you think so?

In popular culture, it's become very strongly associated with sci-fi-ish prosthetics and implants. I've had arguments with people who insist that "artificial intelligence" means sci-fi robots exclusively, despite the actual field of study being older than they are, and it's 10 times worse with "cybernetics."

It's a shame, there's nothing wrong with the term or the people who coined and studied it, but I really think trying to hold on to it would hold the field back unnecessarily.

I have a physical copy of this! My parents got it for me at a garage sale in the 80s when I was a little nipper.

It has a great programming language “tower of babel” illustration.

It wasn’t until years later that I realized it was a Soviet science ecosystem book. Some of the content should have been a giveaway …

Beautiful design, illustrations, and typography.

That's an understatement. It's utterly brilliant.

This is the kind of book I'd rather have physically, preferably a copy from the time it was published.

Had the original (I think it was a gift), one of my favorites, got me interested in computers… There were things you couldn’t (easily) do or have in the Soviet Union, but learn things, OMG…

There's a section describing how perceptrons work. :)

HELLL YAAA

Reply on news.ycombinator.com