Slashdot Mirror
Cybersyn And Early Uniminds
An anonymous reader writes "The Guardian Website is running a story on Cybersyn. An experimental computer network based on cybernetic principles that was used by Chile's revolutionary government between 1971 and 1973 to provide a real-time, decentralized form of economic analysis in the nationalized sector of the Chilean economy. The network has been described as Chile's Internet. There is a photo of the control room which looks something like the deck of the Starship Enterprise.
The whole thing was the brainchild of Stafford Beer, a sort of British Buckminster Fuller. All very Orwellian and Big Brother, the whole experiment was brought to an end by the CIA sponsored coup d'etat on the September 11th, 1973."
I shall go on.
Cybersyn is the implementation of Stafford Beer's "Viable System Model" which is modelled on the working of the human nervous system.
Totalitarianism would be like if the brain demanded to know every detail of what the hand was doing.
The body doesn't work this way - hand control (for instance) is decentralized, with part being controlled by local "muscle memory", part being controlled where the nerves meet the spinal cord, and part being controlled at different levels of the brain. In order to prevent information overload the information passing upward is filtered at every stage to remove redundancy and irrelevancy.
In Cybersyn, the workings of actual factories was monitored by a couple of IBM/370s (if memory serves) but only statistically, to throw up warning events if the stats went out of whack. These warning events would be passed up to the "industry level", where they would mostly be absorbed since the industry may be operating within tolerances even if the individual factory isnt. Only if the industry had problems would signals be passed to the "master control room".
A highly moving aspect of the whole tale is that the "master control room" is a logical neccessity, but when Beer pointed it out on a system diagram to Allende, his immediate assumption was that the box represented THE PEOPLE.
Actually the Cybersyn implementation couldn't be as decentralized as Beer wanted, because Chile could only afford two computers which, by their nature, had to be centralized processing units. In today's world things would be a lot different, and no doubt Beer would advocate open source as the one way to enforce that the government couldn't be collecting information it shouldn't.
I shall go on. Cybersyn is the implementation of Stafford Beer's "Viable System Model" which is modelled on the working of the human nervous system. Totalitarianism would be like if the brain demanded to know every detail of what the hand was doing. The body doesn't work this way - hand control (for instance) is decentralized, with part being controlled by local "muscle memory", part being controlled where the nerves meet the spinal cord, and part being controlled at different levels of the brain. In order to prevent information overload the information passing upward is filtered at every stage to remove redundancy and irrelevancy. In Cybersyn, the workings of actual factories was monitored by a couple of IBM/370s (if memory serves) but only statistically, to throw up warning events if the stats went out of whack. These warning events would be passed up to the "industry level", where they would mostly be absorbed since the industry may be operating within tolerances even if the individual factory isnt. Only if the industry had problems would signals be passed to the "master control room". A highly moving aspect of the whole tale is that the "master control room" is a logical neccessity, but when Beer pointed it out on a system diagram to Allende, his immediate assumption was that the box represented THE PEOPLE. Actually the Cybersyn implementation couldn't be as decentralized as Beer wanted, because Chile could only afford two computers which, by their nature, had to be centralized processing units. In today's world things would be a lot different, and no doubt Beer would advocate open source as the one way to enforce that the government couldn't be collecting information it shouldn't. (Repeated because this topic finally made me get an account)
If you want to know about the chilean guy who was behind this, and what he is up to now, check this article.
For those of you already complaining about how a bloody coup thwarted this clearly great idea, please read that article. It is very politically biased, but it shows how this guys ideas have evolved over time, and I would dare to say he wouldn't think of building such a clearly useless system now.
A system like that cannot take individual human actions into account, it cannot deal with subjective market decisions, it cannot handle human relations. A professor at Universidad de Chile (the one the submitter mentions) told us about this system years ago, and how it seemed to be such a great idea for managing coal production (for example)... until it had to deal with a coal miners strike...
If you want to know why such a centralized system will never be useful check econlib, you might learn a thing ot two.
By the way, I'm chilean.
Cybernetics is the application of control processes from biological systems to artificial systems.
You're thinking of bionics. (Although the definition you give isn't exact for that, either). Cybernetics is the study of control and communication in both living and non-living systems.
Here are the dictionary links:
bionics
cybernetics
(Triva note: the term "cybernetics" was coined by Norbert Wiener, "bionics" was coined by Dr. Jack Steele -- my father-in-law)
-- Alastair
I bet "Cybernetic Principles" sounded really groovy in 1971
True. But when Staff wrote 'Cybernetics and Management" in 1959, the idea that you could apply Wiener's 1948 observational theories to real enterprises, let alone an entire national economy, has got to have been one of the all-time crazy ideas. Like Team Syntegrity (part of the Viable Systems Model, kicking off from the idea that every imaginable system can in some sense be modelled as an icosahedron), based on Buckminster Fuller's idea that 'all systems are polyhedra' - nuts perhaps, but terribly terribly useful and possibly *the* most complete model of 'organisations' (whatever they are, read the book!) ever constructed.
Here's a lecture(pdf) Staff gave in 1973 looking back at his work in Chile. And nearly two decades later, here's 'world in torment' which gives both a lovely flavour of what Staff was all about and a frightening summary of where the world may be going.
NOTHING has ever hit me with the same combination of 'wow' and 'my brain is burning' than Staff's Viable Systems Model seminars, sorry, Syntegrations. But like the man used to say, 'you need big words for big ideas. And you should find it hard to understand.' And you just knew you were being inspired by one of *the* great minds.