Features, Future Please

Future, Please: Cybernetic Society

July 20, 2013
Cybernetics Norbert Wiener

Norbert Wiener, author of “Cybernetics,” a 1948 book in which he develops a theory of communication and control.

Many beloved fictional settings feature societies free from the corruption, meaningless toil, and irrational wastage plaguing modern human societies. These planned societies don’t promise a perfect world – there are no interesting stories in perfect worlds – but offer the possibility of a rational one.

Our own attempts to design centrally planned economies yielded only brittle, crushingly totalitarian states, Stalinist nightmares of fiat rule, corruption, and dehumanization. Yet the dream persists: a planned, smoothly-functioning world, responding rationally to evolving conditions, shepherding resources for the benefit of humanity.

Can engineers do better? As it turns out, we can – and almost did, 40 years ago.

 

The Foundation and the Culture

Many science fiction fans advance Star Trek as an example of such a planned, internally harmonious society. While Trek is many things, it’s not the best example of a cashless utopia – money, graft, and greed rear their heads the moment our crew leaves the ship.

Star Trek Utopia

Star Trek’s crew was not quite a Cybernetic Society.

Better representations can be found in the works of Isaac Asimov and Iain Banks. In Asimov’s Foundation and Robot stories, we see rational, thoughtful rule through the application of advanced mathematical models and carefully engineered management.

The Robot stories – by which I mean the nine tales comprising I, Robot and the two novels, The Naked Sun and The Robots of Dawn – portray the evolution of robots and AI mainframes into the guiding instruments of human society. Particularly, in the final tale of the I, Robot collection (“The Evitable Conflict”), we see Asimov’s AI management platforms quietly transition into leadership roles through generalization of his First Law of Robotics – the AI have concluded that not taking control of human society, alleviating suffering and corruption, would be tantamount to allowing humanity to come to harm by inaction.

In his Robot novels, on the other hand,  the control of human society by its robotic servant class comes about as an emergent consequence of individual robots independently applying the Three Laws (well, Four) to their own areas of responsibility. A careful re-reading of those books, in our age of ubiquitous software tools and tailored information retrieval, may prove them startlingly prescient.

In Iain Bank’s Culture series we see a full-fledged, cashless, post-scarcity utopia in effect. Citizens of the Culture live in egalitarian, cushioned comfort, pursuing their individual dreams and potential in as perfect a vision of future society as I’ve ever encountered.

Iain Bank's Culture

Citizens of the Culture aren’t all biological, of course. Any self-aware mind, regardless of intelligence, capability, or processing substrate, is treated as an equal person. This includes biological, human-equivalent races, as well as drones (small, mobile AI platforms) and capital-M Minds – intelligences animating the city-sized ships, massive habitats, and facilities that comprise the physical structure of the Culture.

The egalitarian core of Cultural values makes their rule of a gentle sort, but Banks makes a careful point by placing Minds at the center of every habitat and ship in his novels. The human-equivalent citizens of the Culture literally live inside their AI “masters”, who oversee the billion details of life support, economic necessity, and social health of the citizens under their care.

The majority of the Culture’s citizens live in massive Orbitals, controlled by a central Mind so powerful it can speak to and relate with each citizen individually. When I first encountered these Minds, I was immediately reminded of Richard Brautigan’s 1967 poem, All Watched Over by Machines of Loving Grace:

I like to think (and
the sooner the better!)
of a cybernetic meadow
where mammals and computers
live together in mutually
programming harmony
like pure water
touching clear sky.

I like to think
(right now, please!)
of a cybernetic forest
filled with pines and electronics
where deer stroll peacefully
past computers
as if they were flowers
with spinning blossoms.

I like to think
(it has to be!)
of a cybernetic ecology
where we are free of our labors
and joined back to nature,
returned to our mammal
brothers and sisters,
and all watched over
by machines of loving grace.

In terms of social progression, one can draw an almost straight line from the enlightened AI management and emergent supervision of the Robot stories to the fully equal, post-scarcity utopia of Bank’s Culture series. Lacking strong AI, however, it’s easy to conclude that the advantages – in terms of increased liberty and the elimination of scarcity through central planning – of the Culture’s controlled economy are out of our reach.

Interestingly enough, the Robot series sketched out a path towards just such a benevolent “garden” economy 60 years ago – and real-world engineers were ready for field tests in the early ’70s.

Cybernetics and Feedback Systems

Artificial Intelligence vs Cybernetics

Image copyright Paul Pangaro http://www.pangaro.com/published/cyber-macmillan.html

Cybernetics is a multi-disciplinary approach to systems theory focusing on the interplay of feedback and control in goal-oriented systems. The term has been in use since Aristotle and comes to English from the Greek “kybernētēs”, meaning either a pilot, governor, or rudder. Aristotle, and latter Enlightenment thinkers, occasionally used the term to organize reflections on the dynamics of social control and government.

In engineering, parallel work on self-regulating mechanisms in the West similarly dates back to the Classical period. Nineteenth century applications borrowed the vocabulary of cybernetics without merging the disciplines. Centrifugal feedback valves (“governors”) were incorporated into the Watts steam engine, for example. While the engineering of self-regulation through feedback mechanisms continued, the principle continued to see its widest application in the biological and social sciences. Alfred Russel Wallace, in his 1858 paper “On the Tendency of Varieties toDepart Indefinitely From the Original Type”, identified negative feedback as the core mechanism of natural selection, and essentially cybernetic analytical tools were used to explore animal anatomy and behavior.

Its modern emergence as a robust, analytical approach originates with negative feedback control systems in electrical engineering, specifically in control circuits for radar antenna and anti-aircraft gun platforms. The principles developed in MIT’s Servomechanisms Lab grew into a wider exploration of the mathematics and engineering of dynamic control systems more generally, with application to biology, network systems theory, mechanical engineering, corporate governance, and roboitcs… As an analytical framework, Cybernetics (the term proposed by Norbert Wiener) proved applicable to any complex, self-regulating system.

Its application to industrial management and corporate governance began almost immediately, during the mid-20th century, but no one seriously proposed building a national government rooted in Cybernetic principles. While these tools put the goal of a rationally managed, engineered society within theoretical reach, 1950s-era technology and the prevailing social structure of Western nations made their practical application at that scale unlikely. Cybernetic social control systems were limited to industrial and corporate settings. Developing a planned, rational economy on the national scale was unthinkable, for all practical purposes.

Until 1971, that is, when Allende went and did it.

The Cybersyn Project

Project Cybersyn

Anthony Stafford Beer made his professional reputation applying cybernetic management systems at United Steel and the International Publishing Corporation. His theoretical work – the Viable Systems Model – informed the practical development of nested feedback systems in organizational management by applying lessons learned from study of the brain and nervous system in animal behavior.

The Viable Systems Model maps an organization’s feedback and self-regulatory structures as a set of of five interactions subsystems:

  • System 1 comprises the day-to-day operations of the system; workers at a factory, engineers designing tests, truckers making deliveries, and so on.
  • System 2 is the networking and communication structure between Systems 1 and 3, enabling information flow and coordination of shared resources.
  • System 3 is a the set of policies and immediate supervisory bodies which deal with evolving “best practices” in System 1 activities and interface with higher systems.
  • System 4 gathers information on the wider context in which the organization exists, alert for feedback generated in response to the organization’s activities and changing conditions to which adaptation will be necessary.
  • System 5 coordinates and determines policy for the organization.

Beer’s successful application of Viable Systems Models to large corporations drew the attention from the Chilean government of Salvador Allende. They were intrigued by the possibility of developing a cybernetic management architecture for the entire, state-run economy of Allende’s democratic socialist government; in this way, a planned economy could be realized without the brutal coercion of Stalin or Mao. The feedback structures of cybernetic systems would, in theory, allow self-regulation within the context of a planned economy.

Beer helped plan and partially implement Project Cybersyn, the world’s first Cybernetic controlled economy. In 1971, he began wiring the State-controlled industries together into a neural network on a backbone of Telex machines.

Cybersyn consisted on four networked systems: Cybernet, Cyberstride, CHECO, and the Ops Room.

  • Cybernet was the network of Telex machines connecting state-run factories and systems across the nation.
  • Cyberstride was a system of economic and statistical models which accepted input from Cybernet – output, materials cost and supply, worker performance, and so on – and pushed alerts up and down the Cybersyn network. Cyberstride served as an automated operations manager, in effect, alerting the workers at state-run enterprises of potential issues and escalating those issues if they remained un-addressed after specified intervals.
  • CHECO was the overall model of the Chilean economy, the Chilean Economic Simulator. This processed information and alerts from Cyberstride and developed projections for use in policy development and economic forecasting.
  • The Ops Room was the heart of the Cybersyn network. Managers in the Ops Room, theoretically, could pull information from any other system in the network to develop policy, monitor performance, and take hands-on control in emergencies.

The partially implemented Cybersyn network proved its usefulness during a 1972 garbage workers’ strike, where 40,000 striking workers blocked access to the city of Santiago, triggering food shortages and service interruptions. Using Cybernet and the mostly-implemented Cyberstride, Ops Room personnel were able to feed and serve the entire population of Santiago with 200 hundred trucks in the field.

This incredible success spurred aggressive implementation of the rest of the Cybersyn system. By 1973, Cybersyn reached the advanced prototype stage, with complete implementation expected soon afterwards.

In the Fall of 1973, almost one year after Cybersyn’s remarkable trial by fire during the garbage strike, the Allende government fell to a military coup. The Junta established a brutal regime of authoritarian rule and the Cybersyn network was destroyed.

Why We Can’t Have Nice Things

Interestingly, given the incredible advances in network engineering, computer technology, software agents, and systems theory over the past 40 years, we absolutely could build the Cybersyn network today. In some respects, you’re using Cybernet right now. The financial sector’s high-frequency trading houses are busily wiring Cyberstride together from automated trading algorithms.

What we’re missing is our modern-day CHECO and an Ops Room. The former is possibly beyond our mathematical and modeling ability – look at our never-ending debate over the efficacy of tax cuts, for example – while the latter is only feasible if a centralized controlling body exists. While there are very tight consolidations of financial and political power in our modern world, there is no central governing body to occupy the Ops Room.

Honestly, that’s just as well – it’s difficult to imagine a scenario where a human-controlled Cybersyn network becomes anything but a system of brutal dystopian control.

While the formation of a planning, cybernetic economy on the Cybersyn model is arguably impossible, perhaps horrifying, and we lack the strong AI to pursue a Cultural path, the Robot books of Isaac Asimov may point the way forward for the development of a rational and enlightened planned economy.

As more organizations adopt networked, feedback-driven control structures, the computerized backbone of those structures become smarter, faster, and more ubiquitous. A just, cybernetic society may rise out of the emergent interplay between ourselves and our increasingly intelligent organizational and communications tools; a world ruled not by benevolent AI overlords, but through the shared assumptions coded into our software platforms and bureaucratic systems.