After I wrote the post on Red Star, the first Bolshevik Utopia,https://nightsoflabour.wordpress.com/2010/03/06/mining-potentials-in-capitalism-thought-experiments-in-the-communist-laboratory/

brilliant Owen Hatherley, the author of Militant Modernism, pointed out to me a project called Cybersyn, from the former socialist government in Chile in the early 1970s. I regret that I did not know before this project (http://en.wikipedia.org/wiki/Project_Cybersyn), which was a real-time, computer controlled planning system, coordinating 500 nationalised factories in Chile. I then read the PhD thesis of Eden Miller submitted to MIT in 2005, which follows the traces of the project, the engineers who set it up, Chilean history in the 1970s. The thesis is a history of technology thesis, yet gives brilliant insights regarding possible models and designs which can be put at the service of a socialist practice…which is exactly what the communist laboratory is for. In this post I want to summarise the findings of Dr. Miller, which is the unique authoritative source on the subject and interpret them in my own way, by also offering some comparative outlook with Red Star’s statistical system of labour allocation.

Cybersyn originates in the work of a British professor called Stafford Beer who developed cybernetic ideas during the 1960s. A group of engineers was influenced by those ideas and when they took a position in Allende’s government invited Beer to set up a computerized system to coordinate 500 factories in Chile with the aims to increase worker participation and achieve efficiency at the shop floor. When Beer came to Chile, there was only one computer, which would be the main government computer at the center. The Cybersyn team used a telex network, which had been used to track satellites. Even though telex networks were only able to transmit ASCII characters, they were based on a high speed of information exchange similar to the Internet.

The first component of the system, Cybernet used the existing telex network in order to make every factory communicate with the main computer. Production information was sent from the factories to a telex control room where employees transferred the data onto punch cards and fed them into the mainframe computer for processing. The system was designed as a real-time economic control, similar to the Red Star’s statistical agencies, but in practice data was transmitted once a day. Telex networks are a brilliant bricoleur’s solution to the lack of available technology in the context of a developing country, which is subject to political hostility by imperialist powers. It shows how available resources can be creatively used for new purposes by change-driven individuals in a change-inducing environment.

The second component of the Cybersyn system, Cyberstride produced quantitative flow charts of activities within each factory. It did ‘statistical filtration on the numbers output from the factory models, discarding the data that fell within the acceptable system parameters and directing the information deemed important upward to the next level of management’ (Miller, 2005). The software developed certain methods to identify production trends and if a variable fell outside of the range determined by the system, the system made a warning, which was called an ‘algedonic signal’. The relevant person from the factory emitting the signal was given time and freedom to solve the problem. In case he was unsuccessful, the central management had the right to interfere. This principle was to preserve the autonomy of the lower-levels managers.

According to Miller, Cyberstride could be seen as an instrument to predict and map the behaviour of Chilean factories. The algorithm was based on Bayesian probability theory to foresee industrial performance. Since the government could rely on possible trends, it could intervene in advance. Beer wanted to keep the problem signals as simple as possible: sources of energy, raw materials, worker satisfaction present on a given day.

In the case of Red Star, information coming from the factories was geared to send signals about shortage and supply of labour. As long as the labour needs of factories were determined as a result of changes of technology…etc. the individual workers could make their own choices accordingly. Individual choices would be both the cause and consequence of the relationship between demand and supply. However in the case of Cybersyn, the objective is far more ambitious: One needs to determine variables to control, detect anomalies (whose degree has to be determined as well). However, similar to Red Star’s statistical agencies collecting data, Cyberstride was performative. Moreover, Cyberstride is much clearer about how this data will be collected and interpreted: algorithms transform the data into inputs, which are supposed to send signals regarding production decisions.

The third part of the Cybersyn project, CHECO (Chilean Economy) aimed at providing simulations of future economic behaviour. CHECO was designed to create the tools for planning and flexibility. But it was not really put into practice.

The fourth component of the system, Opsroom, created a new environment for decision making, which modeled after a British World War II War room. It consisted of seven chairs in an inward facing circle and a series of projection screens, each of them displaying the data collected from the nationalized (Miller, 2005). To enable individuals with minimal scientific training to understand the information, all industries were standardized with a uniform system of iconic representation. One of the walls in the room contained four screens, which represented structural information. The large screen contained instructions for changing the images displayed below, a mix of flow diagrams, factory photographs and unitless mappings of actual and potential production capacities. Two screens recorded algedonic signals on another wall, which represented overall production trends and listed urgent problems which required government intervention. This was based on Beer’s metaphor of the economy as a biological organism: Brain (government) intervening from the Opsroom in case problems in the lungs (factories) are not solved.

Miller shows, in her thesis, how ideas similar to Cybersyn were implemented in Soviet Union after several years of the publication of Red Star, when the utopia of Bolshevik Revolution became a reality. For instance, a complex three-tiered computer network that would use thousands of local computer centers collecting primary information was built. Those local centers would be linked to 30 to 50 computer centers in major cities and then all information would travel to one central government in Moscow. It was realized that this scheme would require the manipulation of fifty million variables more than the three thousand variables required to manage the Chilean economy with Cybersyn. Soviet economists tries to simplify this problem by using indirect centralization in which the state was responsible for determining optimal prices and efficiency levels and companies were allowed to make their own decisions. According to Miller, this solution seems very similar to Cybersyn, but there were important differences. In its infancy, the Chilean planning was more receptive to new ideas and challenges. The state development agency CORFO underwent a series of changes that allowed it to grow in different areas (public and mixed property) and to increase new management capabilities (such as adding of new layers of bureaucracy to pre-existing administrative hierarchy) In the Soviet context, factory autonomy contradicted the Soviet economic theory and threatened the centralised power of the State. Rather than being instruments of change, computer technology strengthened centralised control in Soviet Union. In other words, I would argue, it was not simply the combination of creative ideas of a crazy British professor and the dedication of a socialist team, which was behind the transformative capacity and efficiency of Cybersyn. Soviet Union had much more capacity to enact those. Rather, it was being open to new things, a critique of centralized planning of already existing socialist experiences and a very strict emphasis on worker participation which pushed individuals to think differently while playing with similar instruments. A different mind set made a difference in Chile as compared to Soviet Union.

Chilean system aimed to improve the efficacy of human interventions in the factory system. Automated functions were aimed to increase participation and intervention. The emphasis on worker control and participation was a recurrent theme in the debates on and implementation of Cybersyn. In the words of the socialist engineer Flores working for Allende:

‘For me what socialism meant is how do you combine autonomy of individuals with community? The classical Marxist idea, they call democratic centralism. But my impression at the beginning was that this was pure lip service and nothing concrete. That’s what brought Beer to me. He found this relationship between autonomy of the unit with the intelligence of the whole. And in that sense there was a connection. I never found anything like it in any other place’ (Miller, 2005).

Overcoming of the labour alienation was a major concern for both Beer and the Chilean socialists. Beer aimed at increasing the worker’s capacity to contribute both physically and mentally to production process. The form of worker participation was a hotly debated issue in Chile. The Christian Democrats advocated the creation of “worker’s enterprises” (empresas de trabajadores) inspired by the Yugoslavia experience: workers would share ownership of their firms and split the profits-an approach that provoked disdain from Allende. According to the president, making the workers “shareholders would be to convert them into pseudo-capitalists’ (Miller, 2005).

Beer envisaged simplified versions of the Operations Room within nationalized industries by forming a space for workers to meet, monitor production and contribute to the decision-making process. Those rooms used chalk on a blackboard and the methods of collecting and presenting information developed by the Cybersyn team. Beer wanted workers to contribute their tacit practical knowledge of production activities to the process of modeling their factories  (Medina, 2005). In this plan workers would assist operations research scientists and engineers in order to choose the key indicators and thresholds of production. Beer later invited workers to take a greater role in this process. ‘There is no-one better qualified to model a plant, than the man whose life is spent working in it.’ Beer had confidence in workers’a ability to master the tools designed by the team (software, simulators, flowcharts and production indices associated with the system) (Miller, 2005).

The emphasis of worker participation has to be taken with caution, however, for at least two reasons. First, tools and instruments we are discussing here are powerful for emancipatory purposes since they have no inherent relation to capitalist society. Yet, this strength, to reiterate, is also their weakness. If they can be used for emancipation, they can also be used for oppression. That is why there emerged several warnings about the possibility to use Cybersyn as a mechanism to oppress workers, by imposing repressive controls and exerting pressure and sanctions. There were even talks about to what extent Cybersyn was a version of Big Brother.

Second and more importantly, while criticizing alienated labour, Marx did not necessarily envisage a new society where workers will take part in every stage of production, management and marketing of goods and services. The ultimate goal was to reduce working hours so that workers can have more time for other activities including fishing, philosophizing, painting and hunting (see also Paul Lafargue’s work). In other words having control over production is not the opposite of alienation. It is a first step of a broader process where increased automation of several human functions can decrease factory time.

From a very positive yet still critical glance, what do Cybersyn and Red Star tell us in common? First, both of them have the common feature of creating a system, a network, which transmits signals, like a neuron system. The channels are very well connected so that in each specific locality individuals can make their decisions (workers’ decisions to choose their own jobs for the Red Star, factory management and workers to make decisions for production for Cybersyn), but not independent of the central brain (supply and demand functions for Red Star, central government computer for Cybersyn).

Second, for both systems to work for communist goals, there needs a set of conditions, which are discernible from the building of the system. In other words design of communist objects at each stage makes us ask a number of questions. For instance, in the case of allocation of labour, it is apparent that the ability of individuals to have diverse skills is a must for this system to work properly. So a system of education is a pre-requisite to enable individuals to be equipped with necessary skills so that they can transfer their labour between different sectors.

Third, both systems use government and market instruments for their own purposes. In the case of Red Star, statistical agencies work both like a government and market institution. The supply and demand functions emerge from the actions of individuals, so one can talk about a labour market, but which does not behave as a capitalist market. In the case of Cybersyn, a form of decentralized planning for the nationalized industries is implemented, in a way to give government the control of last resort, yet trying to solve the Hayekian problem of knowledge dispersion.

Fourth, in both systems information technologies are put at the service of solving the problems of freedom, autonomy and control. In the case of the Cybersyn, the factory is allowed to solve the problem in a given amount of time. Only after this time period elapses, the government has the right to interfere. In the case of Red Star, the system is more impersonal. Individuals, by simply making choices contribute to the emergence of a supply and demand function, which then gives signals regarding shortage and supply of labour. Individuals meet collective needs by acting according to their wishes. Algedonic signals in the case of Cybersyn and the changing numbers on factory tables sent by the statistical agency in the case of Red Star make those solutions possible.

Yet, technology was almost absent in the period of Red Star and still limited in the time of Cybersyn. Today we have amazing software programmes used by the big corporations and governments. Think that, for one moment, those programmes are aimed at doing what Cybersyn and Red Star aim at doing. Computerised systems and softwares are not simple technological tools to speed up or to replace human labour. They are important to amplify and make more efficient parallel thinking processes to develop solutions to complex problems. 

Eden Miller (2005) Politics, Ideology and Computation in Chile, 1964-1973, Unpublished PhD Thesis, MIT.

Advertisements