Evolution and control of large cybernetic systems

Analogy considerations between cybernetic systems allow important conclusions for the evolution and control of a social order. I deem it imperative for a social science of the future to extensively analyze existing functional systems. The insight into the fact that an optimum control system evolves by itself will govern the logical thinking of such a science. We have to create the conditions under which the society we want to live in can evolve. We must not try to create a control mechanism but we have to set the conditions under which the proper aspects can evolve. It is doubtful that organizations greedy for power will have the "guts" to initiate such an evolution and to draw potential consequences. In most of the states currently existing a reorganization of power structures will be the prerequisite for such a course.

The major problem of human society is that conscious systems begin to consider the society as their environment and actively influence it. Due to the high evolutionary speed, the problem is acute. It may also be existing in a few animal societies. But the control level "animal society" has its elements in a firm grip. Only in the further course of increase in evolutionary speed and the system´s gain in freedom will these encroachments into society become more successful.

And there is a second system making an assault upon society: the group. It also constitutes an autonomous system attempting to restructure the society according to its intents. Seen from this perspective, Rousseau´s demand to protect the society from group interests turns out to be highly topical again. Accordingly, the evolution of society seems to have been put at a disadvantage, compared to the systems which comprehensively sum it up to a new system. In order to successfully research the evolution of large cybernetic systems it is necessary to accelerate the search for evolutionary criteria as general as possible.

Larger cybernetic systems consist of a hierarchy of control loops, with each loop fulfilling a certain function. In the course of evolution, these individual functions are being built up in the system and reflect environmental conditions. Control loops of a more general nature are of a more recent date and access older lower levels as command variables. This means they don't control an output function; much rather, they provide the command variable affecting the adjustment of the output size. This can be observed in technical and biological systems. With current legislation, this simple fact is stubbornly ignored. It seems that word hasn't gotten around in politics yet about the truism that one should regulate that which is supposed to be regulated. Politicians consistently attempt to regularize, bar, and restrict seemingly negative events with the full force and power of the law. Existing balances are constantly ignored and/or fought against with utter enthusiasm. There seems to be no understanding and insight into the fact that this imposes an additional, mostly unnecessary control effort upon society.

The utilization of existing control loops and the setting of their nominal value is considerably more effective than stubborn regularization.

In the same way, fighting against societal phenomena for reasons of a religious, ideological, and so-called moral nature imposes a heavy burden upon society and individuals. And what for? Just for the purpose of restricting the freedom of individuals. Intolerance and exploitation are the true reasons behind the pretense that the intent is merely to protect people from something. In this way, self-protection capabilities are being reduced. Prices for illegal commodities (drugs, alcohol) are soaring. Customs duties, subsidies, and financial support is needed to conceal the consequences of these arbitrary encroachments. The pressure in the direction of restriction increases. Autonomous systems are evolving, undermining the restrictions, thus allowing the overall system to evade the course of events.

Only a free evolution can lead to a full evolvement of a system's performance capability and avoid ineffective, strenuous conflicts within the system. As it is the case with large systems in particular that the function analysis determines the efficiency of interferences into the control mechanisms, it shall be shown that only those interferences can be successful that utilize the adaptivity and function of the system. Opposing basic system functions is the most nonsensical method for the control of complex systems.

In the following, the criterion "economy" shall be addressed. The criterion of economy can be identified in any other criterion. It seems to be a component part of a general evolutionary criterion. The efficiency of an evolution relates directly to speed and material efforts. Not just the preference of effective systems, but also the autonomous drive of more highly organized systems towards comfortable paths is playing a role here. This laziness is also the product of an evolutionary criterion and highlights the major significance of the criterion "efficiency".

This shouldn't come as a surprise, as the potential reduction of physical systems takes the same direct and shortest path as well. This basic attribute of adapting to physical fields can possibly be found again in the adaption to informational fields at a higher level in the criteria "speed" and "efficiency".

The system is challenged with the task to solve all assignments with a high performance/effort quotient, that is, in an economical and efficient way. The optimization criterion is therefore clear and obvious. But how does the flow of information run? For each performance the system has to engage in the exact effort has to be determined. The performances thus are comprised of different efforts, each of which represents another performance in itself. For a totals formation which is supposed to satisfy the above-mentioned criterion, all efforts have to be correlated. A gauge is needed to assess each single effort.

Thus we are confronted with the task to assess the correlation between all performances on a value basis. This assessment process draws its information from a criterion. This criterion is the optimization criterion the entire process is subject to. It contains the value correlations between all partial performances. The optimization process now attempts to correlate the partial performances according to the criterion. The necessity of an assessment process thus ends in the confrontation of the partial performances and/or sub- systems which render performances.

Here we can already see that an even more general view of the term "system" could be a sensible approach. All performances rendered in a society are subject to an evolution. Programs are also systems processing information, independent of the hardware realizing or storing these systems. In an assessment process, a contest of performances takes place as well. Constructions, physical systems, have to be assessed. Furthermore, performances of any kind of have to be assessed. Even news, as performances of a purely informative character, require a correlation by value to other performances in the overall system.

Higher-level systems can deal with the contest with all active means at their disposal. The fight therefore is an assessment process. Just like in a fight, the assessment of all performances result in an abstraction which also is a demand of the criterion efficiency. A contest carried out via information processing requires considerably less energy and material than a contest carried out at the physical level. However, this fight can only take place at a new system level. This is also the reason why, in the case of collapse or dysfunctions at this system level, a backslide into physical fight takes place. As the ways and means available to a system at this level do not necessarily have to correspond to the physical ways and means of the system, there is always the danger of such a backslide whenever a sub-system feels underrated.

The necessity of the abstraction process leads to a new quality of the contest and its assessment.

The established value is expressed in the price which has to be paid for this particular performance to other performers. It is the basis for the inception of a monetary system, an essential step in the abstraction process. The money is subjected to the same assessment process as all other objects. Money turns into a commodity. Just like a commodity, its value is relative. It is a more suitable gauge than any other commodity because it is not subject to larger fluctuations. In order to eliminate influencing variables, abstraction went on from coinage to paper money, and it will continue until all that is left his numbers in a computer. There is no necessity for the physical existence of a gauge; in the case of dysfunctions, a financial system can resort to previous states, for instance a barter system.

As money can't be used for anything else than exchange, its value only depends upon the concentration on the market. However, subjective factors affect the price of any commodity. They are part of the criterion. Naturally, expected changes in concentration have a negative effect upon the stability. The demands "reaction rate" and "stability" cannot be both met at the same time. The faster a system reacts to influences the more it tends to instabilities and dysfunctions. This leads to the task to adapt one's own reaction rate to the respective requirements. The fixing of interest rates has to be oriented to the accrual of the total value of all available performances in a society so as to exclude an additional disruptive factor.

If the market, the assessment process, is in balance no value gradient exists; in that case, any barter action ending at the object of barter it originally started from, via any number of stations in a short time span, must result in the same amount of the object of barter. The assessment depends upon time and place. A dynamic balance has come into existence. This does not take the effort applied for the barter process into consideration.

In a dynamic system, pressure creates counter-pressure - a fact almost unknown to politicians. Accordingly, that is what laws look like: bans, prohibitions, demands, penalties. What is needed here is an insight into the dynamic correlations. A consistent attempt is undertaken to control society with harsh and strict laws. The attempt to interfere into economical processes with violent measures of this kind is even more futile. The dynamic system will try to evade the pressure. If it is being deprived of all opportunities to do so, it is also being deprived of its performance capabilities - something that can lead all the way to the paralysis of an entire national economy.

This would look different if the existing balances would be utilized. The ever-present market, the balance between effort and benefit, provides the basis.

For instance, a supplementary tax could be levied on processes harmful to the environment, because the society has to carry the resulting burden. However, a fixed mechanical tax doesn't pay. The tax has to be oriented to the effective damage - for all processes alike! It has to be taken care, though, of a continuous adjustment to the ever-changing conditions so that no additional, unnecessary, and therefore harmful distortion of the criterion takes place! Only then will those processes with less resulting damage be preferred, and processes with a high potential of environmental damage will only be implemented when it comes to socially important tasks capable of realizing corresponding prices.

The preferential treatment of individual processes would also be wrong in this case, because the formation of new processes is not being promoted from the beginning. Consequently, society has to superimpose its demands onto the existing conditions. Whatever society considers to be harmful has to be charged with a debit to the polluter/waste producer. Some sort of "negative feedback" has to be arranged for those processes and/or types of behavior acting as a burden for society. This brings about stability for the overall system. The stability cannot be attained with other processes that suffer, in part, from regenerative feedback. In short: exertion of positive influence upon a process, for instance a subsidy, places all other processes at a disadvantage, including those which were not yet capable of evolvement. Exertion of negative influence, for instance a tax, does not exert an influence upon the balance of processes not affected.

Perhaps one day it will be possible to simulate societal processes on large computers. I personally deem it possible to examine societal processes in this way, and to arrive at statements about general criteria. In addition, an examination of the stability of large systems and the stability of system/environment relationships could be conducted in this way.

The complexity of technical systems will increase further in the future. Along this course, we can assume that systems will come into existence soon which can generate simple technical systems. Computer-aided drafts of control systems, or construction per selection is something which is already common practice today. In addition, programs are already in existence that are capable of tactical and strategic considerations, and which can be produced applying the development approach.

Looking at the future, it can be assumed that complex programs cannot be programmed directly any longer because it will become impossible to manage an analytical compilation and implementation of all existing conditions. In particular, flexibility and adaptability are hard to project into a system. It could be an advantage to utilize the natural process of evolution on a large scale. Construction via selection is already common practice. Why should these possibilities not be an aid to systems processing information so as to attain a new quality level? The experiences gained in this way would be urgently needed in order to resolve societal problems.

What is needed is a science addressing the control of large systems. So far, cybernetics has restricted itself too strongly to controls and regulations, research, and mathematical descriptions of sub-systems. It was only interested in the "how?" What was missing out from my viewpoint was the answer to the question why something works the way it does, and why it has evolved in this particular way. These questions, however, are of major significance for an evolutionary theory. An extension of cybernetics to evolutionary processes would be a sensible approach. I understand cybernetics to be the philosophy of natural sciences. Meanwhile, the major significance of information flows should be beyond dispute.

Applied in this way, cybernetics could introduce a standardized and unified approach, oriented to information flows, to the natural sciences and act as link between physics, biology, and sociology. The evolutionary path is the only sensible approach to improve a control system. Complex structures cannot be erected in a short amount of time, even if all the required information is available. Large time constants are accountable for the fact that the system cannot swing abruptly into a stable operational state. The "revolutionary method" which is often understood to mean that the old has to be smashed so as to create something new in its place bears major disadvantages. Revolution means overthrow - and not annihilation! And it should be understood in this way because otherwise the established order of the system is being ignored. If the operation would have a positive outcome significantly larger amounts of information than usual would have to be managed. With regard to stability and system start-up, enormous difficulties can be expected as well. For these reasons, an evolutionary redesign of large cybernetic systems would be much more sensible. Natural independent processes are to be utilized.

A system is not located in a stable environment for the entirety of its lifetime. It can happen to any system to be confronted with sudden, unpredictable environmental changes, catastrophes, or assaults from other systems. For emergencies of this kind, each system needs control mechanisms enabling the system to survive in such cases. The system must be enabled to focus all available forces on the most important tasks. The structure must be memory stored in a reliable manner for times after this emergency. The balances of power have to allow for a return to this structure. Biological systems take these facts into account. For instance, they have escape reactions at their disposal. There is a so-called critical reaction enabling an animal under pressure to attack a dreaded enemy. Control functions have to be available for rare situations as well. Consequently, a society should establish functions for potential cases of need enabling survival in crisis situations. However, the concentration of power associated with this poses a danger to society. A notorious example is the apparent (but actually non-existent) threat to the survival of a nation evoked by its military dictatorship merely serving as justification for emergency laws and even more restrictive legislation. The arbitrary measures which then become possible sensitize the population and thus create a real internal threat because opposition is the only way out. Concentrations of power for such purposes have to be strictly separated from the customary control structure.

The concentration of power has to be temporally restricted from the outset or via periodic polls. If such a system succeeds in making itself independent then that particular society is doomed. The problem of how power can be concentrated and yet remain in the hands of the people has to be resolved in a reliable manner, first of all.

The apparatus wielding power in the case of emergency laws coming into effect must not be in a position where it can bring about the situation by itself. The urge to extend its sphere of influence is native and inherent to every active system. It is part of a general evolutionary criterion. Officials and functionaries carry this feature into the apparatus; the apparatus does not have to acquire it first. The people building it up project their acquired features into the system which they consider their instrument. Consequently, the selection criteria for executives can be found in the systems these persons are building up. The necessity of a strict separation between build-up of a control apparatus and its use becomes clear and obvious here. The demand for separation of powers in a society can easily be substantiated with this.

Therefore, the selection of the criterion establishing the apparatus will be responsible for its future appearance. This selection thus is the original source of information, the original criterion. If the society is supposed to be there for the people and not vice versa we have to arrive at the conclusion that only the population can establish the criterion for the evolution of the state apparatus and its behavior in cases of emergency. Only the common will of the people can determine the societal development if that society is not to fail due to internal frictions.