In the natural science of the first half of our century, physics was the leading direction. Since the 1950s, along with physics, chemistry, and biology, cybernetics began to exert increasing importance and influence on the development of science and the whole way of our life. Cybernetics is becoming the most important factor of the scientific and technical revolution at the highest stages of its development.
Cybernetics arose at the junction of many areas of knowledge of mathematics, logic, semiotics, biology and sociology.
The generalizing nature of cybernetic ideas and methods brings together the science of control, which is cybernetics, with philosophy.
The task of substantiating the initial concepts of cybernetics, especially such as information, control, feedback, etc., requires an exit to a wider, philosophical field of knowledge, where the attributes of matter, general properties of motion, and laws of cognition are considered.
Cybernetics itself as a management science gives a lot to modern philosophical thinking. It allows you to more deeply uncover the mechanism of self-organization of matter, enriches the content of the category of relationships, causality, allows you to study in more detail the dialectic of necessity and chance, possibility and reality. The way is open for the development of “cybernetic” epistemology, which does not replace dialectical materialism with the theory of knowledge, but allows to clarify a number of important problems in the light of the science of management.
Having arisen as a result of the development and mutual stimulation of a number of, in the recent past, weakly interconnected disciplines of a technical, biological, and social profile, cybernetics has penetrated many spheres of life.
Such an unusual “biography” of cybernetics is explained by a number of reasons, among which there are two.
First, cybernetics has an extraordinary, synthetic character. In this regard, there are still differences in the interpretation of some of its problems and concepts.
Secondly, the fundamental ideas of cybernetics came to our country from the West, where they were from the very beginning influenced by idealism and metaphysics, and sometimes ideology. The same, or almost the same thing happened with us. Thus it becomes obvious the need to develop the philosophical foundations of cybernetics, the coverage of its main provisions from the position of philosophical knowledge.
Understanding of cybernetic concepts from the position of philosophy will contribute to more successful implementation of theoretical and practical work in this area, will create better conditions for effective work and scientific research in this field of knowledge.
Cybernetics as a promising area of scientific knowledge attracts increasing attention of philosophers. The positions and conclusions of cybernetics are included in their fields of knowledge, which largely determine the development of the modern theory of knowledge. As rightly noted by domestic researchers, cybernetics, the achievement of which is of great importance for the study of the cognitive process, in its essence and content should be part of the theory of knowledge.
The study of the methodological and epistemological aspects of cybernetics contributes to the solution of many philosophical problems. Among them are the problems of a dialectical understanding of simple and complex, quantity and quality, necessity and chance, possibility and reality, discontinuity and continuity, part and whole.
For the development of mathematics and cybernetics themselves, the application to the material of these sciences of a number of fundamental philosophical principles and concepts, an application that necessarily takes into account the specifics of the relevant fields of scientific knowledge, is of great importance. Among these principles and concepts, the position of reflection, the principle of the material unity of the world concrete and abstract, quantity and quality, a normal and meaningful approach to knowledge, etc. should be emphasized.
Philosophical thought has already done a lot in analyzing the aspects and the theoretical-cognitive role of cybernetics. It was shown how philosophically promising to consider in the light of cybernetics such questions and concepts as the nature of information, purpose and focus, the ratio of determinism and theology, the ratio of discrete and continuous, deterministic and probabilistic approach to science.
It should be said about the great value of cybernetics for the construction of a scientific picture of the world. Actually the subject of cybernetics – the processes occurring in control systems, the general laws of such processes.
Phenomena that are displayed in such fundamental concepts of cybernetics as information and control, take place in organic nature and social life. Thus, cybernetics can be defined as the science of control and communication with wildlife in society and technology.
One of the most important questions around which philosophical discussions are going on is the question of what is information, what is its nature? To characterize the nature of information processes, it is necessary to briefly consider the natural basis of any information, and such a natural basis of information is the intrinsic property of reflection inherent in matter.
The provision on the indissoluble connection of information and reflection has become one of the most important in the study of information and information processes and is recognized by an absolute majority of Russian philosophers.
Information in nature, in contrast to non-living, plays an active role, as it participates in the management of all life processes.
The materialistic theory of reflection sees the solution of new problems of science and, in particular, such a cardinal problem of natural science as a transition from inorganic to organic matter, in the use of the methodological basis of dialectical materialism. The problem is that there is a matter capable of sensing, and matter created from the same atoms and at the same time not possessing this ability. The question, therefore, is posed quite concretely and, thereby, pushes the problem to a solution. Cybernetics is closely involved in the study of mechanisms of self-regulation and self-government. At the same time, while remaining methodically limited, these achievements left open a number of problems, which the internal breaking of cybernetics led to.
Consciousness is not so much a product of the development of nature as a product of human social life, social labor of previous generations of people. It is an essential part of human activity through which human nature is created and cannot be accepted outside of this nature.
If in machines and in general in inorganic nature the reflection is a passive, dead physico-chemical, mechanical act without generalization and penetration into the essence of the generalized phenomenon, then there is a reflection in the form of consciousness, then, according to F. Engels, “knowledge of highly organized matter itself, penetration into the essence , the law of the development of nature, objects and phenomena of the objective world “.
In the car, the reflection is not consciously, since it is carried out without the formation of ideal images and concepts, but occurs in the form of electrical impulses, signals, etc. Since the machine does not think, this is not the form of reflection, which takes place in the process of human cognition of the surrounding world. The regularities of the reflection process in the car are determined, first of all, by the regularities of the reflection of reality in human consciousness, since the person creates the machine for a more accurate reflection of reality, and not the machine itself reflects reality, but the person reflects it with the help of a machine. Therefore, the reflection of reality by the machine is an integral element of the reflection of reality by man. The emergence of cybernetic devices leads to the emergence not of a new form of reflection, but of a new link mediating the reflection of nature by man.
Commonality of thinking with the ability of reflection is the objective basis for modeling the processes of thinking. Thinking is associated with the creation, transmission and transformation of information, and these processes can occur not only in the brain, but in other systems, such as computers. Cybernetics, establishing a relationship between reflection, sensation and even thinking, takes a certain step forward in solving the problem posed. This relationship between thinking and other properties of matter arises from two fundamental principles of the materialist dialectic of the principle of the material unity of the world and the principle of development. However, one cannot absolutize or deny this relationship. Thinking is human quality and different from cybernetic.
Despite the qualitative difference between the machine and the brain, there are general laws in their functions (in the field of communication, control, and control), which cybernetics studies. But this analogy between the activity of the automatic and the nervous system, even in terms of information processing, is relatively arbitrary and cannot be absolutized. And in this regard, it should be noted that some studies on cybernetics, especially those performed in the initial period of its development, were characterized by mechanistic and metaphysical trends, although in appearance they appeared to appear diametrically opposed.
There was a failure to take into account the qualitative differences between inanimate matter and the thinking brain, erased every line between the knowing subject and the object of the material world. Since modern computers are universal and capable of performing a number of logical functions, it was argued that there is no reason not to recognize this activity as intellectual. It was allowed to create an artificial intelligence or a machine that would be “smarter” than its creator.
There were raised other issues related to the possibility of such a machine. Can a car completely, in all respects, replace a person? Are there any limits to the development of cybernetic devices? Of course, these questions have not lost their relevance. It would be premature to write them off into the archive of loosely posed questions, because the line of conflict between various philosophical schools, materialism and idealism, concerning the main issue of philosophy, runs through them.
In other words, this is one of the aspects of the modern historical form of the main issue: the essence of human consciousness and its relation to the functioning of cybernetic devices.
Currently, there is a discussion about the prospects for the development of cybernetic machines and their relationship with the human mind.
To create a machine that functions as a brain, it is necessary to create a substance with properties or the like of highly organized protein matter, which forms the brain. Indeed, such a machine will function “like a brain,” but it will function, not think. In order to think, matter must exist not only in economic but also in social form. A replacement of inorganic content with organic does not, and moreover, as a result of this replacement, one of the main advantages of an electronic machine will lose its speed.
Considering the possibility of creating artificially, on the basis of modeling, a thinking being, it is necessary to dwell on two aspects of this problem.
First, cybernetics does not model all brain functions, but only those related to receiving, processing and issuing information, i.e. functions that are amenable to logical processing. Yet the other, infinitely diverse functions of the human brain remain outside the field of vision of cybernetics.
Secondly, from the point of view of the modeling theory, it makes no sense at all to speak of the complete identity of the model and the original.
The identification of human and “machine” mind occurs when the subject of thinking is replaced by some material system capable of reflecting. The only subject of thought is a man armed with all the means at his disposal at this level of his development. These tools include cybernetic machines, in which the results of human labor are materialized. And, like any instrument of production, cybernetics continues and enhances the capabilities of the human brain. A person will transfer to the machine only some functions performed by him in the process of thinking. Thinking itself as a spiritual production, the creation of scientific concepts, theories, ideas, which reflect the laws of the objective world, will remain with the person.
Until now, the dialectical materialist understanding of thinking has relied mainly on generalized data from psychology, physiology, and linguistics. These cybernetics allow us to raise the question of a more specific understanding of thinking.
Cybernetics does not aim to “replace” a person or “replace” his thinking. It only gives new arguments in favor of the dialectical materialist idea of the machine – the human assistant.
Cybernetics leads to the materialistic conclusion that when deciding the question of fundamental and real issues of machine modeling of thinking processes, one should first of all take into account the social determination of thinking, consciousness, and mental life of a person.
Modeling as a research method is characterized by mediated practical or theoretical study of the object. In this case, it is not the object that is studied, but an auxiliary artificial or natural system, which is in objective accordance with the object under study, capable of replacing it in a certain respect and giving information on the modeled object itself during its study.
From the epistemological point of view, the essence of modeling lies in the indirect knowledge of the object of interest to us, i.e. by model we judge some properties of the original. With the help of modeling, new phenomena are learned on the basis of those already studied. The cybernetic approach means the modeling of the processes of human intellectual activity from one particular side, namely, at the level of elementary information processing processes.
The nature of thinking, the mystery of consciousness, the mystery of the mind, all this is undoubtedly one of the most exciting human problems. The popularity of cybernetics, the unrelenting interest in it from the widest circles, is largely due to its close connection with this “eternal” problem. From the very moment that man began to think about the problem of thinking, in approaching it there are two main diametrically opposite directions: materialism and idealism. Idealism proceeds from the recognition of thinking as a certain special entity, radically different from matter, from all that we are dealing with in the external world. Materialism, on the contrary, asserts that “… that material world, which we sense perceptible to us, which we ourselves belong to, is the only real world and our consciousness and thinking, however supersensitive it may seem, are the product of a material, bodily organ of the brain [one].
This basic thesis of materialism in the interpretation of thinking receives from cybernetics a new (and in a certain sense decisive) proof. The essence of the matter is as follows. Since its inception, natural science has provided a continuously growing argument in favor of the materialistic concept of thinking. The data of physiology, evolutionary biology, psychology from the most diverse aspects substantiated the thesis of materialism. But all these data deal with one object – the brain, with its inherent ability of thinking, which is already in a ready-made form given by nature.
There is always a “loophole” for an idealistic doubt about the thesis that the brain is an organ of thought.
The American psychologist and pragmatist philosopher W. James at the end of the last century tried to substantiate this point with the greatest clarity. James does not dispute a single statement of physiology that establishes a link between the processes that we subjectively recognize as thinking, and the material processes that occur in the brain. But (and this is the meaning of James’s arguments) from a logical point of view, this connection does not mean that the brain is an organ of thought; any data of physiology proves only the existence of conformity and nothing more.
Ultimately, the ultimate judge of scientific concepts is practice. “If we can prove the correctness of our understanding of this phenomenon of nature by the fact that we ourselves produce it, call it from its conditions, force it to serve our purposes, then the Kantian elusive” thing in itself “comes to an end.” [2] This argument of artificial reproduction was absent in traditional philosophy and cybernetics gives it, regardless of the outcome of the debate about the possibility of creating artificial intelligence comparable to human intelligence.
On the basis of what has already been achieved, it can be argued that a number of thinking functions, previously considered the exclusive possession of the living brain, are artificially reproduced by cybernetic devices. This is the great importance of the philosophical result of cybernetics, which can be stated today.
So, the concrete-scientific substantiation of the materialistic concept of thinking, the practical proof that thinking is a function of a highly organized material system is the most important philosophical conquest of cybernetics. But cybernetics goes further and poses the question, with which we fall into the abyss of disputes, the question of the possibility of “artificial intelligence”, “machine thinking”, “cybernetic mind”, etc. It reveals the full range of views, ranging from “extremely optimistic” to “extremely pessimistic” about the possibility of thinking machines. The argument in favor of the pessimistic view is usually twofold: either the authors proceed from the particular substantive nature of thinking, or from its particular qualitative specificity. The truth is not entirely clear, what distinguishes the first from the second.
It seems the most reasonable position, which can be called “moderately optimistic”: not today there are no insurmountable, fundamental barriers to the creation of artificial devices with intelligence. But on this path there are enormous difficulties, which are by no means diminishing with the rapid development of cybernetics (for example, machine translation), although 10 years ago most of the experts drew the most optimistic prospects for the very near future; but the task turned out to be much more complicated than it seemed at first.
In addition, there is no reason to believe that insurmountable obstacles will not appear in the future.
Our knowledge includes both a set of scientific theories and empirical information, as well as general philosophical principles. From the available scientific theories and empirical data, “extremely pessimistic” conclusion does not follow. Arguments against the possibility of artificial intelligence, based on existing scientific theories and empirical data, can be called “specific” arguments. Usually they consist in pointing out any specific actions of thinking that no cybernetic device can perform. However, all such arguments were refuted in the course of cybernetics development. Moreover, there is the McCulloch Pits theorem, which reduces the question of the performance of any brain function to the question of the knowability of this function. Without becoming agnostic, it’s hard to be adherent to “concrete” arguments.
The idea of artificial intelligence is often declared mechanistic on the grounds that the work of a computer is governed by the laws of electrodynamics, and this means that the higher (thinking) is reduced to the lower (physical processes in a computer). However, the original premise is incorrect.
Computer operation is not governed by the laws of electrodynamics. These laws govern the operation of individual elements of the machine. According to the physical laws of the computer works only in the sense that, for example, it converts electrical energy into heat. After all, the essence of the work is not in this transformation, but in the fact that it performs certain arithmetic-logical operations. The machine deals with information and works according to the laws of information transformation, i.e. according to the laws of cybernetics. Therefore, if we consider these processes from the position of the mechanism, you inevitably find yourself in the positions of mechanism, since
there is a reduction of more complex processes of information processing to more simple. This is the same as saying that the work of the brain is reduced to biochemical and biophysical processes. In fact, these processes occur at the level of nerve cells, and at the level of information processing processes, there are other laws, the laws of which are not equivalent at all.
From this point of view, the work of a computer should be regarded as the work of a system for processing information. The thesis of artificial intelligence is also attributed to the rejection of the ideal nature of consciousness and the accusation of vulgar materialism. It can be shown that it is not. Without addressing the question of the structure of information, which is a measure of the ordering of the process and constituting its internal wealth, we will characterize the external or relative information, always related to the ratio of the two processes. Let there be processes A and B with the set in some way of the ordered states A1 … An and B1 … Bn. If each Bi corresponds to a certain Bi and the relationship between states A is isomorphic to states B, then we can say that process B carries information about process A. This information is not in B or A, but it exists in relation to these processes to each other . Taken by itself, this information is as objective and material as any other properties and relations of objects or processes.
Now take the many states of our brain in the process of functioning. The brain reflects the external world, which means that there is a correspondence between the set of states of the brain elements and the set of states of external processes, i.e. the brain has information about external processes. This information is enclosed and not enclosed in the brain, because no matter how much we explore the brain, except for the electrical, chemical, and other characteristics of the neurons, we will not find anything there. It is necessary to consider the connection of the brain with the outside world. It is in this and enclosed information, the carrier of which are neurons.
The information with which the brain works is the ideal side in its work, and thus the ideal does not exist as a special object or substance. It exists as a side of brain activity, which consists in establishing links between a variety of states of the external world and the brain. Ideal information of the human brain is basically the same character as relative information in general.
At a certain stage of the historical development of matter, there was a qualitative leap, as a result of which information, becoming a property of the brain, acquired the character of ideal information. If we recognize in cybernetic systems the possibility of achieving complexity comparable to the complexity of the brain, then it is necessary to recognize the existence of features in such systems that we call ideal.
A number of authors declare the thesis of artificial intelligence contrary to the thesis of the social nature of consciousness and thinking. But here lies the mistake of the lack of distinction between the naturally historical origin of thinking and the conscious reproduction of it by man in a universal computer. In the second case, the machine does not become a social being, but the person, having understood the essence of thinking, recreates him in the machine. If the social nature of thinking is natural and cognizable, then it can in principle be artificially reproduced.
Man, besides, is not only a natural being, his main characteristics are a product of social, and not purely biological, development. This means that a person’s thinking cannot develop in isolation, for this it is necessary that a person be included in society.
First, for the emergence of thinking, you must have a language, which is possible only in society. Secondly, from a cybernetic point of view, the “rationality” of a machine is determined by the amount of information processed, therefore even a powerful system that has got into the information-poor environment cannot become sufficiently “reasonable”. A vivid example is children who grew up outside of society, for example, in the forest. For a person, the necessary condition for his development was functioning in society, since society in its informational parameters is an extremely rich medium.
All this makes it possible to understand that the thesis about the social nature of thinking does not contradict the thesis about artificial intelligence. A cybernetic system that has sufficient power to fully use its capabilities should be placed in an information-rich environment, forming together with the creators a kind of symbiosis, called “integral intelligence”. The principle of impossibility of cybernetic intelligence rigidly binds a certain kind of functioning to a strictly defined substrate (brain). This poses the philosophical problem of the relationship between function and substrate.
Philosophical analysis of the trends of modern scientific knowledge makes it very unlikely (but does not exclude) the conclusion about the rigid attachment of thinking to the brain. It is because of this that the “extreme pessimist” denies the possibility of the presence of intelligence in a cybernetic device.
He unconditionally connects thinking with one, strictly defined substrate – the human brain, and does not accept attempts to define thinking without connection with the structure of the thinking system. In his opinion, this is a reduction of thinking only to the information side, while thinking is the ability that has arisen in biological creatures. Thus, thinking can only be called something, then carried out only by the human brain, but this is not an acceptable solution to the problem.
Of course, thinking is a function of highly organized matter and is determined by the structure of the system. But from the epistemological point of view, knowledge of a function is derived from knowledge of the structure, and knowledge of the structure is a conclusion from an ever more complete study of the ways of functioning.
If we imagine many different systems that perform the function of thinking, then it is the identification of the invariant aspect of these systems that will be the revelation of the structure that lies in the process of thinking. Of course, it may turn out that this structure is rigidly connected with a strictly defined substrate, but this thesis should be the result of scientific research, and not the initial premise.
The question of the rigid connection of thinking with a strictly defined substrate is connected with the question of the role of substrate methods in general. There is no doubt about the leading role in the modern natural science of functional and structural methods. While science was dealing with directly sensed objects, it could come from a substrate point of view. Its essence lies in the fact that the object has a set of characteristics expressing its nature, the properties of the material from which it is made. Knowing these characteristics, you can study the behavior of the object.
The material, the substrate is primary; movement, behavior is secondary. This point of view forms the content of so-called mythical substantialism.
Already in the 19th century, the limitations of this concept were revealed by dialectical materialism, which showed that “only in motion does the body discover that it is …
The knowledge of various forms of movement is the knowledge of bodies. ” From this, of course, it does not follow that only movement exists and there is no substrate at all. This only implies the illegitimacy of using the primary-secondary relationship to characterize the connection of movement (behavior) and substrate in terms of their real existence.
It also follows that in the gnoseological sense, behavior is indeed primary with respect to the substrate and knowledge of the substrate contains nothing else but continuously expanding ways of studying objects.
The dialectical-materialistic concept of thinking understands the latter as a property of a highly organized matter in a special way. It does not contain any restrictions on specific characteristics and opens up immense prospects for the study of these characteristics. Cybernetics achieves some results on this path.
Nowadays, going under the sign of accelerating scientific and technological progress, the automation of intellectual activity is becoming an urgent problem.
According to the position of the Soviet cybernetics expert I. A. Poletaev, we are entering the era of “intersection of curves”. Extrapolating to the foreseeable future the current trends in the development of society can come to paradoxical results. Now the number of people employed in the field of management and services is growing faster than the number of people directly involved in production. And it happens so quickly that after a while the number of people employed in the non-production sphere and, in particular, in science will be close to the total population of the Earth.
The rapid increase in the flow of processed information, where it was almost never (trade, banking), will also lead to significant changes in working methods and will require automation, and possibly intellectualization.
By intellect we mean the ability of any organism (or device) to achieve some measurable degree of success in finding one of the many possible goals in a wide variety of environments. We will distinguish knowledge from intellect, bearing in mind that knowledge is useful information accumulated by an individual, and intelligence is his ability to predict the state of the external environment in combination with the ability to transform each prediction into a suitable reaction leading to a given goal. The definition of artificial intelligence is given differently. It is believed that it will be possible to talk about the implementation of artificial intelligence only when the machine starts to solve problems that cannot be done by humans, and this will not be done as a result of high speed, but as a result of applying the new method found. However, not everyone agrees. In most cases, at the current initial stage of research on artificial intelligence only commensurate with the results obtained by man, and not so original.
It is customary to distinguish three main ways of modeling intelligence and thinking: – classical, or (as it is now called) bionic; – heuristic programming; – evolutionary modeling. Consider them in this sequence.
BIONIC MODELING
Direct modeling of the human brain (i.e., modeling of each nerve cell and the connections between them) in order to create automata with intelligence is extremely difficult. The brain is the most complex and only partially studied structure. The most complex interweaving of connections of the cerebral cortex practically defies deciphering. We only know the approximate location of the areas of the brain responsible for a particular function. At present, the principle of operation of the brain elements of neurons, whose numerous connections are externally chaotic, is not known either. Attempts to model the work of the brain by connecting multiple processors like a neural network have shown that some increase in the speed and flow of the processed information goes only to the level of one to two dozen processors, and then a sharp drop in performance begins. Processors seem to be “lost”, stop controlling the situation or spend most of the time waiting for a neighbor. Some success was achieved only in devices operating in the “two-dimensional version”, i.e. processing not sequential, but parallel information, for example, in pattern recognition systems. In them, one data plane simultaneously interacts with another, and the number of units of information can reach several millions. Thus, there is a one-time coverage of the object under study, rather than a consistent study of its parts.
HEURISTIC PROGRAMMING
The second approach to solving the problem of artificial intelligence is associated with heuristic programming and solves problems that, in general, can be called creative.
The practicality of this method is to drastically reduce the options required when using the trial and error method. True, there is always a chance of missing the best solution, so they say that this method offers solutions with a certain probability of correctness.
Two methods are commonly used: a method for analyzing goals and means and a planning method. The first is the selection and implementation of such operations, which consistently reduce the difference between the initial and final state of the task. In the second method, a simplified formulation of the original problem is developed, which is also solved by analyzing the goals and means.
One of the options obtained gives a solution to the original problem.
EVOLUTIONARY MODELING
The third approach is an attempt to model not what is, but what could have been if the evolutionary process were guided in the right direction and evaluated by the proposed criteria.
The idea of evolutionary modeling boils down to an experimental attempt to replace the process of modeling human intelligence with modeling its evolution. When modeling evolution, it is assumed that rational behavior involves a combination of the ability to predict the state of the external environment with the ability to choose the response to each prediction that most effectively leads to the goal.
This method opens the way to automating the intellect and getting rid of routine work. This frees up time for the problem of target selection and identifying environmental parameters that deserve research. This principle can be applied for use in diagnostics, management of unknown objects, in game situations.
So, there are three ways of modeling intelligence: bionic, heuristic, and evolutionary. Depending on the means used, three phases can be distinguished in research. The first phase is the creation of devices that perform a large number of logical operations with high speed.
The second phase includes the development of problem-oriented languages for the equipment used in the equipment created in the first phase. The third phase is most pronounced in evolutionary modeling. During the development of this phase, there is no need for a precise formulation of the problem statement, i.e. The problem can be formulated in terms of goal and eligible costs, and the solution method will be found independently using these two parameters.
Works on artificial intelligence are closely related to the philosophical problem of cybernetic modeling. These works are often associated with building an exact copy of the human brain. However, this approach can be called “non-cybernetic”. What are the features of the cybernetic method of thinking, what questions does cybernetics contribute to human cognition? In his “History of Western Philosophy”, B. Russell raises the question of the factors that allowed Europeans to create a type of culture in which science took the leading place. The reason for this Russell sees, as he puts it, in two great intellectual inventions: the invention of the deductive method by the ancient Greeks (Euclid) and the invention of the experimental method in the Renaissance (Galileo).
It was these two intellectual inventions, the deductive method (and thus mathematics) and the experiment that made it possible to create classical science. To these two basic intellectual tools, modern development of knowledge adds a third — a mathematical model and mathematical modeling. Combining deductive constructions of mathematics with data obtained by an experimental method creates natural science, in the center of which is the concept of scientific law.
The totality of laws is the main content of natural science; their establishment is his main task.
The law claims to be an exact (within this level of knowledge) description of the course of phenomena. The law is either right or wrong, meaningless to talk about good and bad laws. The model in this respect is contrary to the law. A model can be bad or good, it does not claim to reproduce the complex system accurately, but is limited to the description of individual aspects, and for the same aspect, models that simultaneously have the right to exist can be proposed.
In the study of complex systems (including diffuse – it is impossible to isolate individual parts without damaging the system), the formulation of relatively simple laws turns out to be impossible and is replaced by the construction of sketched models.
Figuratively speaking, here we are dealing with a mathematical description that resembles modern abstract painting. It can be said that attempts at a realistic description of complex systems are illusory, such a description would not be perceived because of excessive complexity.
This does not mean that the category of law loses its meaning in science, but the fact that, in addition to the previously known intellectual tools – strict deduction and experiment, a third tool is born mathematical modeling, in which mathematics appears in a new way and a computer experiment appears which play different models, followed by comparison with a real experiment. The path that cybernetics suggests is to build sketch models covering an increasingly wide range of thinking functions. The tasks of uncovering “the essence of thinking” in the forehead are not set, but the task is to construct sketch models that allow to describe its individual sides, its individual functions are reproduced and, moving in this direction, build systems more and more approaching the human brain.
The absence of a rigid connection of the mode of functioning (behavior) with a strictly defined substrate means that if two systems exhibit the same behavior in a fairly wide area, then they should be regarded as systems similar, similar in this mode of behavior.
Sometimes there is a statement that cybernetic modeling is generally not applicable to the study of thinking, because modeling is based on the concepts of conformity and isomorphism, and thinking is a purely human ability, allegedly unable to be described on the basis of the concepts of conformity. It is sometimes said that the understanding of cognition, of thinking as the correspondence of an image to an object means nothing less than a dualistic point of view, which externally compares an object and an image.
Understanding of consciousness as a reflection inevitably means understanding it as a correspondence arising in the course of the organism’s adaptation to the environment. Moreover, this correspondence is not just an external correspondence of a thing and an image as an independent object in relation to a thing. This would indeed be a dualistic point of view, but it cannot monopolize the notion of conformity. Materialism understands an image that is ideal precisely as the correspondence of certain states of the brain to certain states of the external world. This correspondence and carries information about the outside world.
In the above statement, no distinction is made between information modeling of information processes and information modeling of non-information processes. The information model of the device will not work, but will only model the work, but with respect to thinking this thesis seems to be controversial. In relation to information processes, their modeling is functionally complete, i.e. if the model gives the same results as the real object, their distinction loses meaning.
Many of the controversies surrounding the “cybernetics and thinking” problem have an emotional background. The recognition of the possibility of artificial intelligence seems to be something degrading to human dignity. However, questions of the possibility of artificial intelligence cannot be confused with the question of the development and improvement of the human mind. Of course, artificial intelligence can be used for unsuitable purposes, but this is not a scientific problem, but rather a moral and ethical one.
However, the development of cybernetics raises a number of problems that still require close attention. These problems are connected with the dangers arising in the course of work on artificial intelligence.
The first problem is related to the possible loss of incentives for creative work as a result of mass computerization or the use of machines in the arts. Recently, however, it has become clear that a person will not voluntarily give up the most qualified creative work, because he is attractive to man himself.
The second problem is more serious in nature and has been repeatedly pointed out by such specialists as N. Wiener, N. M. Amosov, I. A. Poletaev, and others. It consists of the following.
Already there are machines and programs that are capable of self-learning in the process of work, i.e. improve the efficiency of adaptation to external factors. In the future, perhaps there will be machines with such a level of adaptability and reliability that the need for a person to intervene in the process will disappear. In this case, it is possible for a person to lose his own qualities responsible for finding solutions.
There is a possible degradation of a person’s ability to react to changes in external conditions and, possibly, the inability to take control in the event of an emergency. The question arises about the feasibility of introducing a certain level in the automation of processes associated with severe emergencies. In this case, the person “supervising” the control machine will always have the ability and the reaction to influence the situation in such a way as to extinguish the flare-up emergency situation. Such situations are possible in transport, in nuclear power. It is worth noting such a danger in the strategic missile forces, where the consequences of a mistake can be fatal. A few years ago, the United States began to introduce a fully computerized system of launching missiles on the commands of a supercomputer that processes huge amounts of data collected from around the world. However, it turned out that even with repeated duplication and rechecking, the probability of error would be so great that the absence of a controlling operator would lead to an irreparable error. From the system refused.
People will constantly solve the problem of artificial intelligence, constantly facing new problems. And, apparently, this process is endless.