St. Petersburg, 2012
The text is not about philosophy on the Internet and not about the philosophy of the Internet - philosophy and the Internet are strictly divorced in it: the first part of the text is devoted to philosophy, the second - to the Internet. The concept of "evolution" acts as a connecting axis between the two parts: we will talk about philosophy of evolution and about Internet evolution. First, it will be demonstrated how philosophy - the philosophy of global evolutionism, armed with the concept of "singularity" - inevitably leads us to the idea that it is the Internet that is the prototype of the future post-social evolutionary system; and then the Internet itself, or rather the logic of its development, will confirm the right of philosophy to talk about seemingly purely technological topics.
Technological Singularity
The concept of "singularity" with the epithet "technological" to denote a special point on the time axis of the development of civilization was introduced by mathematician and writer Vernor Vinge. Extrapolating Moore's well-known law, according to which the number of elements in computer processors doubles every 18 months, he made the assumption that somewhere around the year 2025 (plus or minus 10 years), computer chips should be equal in processing power to the human brain (of course, purely formally - according to the expected number of operations). Vinge stated that beyond this border something inhuman, artificial superintelligence awaits us (humanity), and we should think carefully whether we can (and should we) prevent this attack.
Evolutionary planetary singularity
The second wave of interest in the singularity problem arose after several scientists (Panov, Kurzweil, Snooks) carried out a numerical analysis of the phenomenon of evolution acceleration, namely, the reduction of periods between evolutionary crises, or, one might say, “revolutions” in the history of the Earth. Such revolutions include the oxygen catastrophe and the appearance of nuclear cells (eukaryotes) associated with it; the Cambrian explosion is a rapid, almost instantaneous by paleontological standards, formation of various types of multicellular organisms, including vertebrates; moments of the appearance and extinction of dinosaurs; the birth of hominids; neolithic and urban revolutions; the beginning of the Middle Ages; industrial and information revolution; collapse of the bipolar imperialist system (collapse of the USSR). It was shown that the listed and many other revolutionary moments in the history of our planet fit into a certain pattern-formula that has a singular solution around 2027. In this case, unlike Vinge's speculative assumption, we are dealing with a "singularity" in the traditional mathematical sense - the number of crises at this point, according to the empirically obtained formula, becomes infinite, and the gaps between them tend to zero, that is, the solution of the equation becomes indefinite.
It is clear that the indication of the point of evolutionary singularity hints at something more significant than a banal increase in computer performance - we understand that we are on the verge of a significant event in the history of the planet.
Political, cultural, economic singularities as factors of the absolute crisis of civilization
The peculiarity of the nearest historical period (the next 10-20 years) is also indicated by the analysis of the economic, political, cultural, scientific spheres of society (carried out by me in the work”): the extension of existing development trends in the context of scientific and technological progress inevitably leads to “singular” situations.
The modern financial and economic system, in fact, is a tool for coordinating the production and consumption of goods spaced apart in time and space. If we analyze the trends in the development of networked means of communication and automation of production, we can come to the conclusion that over time, each act of consumption will be as close as possible in time to the act of production, which will certainly eliminate the very need for the existing financial and economic system. That is, modern information technologies are already approaching a level of development when the production of a particular single product will be determined not by the statistical factor of the consumer market, but by the order of a particular consumer. This will also become possible as a result of the fact that the natural reduction in the cost of working time for the production of a single product will lead to a situation where the production of this product will require a minimum effort, which is reduced to the act of ordering. Moreover, as a result of technological progress, the main product is not a technical device, but its functionality - a program. Consequently, the development of information technologies indicates both the inevitability of an absolute crisis in the future of the modern economic system, and the possibility of unambiguous technological support for a new form of harmonization of production and consumption. It is reasonable to call the described transitional moment in social history an economic singularity.
The conclusion about the approaching political singularity can be obtained by analyzing the relationship between two administrative acts spaced apart in time: making a socially significant decision and evaluating its result - they tend to converge. This is primarily due to the fact that, on the one hand, for purely industrial and technological reasons, the time interval between the adoption of socially significant decisions and the receipt of results is steadily decreasing: from centuries-decades earlier to years-months-days in the modern world. On the other hand, with the development of network information technologies, the main problem of management will not be the appointment of a decision maker, but the evaluation of the effectiveness of the result. That is, we inevitably come to a situation where the opportunity to make a decision is provided to everyone, and the evaluation of the result of the decision does not require any special political mechanisms (such as voting) and is carried out automatically.
Along with technological, economic, political singularities, one can also speak of a quite unambiguously manifested cultural singularity: the transition from the total priority of successive artistic styles (with a decreasing period of their prosperity) to the parallel, simultaneous existence of the entire possible variety of cultural forms, to the freedom of individual creativity and individual consumption of the products of this creativity.
In science and philosophy, there is a shift in the meaning and purpose of cognition from the creation of formal logical systems (theories) to the growth of an integral individual understanding, to the formation of the so-called post-scientific common sense, or post-singular worldview.
Singularity as the end of an evolutionary period
Traditionally, the conversation about the singularity - and the technological singularity associated with fears about the enslavement of man by artificial intelligence, and the planetary singularity, derived from the analysis of ecological and civilizational crises - is conducted in terms of a catastrophe. However, proceeding from general evolutionary considerations, one should not imagine the coming singularity as the end of the world. It is more logical to assume that we are dealing with an important, interesting, but not unique event in the history of the planet - with the transition to a new evolutionary level. That is, a number of singular solutions arising from the extrapolation of trends in the development of the planet, society, digital technology indicate the completion of the next (social) evolutionary stage in the global history of the planet and the beginning of a new post-social one. That is, we are dealing with a historical event comparable in significance to the transitions from protobiological to biological evolution (about 4 billion years ago) and from biological to social evolution (about 2,5 million years ago).
In the mentioned transition periods, singular solutions were also observed. Thus, during the transition from the protobiological stage of evolution to the biological stage, the sequence of random synthesis of new organic polymers was replaced by a continuous regular process of their reproduction, which can be referred to as "singularity of synthesis". And the transition to the social stage was accompanied by a “singularity of adaptations”: a series of biological adaptations grew into a continuous process of production and use of adaptive devices, that is, items that allow you to almost instantly adapt to any changes in the environment (it got colder - put on a fur coat, it started to rain - opened an umbrella). Singular trends indicating completion social stage of evolution, can be interpreted as a "singularity of intellectual innovations". In fact, over the past decades, we have been observing this singularity as the transformation of a chain of separate discoveries and inventions, previously separated by significant periods of time, into a continuous stream of scientific and technical innovations. That is, the transition to the post-social stage will manifest itself as a change from the sequential appearance of creative innovations (discoveries, inventions) to their continuous generation.
In this sense, to some extent, we can talk about the formation (namely, the formation, and not the creation) of artificial intelligence. To the same extent as, say, social production and the use of adaptive devices can be called "artificial life", and life itself from the point of view of the continuous reproduction of organic synthesis - "artificial synthesis". In general, each evolutionary transition is associated with ensuring the functioning of the basic processes of the previous evolutionary level in new non-specific ways. Life is a non-chemical way of reproducing chemical synthesis, mind is a non-biological way of providing life. Continuing this logic, we can say that the post-social system will be an "unreasonable" way to ensure the intellectual activity of a person. Not in the sense of "stupid", but simply in form not related to the rational activity of a person.
On the basis of the proposed evolutionary-hierarchical logic, one can also make an assumption about the post-social future of people (elements of the sociosystem). Just as bioprocesses did not replace chemical reactions, but, in fact, they were only a complex sequence of them, just as the functioning of society did not exclude the biological (life) essence of a person, so the post-society system will not only not replace human intelligence, but will not surpass it. The post-social system will function ON the basis of human intellect and TO ensure its activity.
Using the analysis of patterns of transitions to new evolutionary systems (biological, social) as a method of global forecasting, we can indicate some principles of the coming transition to post-social evolution. (1) Preservation and stability of the previous system during the formation of a new one - man and humanity, after the transition of evolution to a new stage, will retain the basic principles of their social organization. (2) Non-catastrophic transition to a post-social system - the transition will not be manifested in the destruction of the structures of the current evolutionary system, but is associated with the formation of a new level. (3) The absolute inclusion of the elements of the previous evolutionary system in the functioning of the next - people will ensure the continuous process of creation in the post-social system, maintaining their social structure. (4) The impossibility of formulating the principles of a new evolutionary system in terms of the previous ones - we do not have and will not have either the language or concepts for describing the post-society system.
Post-social system and information network
All the described variants of the singularity, indicating the coming evolutionary transition, are somehow connected with scientific and technological progress, or rather with the development of information networks. The technological singularity of Vinge directly hints at the creation of artificial intelligence, a supermind capable of absorbing all spheres of human activity. The graph describing the acceleration of planetary evolution reaches a singular point, when the frequency of revolutionary changes, the frequency of innovations, presumably becomes infinite, which, again, is logical to associate with some kind of breakthrough in network technologies. Economic and political singularities - the combination of acts of production and consumption, the convergence of the moments of decision-making and evaluation of its result - are also a direct consequence of the development of the information industry.
The analysis of the previous evolutionary transitions tells us that the post-social system should be implemented on the basis of the main elements of the social system - individual minds, united by non-social (non-production) relations. That is, just as life is something that needs to provide chemical synthesis by non-chemical methods (by reproduction), and the mind is something that needs to ensure the reproduction of life by non-biological methods (in production), so the post-social system should be thought of as something that needs to ensure reasonable production by non-social methods. . The prototype of such a system in the modern world, of course, is the global information network. But it is precisely the prototype - in order to break through the point of singularity, it itself must still go through more than one crisis in order to transform into something self-sufficient, which is sometimes called the semantic web.
Many Worlds Truth Theory
To discuss the possible principles of organization of a post-society system and the transformation of modern information networks, in addition to evolutionary considerations, it is necessary to fix some philosophical and logical foundations, in particular, those relating to the relationship between ontology and logical truth.
In modern philosophy, there are several competing theories of truth: correspondent, authoritarian, pragmatic, conventional, coherent and some others, including deflationary, which denies the very necessity of the concept of "truth". It is difficult to imagine this situation as resolvable, capable of ending with the victory of one of the theories. Rather, we must come to understand the principle of the relativity of truth, which can be formulated as follows: the truth of a proposition can be stated only and exclusively within the boundaries of one of the many more or less closed systems, which in the article "» I suggested calling logical worlds. For each of us, it is obvious that in order to assert the truth of the sentence we uttered, stating a certain state of affairs in personal reality, in our own ontology, we do not need to refer to any theory of truth: the sentence is true simply by the fact that it is embedded in our ontology, in our logical world . It is clear that there are also supra-individual logical worlds, generalized ontologies of people united by one or another activity - scientific, religious, artistic, etc. And it is obvious that in each of these logical worlds the truth of sentences is fixed in a special way - according to the way they are included in a specific activity. It is the specificity of activity within a certain ontology that determines the set of methods for fixing and generating true sentences: in some worlds, the authoritarian method prevails (in religion), in others, coherent (in science), in others, conventional (in ethics, politics).
So, if we do not want to limit the semantic network to a description of only one sphere (say, physical reality), then we must initially proceed from the fact that it cannot have one logic, one principle of truth - the network must be built on the principle of equality of intersecting, but fundamentally not reducible to each other logical worlds, reflecting the multitude of all conceivable activities.
Activity ontologies
And here we move from the philosophy of evolution to the evolution of the Internet, from hypothetical singularities to the utilitarian problems of the semantic web.
The main problems of building a semantic network are to a large extent connected with the cultivation by its designers of a naturalistic, scientistic philosophy, that is, with attempts to create the only correct ontology that reflects the so-called objective reality. And it is clear that the truth of sentences in this ontology should be determined according to uniform rules, according to the universal theory of truth (which is most often understood as correspondence theory, since we are talking about the correspondence of sentences to some “objective reality”).
Here the question should be asked: what should the ontology describe, what is the “objective reality” for it, to which it should correspond? Some indefinite set of objects called the world, or a specific activity within a finite set of objects? What interests us: reality in general or fixed relationships of events and objects in a sequence of actions aimed at achieving specific results? Answering these questions, we must necessarily come to the conclusion that ontology makes sense only as a finite and exclusively as an ontology of activity (actions). And consequently, it makes no sense to talk about a single ontology: how many activities - so many ontologies. Ontology does not need to be invented - it must be identified by formalizing the activity itself.
Of course, it is clear that if we are talking about the ontology of geographical objects, the ontology of navigation, then it will be the same for all activities that are not focused on changing the landscape. But if we turn to areas in which objects do not have a fixed reference to space-time coordinates, are not related to physical reality, then ontologies multiply without any restrictions: we can cook a dish, build a house, create a training methodology, write a program political party, to combine words into a poem in a non-finite number of ways, and each way is a separate ontology. With this understanding of ontologies (as ways of fixing a specific activity), they can and should be created only in this very activity. Of course, provided that we are talking about activities directly performed on a computer or recorded on it. And soon there will be no others at all; those that will not be “digitized” should not be of particular interest to us.
Ontology as the main result of activity
Any activity consists of separate operations that establish links between objects of a fixed subject area. The doer (hereinafter we will traditionally call him a user) over and over again - whether he writes a scientific article, fills in a table with data, draws up a work schedule - performs a completely standard set of operations that ultimately lead to the achievement of a fixed result. And in this result he sees the meaning of his activity. But if you look from a position not locally utilitarian, but systemically global, then the main value of the work of any professional lies not in the next article, but in the method of writing it, in the ontology of activity. That is, the second basic principle of the semantic network (after the conclusion “there must be an unlimited number of ontologies; how many activities, so many ontologies”) should be the thesis: the meaning of any activity lies not in the final product, but in the ontology fixed in the course of its implementation.
Of course, the product itself, say, an article, contains an ontology - it, in fact, is the ontology embodied in the text, but in such a frozen form, the product is very difficult to ontologically analyze. It is on this stone - the fixed end product of activity - that the semantic approach breaks its teeth. But after all, it should be clear that it is possible to reveal the semantics (ontology) of a text only if one already possesses the ontology of this particular text. It is difficult even for a person to understand a text with a slightly different ontology (with changed terminology, conceptual grid), and even more so for a program. However, as is clear from the proposed approach, there is no need to analyze the semantics of the text: if we are faced with the task of identifying a certain ontology, then there is no need to analyze a fixed product, we need to refer directly to the very activity during which it appeared.
ontological parser
In fact, this means that it is necessary to create a software environment that would simultaneously be a working tool for a professional user and an ontological parser that captures all his actions. The user is not required to do anything more than simply work: plan the text, edit it, search for sources, highlight citations, place them in the appropriate sections, make footnotes and comments, organize an index and thesaurus, etc., etc. Maximum additional actions is to mark new terms and link them to the ontology using the context menu. Although any professional will only be happy with this additional "load". That is, the task is quite specific: it is necessary to create such a tool for a professional in any field that he could not refuse, a tool that not only allows you to perform all standard operations for working with all kinds of information (collection, processing, configuration), but also automatically formalizes the activity, builds the ontology of this activity, and corrects it when “experience” is accumulated.
Universe of objects and cluster ontologies
It is clear that the described approach to building a semantic network will be truly effective only if the third principle is fulfilled: software compatibility of all created ontologies, that is, ensuring their systemic connectivity. Of course, each user, each professional creates his own ontology and works in its environment, but the compatibility of individual ontologies according to the data and the ideology of the organization will ensure the creation of a single universe of objects (data).
Automatic comparison of individual ontologies will allow, by identifying their intersections, to create thematic cluster ontologies – hierarchically organized non-individual structures of objects. The interaction of an individual ontology with a cluster one will significantly simplify the user's activity, guide and correct it.
Uniqueness of objects
An essential requirement of the semantic network should be to ensure the uniqueness of objects, without which it is impossible to implement the connectivity of individual ontologies. For example, any text should be in the system in a single copy - then each link to it, each citation will be recorded: the user can track the inclusion of the text and its fragments in certain clusters or personal ontologies. At the same time, it is clear that “single instance” does not mean storage on one server, but the assignment of a unique identifier to an object that does not depend on location. That is, the principle of the finiteness of the volume of unique objects should be implemented with the multiplicity and non-finality of their organization in the ontology.
Usercentrism
The most fundamental consequence of the organization of the semantic network according to the proposed scheme will be the rejection of site-centrism - the site-oriented structure of the Internet. The appearance and presence of an object in the network means only and exclusively the assignment of a unique identifier to it and inclusion in at least one ontology (say, the individual ontology of the user who placed the object). An object, for example, a text, should not have any address on the Web - it is not tied to either a site or a page. The only way to access the text is to display it in the user's browser after finding it in some kind of ontology (either as an independent object, or by reference or quote). The network becomes exclusively user-centric: before and outside the user’s connection, we have only a universe of objects and a set of cluster ontologies built on this universe, and only after connection, the universe is configured with respect to the structure of the user’s ontology - of course, with the possibility of free switching of “points of view”, transition to positions of other, neighboring or distant ontologies. The main function of the browser is not displaying content, but connecting to ontologies (clusters) and navigating in them.
Services and goods in such a network will appear as separate objects, initially inscribed in the ontology of their owners. If the need for a particular object is tracked in the user's activity, then if it is available in the system, it will be automatically offered. (In fact, contextual advertising is currently operating according to this scheme - if you were looking for something, you will not be left without offers.) On the other hand, the very need for a new object (service, product) may be identified when analyzing cluster ontologies .
Naturally, in a user-centric network, the proposed object will be presented in the user's browser as an embedded widget. To view all offers (all manufacturer's products or all author's texts), the user must switch to the supplier's ontology, which systematically displays all objects available to external users. Well, it is clear that the network immediately provides an opportunity to get acquainted with the ontologies of cluster producers, as well as, most interestingly and importantly, with information about the behavior of other users in this cluster.
Conclusion
So, the information network of the future is presented as a universe of unique objects with individual ontologies built on them, combined into cluster ontologies. An object is defined and available on the web for the user only as inscribed in one or more ontologies. Ontologies are formed mainly automatically by parsing the user's activity. Access to the network is organized as the existence / activity of the user in their own ontology with the possibility of expanding it and moving to other ontologies. And most likely it is already difficult to call the described system a network - we are dealing with a kind of virtual world, with the universe only partially presented to users in the form of their individual ontology - a private virtual reality.
*
In conclusion, I would like to emphasize that neither the philosophical nor the technical aspect of the coming singularity is relevant to the problem of the so-called artificial intelligence. The solution of particular applied problems will never lead to the creation of what could be fully called intelligence. And the new, which will constitute the essence of the functioning of the next evolutionary level, will no longer be intelligence - neither artificial nor natural. Or rather, it would be more correct to say that it will be intellect to the extent that we can understand it with our human intellect.
When working on the creation of local information systems, one should treat them only as technical devices and not think about philosophical, psychological, and, moreover, ethical, aesthetic and globally catastrophic moments. Although both humanists and techies will undoubtedly do this, their reasoning will not speed up or slow down the natural course of solving purely technical problems. Philosophical understanding of both the entire evolutionary movement of the World and the content of the coming hierarchical transition will come with this transition itself.
The transition itself will be technological. But it will not happen as a result of a private brilliant decision. And by the totality of decisions. Breaking critical mass. The intellect will embody itself in the hardware. But not private intelligence. And not in a specific device. And he will no longer be an intellect.
PS Attempt to implement the project (option after initial testing).
Literature
1. Vernor Vinge. technological singularity,
2. A. D. Panov. Completion of the planetary cycle of evolution? Philosophical Sciences, no. 3–4:42–49; 31–50, 2005.
3. Boldachev A.V. . SPb., 2008.
4. Boldachev A.V. . SPb., 2008.
5. Boldachev A.V. Innovations. , St. Petersburg: Publishing House of St. Petersburg. un-ta, 2007. - 256 p.
Source: habr.com