Ascending from Abstract Mind Concept to Concrete AI Project
Abstract. The knowledge gained by The Humanities about human mind during many years should not be lost and can be used to format projects to create artificial intelligence (AI). This effort needs special attention as to properly switch between epistemological procedures used in Arts and Sciences which is described as ascending from abstract to concrete. Such a switch requires lots of preparatory work including taxonomies of reasoning types, thinking tools, and problems' types where the said taxonomies are specifically designed from the hard sciences' point of view. The major challenge is to find organizational resources and develop new team thinking patterns to unite experts from various disciplines to work as One Mind Team.
1 Introduction
The dream to create Artificial Intelligence excited many technically minded people since the times the word 'robot' was invented in 1920 [1] (it is a play with one of the first examples of artificial human-like beings in art and literature), Norbert Wiener started the computational revolution in 1948 [2], and Isaac Asimov wrote his series 'I, the robot' (1950) [3]. By this time it is over fifty years now that many projects to construct AI have been started and closed. Yet there are no tangible results. The discrepancy between the effort made and the outcomes gained indicates there is a problem there. Based on the systems approach we have to outline the problem in general first in order to handle it properly. This means it is high time we have to review the basics of the effort.
During these last fifty years many physicists presented new frames of reference, new world pictures; 'dialogue-based thinking method', 'strings theory', 'singularity', various shades of constructivism, views based on quantum mechanics' concepts, and many other thinking efforts to find better ways to explain how the world works were made by people who were educated, trained, and provided for their living as scientists.
2 How to Transfer Knowledge about Mind from the Humanities to Hard Sciences?
It is also widely known that the amount of knowledge in each scientific field, in every discipline increased so much that normally one person cannot excel in more than one field of knowledge. Here comes a question: can an IT specialist review one's intellectual implements used to work on creating AI or vice versa: can a mind specialist properly review the guidelines for a computer based AI project?
Anyone who is familiar with managing projects knows how important it is to make sure once created knowledge is preserved and being properly used. It also seems quite obvious that an IT expert cannot be proficient enough with psychology, philosophy, KM, and other mind related disciplines at the same time. We propose to work on a know-how to organize team effort in a way that a group of experts in various fields can work together as One Mind equally knowledgeable in a number of fields. This article presents several ideas that might interest IT professionals enough to join our efforts in building such a multi-expert One Mind Team.
3 Reviewing the Basics
One of the KM rules says that it is important to voice all the banal basics when discussing a joint project no matter how banal they might seem to some people. As a matter of fact practically all the ideas and concepts mentioned in this article are not new; all of them have been discussed before in human history (please, see the reference list). Every generation of thinkers has to repeat them anyway because they still get a new shade of meaning as they are viewed under new circumstances and for new projects. Another point for repeating is that concrete people as 'knowledge containers' endure about 25 to 50 years in this capacity; this means that each generation has to repeat, re-view and re-process all the thousand years old knowledge to make sure it stays intact.
The history of resemblance between a solution proposed by Nature and a technological one demonstrates huge difference between the two, as in “bird -- air-plane”, “fish - ship”. In order to be able to compare the two solutions we need to have a concept of the basic skill or quality that we compare. This takes us to the question: what Intelligence is?
How can we describe Intelligence or Reasoning using the range of notions and worldviews most common within sciences rather than humanities?
For your information: this kind of work, i.e. a scientific outlook for human mind, has already been done more than two hundred years ago by a member of Encyclopedia team, namely by Marie Jean Antoine Nicolas Caritat, marquis de Condorcet (1743 1794) in his book “Sketch for a Historical Picture of the Progress of the Human Mind”[4].
He describes the history of our civilization via demonstrating how ancient philosophy broke down into sciences, Maths, and metaphysics; he states that progress in technology, education, and social justice is the change of the human mind, our intellectual faculties. He argues that the proof of progressive change in human mind is that nowadays (i.e. in 18th century) a young student can do the same calculations that only Newton was capable of long ago. Marquis de Condorcet also notices that there is a difference “... between what belonged properly to the progress of the art itself and what was due only to the talent of the individual artist.”[4] .
4 Macro- and Micro-worlds in Arts
Condorcet's statement above points to a huge problem in scientific observation of Human Mind: normally we witness micro-phenomena (“an individual artist”) as inter-human relations and have to relate them to our macro-world (“the art itself”). In other words our everyday immediate personal experience with monitoring one's flow of consciousness is routinely generalized by the concepts at use as inherent to a different phenomenon: Human Mind. Consider this: a physicist has dropped in size close to that of an atom (micro-world) where quantum mechanics rule; yet that physicist is challenged to draw Newton laws intrinsic to the macro-world based on one's micro-observations only! This is exactly what happens with any effort to discuss Human Mind: we are confused by the nature of the facts available via individual perception and the scope of the generalized conclusions we tend to make based on them. Quite interesting it is that this confusion between macro-and micro-worlds does not exist in physics: everybody normally knows where to apply Newtonian laws and where to use quantum mechanics.
5 A Method in Logic Called 'Ascending from Abstract to Concrete'
When observing falling metal balls as Galileo once did we do not have to think about relating the nature of the directly observed objects to the domain of the rules that govern them: they are of the same domain already. When relating our introspection of personal thinking experience to the Nature of Human Mind we have to perform an additional procedure of going from one domain to another. This procedure has been clearly described in a number of papers about ascending from abstract to concrete (AAC) [5, 6, 7, 8]. According to G.P. Schedrovitsky: “The core concept of the method called ascending from abstract to concrete maintains that every next epistemological procedure depends on the character and results of the previous one which creates reverse dependence (though at a different level) of every procedure and its results from all the following ones.” [7] In a very rude approximation we can illustrate AAC by the following example: when told to perform the sum of “1 plus 2” we have to ask back: what are the qualities of those units that we add to each other? The sum result will be different depending on that additional piece of information: if we add a fox and a rabbit into one cage the result is “1”, if we add a litre of water to 2 litres of gravel the resulting volume will stay the same; if we add alcohol and water together the solution volume and the content will be different.
To summarize this part: if we consider the works of Human Mind of our macro-world in a scientific most rigid and consistent systemic way one will certainly notice that Human Mind consists of more than one person and is organized with regards to time, place, structure, processes, and material in a very complex manner. In a metaphoric way this concept was worded by constructivism and system-reasoning-activity methodology (SRAM) as “a lone person does not think” and “reasoning belongs outside human brain”. [9, 10]
There is no doubt the above mentioned maxims strike contrary to what any average scientist would keep in mind while considering AI-related problems which certainly results in how they would formulate their research objectives and choose methods. Based on the results to date that average concept of what Human Mind is needs reviewing.
The endeavour of reviewing one's basics involves entering The Second Nature [11], the world of human activities which suggests new concepts, recharging one's value system and judgemental patterns located within our mainstream logic [12]. All of these scientific resources are being researched and developed as you are reading these lines. Probably the major challenge is in finding organizational solution to merge several scientific fields in order to work on one set of problems together, as One whole Mind. There is a number of established patterns of helping a team of experts brainstorming one problem together. Here are some to mention: those developed by Stafford Beer [13], Jim McCarthy [14], S. Pereslegin and his 'knowledge reactor' [15](there is no publication on the topic by the author other than some information on his website), and the most successful and deeply developed one by Moscow Methodological Circle - Organizational Activity Games (OAG) [16].
6 Detailing the Basics
Professor O.S. Anisimov interprets the core idea of ascending from abstract to concrete as detailing [8]. Well, really, after we figure out the general state of affairs we should provide more concrete staff, explain our statement using examples which is dwelling into details. In this light the whole idea of merging the knowledge gained by various mind related disciplines with approaches, patterns, and epistemological procedures inherent in sciences can be interpreted as an instance of that very method described above.
Let us get more specific with this step, detailing.
According to the traditional train of thought by Hegel-->Marx-->Vygotsky-->SRAM[17, 11, 18, 10] Reasoning is one of human activities; one of the same level as Production, Education, and Leisure [19]. According to this definition it cannot become artificial in principle, because 'artificial' means 'non-human'. Consequently, in order to proceed with an AI project, we should think of which part of that Reasoning Activity can be performed by non-human means. Looking at the components of the activity in focus [16] one can easily identify the needed part: the 'Pure Reasoning' belt where operations on signs are performed.
Let us look closer at this unit.
Basically this unit is our Speech Consciousness.
The next question is: how does it work?
In a sentence: it is a joint venture of Words Processing Center and our Perception-Value-Emotions-Willpower Center, where the former Center is very new to the animal world, unique only to humans and does not show traceable links to the latter Center [20]. This is what happens when we switch from theory to practice: we certainly can single out Pure Reasoning as an abstraction. But the challenge is to ascend from this abstract notion to its real life mechanism. The real life Reasoning mechanism is called Speech Consciousness. And Speech Consciousness is inseparable from our bodily emotions, judgements, values, and will-power.
Current IT industry is quite capable of doing calculations and even composing texts that make sense to humans. Yet all these 'operations on signs' are not equal to human-like acts because our artificial intelligent devices do not have will-power nor judgemental opportunities. Within a human our Perception-Value-Emotions-Willpower Center serves as the engine that drives Speech Consciousness throughout many stages and at many levels while composing texts. This fact stumbles our discussion of AI; first we need to be clear on the role of will-power and senses within our intelligence, and then we can move on with designing AI equal to human-like thinking or there might be a different route to success.
There is a chance that our 'artificial' reasoning needs to fill in much more small lacunae in an environment where AI may flourish, that IT-based 'reasoning software' has to go through evolution of creating huge number of various gadgets. During the next 10-15 years we might witness engineering a number of different gadgets that will supplement existing IT capabilities; and at some point they will come together as one intelligent machine, intelligent in the Asimov sense of 'robot'.
7 Further Detailing Requires More Preparatory Work
There is one more important aspect to the AI basics. It is vital to address it as we are switching from one epistemological set of procedures to another. The Humanities in a broader sense of the word, disciplines related to human world, have a long history of understanding this domain of our knowledge. Every researcher in the related field understands that the concept of Mind used during discourse varies depending on the academic school. Once we formulate our problem as AI, or Human Mind when discussing it with scientists, people from Maths, IT, or engineering this aspect is not mentioned. The most accepted view is that when presenting our problem as creating an AI we are assuming to create a gadget that will act as one's brain.
Contrary to this assumption as we discussed above one gadget - even the most perfect one! -- cannot duplicate Reasoning in principle. Most probably computers can only substitute some steps within the totality of human reasoning activity.
On top of that there are many different kinds, types, levels, and patterns of Reasoning. S.Pereslegin suggested a taxonomy of Reasoning based on its organization and complexity. Here is the list from his most recent book: common sense, academic reasoning, reasoning based on natural sciences, humanitarian, juridical, dialectical, technological, system dialectical, methodological dialectical, trialectical, complex, and sublimated reasoning [21]. The book can be found only on-line.
We need to have a mainstream Reasoning taxonomy so that we can discuss this abundance of possibilities meaningfully within broader public. This taxonomy has to be supplemented by taxonomies of problems and Intellectual Implements (aka 'thinking tools') which was also earlier suggested in our articles. [22, 23]
8 Conclusion
Real life examples of Reasoning instances vary from extremely primitive to overlay complex. From this point of view people have already invented and are using bits of AI at various stages of the full Reasoning cycle. The most success is demonstrated in calculations; but we still cannot create a software that can solve a Math problem that has not been solved by humans yet.
As our current society is getting more complex we witness emergence of a new type of problems that we could not approach consciously before, and namely problems related to the life of organizations, communities, and various aspects of their activities. A new type of a problem to solve requires new ontologies, procedures, notions. This range of problems has to be discussed so as to create a set of notions, cloud of meanings with regards to Mind and problem solving. Right now we experience difficulties with describing one's actions to successfully solve a problem. In other words, some problems are solved but people who do it are not able to adequately describe their actions, they cannot explain what exactly they did. The skill of explaining one's actions is called 'reflection' and it is one of the major intellectual implements.
In order to properly outline an AI project we need to get a better picture of what Reasoning is as an activity while keeping in mind that we do that specifically for that project, which parts of it we can and should support with our gadgets, and what the requirements for those gadgets would be in terms of their role, functionality, and composure.
References
1. Čapek, Karel. R.U.R. (Rossum's Universal Robots) (Rossumovi univerzální roboti), 1920.
Wiener, N. Cybernetics: Or Control and Communication in the Animal and the Machine. Paris, (Hermann & Cie) & Camb. Mass. (MIT Press) 1948.
Asimov, A. I,Robot. 1950
Condorcet, Marie Jean Antoine de. Sketch for a Historical Picture of the Progress of the Human Mind, trans. June Barraclough (London: Weidenfeld & Nicolson, 1955)
Zinoviev, A.A. Ascending from abstract to concrete (material of “Capital” by K.Marx). 1st level PhD thesis. Moscow University. Зиновьев А.А. Метод восхождения от абстрактного к конкретному (на материале «Капитала» К.Маркса). Автореферат канд. дисс., МГУ, 1954.
Ilyenkov, E.V. Dialectics of abstract and concrete within scientific theoretical reasoning. Ильенков Э.В. Диалектика абстрактного и конкретного в научно-теоретическом мышлении»: «Российская политическая энциклопедия» (РОССПЭН); Москва; 1997
Schedrovitsky, G.P. The core concept of the ascending from abstract to concrete method. Щедровицкий Г.П.. Общая идея метода восхождения от абстрактного к конкретному 1965.
Anisimov, O.S. A.A. Zinoviev on the 'Ascending' method. Анисимов О.С. А.А. Зиновьев о проблеме метода «восхождения». M., 1974.
Cull J. The Circularity of Life: An Essential Shift for Sustainability. E-book. 2013.- Piskoppel, A.A. Science, Reason, and Knowledge in SRA-Methodology. Пископпель А.А. Наука, мышление и знание в СМД-методологииM., 1999.
Marx, K. Capital. Маркс К. Капитал.
Chumakin, M. On modeling links between values and and rational argumentation within decision-making process. St. Petersburg Institute of International Economic Relations, Economics and Law, Novokuznetsk department. Чумакин, М. О моделировании связи ценностей с рациональной аргументацией в процессе принятия решений. Сборник научных статей по итогам Международной научно-практической конференции. Новокузнецк, Санкт-петербургский институт внешнеэкономических связей экономики и права, Новокузнецкий филиал, 2011
Beer, Stafford. Cybernetics and management. 1959.[London] English Universities P. [1971, c1967]
McCarthy, Jim. The Core System. Version 1.5. Core Adoption Records C/O McCarthy Technologies, Inc., 2004
Pereslegin, S. A description of the knowledge reactor. 2008. http://znatech.ru/proekty/tehnologiya/metodologicheskoe_opisanie_tehnologii_znanievyj_reaktor/
Tschedrovitsky, G.P., Kotelnikov, S.I. Organizational development game as a new organizational form and method for developing team reasoning activity. Щедровицкий Г.П., Котельников С.И. Организационно-деятельностная игра как новая форма организации и метод развития коллективной мыследеятельности. 1983
Hegel, W. The science of logic. Cambridge ; New York : Cambridge University Press, c2010.
Vygotsky, L.S. Psychology of human development. Выготский Л.С. Психология развития человека. - М.: Изд-во Смысл; Эксмо, 2005.
Tschedrovitsky, G.P. ‘Project on Researching Pedagogics. (Methodological Analysis), Щедровицкий Г.П. Система педагогических исследований.(Методологический анализ). Монография, 1968.
Chumakin, M. A Model of Speech Consciousness. Kuzbass State Pedagogical Academy. Чумакин М. Вариант модели речевого сознания, Кузбасская государственная педагогическая академия, Новокузнецк, 2014.
Perslegin, S. Summa Strategia. St.Petersberg. 2013. http://future-designing.org/proekti/uchebnikpostrategii.html
Chumakin, M. On reasoning implements' taxonomy. Novokuznetsk Miningtransport College, Novokuznetsk. Чумакин М. О классификации инструментов мышления. «Новокузнецкий горнотранспортный колледж», Новокузнецк, 2012.
Chumakin, M. Towards the notion of Intellectual Implements. Novokuznetsk, St. Petersburg Institute of International Economic Relations, Economics and Law, Novokuznetsk department, Сборник научных статей по итогам Международной научно-практической конференции. Новокузнецк, Санкт-петербургский институт внешнеэкономических связей экономики и права, Новокузнецкий филиал, 2013.