The use of psychological knowledge in the construction of artificial personality Bashny.Net

This is my first article on the Habré will be grateful for the help in the design. This introductory article, here I present my vision of the problem of artificial intelligence, and propose a cycle of texts that will expand my approach. We are talking about the use of theoretical psychology in AI.

For me, programming and allied area - primarily a tool that is used to improve quicker, more efficient processes of the real world. Throughout my work I had to take some subject area, to build its model and implement it in the form of a computer program. Therefore programmers I appreciate not only the skill to create a quality system, but also the ability to understand any subject area and build its model, which satisfies the conditions of the problem.

After 10 years of work, I realized that I was not so interesting to program objectives, which descends from above uncle, write the code, which will make more money on derivatives, or how to connect the two trading systems, and I want to put myself to the task and their implement. Now I write (June protection) master's thesis in psychology and I have my own task, it can be said, the task of a lifetime.

Even before I went to the department of linguistics department of artificial intelligence in the Russian State Humanitarian University after reading Zelazny I wanted to create something like a ghost wheels, artificial personality that could manipulate the physical properties of the world, developing and becoming more human. Many years later I realized for myself that all development disciplines related to artificial intelligence, does not converge to a point which is so important for me.

Stanislaw Lem (in my "Summa Technologies") expressed to me the idea. It is quite simple and true, even now, in my opinion. Humanity is in some other way from all the hopes and dreams of science fiction of the time. People make more and more advanced stone axes. After all, any most sophisticated achievement of smart technologies - it is very, very complex, often self-contained, but just tools - machines from Google, robots on the assembly line, even talk to Siri and its analogues, especially if you yourself participated in the development of such machines is simply operating goals, which put them man.

But we want the other, and do not think it's just me, judging by movies like "Her" or book of Isaac Asimov's "I, Robot", I think any lead here a thousand examples. Is the dream of creating their own kind, about the substance that will somehow relate to the world, and not just a treat, but also to experience this attitude. It is a creature that will not serve the purposes of man, but the set itself the goal.

There are very strong obstacle stands the concept of "artificial intelligence." Now for me it is a set of related technologies and knowledge that are good at and deal effectively with difficult tasks, many of which a person decides pretty good. And it is very strange to see the enthusiastic comments on the fact that soon the technical singularity, I can not imagine that this is possible in the near future.

After all, we are talking about intelligence. That creature who wants to create, it is still like a man who can not be reduced simply to the intellect. Even in this branch of psychology is, concepts such as emotional intelligence (the ability to recognize and manage their own and others' emotions), is a dialectical logic (the ability to solve problems in logic without the excluded middle, a very interesting work N.E.Veraksy with an attempt to construct a mathematical model) , there is the concept of will, finally. Man - this is primarily a potentially free person, and it is not confined to mental models.

So I formulated the problem yourself so - to develop a model of artificial personality. To better understand how the human psyche is arranged, I very much zadaunshiftilsya, and went to study at the Russian State Humanitarian University Masters fukultet psychology, as well as Gestalt therapist MIGIP (learning without practice, in my opinion, is not worth a damn).

As I now seem to have two destinations, which anyway can come to create an artificial personality. The movement can be performed from two sides - from the neurophysiology that better and more detailed versed with the processes occurring in our nervous system (I want to mention here, and the article by Alexei Redozuba with whom we have a theoretical protivorchiya but whose model of emotions in general is confirmed in my master's thesis), and from the theoretical psychology, which tries to figure out how to build a model that describes the human psyche.

Neurophysiological approach seems to me reliable, but there is a suspicion that the level of processing power that is required for full-fledged research and experimentation, humanity will not soon reach, possibly, the development of quantum computers, it can somehow speed up, but it is not in a hurry.

Psychological approach is closer to me. I think this is akin to finding a solution or a complete listing, or search for practical approaches that will get the solution given task faster. Because the problem I stated in terms of psychological (personality), it may be worthwhile to understand that humanity has dug up many years of informal existence of psychology and build a model that will lead to the desired solution, although, of course, in fact, I'm sure the decision will be somewhere in the junction.

Mainly it seems to me here that I have not met a fundamental objections to the possibility of creating an artificial personality. Existence of us, as individuals, are not caused by the fact that we are embodied in matter, although strongly conditioned by the laws themselves unit of matter, there is no reason that these effects could not be simulated in a computer system. The question remains how to implement cultural influence about him, too, need to talk.

When I began to learn and understand what people naizuchali in psychology, I came in silent horror. I came across an infinite number of theories, approaches and techniques that were arguing with each other, and each of them had no way to once and for all to prove their point of view, because the experiment itself, a key part of classical science, was easily susceptible to disturbances . This leads to the fact that people are even quite often challenge the very meaning of the existence of such sciences as psychology.

Reading Habr, I often come across the fact that people do not distinguish his experience of psychology from its scientific understanding, often speaking utter nonsense about the basic concepts for which there are just some consensus in the scientific world. Conversely, I think in psychology sorely lacking programmers who are not allowed to be expressed so clearly and would have found a way to operationalize and formalize concepts.

So I thought it would be cool to create a series of articles in which habrasoobschestvo could familiarize themselves with the various aspects of psychological knowledge associated with the specific purpose of creating an artificial person, and maybe in the future would lead someone of them in psychology, in order to shake up this science and to cause it to further growth. Of course, this is all my personal choice, but there is always the opportunity to see what is next.

I am also sure that the expansion of horizons habrachitateley in theoretical psychology will give them additional opportunities to work on these areas of IT, as artificial intelligence, creating recommender systems and game development.

I plan to write articles on the following topics (not necessarily in this order):
Consideration of the most frequent disputes about psychology - psychology is not a science that actually studied psychology, common misconceptions about the human psyche (psychophysiological problem of psychology, the unity of consciousness and activity) - I would like to minimize the typical arguments about psychology
Existing definitions of life and existence of the psyche (Bauer, Leontiev) - without criteria of what I want to get, it is impossible to start a modeling experiment. Conditionality of living matter in a form of psychic laws of the physical world (Belousov-Zhabotinsky reaction, the elements of embryology)
Cultural-historical psychology of Vygotsky - this theory is important to me above all the fact that somehow unites existing psychological theory, and thus, its similarity to expense synergistic constructs seem subservient to mathematical modeling
Activity theory Leontiev - very formal, in my opinion, an attempt to model human activity, I think it can be easily applied to simulate various types of intelligent systems. In my last project finally come to something like this, creating recommender systems.

Would be happy to get feedback, reasoned criticism of the approach to the problem of artificial intelligence and maybe there are some other important topics, articles on the subject which would be interested in reading habrazhitelyam from a man who recently lost in psychology.

Source: habrahabr.ru/post/222145/