How does our brain work and what happens in it? On the basis of what does a person make decisions and how are they conditioned? Vyacheslav Demin, candidate of physical and mathematical sciences, scientific secretary of the Kurchatov complex of NBIKS-technologies of the National Research Center "Kurchatov Institute", spoke about this in a lecture at the educational center "Sirius". "Lenta.ru" publishes excerpts from his speech.
Break through the gap
The brain consists of about one hundred billion neurons, that is, nerve cells that receive and transmit information to each other using electrical and chemical signals through processes (dendrites and axons). By touching, neurons create neural networks. The place of contact is called a synapse. There are about a quadrillion synapses in the brain (a quadrillion is a number followed by 15 zeros, that is, a million billion). This means that each neuron has about 10 thousand connections - a very revealing illustration of how diverse and multifaceted the connections of only one nerve cell can be. A substance that helps convey information is called a neurotransmitter. Science knows several hundreds of such substances.
Vyacheslav Demin
The scientific community approaches the issue of studying the brain from different perspectives. There are neurophysiologists who consider specific processes at the neural level; they can be conventionally called “materialists”. On the other hand, there are neuropsychologists, they can be conventionally called "idealists", in the center of their attention is the world of ideas, the space of human higher cognitive functions responsible for memory and thinking, consciousness and subconsciousness, emotions and decision-making, attitude towards oneself and other people … There is a fundamental explanatory gap between the first approach and the second. It is studied by cognitology, a scientific direction that has recently developed at the junction of neurophysiology and neuropsychology. Apparently, it is cognitology that can lead to a breakthrough in the creation of artificial intelligence in the first place.
Promotional video:
Finding the optimal solution
What is thinking? It is a constant search for the optimal solution to the challenges we face. As a rule, when making even the smallest decision, a person has several options, before each step he finds himself at a fork, and the outcome is not predetermined. The person must make the best move. That is, every second each of us builds a “tree of possibilities” in our heads, and sometimes this tree is incredibly branched.
How to choose the right one, especially if the search algorithm is unknown? The intellect uses so-called heuristics. An example from chess can be used to illustrate this. On the board, such an arrangement of pieces is possible when White, for example, has only the king and pawns, but the pawns are placed so that they do not allow Black to pass. The person immediately understands that under such conditions the most favorable and quite probable outcome of the game for White is a draw.
But the computer program Deep Thought, which later beat the world champion Garry Kasparov, considered situations exclusively from a mathematical point of view. She saw that the white pawn could take the black rook, and this would lead to a noticeable weakening of the opponent and an improvement in the point position. The computer did not realize that with this move it was opening a hole in its defense. As a result, he could no longer count on a draw, received a checkmate and lost the game.
Photo: Carina Johansen / NTB Scanpix / Reuters
Subsequently, programmers introduced an algorithm for actions in such situations into the computer, and the machine no longer made such mistakes. Natural intelligence, unlike artificial intelligence, is able to independently draw conclusions, analyze mistakes and not repeat them.
Knowledge representation
The second aspect of thinking is the representation of knowledge. We all look at the world through the prism of perception and form in our heads a model of a process or object. These views are individual. And when we think, we operate with models, and not with real objective data.
There is a famous joke about a glass half filled with water. The optimist thinks he is half full, the pessimist half empty. But there may be other ideas as well. For example, a programmer will say that the capacity is twice as much as needed. The initial objective data are the same, but the models that people operate on their basis are different. As a result, if a certain task is associated with the order book, then the solutions may differ from each other. It is important to find a suitable representation, in which there is an algorithm that solves the problem. In another, unsuccessful, presentation, the problem may turn out to be extremely difficult or completely insoluble.
Therefore, thinking should be combined with learning, that is, the accumulation of information with subsequent generalization. You can watch the grandmaster endlessly, write down and memorize his moves, then play them back. But that won't teach you how to play chess. On the contrary, attempts to understand the very system or tactics of the game, which give ideas about the general presentation of chess problems by a grandmaster, will eventually yield positive results with time and practice. This is learning.
Types of thinking
How does a person's thinking develop? In childhood - through a visual and effective presentation: "I saw - I did an action." Visual-figurative thinking is gradually formed: “I saw - I remembered or presented related objects or options for actions - I performed an action”. Individual objects are replaced by categories, representations, separate links between them are modeled. The next stage is completely abstract verbal-logical thinking, when for the process of thinking itself there is no longer a need to perform any actions, everything happens in the imagination.
In the middle of the 20th century, German psychologist Wolfgang Keller conducted an experiment. Next to the monkey cage, he put a banana and gave the animals a stick. They almost immediately figured out how to reach the banana with a stick and push it to the cage. This happened due to visual-active thinking: the monkeys took a stick and experimented, quickly finding a solution.
Then the task was complicated: the banana was put further, and the monkeys were given two sticks, from which one could assemble one long one. This puzzle was overwhelming for the vast majority. The monkeys were furious, but could not figure out what to do, jumped around the cage, banged on the bars with a stick.
The smartest sat down, thought and after a while understood what to do. This moment of transition to visual-figurative thinking is called "gestalt switching": the monkey stopped active, but chaotic and ineffective actions and started thinking. In other words, a thought is a “curtailed action,” that is, an action transferred into the imagination.
Photo: Depositphotos
This is how universal thinking arises: if the chosen algorithm does not fit, the brain looks for a new idea and new possible connections, travels along the "tree of possibilities" until it finds a suitable option. The found solution then affects the external environment (your banana) and goes (possibly along with the new found representation) to the knowledge base, enriching the personal experience.
Emotions play an important role in universal thinking. They modulate the target, modify it. Imagine a robot going to complete a task. Suddenly, everything starts to explode ahead. The machine does not feel fear, so neither the goal nor the line of behavior changes. Explosion - the robot is destroyed. And the person in his place would try to save his life in order to then complete the original task.
Where information is processed
The first task of the brain is pattern recognition. What happens if you see, say, a person's face? Information enters the pupil and is projected onto the retina. The signal is transmitted to the primary visual cortex. It is located closer to the back of the head and is responsible for recognizing only the simplest geometric objects, such as, for example, lines with different angles of inclination. Information is filtered out and transmitted to the secondary visual cortex, where more complex patterns are recognized, for example, semicircles.
Further, the processed information is transmitted to the temporal region of the cerebral cortex (this is the so-called ventral path of processing visual information), where such simple elements as the nose, eye, and ear are recognized. How does this happen? There are neurons that only respond to the nose, there are neurons that only respond to the eye, and so on. At the same time, there are neurons without special specialization, and they can respond to both the nose and the eye.
As a result, the activity of the entire set of these cells is transmitted to the orbitofrontal cortex of the brain in the frontal lobes. There, the picture is brought together, and you recognize the face as a whole. As it progresses, the information is compressed, each time it is encoded by a smaller number of neurons - it seems to be archived. In the anterior lobes of the brain, the storehouse of various high-level images is encoded, with which a person ultimately operates.
The brain is not independent in its actions. It is conducted by the thalamus, a paired organ that ends in the midbrain from the spinal cord. In the thalamus, threads are attached to each section of the cortex. Pulling for them, he activates certain areas that are currently responsible for the optimal solution of the current task.
But even the conductor is not independent. The thalamus is controlled by the so-called basal nuclei (ganglia). Key neurons in these nuclei are highly dependent on dopamine, a neurotransmitter that causes acute pleasure in humans.
We are all dopamine addicts, no matter how sad it is to admit it, the basal nuclei want a lot of dopamine all the time. But it stands out in response to the subjective value of a particular decision for which a certain area of the cortex is responsible.
Human brain
Image: Diomedia
If the value of activating a section of the cortex is high, that is, this decision is supposedly optimal for us in the current situation, then more dopamine will be released, and we will experience joy. What determines value? First, our experience. A small child has minimal experience, and he rejoices in almost everything in the world, any cube. By being curious, a person tries different options, reinforces those that bring subjective benefits and, accordingly, dopamine releases, and avoids those that, on the contrary, cause unpleasant or painful sensations. As you mature and gain experience, the bar for value rises.
Second, value is determined by emotions (and not only positive ones): the brighter they are, the higher the value. Hence, another neurophysiological regulator follows - the areas of the brain responsible for emotions (tonsils, hippocampus, anterior and temporal lobes of the cortex, and others).
It turns out that the brain, in the process of finding the optimal solution to the task before it, works as a self-regulating system. On the one hand, it uses knowledge from experience (that is, from the corresponding parts of the cortex), on the other hand, it weighs these decisions through the system of experiencing emotions (including the same and other parts of the cortex and organs of the limbic system of the brain). All this is collected by the basal nuclei, and through the thalamus, the “go-ahead” is given to activate the area of the cortex that brings the greatest reward to dopamine neurons in the basal and other brain structures.
Cerebellum
The cerebellum plays an extremely important role. It is believed to be responsible for the coordination of movements, the sense of balance and balance. But it is known that in the cerebellum, which accounts for only about 10 percent of the brain's volume, for some reason there are about twice as many neurons as in the rest of the brain - 70 billion versus 30. Is it really that so many nerve cells are needed only to coordinate movements?
Scientists have only recently begun to understand that the cerebellum is responsible not only for movements, but in general for all automatisms, including "curtailed actions" - patterns of mental responses from the knowledge base. For example, for a trained athlete it will not be difficult for a trained athlete to perform a back somersault with a 360-degree screw. He will do this without hesitation, because his cerebellum will extract information from the storage at the right time, the brain will receive the necessary commands, and the body will perform this acrobatic element automatically. The athlete practically does not think, his subconscious is working.
The same seems to be the case with other automatisms, for example, with speech. A person thinks in higher images, and the cerebellum itself decides how best to clothe it in a means of communication. At the same time, of course, long and reliably established speech processing centers in the cerebral cortex are involved, but in close connection with the cerebellum, which continuously offers ready-made, to automatism solutions worked out and / or corrects inevitably arising errors in accordance with them.