Thinking and the principle of "mad programmer"
I will try to disrupt the sequence of the narrative, which was originally planned. After describing the Emotional Computer, I wanted to write about the device of associative memory, but from the comments I understand that readers are already tired of boring lectures and are waiting for action.
First, tell the story. I had a friend, a talented programmer. The man is really not ordinary. When he studied at the St. Petersburg Polytechnic, he was a Lenin scholarship. Maybe someone remembers such a computer BK-0010, and so the best games were written by him, and he wrote them in codes. For the first channel I wrote a “schedule” on the Amiga, under which the weather forecast was turned.
He lived brightly, abused everything that was possible, and even more so than not. Then, something happened to the health in general and to the head in particular. He now lives, practically, without leaving home. His main occupation of the conversation with the TV. At the same time, he has four televisions turned on, all four he tunes to conversational programs. Then it happens like this. He utters the phrase. And the TVs "respond" to it. That is, they, TVs, simply give out a lot of diverse text, and he “on the fly” constructs the “answer”, which consists of these television words, but at the same time fits into the context of the conversation. So he, day and night, communicates, as he says, with "the world's best interlocutor."
')
It is not commonplace that it is:
He assures that with some training, anyone can learn to talk like that. I have not tried, I'm afraid, suddenly sit down.
So this principle, which I call the “principle of a mad programmer”, is surprisingly similar to how, from my point of view, the human thought process proceeds.
Since I missed the description of the associative memory, I’ll just give you the main points. If there is interest, I will decipher them later.
Imagine that we have a very good sensory system that receives information from the outside world and translates it into a very convenient form. For example, the sound subsystem gives us ready-made phonemes, from which you can jokingly combine words, she herself selects the signals of laughter, crying, pain, pain and everything else, recognizes the sounds of nature. To set unconditioned reflexes, we do not need to philosophize slyly. If we want to get a reflex fright, then by ready-made signs (unexpected sound, someone's frightened scream, fast movement in the peripheral area of ​​vision, etc.) we will specify the triggering of the corresponding reflex.
Imagine that we have a limbic system, which, by a set of signs at the output of the sensory system, forms a reflex response of emotions and sensations.
Imagine that we have an associative memory that:
Memories of the first level are formed on the basis of the sensory layer when the emotions tied directly onto the sensory layer trigger (for the most part we call these emotions sensations).
The formation of the upper levels is delayed in time and begins with the accumulation of sufficient experience in the first layer.
There are “tags” that allow us to identify certain integral events or phenomena. "Tags" can be quite diverse, they all cause an answer to the sensory layer. However, we can not feel the presence of the label and not guess about its existence. A vivid example of "tags" - pheromones.
Pheromones are biological markers of their own species, volatile chemosignals that control neuroendocrine behavioral reactions, developmental processes, and many processes associated with social behavior and reproduction (Wikipedia).
Pheromones identify for us who is a man, who is a woman, trigger emotions associated with sexual desire, define a "leader" and much more.
“Tags” are required for the “imprinting of emotions” procedure. The essence of imprinting is that the “label” causes the appearance of emotion, and memory is shaped by a memory in which the arising emotion and object or situation are linked, in its reflection on the sensory and subsequent levels.
The result of imprinting is that the phenomenon, through its signs, is associatively associated with a certain emotion. Subsequently, the emotion arises regardless of the presence of the "tag", but as a reaction to the signs of the phenomenon. Thus, emotions are “launched” for which we do not observe an immediate sensory cause. Actually, where we can trace the receptor source, we call sensation, and where there is no emotion.
The formation of the upper levels of memory occurs on the basis of the underlying, under the influence of the corresponding emotions. You can separately highlight the emotion that makes, perhaps, the most significant contribution - curiosity. Positive transition of emotional state - satisfaction of curiosity. To implement the emotions of curiosity, a separate functional system is needed.
As a result of the formation of memory, the following property arises in it: the picture of the external world causes a certain activation of the groupings of neurons associated with memories containing elements of this picture.
Simply put, when something happens to us, we display it with all of our previous memories.
Well, now let's move on to the principle of “crazy programmer”, which is also “the principle of associative blurring”.
We continue to simulate the brain. Let's make an associative memory such that it serves three processes in parallel. That is, we will model the memory in three “layers”, it will be the same memory with the same connections, but in different pictures of activity (do not confuse with memory levels).
The state of our brain will change rhythmically, there will be a change of "frames". There may be several processes with different frequencies. For simplicity, consider the overall rhythm.
Information from the outside world enters the associative memory and causes the activation of those neurons that “recognize” familiar signs. There is a picture of the activity of memory, corresponding to what we perceive. This picture describes what is happening around in terms of sensory perceptions and our experience gained in memory. This picture is not accidental, it corresponds to some existing event, "it is comprehended." We will interpret this state of activity of memory as a state of the “first layer”.
Now imagine that in parallel there is a picture of activity, describing what we are thinking about. These are the images that we represent, the inner speech that we utter to ourselves and other "fantasies." Now we are talking about a single frame of such “thoughts”. The picture of our thought is similar in structure to the picture of the external world and, like that, it corresponds to a certain picture of the activity of associative memory. We will speak of this as the state of the “third layer”. "Second layer" is reserved.
Emotions will give us an assessment of the state of the first layer and separately assess the state of the third layer.
We introduce the “focus of attention” characteristic, which will describe that “we have priority over” layer one or layer three. We will make the “attention focus” a function: its previous state, the emotional evaluation of layer one and layer three. When changing states, we will keep the “focus of attention” if there is no reason to change it, and we will adjust it in the direction of that layer, the final emotional assessment of which will prevail.
Let us collect at the sensory level a picture caused by external information and “projected” from layer three, which corresponds to our current “frame” of thinking. The power of “projection” will be correlated with the state of “focus of attention”. We can get the range from "we are closely watching what is happening, with bated breath," to "distracted and gone into himself."
The current emotional state is defined as a weighted sum of the state of layers one and three, taking into account the focus of attention.
We produce an act dictated by memory (see "Emotional computer").
Create a new memory that takes into account our current state and its emotional evaluation. A new memory can reflect real experience, or our imagination and its emotional evaluation.
Note that the “collected” earlier picture and the state of emotions correspond to our “awareness” at the current moment.
Now let's place, conditionally, between the first and the third “second layer”. Let's start the procedure of “associative blurring”.
We will activate all the memory elements associated with those that are active in the first and third layers. We will take into account the power of activity and the focus of attention. We get some kind of blurred cloud of associations. This cloud will not be random, the concepts (memories) that form it will be an associative consequence of what we have recognized in the outside world and our current “thought”. However, this blurred cloud will be "meaningless." The picture of activity will not correspond to any “meaningful” memory or fantasy. It will be a certain analogue of speech, heard immediately from four televisions. Now we need a "crazy programmer" who will subtract from this cloud, the "next phrase."
Such an element will be the mechanism of emotions. Blurred picture will trigger the trigger of various conflicting emotions. When many signs are active, there is a great chance that among them there will be those that are needed to activate a certain emotion. Now choose the strongest emotion and go backwards. Leave active only those concepts that caused its operation. The final picture will have an interesting property - it will be “comprehended”.
Actually, we will rewrite this picture in layer three and call it our next thought.
Along the way, for a short period of time the blurred associative picture caused the appearance of a “cloud of emotions”, this “parasitic background” can be called a “mood”. Its properties are surprisingly consistent with the properties of mood in humans.
Finish the beat. And repeat all over again.
That's where you can simulate thinking.
In order for the narrative not to grow, I had to omit many significant moments, but I managed to convey the main direction of thought.
Finally, for example, I will give you how two basic approaches to solving problems are implemented in this scheme.
Logical thinking. We know certain rules, we narrow the associative area, leaving the concepts relating to the subject area we need. With associative blurring, the set of associations is not large and the following thought is practically determined by a “strong” associative transition.
Creative thinking. Logical paths leading to the goal is not visible. We begin the associative search. We are doing a broad associative blur, there gets everything that can push us to the answer. Associations caused by the outside world bring in a random component that can be decisive. If you can find an answer from a set of concepts, that is, to put together an image that will evoke the “eureka” emotion, then we may find it. If not, we will continue to cast a net. We will find "intermediate islands". They will help us to form new pictures of associative blurring in which all the necessary concepts are possible will be gathered.
Addition.
Trying to simplify, I, probably, in vain lowered one more process. The picture in a layer one causes an emotional assessment, this assessment on the strongest emotion “simplifies” the picture of this layer, leaving only what is connected with it. The full picture of what is happening remains at the sensory level, and on the layer one there is an accentuated part of it. This state of layer one corresponds to the selection of what our attention is directed at. You can “parallelize” in this way several emotions by entering several additional layers.
I would add that this process really wants to be linked with the alpha rhythm, and the process taking place in layer three with the beta rhythm.
First, tell the story. I had a friend, a talented programmer. The man is really not ordinary. When he studied at the St. Petersburg Polytechnic, he was a Lenin scholarship. Maybe someone remembers such a computer BK-0010, and so the best games were written by him, and he wrote them in codes. For the first channel I wrote a “schedule” on the Amiga, under which the weather forecast was turned.
He lived brightly, abused everything that was possible, and even more so than not. Then, something happened to the health in general and to the head in particular. He now lives, practically, without leaving home. His main occupation of the conversation with the TV. At the same time, he has four televisions turned on, all four he tunes to conversational programs. Then it happens like this. He utters the phrase. And the TVs "respond" to it. That is, they, TVs, simply give out a lot of diverse text, and he “on the fly” constructs the “answer”, which consists of these television words, but at the same time fits into the context of the conversation. So he, day and night, communicates, as he says, with "the world's best interlocutor."
')
It is not commonplace that it is:
- self talk;
- randomized external source;
- the space of possible answers is narrowed by a set of television words.
He assures that with some training, anyone can learn to talk like that. I have not tried, I'm afraid, suddenly sit down.
So this principle, which I call the “principle of a mad programmer”, is surprisingly similar to how, from my point of view, the human thought process proceeds.
Since I missed the description of the associative memory, I’ll just give you the main points. If there is interest, I will decipher them later.
Imagine that we have a very good sensory system that receives information from the outside world and translates it into a very convenient form. For example, the sound subsystem gives us ready-made phonemes, from which you can jokingly combine words, she herself selects the signals of laughter, crying, pain, pain and everything else, recognizes the sounds of nature. To set unconditioned reflexes, we do not need to philosophize slyly. If we want to get a reflex fright, then by ready-made signs (unexpected sound, someone's frightened scream, fast movement in the peripheral area of ​​vision, etc.) we will specify the triggering of the corresponding reflex.
Imagine that we have a limbic system, which, by a set of signs at the output of the sensory system, forms a reflex response of emotions and sensations.
Imagine that we have an associative memory that:
- - able to memorize the current picture at the output of the sensory system, as well as to memorize our actions and then what kind of change in the emotional state this was accompanied by;
- - able to manage our behavior, as described in the post "Emotional computer";
- - has a multi-level structure.
Memories of the first level are formed on the basis of the sensory layer when the emotions tied directly onto the sensory layer trigger (for the most part we call these emotions sensations).
The formation of the upper levels is delayed in time and begins with the accumulation of sufficient experience in the first layer.
There are “tags” that allow us to identify certain integral events or phenomena. "Tags" can be quite diverse, they all cause an answer to the sensory layer. However, we can not feel the presence of the label and not guess about its existence. A vivid example of "tags" - pheromones.
Pheromones are biological markers of their own species, volatile chemosignals that control neuroendocrine behavioral reactions, developmental processes, and many processes associated with social behavior and reproduction (Wikipedia).
Pheromones identify for us who is a man, who is a woman, trigger emotions associated with sexual desire, define a "leader" and much more.
“Tags” are required for the “imprinting of emotions” procedure. The essence of imprinting is that the “label” causes the appearance of emotion, and memory is shaped by a memory in which the arising emotion and object or situation are linked, in its reflection on the sensory and subsequent levels.
The result of imprinting is that the phenomenon, through its signs, is associatively associated with a certain emotion. Subsequently, the emotion arises regardless of the presence of the "tag", but as a reaction to the signs of the phenomenon. Thus, emotions are “launched” for which we do not observe an immediate sensory cause. Actually, where we can trace the receptor source, we call sensation, and where there is no emotion.
The formation of the upper levels of memory occurs on the basis of the underlying, under the influence of the corresponding emotions. You can separately highlight the emotion that makes, perhaps, the most significant contribution - curiosity. Positive transition of emotional state - satisfaction of curiosity. To implement the emotions of curiosity, a separate functional system is needed.
As a result of the formation of memory, the following property arises in it: the picture of the external world causes a certain activation of the groupings of neurons associated with memories containing elements of this picture.
Simply put, when something happens to us, we display it with all of our previous memories.
Well, now let's move on to the principle of “crazy programmer”, which is also “the principle of associative blurring”.
We continue to simulate the brain. Let's make an associative memory such that it serves three processes in parallel. That is, we will model the memory in three “layers”, it will be the same memory with the same connections, but in different pictures of activity (do not confuse with memory levels).
The state of our brain will change rhythmically, there will be a change of "frames". There may be several processes with different frequencies. For simplicity, consider the overall rhythm.
Information from the outside world enters the associative memory and causes the activation of those neurons that “recognize” familiar signs. There is a picture of the activity of memory, corresponding to what we perceive. This picture describes what is happening around in terms of sensory perceptions and our experience gained in memory. This picture is not accidental, it corresponds to some existing event, "it is comprehended." We will interpret this state of activity of memory as a state of the “first layer”.
Now imagine that in parallel there is a picture of activity, describing what we are thinking about. These are the images that we represent, the inner speech that we utter to ourselves and other "fantasies." Now we are talking about a single frame of such “thoughts”. The picture of our thought is similar in structure to the picture of the external world and, like that, it corresponds to a certain picture of the activity of associative memory. We will speak of this as the state of the “third layer”. "Second layer" is reserved.
Emotions will give us an assessment of the state of the first layer and separately assess the state of the third layer.
We introduce the “focus of attention” characteristic, which will describe that “we have priority over” layer one or layer three. We will make the “attention focus” a function: its previous state, the emotional evaluation of layer one and layer three. When changing states, we will keep the “focus of attention” if there is no reason to change it, and we will adjust it in the direction of that layer, the final emotional assessment of which will prevail.
Let us collect at the sensory level a picture caused by external information and “projected” from layer three, which corresponds to our current “frame” of thinking. The power of “projection” will be correlated with the state of “focus of attention”. We can get the range from "we are closely watching what is happening, with bated breath," to "distracted and gone into himself."
The current emotional state is defined as a weighted sum of the state of layers one and three, taking into account the focus of attention.
We produce an act dictated by memory (see "Emotional computer").
Create a new memory that takes into account our current state and its emotional evaluation. A new memory can reflect real experience, or our imagination and its emotional evaluation.
Note that the “collected” earlier picture and the state of emotions correspond to our “awareness” at the current moment.
Now let's place, conditionally, between the first and the third “second layer”. Let's start the procedure of “associative blurring”.
We will activate all the memory elements associated with those that are active in the first and third layers. We will take into account the power of activity and the focus of attention. We get some kind of blurred cloud of associations. This cloud will not be random, the concepts (memories) that form it will be an associative consequence of what we have recognized in the outside world and our current “thought”. However, this blurred cloud will be "meaningless." The picture of activity will not correspond to any “meaningful” memory or fantasy. It will be a certain analogue of speech, heard immediately from four televisions. Now we need a "crazy programmer" who will subtract from this cloud, the "next phrase."
Such an element will be the mechanism of emotions. Blurred picture will trigger the trigger of various conflicting emotions. When many signs are active, there is a great chance that among them there will be those that are needed to activate a certain emotion. Now choose the strongest emotion and go backwards. Leave active only those concepts that caused its operation. The final picture will have an interesting property - it will be “comprehended”.
Actually, we will rewrite this picture in layer three and call it our next thought.
Along the way, for a short period of time the blurred associative picture caused the appearance of a “cloud of emotions”, this “parasitic background” can be called a “mood”. Its properties are surprisingly consistent with the properties of mood in humans.
Finish the beat. And repeat all over again.
That's where you can simulate thinking.
In order for the narrative not to grow, I had to omit many significant moments, but I managed to convey the main direction of thought.
Finally, for example, I will give you how two basic approaches to solving problems are implemented in this scheme.
Logical thinking. We know certain rules, we narrow the associative area, leaving the concepts relating to the subject area we need. With associative blurring, the set of associations is not large and the following thought is practically determined by a “strong” associative transition.
Creative thinking. Logical paths leading to the goal is not visible. We begin the associative search. We are doing a broad associative blur, there gets everything that can push us to the answer. Associations caused by the outside world bring in a random component that can be decisive. If you can find an answer from a set of concepts, that is, to put together an image that will evoke the “eureka” emotion, then we may find it. If not, we will continue to cast a net. We will find "intermediate islands". They will help us to form new pictures of associative blurring in which all the necessary concepts are possible will be gathered.
Addition.
Trying to simplify, I, probably, in vain lowered one more process. The picture in a layer one causes an emotional assessment, this assessment on the strongest emotion “simplifies” the picture of this layer, leaving only what is connected with it. The full picture of what is happening remains at the sensory level, and on the layer one there is an accentuated part of it. This state of layer one corresponds to the selection of what our attention is directed at. You can “parallelize” in this way several emotions by entering several additional layers.
I would add that this process really wants to be linked with the alpha rhythm, and the process taking place in layer three with the beta rhythm.
Source: https://habr.com/ru/post/105603/
All Articles