What is intelligence?

This question is of concern to many. Theories are constructed with the involvement of various complex concepts, from magic to quantum physics. What if it's easier? The article reflects on some aspects of natural and artificial intelligence, as well as the definition of what intelligence is.
 
Wiki
 
 
The model simulates the simulated system with some accuracy. The greater the accuracy, the more the objects that make up the model will be in the model. Information model has information objects. At the maximum accuracy for each modeled object of the system there will be the corresponding information one. That is, there will be a single mapping. In the future, by the word "model" I mean exactly the mapping of some objects into others, a given correspondence of the real and informational.
 
 
An information system is a system for processing information. Information comes in some way to the input of the information system. To properly process it, the system has an information model.
 
 
Where does the information come from? From some external source. I call it "reality."
 
 
Reality is everything that is outside the information system.
 
 
So, in order to interact with reality, the information system must have some model of this reality. For every real object there must be a corresponding information one, that is, a model of this object. Another information system is also a real object. Accordingly, if we want to somehow interact with it, it also needs a model.
 
 
The need for such a model is clearly visible when developing the API. On the server there is a structure on the output, on the client the same structure at the input. In PHP and JS, the same classes. This is not redundancy, it is the modeling of one system in another.
 
 
 

Intelligence


 
Natural intelligence is also an information system. He has sensors that give incoming information. Based on this information, he builds a model of what happens outside this information system, and makes decisions about actions on the basis of the model. Thus, we can give a definition of what intelligence is.
 
 
Intelligence is the ability of an information system to build a model of reality based on incoming information.
 
 
A computer program is also an information system. Does she have intelligence? No, because the programmer pawns the program. Consider, for example, a program for marking a hard disk. The information on possible formats, the order of execution of commands, response codes is all laid in the source code. The program has a hard disk model, it is updated when receiving responses to commands, but the program does not build it, can not change it.
 
 
Artificial intelligence must build the right model of reality. Including, he should correctly respond to it. At least the same as a person. First, it is a feedback for us, by which we can come to the conclusion that the system has intelligence. You can check this in the debugger, but you need to know what should be there. Secondly, because one of the goals is the fulfillment of human activities in some tasks.
 
 
All animals react to reality in one way or another. Therefore, one can not talk about them whether there is an intellect or not an intellect. You can talk about whether it is more or less, whether it is enough for a certain type of tasks, for constructing certain models.
 
 
Unicellularity has almost no intelligence, because there are no elements that store and process information about the environment. There are only chemical interactions that allow them to function. They can also be considered a model, but not an information model, but a hardware one. Because of its small size and chemical laws, it can not be complicated.
 
 
It is also worth noting that the intellect is closely related to the concepts of object and memory. An object is something that has state and behavior, which we define as one and the same at different times. Memory is needed to store the attributes of objects between these moments, here the model begins. But objects are the topic of a separate conversation.
 
 
 

Sensations


 
People experience pleasant and unpleasant sensations. Consider the concept of pain. Pain is an unpleasant sensation, if present, the brain constantly turns its attention to it. Usually a person can not feel or feel pain at will. Thus at people of unpleasant sensations it is a little, for example, there is still a pavor.
 
 
For an arbitrary information system, one can give such a definition.
 
 
Sensations are a process of perceiving data that enter the information system constantly, require its processing, while the system is designed in such a way that it can not not react to them. Perception is how data affects the model. Their presence can be perceived by the system as positive or negative feedback. It tends to increase the positive connection, to decrease it negative.
 
 
Any signals from the sensors are sensations.
 
 
Can artificial intelligence experience pain?
 
 
First we note that the natural intellect experiences several different sensations for it. Pain is just the name of one of them. It signals damage. Thus, if a certain information system with the intellect gets a sensation with information about the damage, which it determines as unpleasant for it, tends to minimize its influence and can not do it by shutting down, then this is an analogue of pain. It is an analogue, we can only determine the similarity that "Sensation 1" is like human pain, and "Sensation 2" is for the pleasure of something. The main thing here is how the system reacts to these data. How to determine that AI does not lie? Check in the debugger.
 
 
The more features, the more similar, including chemical and electrical processes. With a high probability, animals feel sensations similar to human ones, since they react similarly to humans, they also have nerve cells, which also consist of organic compounds.
 
 
That is, if we make the information model of a person accurate to atoms and electric fields, as well as the surrounding reality with which it will interact, and simulate a cut on the finger, then yes, in this reality it will experience pain. If with the help of sensors it will interact with our reality, then in ours. Since in terms of information processing, it will not be different from other people.
 
 
Is it ethical to make him feel pain against his desire in any reality? I think no. This applies to any other being and any unpleasant sensation. But since this is an information model, it is possible to model another reaction or another way of perception, and then the level of ethics will be different.
 
 
Consider a database management system with monitoring. If a failure occurs, the monitoring subsystem begins to send a notification to the administrator with some frequency with an alarm level. Is it possible to say that such a system is in pain? No. First, it does not have the intellect, the model is embedded in it by the programmer, and there is no such thing as a similar pain. Secondly, although the signal is assessed as unpleasant, it has no desire to minimize it. In general, on the contrary, it was made in order to provide the most accurate model of what is happening to the system. It can be called an analogue of nerve cells, and not a being that has them.
 
 
Consider a bot with AI in a computer game. The game for him is a reality, albeit very limited. For correct actions, information about the damage is required. If the bot receives it at will - performed some actions, drew attention to the wound, began to receive signals about the damage, remembered their parameters, distracted, ceased to act - it is not an analogue of pain. If the bot perceives it as additional information, and not as a negative feeling, the level of which should be minimized, if it does not affect the actions taken for other purposes (for example, there is no desire to pull his hand away if he received a blow on the wound on his hand), too no.
 
 
 

Other concepts


 
What is consciousness? It can be defined as the process of the functioning of the intellect, the process of obtaining information and updating the model of reality. It is closely connected with the concept of "I". "I" is an element in the model that represents the information system itself. Is there a "I" in dogs? Yes, this is the object with which they connect the team "Sit-Lay". This means that dogs have consciousness and intelligence. But to multiply for example dogs do not know how. Perhaps this is due to the fact that for higher levels of abstraction, additional information elements are needed that will observe the lower levels and build their model.
 
 
What is understanding? What does it mean when we say that a person has understood something? This means that he has the right model for what he analyzed. Information elements of the model correspond to elements of the real system. Understanding is building the right model. Why is experience so important? This is a model. It allows you to model processes in an informative form, and therefore plan and choose the right option.
 
 
There are neural networks, they themselves build a model for incoming information. Do they build the right model? The results show that somehow not very. They usually require a special sample, this is also a prepared model. But they can build a model, so they have some level of intelligence.
 
+ 0 -

Add comment