Nikolai Nikolaevich Kretov: What is artificial intelligence?

What is artificial intelligence? Now, perhaps, few people ask such a question. Even a small child will try to answer this question.

Returning to the last century, let's plunge into the atmosphere of creation and discovery by Norbert Wiener of the new science of "Cybernetics", which laid the foundation for the creation of "smart machines". Cybernetics is more about living organisms, man and society than about machines. The machine is rather a tool and model in general cybernetics, rather than a subject of study. Wiener himself thought so. Wiener compared machines created by man and machines created by nature and concluded that machines created by nature (people) are more efficient and adaptable, but machines created by man gave man a tool for natural experiment and mental experiment.

Wiener had no doubts about the possibility of learning machines, as well as living systems, he cited the example of learning playing machines, including chess. But he, of course, did not know, although, I think, he had no doubt that such a supercomputer would be created that would be able to play the most difficult chess games on equal terms with world chess champion Garry Kasparov.

And here's another example: the creation of "water" (Merman) by a group working with computer animation of the underwater world. They modeled the nervous systems of the inhabitants of the underwater world, and the underwater man (Merman) was trained to "think", trying to avoid a collision with a shark, although such actions were not included in the program. What is not reasonable behavior, but it, obviously, was the result of the action of the model of the nervous system of a living being embedded in the program, for whose life program it is common to hide in case of danger.

Without trying to shine with a sophisticated game of natural intelligence, I will give the characteristics of artificial intelligence that were proposed by L.T. Kuzin, and, in my opinion, are the most appropriate:
the presence in them of their own internal model of the external world;
this model provides individuality, relative independence of the system in assessing the situation, the possibility of semantic and pragmatic interpretation of requests to the system;
the ability to deductively infer, i.e., to generate information that is not explicitly contained in the system;
this quality allows the system to construct an information structure with new semantics and practical orientation;
the ability to operate in situations related to various aspects of fuzziness, including the "understanding" of natural language;
ability to adapt.

Agree that these characteristics quite accurately correspond to the characteristics of our intellect.

Of course, the previous phrase is more relevant to a human reader, but if this is read by a search engine indexing robot or a fairly highly organized browser, then I will be happy to listen to your thoughts, dear robot, on this matter, if your operating system has of this pair of quanta of free time.

In the early stages of development, scientists set the task of creating a general intelligence that could operate in any area, but then they came to the creation of systems that solve intellectual problems of any limited areas.

In the 1980s and 1990s, attempts were also made to create universal expert systems that contained databanks of millions of different statements connected by semantic links, in the hope that many tasks could be reduced to a sequence of these statements. Some of these systems are replenished to this day, never having learned to solve problems from the real world.

Many mistakes in the creation of artificial intelligence could have been avoided if one had immediately come to the conclusion that intelligence did not appear in humans, but is a product of the successive evolution of more complex adaptive behavior. Many living creatures, even those with a primitive nervous system, have a peculiar intelligence or the ability to solve their problems and adapt in an unknown environment, the ability to foresee and choose between different options for action, to learn and form the necessary knowledge. Therefore, it is worth paying attention to the models of properties and our smaller brothers.

When scientists from various fields of knowledge realized that they have many of the same questions about the nature of the human mind, and they all use the appropriate methods to solve them. Over the past years, there has been an unprecedented leap in information technology, as well as huge developments in the science of the brain, which provides new opportunities for the development of artificial intelligence.

At present, there is a kind of “linking” of human consciousness with the information computer environment, and many processes of human life are transferred to the virtual information computer environment. The resulting system benefits from the synergistic multiplication of the strengths of each of its constituents. This is the principle of maximum efficiency in the use of computer devices.

Thanks to the Internet, huge amounts of information become available to any user, regardless of location. On the Internet, people carry out their professional activities, study with the help of computer programs, read, get the necessary information, communicate with colleagues and friends... A person interacts with computer systems using the brain and its sensory receptors. It turns out, as it were, a united mind with a common intellect: a person and a machine. This interaction undoubtedly still has a long and fascinating road ahead, and I can't help but follow its development with excitement and exciting interest, trying to anticipate what lies there, behind the next turn in the winding path of progress. And of course, I will continue to share my new thoughts on this and many other interesting topics on the pages of the site with pleasure.