Military Review

Artificial Intelligence. The future of Russia's national security?

85

Ten years of development



It's no secret that artificial intelligence is penetrating deeper into the lives of ordinary people around the world. This is facilitated by the global spread of the Internet, as well as a serious increase in computing power. Neural networks, which have a certain similarity with the human brain, have made it possible to qualitatively improve the work of the software being developed. There are, however, a couple of clarifying points: neural networks are still very far from the level of the human brain, especially in terms of energy efficiency, and the operation algorithms are still extremely difficult to understand.


Money in the artificial intelligence industry, despite some limitations and resonant incidents with autopilot cars, goes a wide river. Last year, according to the approved National Strategy, the market for IT solutions in this area exceeded 21,5 billion. God knows how much, but it will only increase every year, and by 2024, the total AI in the world will be conditionally worth 140 billion, and the potential economic growth from the introduction of AI by this time will reach a decent 1 trillion. dollars. Actually, an attempt to keep up with world trends was the approval by President Vladimir Putin of the October 10 of October 2019 of the aforementioned National Strategy. At the same time, the program itself declares not just reducing the gap with world leaders, but entering the number of top players in this market. And it is planned to do this by the 2030 year. Among the obvious obstacles to this path will be protectionist statements by a number of countries that any Russian software carries a potential danger.

Where are they going to realize the “limitless” capabilities of AI on Russian soil? First of all, it is the automation of routine operations along with the replacement of people in hazardous industries (read: including in the army). Further, serious work is planned with big data that has been recently generated simply like an avalanche. It is assumed that they will be able to improve forecasts for managerial decisions, as well as optimize the selection and training of personnel. Health care with education in ten years will also be active AI users. In medicine, prophylaxis, diagnostics, dosage of drugs and even surgical intervention will be partially or completely given to the machine mind. In schools, AI will be involved in individualizing learning processes, analyzing a child’s propensity for professional activity and early identification of talented youth. In the strategy, one can find a provision on the "development and implementation of educational modules in the framework of educational programs at all levels of education." That is, the basics of AI will be taught at school?

As usual, in addition to the tangible results of the development of AI, the scientific environment will be required to increase the number and citation index of articles by Russian scientists in world specialized publications. And by the 2024 year, that is, very soon, the number of citizens with competencies in the field of AI should increase in Russia. Including this will be realized by attracting domestic specialists from abroad, as well as attracting foreign citizens to work on this topic in Russia.

However, AI has one controversial quality, which the strategy is supposed to solve "by developing ethical rules for human interaction with artificial intelligence." It turns out that the cold calculation of computer intelligence forces it to make biased and unfair generalizations.

AI bias


Among the mass of questions about the functioning of modern AI systems, the currently imperfect algorithms for autopiloting wheeled vehicles, which still do not allow legislatively authorizing their widespread use, are especially notable. Most likely, in the foreseeable future we will not see AI-driven cars on our roads. For this we have inappropriate road conditions, and the climate does not favor using autopilot all year round: dirt and snow will quickly “blind” the sensor systems of the most advanced robot. In addition, the massive introduction of AI will inevitably take away the work of millions of people around the world - they will either have to relearn or spend the rest of their days in idleness. It’s fair to say that various newfangled “Atlas of the professions of the future” sometimes carry outright nonsense: in one of them dated the 2015 year, for example, the profession of accountant, librarian, corrector and tester should have become obsolete. But, nevertheless, the profile of most professions will change, and the negative factor of AI will prevail here. In any case, the prospects for the further implementation of AI in society pose many questions for state regulators. And how to solve them, it seems, few people know.


Another issue that is already looming on the horizon is AI bias in decision making. The Americans were one of the first to encounter this when they introduced the COMPAS system in 15 states to predict cases of relapse. And everything seemed to start very well: it was possible to develop an algorithm that, based on the mass of data (Big Data), was able to formulate recommendations on the severity of punishment, the regime of correctional institutions or early release. The programmers rightly argued that in the pre-dinner time a hungry judge could endure an excessively harsh punishment, and a well-fed one, on the contrary, would be too lenient. The AI ​​should add a cold calculation to this procedure. But it turned out that COMPAS and all similar programs were racist: AI was twice as often mistakenly accusing African Americans of a tendency to relapse than white people (45% vs 23%). AI generally regards fair-skinned criminals as people with a low level of risk, since they are statistically less likely to violate the law - therefore, their forecasts are more optimistic. In this regard, in the United States, voices are heard more and more often about the abolition of AI in resolving issues of bail, sentencing and early release. At the same time, US justice has nothing to do with the program code of these systems - everything is purchased from third-party developers. The software systems that work on the streets of many cities around the world, Predpol, HunchLab and Series Finder, have already statistically reliably proven their effectiveness: crime is declining, but they are not without racial prejudice. The most interesting thing is that we do not know what other “cockroaches” are sewn into the artificial brains of these systems, since many analysis parameters are classified. There are also doubts that the developers themselves understand how the AI ​​makes certain decisions, which parameters it considers key. Similar situations are developing not only in law enforcement and justice, but also in recruiting agencies. In most cases, AI prefers hiring young men, leaving aside the weaker sex and age candidates. It is funny that the values ​​of the West, which they so eagerly propagate there (equality of the sexes and races), are violated by the latest Western achievement - artificial intelligence.


The conclusion from a short digression into the theory and practice of AI is as follows. It’s one thing when our data from social networks and other sources are massively processed for the purpose of marketing or political manipulation, and another thing when the sword of justice is passed into the hands of the AI ​​or, worse, an arsenal of national security. The price of a biased decision increases many times, and something needs to be done. Whoever succeeds will become the real ruler of the 21st century.
Author:
Photos used:
medium.com, unbabel.com
85 comments
Ad

Subscribe to our Telegram channel, regularly additional information about the special operation in Ukraine, a large amount of information, videos, something that does not fall on the site: https://t.me/topwar_official

Information
Dear reader, to leave comments on the publication, you must sign in.
  1. rocket757
    rocket757 11 December 2019 18: 10
    +1
    Artificial Intelligence. The future of Russia's national security?

    If behind their iPhones, the upper ones completely forget that the main thing is always this MAN !!!
    For a long time, they don’t reach us!
    1. Svarog
      Svarog 11 December 2019 19: 17
      +5
      Quote: rocket757
      If behind their iPhones, the upper ones completely forget that the main thing is always this MAN !!!

      Victor, and if you look from the other side .. I think if you cut down all the thought, the Federation Council, the government and replace all of them with AI ... I think there will be more benefits. More economical, exactly! Artificial intelligence does not need to think about how to lobby laws in its favor, how to steal, arrange its relatives ... In general, it is clean about all kinds of obligations and sound decisions can be expected. Another question is what to bring to life, all the same to people, and here you are definitely right. hi
      1. rocket757
        rocket757 11 December 2019 20: 16
        +8
        Vladimir soldier
        Quote: Svarog
        Artificial intelligence does not need to think

        Immediately I remember Soviet humor and anecdotes! About the automated servus in the hotel "Cosmos" (he himself observed) that it was, as it were, automated, but for some reason, NOBODY allowed him to command / control! About the automated system for controlling the quality of parts at the plant in Togliatti and the folded arm of the manipulator!
        In short, as soon as there is even a tiny danger of replacing officials with AI, he will immediately be declared an enemy of the state, people, the fifth and sixth column!
        1. Narak-zempo
          Narak-zempo 12 December 2019 00: 22
          +3
          Quote: rocket757
          About an automated control system for the quality of parts at a factory in Togliatti and a rolled up arm of the manipulator!

          Almost no joke. Mother worked at a brick factory and said that there the emergency gas cut-off drives in the furnaces were screwed. Because a very smart protection system worked when the pressure deviated from the norm, sometimes several times per shift, and then restarting the furnace was hemorrhagic. A couple of times because of this, a fire started, but it was possible to extinguish and hush on its own.
          1. rocket757
            rocket757 12 December 2019 06: 35
            0
            Quote: Narak-zempo
            Almost no joke.

            I had to see this more than once ... and the consequences of raking too.
        2. Fat
          Fat 12 December 2019 21: 16
          +1
          Quote: rocket757
          In short, as soon as there is even a tiny danger of replacing officials with AI, he will immediately be declared an enemy of the state, people, the fifth and sixth column!

          Wow! The social base of the Neoluddite movement was designated)))
          1. rocket757
            rocket757 13 December 2019 07: 46
            +1
            Quote: Thick
            social base of the neoluddite movement

            Nobody could overcome the instinct of self-preservation closer to the sweet feeder!
            Only radical organizational and technical measures ... i.e. fire all naf!
            1. Fat
              Fat 13 December 2019 19: 18
              +1
              Quote: rocket757
              Quote: Thick
              social base of the neoluddite movement

              Nobody could overcome the instinct of self-preservation closer to the sweet feeder!
              Only radical organizational and technical measures ... i.e. fire all naf!

              And where to put all the extra bureaucratic code? On a Far Eastern hectare? With golden rashutes? No, it should be done slowly and quietly, without attracting attention, masking the use of AI with incomprehensible terms such as "digitalization of executive authorities"
              1. rocket757
                rocket757 13 December 2019 19: 56
                0
                Quote: Thick
                , one must slowly and quietly, without attracting attention

                Do you want to do something with the bureaucracy, the quietest ??? Short up?
                Try it, but I DO NOT BELIEVE in all!
      2. Golovan Jack
        Golovan Jack 11 December 2019 21: 05
        0
        Quote: Svarog
        to reduce the whole Duma, SF, government and replace them with all AI .. I think there will be more benefits. More economically accurate

        "Svarog", you haven't read this, of course:

        ... DAD CHARIA (ex. Rezhikhau; name changed to avoid confusion with the name of the supplicant)

        Find, read, recommend. It's about AI just from the classics.

        Very informative yes
      3. Starover_Z
        Starover_Z 11 December 2019 23: 43
        +3
        Quote: Svarog
        Artificial intelligence does not need to think about how to lobby laws in its favor, how to steal, arrange its relatives ... In general, it is clean about all kinds of obligations and sound decisions can be expected.

        After reading the topic, I remembered the story of Ilya Varshavsky "Judge", which I read in my distant childhood.
        He's like an irony about the "work" of AI. Here is a short excerpt from the story:
        I was the first defendant to appear before the Supreme Electronic Judge Donomagi.

        I finished talking.
        The car was silent. Only periodically flashing light on her panel indicated that she continued analysis.
        It seemed to me that the rhythm of her work was different than during the interrogation. Now, in the slow flickering of the bulbs, I even imagined some semblance of compassion.
        “Is it possible,” I thought, “an automaton created to protect the Law of those who distorted my life, touched by my story?
        With a beating heart, in complete silence, I waited for a decision of my fate.
        The hours passed, and my judge was still thinking.
        Finally, the verdict sounded:
        "Execute and pardon posthumously."

        Think Prospects, AI Proponents
        when the sword of justice is handed over to the AI, or worse, the arsenal of national security.
        1. Fat
          Fat 12 December 2019 21: 41
          +1
          Quote: Starover_Z
          Finally, the verdict sounded:
          "Execute and pardon posthumously."

          Perhaps the machine had an excellent sense of humor.
          If the recognizing machine (perceptron) answers any question with the word "Nonsense", this does not necessarily mean a malfunction, perhaps it just has a philosophical mood (c) V. Savchenko
    2. Chaldon48
      Chaldon48 12 December 2019 05: 26
      +2
      When the bias of AI reaches a critical point, a revolutionary situation may arise when the lower classes do not want to put up with the situation of second-class people anymore, and the lazy tops will either completely dumb or decide that they can do anything, it will come to the use of firearms against the protest masses, with all consequences arising from here.
      1. rocket757
        rocket757 13 December 2019 07: 49
        0
        Quote: Chaldon48
        When AI bias reaches a critical point, it may occur

        Life is generally an inert thing, and even in a socially heterogeneous environment, everything can boil for a very long time and without violent consequences, if the pressure is periodically released!
    3. I am
      I am 16 January 2020 19: 54
      0
      The last words that the terminator heard in Russia: - Damn, Tolyan, but there are 50 kilograms of copper ...
      But before these lives this thing still has time to spoil us ........ but you can’t forget about the switch, and about scrap too))))
  2. knn54
    knn54 11 December 2019 18: 11
    +7
    "The electronic brain will think for us, just like the electric chair dies for us."
    Stanislav Jerzy Lets,
    1. d ^ Amir
      d ^ Amir 11 December 2019 19: 19
      +5
      we are not threatened with a revolt of cars !!! in our version it will not think !!! while one part II will require a certificate of compliance with its II certificate, GOST, Customs Code, etc. the second half of the time will not remain thinking
  3. Freemason Mason
    Freemason Mason 11 December 2019 18: 12
    0
    At the same time, the program itself declares not just reducing the gap with world leaders, but entering the number of top players in this market. And it is planned to do this by 2030

    In the West, IT giants are engaged in AI development; they invest money in promising startups working in the same direction. Will Russia succeed in creating companies comparable to Microsoft, Google, etc. by 2030 - time will tell, but I personally can hardly believe it ...
    1. The comment was deleted.
    2. ares1988
      ares1988 11 December 2019 18: 26
      -1
      Well, yes, just remember: whose is the Big Five IT.
      1. eagle owl
        eagle owl 11 December 2019 18: 29
        0
        Indian coders?
        1. ares1988
          ares1988 11 December 2019 18: 40
          -2
          American (non-Russian) companies.
          1. eagle owl
            eagle owl 11 December 2019 18: 44
            +2
            Not American. but transnational. And TNCs are ruining the US economy now. What is Amazon's turnover? Including in the USA? How many taxes have been paid in the USA? NOOL Dollars, zero cents
            1. Victoria-V
              Victoria-V 11 December 2019 21: 39
              -1
              Why deceive ? Do you think fools are sitting here? Nothing was paid to the US federal budget, because in 2017 1 billion (profit 11,2 billion dollars), and in 2018 already 1,18 billion dollars was paid to local budgets (analogues of our municipal budgets), state budgets and international taxes.
              1. eagle owl
                eagle owl 11 December 2019 22: 12
                +1
                Trump accused Amazon of underpayment of taxes and abuse of the US postal service

                https://www.interfax.ru/world/606099
                Amazon does not pay taxes for the second year in a row

                https://zen.yandex.ru/media/id/5a64da12581669506aff4cc1/amazon-vtoroi-god-podriad-ne-platit-nalogi-5c6dcc6eacf80d00b2276f95
                Last week was not the best for Amazon, but the company still has good news - the giant in the field of e-commerce will not pay a single cent of tax on profits in the amount of $ 11,2 billion received last year.

                https://fomag.ru/news/kak-amazon-udaetsya-izbegat-nalogovykh-vyplat/
                1. Victoria-V
                  Victoria-V 12 December 2019 13: 31
                  -1
                  I can only wish you a speedy recovery.
                  1. eagle owl
                    eagle owl 12 December 2019 13: 37
                    -2
                    I cannot wish you mutually - this is not treated. Living in a fantasy world is incurable
            2. ares1988
              ares1988 12 December 2019 07: 27
              0
              Well, how will this help the Russian Federation?) No way. The largest IT companies today are not ours.
              1. Nastia makarova
                Nastia makarova 12 December 2019 12: 18
                -2
                is it google and facebook? pout rating they have
                1. ares1988
                  ares1988 12 December 2019 12: 38
                  0
                  Well, yes, they are exaggerated, and "Gazprom is an undervalued company") I don't know what kind of ratings you are talking about, I meant the well-established term "big five" (when it comes to IT companies). These are Amazon, Google, Microsoft, Facebook and Apple. These are American companies known throughout the world. Which Russian IT companies are known abroad? Is that a kaspersky, but obviously - neither by fame, nor by revenue, he was not even close to the big five. Therefore, it seems to me that by the 2030s, becoming a leader in AI for Russia is one of the Rogozin projects.
                  1. Nastia makarova
                    Nastia makarova 12 December 2019 14: 44
                    0
                    Well, yes, they have a lot of money, but leadership and money are two different things, many areas, and where you can make a discovery and conquer the market and money has nothing to do with it
                    1. ares1988
                      ares1988 12 December 2019 18: 39
                      -1
                      Wouldn't it be difficult for you to give an example of Russian IT companies that "made a discovery" and conquered the market? And then, judging by the fact that in order to maintain their pants, Russian companies by stimulating sales, at least in Russia, have to pass laws on the mandatory pre-installation of Russian software - they are not doing very well with conquering even the domestic Russian market.
    3. haron
      haron 12 December 2019 08: 31
      0
      Quote: Freemason

      In the West, IT giants are engaged in AI development; they invest money in promising startups working in the same direction. Will Russia succeed in creating companies comparable to Microsoft, Google, etc. by 2030 - time will tell, but I personally can hardly believe it ...

      The fact that giants are good is a huge amount of primary information for teaching this "child", but how the information processing is organized and, most importantly, by whom, this is already a disaster. I recommend that the author look at or try to do the simplest operation or evaluate some kind of blood test before writing that even in 30 years, AI-controlled robots will appear in medicine. AND
      If these giants cannot teach a car how to drive, then what else can we talk about. Maybe something needs to be changed at the conservatory. Maybe besides IT specialists it is necessary to attract specialists? (God forbid proffesuru only)
      Particularly "pleased" that having put the steering wheel in his hands, he will take away the work of many, and becoming a table like not many at the opera table. A bit naive reasoning.
  4. The leader of the Redskins
    The leader of the Redskins 11 December 2019 18: 21
    0
    We have long been preparing us for confrontation and coexistence with AI. Some movies are worth:
    "Terminator" (6?) Parts.
    "Out of the car".
    "Better Than People" TV series.
    And others...
    1. dirk182
      dirk182 11 December 2019 18: 50
      -1
      rather to transhumanism
    2. Sergey39
      Sergey39 11 December 2019 20: 07
      +1
      Case for small. AI exists only in fiction.
  5. Amateur
    Amateur 11 December 2019 18: 25
    +2
    the so-called Massachusetts machine is a cybernetic device created at the beginning of the XXII century with "phenomenal speed" and "boundless memory." This machine worked for only four minutes, and then it was turned off and completely isolated from the outside world and is prohibited by the World Council. The reason was that she "began to behave". Apparently, scientists of the future managed to create a device with artificial intelligence (according to the story “The Beetle in the Anthill,” “in front of the astonished researchers, a new, inhuman civilization of the Earth was born and began to gain strength”).

    Br A. and B. Strugatsky
  6. ares1988
    ares1988 11 December 2019 18: 27
    -1
    Oh, well, it’s necessary: ​​how not tolerant AI was)
  7. KCA
    KCA 11 December 2019 18: 37
    +1
    Why did any software products become fashionable to call AI? NO AI and in the near future (100 years minimum) will not be, all programs work according to the algorithms laid down in them, where is the intelligence?
    1. Katanikotael
      Katanikotael 11 December 2019 19: 12
      +3
      The human brain also works according to pre-established algorithms, as you were taught as you think in childhood. A simple example - a person who has lived all his life in a country with left-hand traffic for a very long time adapts to right-hand traffic, and sometimes distracting subconsciously changes lane.
      1. KCA
        KCA 11 December 2019 19: 26
        +1
        No, here’s an example for yourself - it’s adapting, its brain is not altered, but, again, the example is unsuccessful, someone gets used to the left / right movement for a long time, someone drives wagons between France and England and switches immediately
        1. Katanikotael
          Katanikotael 11 December 2019 19: 56
          +3
          The neural network is also constantly adapting by receiving new data, this has nothing to do with "intelligence". Ha, in an attempt to refute my example, you cite professional drivers who have already gained experience in changing input data? No, I wrote about those who for 50 years had experience of driving only in left-hand traffic, how many hours, days, years will they adapt to new conditions to the automatism of old motor skills?
    2. voyaka uh
      voyaka uh 12 December 2019 00: 16
      -1
      "Why has it become fashionable to call any software products AI?" ////
      -----
      Not at all any.
      AIs are only self-learning programs. When a group of programmers
      writes base code. And launches the program. In the process, the program
      begins to add the source code as "experience". Already without the participation of people.
      For example: a team of people wrote 100,000 lines of code.
      The program ran for 3 hours. After stopping, people will see not 100,000 but 200,000 lines
      code. And after a new launch, the program will no longer act as intended
      people, and more diverse and unpredictable.
      1. dauria
        dauria 12 December 2019 12: 21
        0
        Not at all any.
        AIs are only self-learning programs.


        Rough. Any group of neurons in a simple herring does the same thing over millions of years of evolution - it receives signals, transmits signals. If combinations are often repeated, it performs "caching", takes over part of the processing, speeding up the process and forming a conditioned reflex.
        Here it is different - for millions of years, not only bones, muscles, organs of individuals were taken away. But also the software. And it was stored in the DNA. And in the human brain, there are a bunch of "ancient" layers, along with "ancient" software from the first cartilaginous fish. Your computer has a 2nd 10th BIOS interrupt function that does. the same as in the distant 1981.
        We are not born as humans. We become them by downloading a high-level program, growing up in a herd of people. It is the way of processing with words, the correspondence of words (or simplified pictures, gestures) to objects, actions that is our mind. But it is loaded onto a low-level software environment, selected by nature for millions of years. And it is much larger in size and complexity than our "upbringing". A monkey or a crow is well trained, but still "not intelligent" (controversial statement) We would have to copy a cockroach for a start with its senses, reflexes, ability to reproduce and survive.
        1. voyaka uh
          voyaka uh 12 December 2019 12: 35
          -1
          "We would have to copy a cockroach for a start with its senses, reflexes" ////
          ----
          So at the beginning the creators of AI thought.
          They tried to copy the brain of a person or a living being. A sort of "wise man who knows the answers to all questions." And nothing worked.
          Then they changed the concept: create a specialized AI.
          And things went smoothly.
          For example, the game of go. The most intellectual game. The game of mathematicians. Chess in comparison with it is like tic-tac-toe.
          The method of enumerating options in go is impossible. The supercomputer will think over each move for months.
          Make an AI playing go. For the usual reinforced transonal computer. And after an hour of training and a dozen real games with losses (like training) against major players - AI smashes almost all the gurus, champion, mathematicians, almost dry. With all their human experience, intuition, cunning, surprise.
          Nothing "human" - neurons, genes, etc. - helps.
          AI stably and quickly (it takes minutes or seconds to move) wins.
          1. dauria
            dauria 12 December 2019 13: 39
            0
            AI stably and quickly (it takes minutes or seconds to move) wins.


            Well, the cold heading press stamps more nails than any blacksmith. This program is just a reflex. And if it modifies itself - then a conditioned reflex. In many cases, a reflex is enough - the bird flies better than any pilot. He is also looking for prey, a female, builds a nest, remembers the appearance of enemies and dangerous places - this is instinct (a more general program, but made from reflexes and able to turn it on and off as needed). Next comes the herd, the need for interaction of members. And sharing information. But the low-speed channels - sound, gestures. This is where the need for programs that process words again into pictures appears, their logical processing - and the result again into words. Marx is wrong. It was not labor that made the man out of the monkey, but the competition between the herds of monkeys. And the selection of those herds that had an advantage - the exchange and storage of information is no longer in the genes, but in the herd.
            Whether we like it or not, we will have to follow nature along this route. And to understand why and how, when transmitting only ten bytes of a "green ball" in your head, you have a picture of megabytes, in addition, clinging to adjacent layers .. Otherwise, it will not be intellect, but a rubber inflatable doll.
            1. voyaka uh
              voyaka uh 12 December 2019 13: 47
              -1
              "This program is just a reflex." ///
              ---
              There are no reflexes in AI. Only program code. But he replayes all human reflexes, ingenuity, intuition, life experience, genetic experience, evolution, spontaneity and cunning.
              Read something about game go. I play it (at a primitive level). And you will understand my previous post.
              ---
              A person tends to overestimate himself: "Oh, I am such a complex, spontaneous, sensitive being. I have such memories, experiences .."
              This is garbage in the course of our evolution. It is caused by the biological evolution of man as a species.
              But it has nothing to do with intelligence.
              Therefore, AI developers do not need to follow the human. On the contrary, AI smashes a person, because he does not have evolutionary garbage in his head.
              1. dauria
                dauria 12 December 2019 14: 23
                +1
                There are no reflexes in AI. Only program code.


                Come on. In any of your neurons there is a program code. Sew a toe from your toe to the place of your thumb - after a while you will get used to it. Even in both directions - signals from the brain and back will pass. But the surgeon did not think about the pinout of the connector, the type of interface, the method of encoding. Your pinky finger has more "AI" than Bill Gates has written the codes in the life of Microsoft.
                The rudiments of natural evolution in programming and "evolutionary rubbish" are still visible - no one writes programs from scratch in machine codes. Some of the programs in the operating system hang as a dead obsolete load. We are stomping the same road that nature has blindly stomped for 200 million years.
    3. Narak-zempo
      Narak-zempo 12 December 2019 00: 46
      +1
      Quote: KCA
      Why did any software products become fashionable to call AI? NO AI and in the near future (100 years minimum) will not be, all programs work according to the algorithms laid down in them, where is the intelligence?

      Why is a hemodialysis machine called an "artificial kidney"? It's never a kidney, because a living kidney is orders of magnitude more complicated than any dialysis machine and, in addition to purifying blood, performs a bunch of functions.
      Any computer is a machine that performs an awesomely intelligent task - operations with numbers, and very complex ones. Despite the fact that some carriers of "real intelligence" can hardly master the multiplication table at school.
      It's just that here the word "intellect" acquires an almost magical meaning as a synonym for the essence of man (if you will, "soul" in religious systems). It is necessary to treat this more simply. If it was possible to automate a process that previously required the intellectual efforts of a person, we can absolutely rightfully speak about artificial intelligence of varying degrees of advancement.
      1. haron
        haron 12 December 2019 08: 49
        0
        Quote: Narak-zempo
        Quote: KCA
        Why did any software products become fashionable to call AI? NO AI and in the near future (100 years minimum) will not be, all programs work according to the algorithms laid down in them, where is the intelligence?

        Why is a hemodialysis machine called an "artificial kidney"? It's never a kidney, because a living kidney is orders of magnitude more complicated than any dialysis machine and, in addition to purifying blood, performs a bunch of functions.
        Any computer is a machine that performs an awesomely intelligent task - operations with numbers, and very complex ones. Despite the fact that some carriers of "real intelligence" can hardly master the multiplication table at school.

        Well, about hemodialysis and kidney, the answer is simple. Hemodialysis has not yet been able to make compact and connect to the forehead constantly so that it can perform a regulatory function. Or they don’t want to do it, huge amounts of money are being twisted there, drugs and oil were not nearby.
        Although doctors all kidney function algorithms have long been calculated, but like the brain. So the problem is just in the people creating its algorithms.
        1. Narak-zempo
          Narak-zempo 12 December 2019 09: 26
          0
          Quote: haron
          but like the brain

          You're wrong.
          1. haron
            haron 12 December 2019 13: 49
            0
            Quote: Narak-zempo
            You're wrong.

            Naturally wink if we remove the concept of "norm" for the brain, then I’m definitely mistaken .... and I would be glad that such a concept, if not removed, then seriously expanded.
  8. Pashhenko Nikolay
    Pashhenko Nikolay 11 December 2019 18: 50
    +1
    Do you say AI is a racist? I think Western AI and Russians will not like. And Israeli Arabs. Well, ours, if it certainly does ever appear, it will not like the State Department and the Jews a little bit)))
    1. Katanikotael
      Katanikotael 11 December 2019 19: 16
      0
      This is not an AI racist, but those who programmed it, but the article says that many analysis parameters are classified.
    2. bk0010
      bk0010 11 December 2019 21: 22
      0
      Yes, and the statistics are racist. Neural networks are such a thing - that at the input, then at the output, no minority rights will help here (well, until they are reflected in the input data).
  9. Sergst
    Sergst 11 December 2019 18: 53
    +4
    Quote: KCA
    Why did any software products become fashionable to call AI? NO AI and in the near future (100 years minimum) will not be, all programs work according to the algorithms laid down in them, where is the intelligence?

    I agree! Instead of the word "computer" they began to use "AI", and everything is as far away from intelligence as it was 50 years ago.
  10. Sergey Averchenkov
    Sergey Averchenkov 11 December 2019 19: 11
    -7
    Let's be very simple ... What is artificial intelligence? Will he be able to solve some problems? He will be able to figure out what to do with Ukraine? I think no. I think we’ll tie it up.
    1. sergo1914
      sergo1914 11 December 2019 22: 50
      +2
      Quote: Sergey Averchenkov
      Let's be very simple ... What is artificial intelligence? Will he be able to solve some problems? He will be able to figure out what to do with Ukraine? I think no. I think we’ll tie it up.


      There is no need for AI. The bolt must be driven to Ukraine. And that’s all. But you and the like cannot be stopped. There will be no Ukraine, you will solve the problems of the penguins of Antarctica. Fields and Iskander. Sofa Bismarck, damn it.
      1. Sergey Averchenkov
        Sergey Averchenkov 11 December 2019 23: 14
        -5
        May I start from the end? Couch? Have you, namesake, ever killed something? I’m not even talking about a man. Have you ever shot a bear, a dog, a cat, a capercaillie, a grouse? Not to shoot, but to eat? No? Yes? Well, maybe a little pig was flunked in the yard for his daughter's wedding? AND?
        I don’t want to decide anything at all about Ukraine — I just want to destroy it (as a state naturally, although some of its inhabitants), and believe me, I have more than good reasons for this - personal ones.
        There are none like me. At least here I did not see them.
        And on the account of AI, I already said that without self-development, nothing can be considered intellect, I can’t imagine a self-developing computer ... Even if it develops, but only along the path that they wrote in the program and then how.
        1. sergo1914
          sergo1914 12 December 2019 14: 13
          +2
          Quote: Sergey Averchenkov
          May I start from the end? Couch? Have you, namesake, ever killed something? I’m not even talking about a man. Have you ever shot a bear, a dog, a cat, a capercaillie, a grouse? Not to shoot, but to eat? No? Yes? Well, maybe a little pig was flunked in the yard for his daughter's wedding? AND?


          Are you sharing the problem now?
          1. Sergey Averchenkov
            Sergey Averchenkov 12 December 2019 16: 37
            -2
            I hate Americanisms like problems ... Is there anything to answer on the topic?
  11. S-400
    S-400 11 December 2019 19: 27
    0
    There is no "crazy intelligence" :)) Under the guise of AI, kindergarten toys called "neural networks" are being sold to us, which supposedly "work like a human brain", although in fact these are just slightly more complex algorithms than the good old perceptron , to-ry on the mail reads numbers :) Until the present AI, mankind still has to walk and walk for 300 years, but so far even no ideas are visible on the horizon: ((
    1. Gost2012
      Gost2012 11 December 2019 21: 37
      +3
      de facto AI already exists now, and neural networks work exactly like a living being, i.e. are able to make decisions based on experience. For example, in network search, neural networks are trained on the example of behavior, conventionally, 1000 people, then the self-learning network, using the experience gained, begins to make decisions on the delivery of content and advertising. Therefore, even programmers themselves do not always know what the output will be - there is no linear "if-then" algorithm.
      There is no need to confuse two points - directly AI and the possibility of self-awareness of AI, which you obviously mean. Modern AI is not capable of self-awareness, mainly due to technological capabilities, but this while, and I really hope that this will never be allowed. An AI capable of self-awareness will never be similar to or close to a human, or any other living being, due to at least the fact that we are formed with touch, sight, hearing, movement, pain, pleasure, etc.
      AI will not be able to receive and simulate this in any way, but our logic is laid in it, since we don’t have another one, and we can’t lay anything else. If he can realize himself, no one can say what will come of it and how it will end for us. I’m not even sure that work in this direction is underway, since humanity does not need a consciousness-conscious AI.
      1. S-400
        S-400 12 December 2019 15: 44
        +1
        Quote: Gost2012
        neural networks work just like a living being, i.e. able to make decisions based on experience

        This is a rare nonsense, which was launched a hundred years ago by one enthusiastic fool :( I am usually the same enthusiastic fools who come to conferences on signal processing and tell them how with the help of another super-duper-multilayer-recurrent. ..- neural networks manage to detect a signal when SNR is below 13dB (spoiler: this is not possible in principle :)), causing homeric laughter in the audience, I ask: "Will the kind speaker cite a link to serious research, where it would be proved that
        neural networks work just like a living being
        ?

        I will upset you very much: if you master the last edition of the monumental work: "Brain. Mind, Behavior" (although, you can hardly master :)), you will understand a simple truth: in terms of studying the work of the brain, we are in 2019. about the same place as Ramon-i-Cajal. Not only do you personally have no idea how the brain works, but even neuroscientists have a very, very superficial idea of ​​it.

        Moreover, to hell with it with the brain, modern science has a very vague understanding of how the unfortunate neuron works (that primitive "electrical", like worms, that more advanced "chemical", like in our head). So declare that
        neural networks work just like a living being
        - it is necessary to have a lot of courage, or arrogance, or the ability to lie for any reason :(
        1. Gost2012
          Gost2012 12 December 2019 17: 04
          0
          If I were you, I would have avoided such categorical formulations.
          I want to draw attention to the fact that I have never said anywhere that everyone knows exactly and reliably how the human brain works.
          "it works exactly like that of a living creature" - meant the very principle of work, and precisely self-teaching. The design and management of neural networks is now taught at universities, and this is not considered the most difficult discipline, I know this from my children, who have already received such an education.
          And once again I highly recommend you to be restrained in formulations - the effect may turn out to be opposite to what you want, for example, I am inclined to agree with the post below.
          1. S-400
            S-400 13 December 2019 17: 20
            0
            Knowledge, education, experience and scientific degrees, you know, sometimes oblige "to be categorical" and set the brains of enthusiastic youngsters (and not only youngsters, alas).

            And about the post below, well, if you are tempted to translate the discussion into the plane of stupid jokes and personal insults - i.e. will go down to the level of Internet gopniks with a couple of education classes - it's up to you. I thought that I was discussing with an intellectual, I'm sorry that I was mistaken :(
    2. sergo1914
      sergo1914 11 December 2019 22: 53
      0
      Quote: S-400
      There is no "crazy intelligence" :))


      Some individuals are deprived of the natural.
      1. S-400
        S-400 12 December 2019 15: 47
        -3
        And, of course, judging by the stars, the next with brains broken off in the army. Sorry :)
  12. faterdom
    faterdom 11 December 2019 19: 55
    +4
    The task of creating artificial intelligence is absurd, like a spherical horse in a vacuum. What we all like to do for non-virtual money all sorts of Rusnano and Skolkovo.
    We also bring natural intelligence with the help of the Unified State Examination: a stupid consumer who knows nothing in the humanities literature and history, and especially in physics and mathematics. Which then goes to study en masse in lawyers-economists - people who do not know how to produce anything, but ardently want to consume.
    But AI, that is, the need for it should be determined by pressing tasks, for example, some kind of combat systems that are too complex and multifactorial to be controlled only by the human mind. Logistics, for example, Russian Railways. And these two tasks are very different, and they need specialized algorithms, and methods for solving them.
    Whom it is easy to replace even with the Spectrum ZX, at the same time saving a lot on money and electricity - these are both chambers of our so-called. "Parliament is no place for discussion", the economic bloc of the government.
    1. Bersaglieri
      Bersaglieri 12 December 2019 13: 14
      +1
      They were ahead of Skolkovo and Nanoshaman.
  13. 7,62x54
    7,62x54 11 December 2019 20: 20
    0
    AI needs to be implemented across the power vertical. Federation Council, Parliament, Cabinet of Ministers, Constitutional Court, prosecutor's office, courts, governors, plenipotentiaries. AI only needs to be flashed with humanity and integrity. The effect will be frantic.
  14. boriz
    boriz 11 December 2019 21: 06
    +1
    Before raving about AI, one must revive its electronic industry, the production of its elemental base. What, we have few machines that are disconnected on command because of the cordon? Is it necessary that the entire defense and current life of the country can be paralyzed at the click of a finger across the ocean?
    And the revival of the EP for us is now an insurmountable problem. First you need to remove the liberals from power.
    1. bk0010
      bk0010 11 December 2019 21: 30
      0
      To revive the EP (under capitalism), a market is needed. Without it, do not revive anything: too expensive production, too large volumes must be sold in order not to fly into the pipe, even a few plants produce too much - there will not be enough space on Earth for even a dozen. At least for special applications we will learn how to properly produce electronics, albeit 10 times more crude technical process than actual, but at a reasonable price.
      1. boriz
        boriz 11 December 2019 21: 35
        0
        When it comes to defense, the small series series is offset by price. And when the technology, in principle, appears in the country, then we can already multiply it, talk about volumes. It is not necessary to make iPhones by ourselves, but we will have to do the technology that ensures the defense and security of the country ourselves. And the sooner we start, the less losses there will be.
        1. bk0010
          bk0010 11 December 2019 21: 46
          +1
          The technology is there: back in 2009 we bought an old plant from AMD (0.13 microns). Micron had 0.18 microns. Much use? Small series - high price - no development. It is necessary to break out of this vicious circle. Until there are many products (not in pieces, but in types), specialists begin to appear, the know-how of the technical process will not develop (even if it is rough at first, we would have mastered this to the full), our silicone compiler will not appear, there will be no specialists who can create "our Kvalkom", whose products will expand the scope of our electronics, etc.
    2. Nastia makarova
      Nastia makarova 12 December 2019 12: 22
      +1
      ep never to revive !!! will only have to buy
  15. Ru_Na
    Ru_Na 11 December 2019 21: 21
    +3
    Artificial intelligence, funny, humanity does not yet understand how the intelligence of people works, but is already thinking about AI and already supposedly creating it, well, well!
  16. Private89
    Private89 11 December 2019 21: 25
    +2
    At the pace we are killing education, soon there will be problems with human intelligence.
    1. Vadim237
      Vadim237 11 December 2019 23: 25
      0
      No killing of education is observed in Russia: schools, lyceums, quantoriums are being built all over the country, ours are taking gold at world competitions, creating their inventions, huge amounts of information are available for self-education.
    2. AK1972
      AK1972 12 December 2019 13: 00
      +2
      At the pace we are killing education, soon problems with human intelligence will be.

      They are already there. I will give two examples:
      1. A design engineer (not a bachelor, certified engineer) graduate of RUDN University comes to work for me. I drew him a sketch of the simplest sleeve with an internal thread M6-6H (techies will understand what 6H is). I say - draw in AutoCAD. And this miracle declares me: "So 6-6 is the same 0 will work!". They kept him for a year, they thought to teach him something, but it didn't work out. They offered to write a statement. wrote and left for Australia. Now there is one more ram.
      2. The beginning is the same. My task is to calculate the volume of the tank. From the eyes I realized that the task is from the category of super-difficult. I explain: you multiply LxWxH. Question: "Is it really that simple?" By God, I didn't think of a word. And you, Vadim 237 say "schools, lyceums, quantoriums"
  17. iouris
    iouris 12 December 2019 01: 21
    0
    Car autopilot is not AI yet. And what is meant by AI is a separate problem, i.e. as far as one can judge, the scientific task is not yet solved, therefore, everyone is free to understand by AI what they can imagine by the power of their own intelligence. Many have problems with this.
    The future of national security depends on the question of the future of the state. What is characteristic, development is not always progress. And it needs progress.
    They will invent the AI ​​and tell the very big boss that the AI ​​is able to replace it, because the boss to him (AI) is not good at notation. What is the reaction of the boss?
  18. Comrade Kim
    Comrade Kim 12 December 2019 11: 27
    -1
    Quote: Gost2012
    de facto AI is already there, and neural networks are working

    The question is how we (AI) will use it.
    It is highly likely that all resources will be thrown to total monitoring of citizens, for example, as in China. Analysis of behavior in social networks (with subsequent analysis), the degree of loyalty to the current government, etc.
    On the basis of the data obtained, an individual’s life will be corrected according to the principle: loyal — get a cookie (promotion, credit at a lower%, etc.), you criticize the regime — the life status of an outcast, a loss of rights, and an anal test every morning by the boss.
  19. Bersaglieri
    Bersaglieri 12 December 2019 13: 13
    +1
    So far, AI is not the future, but the present, "cutting" funds on this subject by characters without specialized education :)
  20. prior
    prior 12 December 2019 13: 42
    0
    Artificial Intelligence is "digital handcuffs for living people."
  21. Plate
    Plate 12 December 2019 14: 30
    0
    But it turned out that COMPAS and all similar programs were racist: AI was twice as often mistakenly accusing African Americans of a tendency to relapse than white people (45% versus 23%).

    Maybe the AI ​​said so because it is so? Why should he lie? No reason. He is free from all prejudices (unless, of course, they were specially laid down in him).
    In most cases, AI prefers hiring young men, leaving aside the weaker sex and age candidates.

    And again I can only say the same thing. Why would a computer lie or use some kind of prejudice? It turns out that the traditions that have developed over centuries, however cruel and unfair they may seem to us at first glance, can be quite fair and effective? It seems like wink From my point of view, it is understandable. Just as biological evolution, in the course of cruel selection, leaves only what works, so social evolution leaves only the most effective traditions, cultural customs ... We tried to move away from this, but then the AI ​​invented by us came and said that to move away from "outdated unnecessary dogmas "ineffective.