Time and here we hear that there are a lot of counterfeit products on the marketplaces. Yes, here, i would operate with a word. Its just a fake when were moving down this long road of 400 million e product cards. We hope that e body certification at the time of certification will be precisely verified. Eh, is it that production and dont they make a fake there . Because when he checks the production before starting to issue a certificate, he must request the relevant papers and make a decision. And its really the designers there, that is, the countries that toy is from that side, yes or those rims, its from that side or in russia it doesnt matter. That is what they check. This is the sign of a fake. Thats why we here are one of the parts of this process, including, uh, 400 million cards yes, yes, only goods in the country are being sold by some problems and require some adjustments in terms of working with data. Yes, we are dealing with systems. Every day, in general , from 300 to 350 certificates appear in the country every day. This is a large number, even if compared with the countries of kyrgyzstan, armenia , kazakhstan, belarus, then we have the largest number of certificates, these are the protocols that i told about when they tear the bear there. This is then all fixed in the protocol of such protocols per day 10,000 appear in the system. And if we add verification of video fixing, ah, video fixing is carried out, when we lend to those who are involved in this, if we are now on the verge of making it is this decision that we are already making trial ones. Uh, measurements of what kind of roller will turn out. When your machine wears something for several days, if such a machine rubs the fabric, for example, it washes correctly. Thats how long this place will take. This issue is being studied right now. A so in the system. Eh, here. Eh, now there are about 50,000 documents a day, and this needs to be dealt with constantly by the system. It is necessary to develop , build up and ensure continuity. It is necessary to make digital transparency of all certification testing processes, and even accreditation of these persons, so that other authorities can see it on the marketplace. We saw the same fcs, uh, citizens could use it. Here, this is very important. Thank you very much for the conversation. Oh thank you. Hello russia decades of science and Technology Television company ntv state Corporation Rosatom presents a Science Program and a program on how science will change our lives in the next 10 years. Its already starting. That s it, and im ekaterina shugaeva, and in 10 years Artificial Intelligence will gain free will. Real intelligence is already beginning to acquire free will, but i invited optimists who believe that in 10 years Artificial Intelligence will really gain free will. This is taras barkhatov philosopher, good evening and units larionov an expert in the field of neuromorphic computing. Hello, i have a tradition for those skeptics konstantin shurinov an expert in Information Security and Natalia Galkina an expert in neurotechnology. Good evening at the end of the program our guests and optimists will make a prediction. What is the probability in 10 years . Artificial intelligence will really gain free will about the history of Artificial Intelligence , our expert professor kapustin will tell Artificial Intelligence is a device that can imitate Human Behavior and at the same time learn using existing data, the actual term Artificial Intelligence, mathematician invented. Alan turing. By the way, the same one that cracked the enigma encryption system, he first spoke about the possibility of the existence of machines capable of learning from their own experience. Artificial intelligence reprimands its birth since 1958, when a device imitating human Neural Networks appeared, this technology was called the embryo of a computer that should be able to speak , see, write, and, most importantly, reproduce itself and be aware of its existence. That is , from the very beginning, the computer was required functions unique to humans. And while the words Artificial Intelligence. They came up with later machines made of iron and silicon immediately began to attribute inhuman properties. Denis aleksandrovich what do you think, what is free will, in general, in the last century , heisenberg proposed the socalled uncertainty principle, which showed that the nature of the world. In fact, probabilistic and in the world there is randomness and this, in fact. This is one of the important postulates of Quantum Mechanics, which has been repeatedly experimentally verified, as well as. Here it turns out that the world is not deterministic. And this is precisely the uncertainty in the world. She, in principle, and allows you to put the question. And free will. If the world were determined, then such a question would be meaningless. Well , now im answering the question directly. What is free will is rather conscious decision making. So what is conscious . Previously , people answered this question very simply; they identified consciousness with the soul ; such a metaphysical object, the soul, the subjectivity of consciousness, was introduced, which, among other things, also possessed will ah, well, is this model interesting . It has great practical value and in everyday life it is still, in principle, applicable. And how can one answer the question, what is free will, for example, from the point of view of biology, but the answer would probably look Something Like this first there was physics, then life appeared and for billions of years of life existence. In principle, there was no free will, and then it suddenly appeared, and this is a certain, but a phenomenon associated with evolution with the complication of behavior. And so, well, we all know that behavior is of two kinds. Well, the first thing i call him is not meaningful. This is when the pupil narrows on the source of light and other behavior is meaningful. And uh, when u the subject who makes the decision has a model of the world in this model of the world, huh . There is the ability to make predictions, uh, and uh, predicting possible outcomes, and we can choose the right one. Thats what, in fact, meaningful behavior from the point of view of biology, you can call free will. However, this definition has no practical value. And so, basically, no one says, but the real practical value in this concept is invested by philosophers. Eh, when they understand under free will the ideas of personal responsibility, and most importantly, not that there are some rules according to which some decisions are made. And how can we explain these rules to ourselves and others in society, that is, in fact, answering the question. Why did you do that . We form the basis for such a concept as responsibility, it underlies the right, generally speaking, the structure of society is such when i i ask why you ate all the remaining borsch when you speak in terms of determinism, its not me, its the stomach that wanted it. Its all nonsense, in fact you knew who would be responsible for this. You knew that there would be no borscht left, in general, and it was your freedom, it was selfexpression, that is, freedom, philosophers are to blame for everything. Yes, i agree, in general, you can not blame the intestines. Here, free will exists and people eat borsch precisely because of their free will. A natalya valentina and what did a person learn earlier learned to do earlier think or feel. Now, if we say that yes, there is Artificial Intelligence, there is still a soul with free will well, firstly, i want to fill in about free will, and free will today, in our understanding, is inherent only in living beings , even fish and dogs. And i think yes, even fish and dogs, because they are living beings, when we say Artificial Intelligence, then in general apply the concept of free will to Artificial Intelligence today. Its extremely difficult because anyway, these are algorithms that were written by developers. So, of course, feelings come first. Cant Artificial Intelligence feel . Well, you know, how would it be such a conversation, today is already enough, and not complicated, but nevertheless, which all the time there is an emotional revolution. Thats what arose earlier yes, how would a person feel or think before, right . Do you understand . No, well, at first i felt, and then there, and then there i began to think sometime in antiquity. Well lets look now at behavioral economics, tells us that by 90 , we make a decision not rationally and only 10 rationally why because the emotional revolution has absolutely turned the paradigm upside down, and buying to behavior, well, we are talking about the economy, so purchasing behavior, if previously considered in neosmithrian economics, what is man first . He thinks whether ivan weighs all his or not, then he acts and only after that he experiences pleasure on the couch or not, then today in the information flow in which we are with you we live at first spontaneously acting , maybe i was shown an advertisement 10 times, then this sofa, like, like spontaneously bought, maybe , or maybe they just didnt show you a sofa. You saw a sofa that you have never seen and said i want to emotionally absolutely always, it always happens. Yes, i dreamed you went to buy a car there. He saw a completely different one, and there were still huge discounts on it, some kind of red discount. Is this a spontaneous decision . They took me out. No, nothing like you at this moment your brain and your feelings are your desires. In fact, they were completely irrational. Well, that is. Here , maybe it was a little more expensive than your capabilities, but your desire is your emotional reaction. It feels like the whole world. Uh, hmm regulates my emotional desires part by word manipulated by extreme emotionality. And look, well, of course, they dont manipulate, but still, reasonable people, therefore they must think and must be able to stop themselves. But the fact that we live in a Consumer Society is a fact, but we are people and after that we must explain to ourselves, why we did this, you will always find a thousand reasons to explain to yourself why you still chose this sofa, although you followed a completely different one and this is it, but theres a difference between a that can predict, a Probability Theory and classical mathematics and how can act, and mathematics, which is related to behavioral economics, and with algorithms. That can selflearn. I have a reason to interrupt you, because advertising is advertising on ntv. Yes, you wont believe it. Yevgeny petrosyan was admitted to the hospital havent reached yet hospitals. I have already received calls from everywhere, who broke alexander olezhkos leg. I still have at least 80 years left. So you wont wait where Masha Rasputina sleeps a double door, especially since war is such a thing. How much money does Dmitry Dibrov spend, did he find someone and kapron socks . Yes . Its silk, how ivanushki have fun secretly from their wives and lovers. You wont believe it saturday at 20 20 on ntv you are super 60 new season sunday at 22 20. May tinkoff cashback, cashback bitch up to 70 . Participate on may 28, too, i want to relax. I remove and humiliate i fly away, i swim from the mountain, i move out, and open the zone and relax. Book profitable tickets, hotels tours on ozon i dont gossip that they allow themselves . Yes, whatever. Kete misha rosa cellophane sold kombucha natasha pupyrochka former director of the dolce vita no matter to whom you transfer or pay it is important that without commission and restrictions he was the first in russia to abolish the commission for all transfers and payments. Order a free debit alpha card, alpha bank best mobile bank, spicy burgers and spicy snacks. Do it sharp, whatever. Try not to sleep in burger king magnet price what you need candy 35 34. 99 due to work, i lead a sedentary lifestyle debatable on a daily basis, but how to protect joints thanks to the special collagen form of triple helix promotes support, health, joints just one capsule per day to protect the joints and spine. Mi a favorite burger is called. What difference does it make what its called . And how to figure it all out yandex search everything changes everything is located. 99 when you need something check out the weekly promotion ozone choose different products at great prices samsung headphones for 9. 499 rub. Hair mask horsepower for 279 rubles. Meet the new card for life from vtb, share free transfers and feelings. Catch 1,000 rubles. After the first purchase. Get a new card for life on vtb. Ru. Chny crust from 169 rubles. My children are among the large selection of hotels in the yandex travel app and get up to 10 cashback. Yandex travel booking hotels maysky tinkoff cashbackfest super cashback up to 70 . Issued any tinkoff card by participating in the cashback pestle from may 15 to may 28. Under the contract, the military personnel receive Regional Support measures, the status of a veteran of military operations, exemption from land tax, compensation for payment for vouchers to a sanatorium, vacation twice a year, and other benefits from the state. On the air is a program about how science will change our lives in the next 10 years. In 10 years , Artificial Intelligence will gain free will. Scary. Yes, but we will find out, because at the end of the program of our guests, optimists and skeptics will predict. With what probability, in 10 years , Artificial Intelligence will really gain free will, but dont worry, dont believe in this, our expert professor kapustin, a computer is just a bunch of all sorts of pieces of iron that, according to many, can replace a person. Already today, Artificial Intelligence so far exceeds our abilities in most areas. Well, what makes you see Artificial Intelligence as the main enemy. Should we develop nonhuman minds that may eventually surpass us and outnumbered and outwitted and generally make us tender and replace us whether we should risk losing control of our civilization precisely these questions. Or rather, the answers to them make the developers of Artificial Intelligence think about the limitations associated with the free will of their wards. Just imagine you are driving a car controlled by Artificial Intelligence. And suddenly, for no reason at all, she decides to drive through a red light, or even stop to stop in the middle of the road, and your home robot vacuum cleaner refuses to do cleaning because hes just in a bad mood. Well, ill finish the jokes of techies, agnostics have created an artificial superflyer, endowed with free will. They ask a question. And if god and Artificial Intelligence answers them now there are idiots. Here we have such a cheerful optimistic professor. Yes, taras aleksandrovich let s try to finally deal with free will. What sciences are being studied today, in general, are there any, i think we can recommend for the nobel prize. If we finally figure it out, right . Yes yes then there is an end here, of course, the main problem is what to put in these two words . Yes, we , in general, not about freedom, not about will, do not fully understand, not about that other place, of course, that by intellect, yes, we have a similar story there too. In general, some say so, some confusion in this matter, but nevertheless, and by and large, when we talk about freedom, we mean together or separately. A three aa lines of discussion. Yes the first is what colleagues have already mentioned, and the random factor, and the second line is the factor, subjective awareness. Yes, but the ability to be aware of ones own actions, including perceiving oneself as partially or fully responsible. Yes. That is, as the source of these actions. Here are respectively the second line of discussion and the third line. Maybe the most interesting. Well, the most speculative is actually the ability to produce new creativity. Yes, and accordingly, a to one degree or another. These three factors, together or separately, make up the objects of interest. On in fact, almost all sciences to a slightly lesser extent, of course. A little more public, but one way or another, the problems of free will pop up in physics and e, they remembered heisenberg yes , they remembered the creators of the copenhagen interpretation of Quantum Mechanics and , accordingly, all ideas related to the underdetermination of the world at the micro level, but they did not remember Albert Einstein who stubbornly said, that you dont know how to count this question, and the problems are not in the structure of the world, but in the incompleteness of the existing theory, that is einstein was just an absolute determinist. Dai einstein insisted that gods own did not play in the field. Yes. Yes, yes, he cant, the team answered that, when will you finally stop telling god what to do with him that is not allowed. Not this particular one, not actually that big one, in context, inside of which uncertainty is the unresolvedness that exists, we turn to the question of what more or less reliable detectors we have in order to detect free will, by and large, everything depends on that we do not have the opportunity to go back in time, that is, in order to make sure that the action could be performed, otherwise it would be a creative action. Be it a physical event and so on. We need to go back and try to do it differently. Uh huh because. This is an impossible task, then by and large , talk about the presence or absence of free will in a radical formulation turns out to be purely speculative, nothing can be proved here, but to the social sciences. For example, or response, yes, but it is critically necessary free will, as a factor of responsibility of the individual for his own actions at the moment when we say, we suspect that a persons actions are determined by external causes, he will have to be released. Responsibility, and this is what scares us the most in the implementation of Artificial Intelligence in everyday life, because now it is free from responsibility, because it definitely did not decide anything on its own. Ah, how interesting. And by the way, mikhailovich well, with a persons free will, its about the same no. All the same, nothing is clear, but lets say already on the horizon. Yes, you understand, yes, yes, lets figure it out. And what is called Artificial Intelligence, the second part of our program is Artificial Intelligence, then this is what Artificial Intelligence is. This is what experts call machine learning, and basically people call it a Neural Network, that is, here are the abacus, this is a device on which you can do addition and subtraction , there is a calculator. On it you can still do the division of the logarithm to take. A Neural Network is a box that can decide differential equation all Neural Network. Its just a lot of counting that differential equations can solve. All. Thats what professor kapustin was talking about, its amazing that Artificial Intelligence. Now he can do some things that a person does not, its amazing how much of a persons actions are just a pattern action, described by simple defurs by people who consider themselves creative, actually reproduce pattern actions. Oh thats why Artificial Intelligence is behind people writes texts, because these texts are not dostoevskys. Its absolutely a pattern. Quite right. And we, that is, without us, Artificial Intelligence does not work the abacus. These can i submit, we interpret. Here are these knuckles, like units, these are like tens, these are like hundreds, what is happening with Artificial Intelligence with universities. Is it possible in these boards . Not the euroset is chasing on a computer in the computer there is a processor on legs, processors , voltage drops, they look Something Like this. I thought that the processor is just 0 or unit, only goes or does not go goes in real life, they look like. This is how we interpret. Its like a step. Well, yes, then the stairs. We interpret as ones or zeros, ones or zeros, we interpret as hexadecimal codes, machine codes , we interpret them as instructions in python or c . We interpret them as some kind of algorithms that are a reflection of our model. Our understanding of how neurons work is related to real neurons to zero or one. Well, in in real life, in 2014, stanford scientists connected electrodes to a rats dendrite. So the dendrite is not dead, and they found that in fact the brain produces 10 times more impulses than previously thought, and they are not digital in nature, that is, our brain is not a digital machine, some biologists believed for 50 years, that is believing that this sequence of interpretation has something to do with the flow is like believing. Here i drew some curves. Here i draw curves. Now ill draw here dashes. And i say, i interpret these curves like a mammoth, and this man and here we are throwing a spear. We all flunked the mammoth, the thing is done now in real life. Well, of course, we will fill it up. This is pure magical thinking. What does cave painting have to do with real mammoth hunting, no what does these impulses have to do with real neurons. There the nerve networks of dogs, rats of man, monkeys have no faith in what can be made of this , any Real Intelligence can come about. Its in pure form of magical thinking intellect. No. Okay, now lets deal with this Denis Alexandrovich and what do you think they are , is free will generally an integral quality of intelligence, including Artificial Intelligence and in general. Is it possible to create an Artificial Intelligence that will not be endowed with free will, huh . With regard to Artificial Intelligence, as has already been noted today and even in the near future, apply the concept of free will to it. Well wrong the way he just doesnt have the intellect because free will. Uh, it is, uh, necessarily an attribute of consciousness. Uh, before talking about free will. Lets make it so that Artificial Intelligence has a consciousness of selfawareness. Corporeality is mortality. We must solve the problem, and responsibility in society and many, many, many more difficulties. So far, the Scientific Community does not see approaches to solving, well, many of the problems i have listed, but, nevertheless, and if in the field of Artificial Intelligence, there is such a popular direction as Adaptive Systems capable of autonomously acting in conditions and changing. The environment, and here, if you put the term free will, this is the most adaptive behavior, then here we already see great progress and, uh, in the field of robotics. We create systems that are able to selflearn to act adaptively in difficult conditions and, in relation to their description, the term free will. Well, now you can think about its application. Well, here it is always here in these e phrases. I am confused by only one word so small for now, but what will happen in the next 10 years, we will find out after advertising advertising on ntv 7 years ago, i buried my daughter ago. I dont want its hard for me to even remember, so they lived a difficult life, but kept their youth in their souls. Have not forgotten how to dream. L you love me, you create sculptures, i pray we sit silent. This is the most beautiful thing i ve heard. You are super 60 new season of the show not only about music, but also about real life on sunday at 22 00. Borschichi premiere tomorrow at 20 00 on ntv when you need something, go to the weekly ozone promotion choose different products at competitive prices for 199 rubles. Floor biofireplace mass for 3. 99 rubles. May tinkov participate until may 28 and my favorite burger is called. What difference does it make what its called . And how to figure it all out yandex search everything is changing everything is a future in which you dont have to choose between form and content, between complete control and absolute freedom. We imagine this is how we create it, a fastbe crossover amoda, t5 with an intelligent allwheel drive system, choosing the path you choose the future will someday end no, the track on the video is endless. You can watch the video endlessly just choose the dragon. Kill on your side. Meet in the marathon of discounts our world megamarket a new part of the blanket pillow 1990 rub. Sber megamarket now, when someone decides to steal your food to burger king, he will definitely burn from spicy burgers and spicy snacks. Make anything spicy and try not to sleep at burger king credit card reasonable 145 days without interest on everything without commission for withdrawals and transfers issue online Renaissance Bank of all kinds for real discoveries , you need time and place among a large selection of hotels in the yandex travel application and get cashback up to 10 , yandextraveler hotel booking. Mom, dad woke up at five. The hostesses left without me. You are my family tradition. We save the years, the children grow up, and pyaterochka continues to delight with discounts and promotions, he just threw off a couple of kilos, and then there are discounts. Mi money on the card immediately deferment of the first payment up to 90 days to issue a cash loan and receive money right now, he is the only one. On the air is a Science Program and a program on how science will change our lives in the next 10 years in 10 years Artificial Intelligence will gain freedom, she said the will and you just dont believe it. Yes . Well, yes, because you know, today you have already heard that in fact he is not an intellect and he is not free. Yes, he is a little wiser, but hes not just super. Smart intellect. Its just that we have become a little stupid, we dont know, i dont know , but professor kapustin believes in this , it scares me very much that lately Artificial Intelligence has constantly kept victories over homa sapiens, computers easily beat people in chess and checkers. I didnt talk about preference and bridge the computer easily won in its game and defeated the best player in the game without any problems. Go will soon get to what where when Artificial Intelligence constantly monitors all of us. Even if you want to forget something to expel from memory, Artificial Intelligence will not let you do this. He will remember everything. But what if Artificial Intelligence wants to go further, after all, even today it is not only follows, but already tries to anticipate our desires and needs. But the question is, how long will Artificial Intelligence be engaged in satisfying precisely our needs. And when he decides that its time to take care of himself, how soon will the nonmaterial substance of life arise , some barren creature with free will and when will mankind be led by a soulless machine . Well, i dont know, it beats, there is Artificial Intelligence in chess in goa in checkers. Let him try to play with me in chapaevo, that well see who won whom you . I never had any doubts, but by the way, our friend pavel has, uh, a pet that he believes has free will. Here, but again and again, and a wonderful manipulator has already learned how to manipulate him in everything. Mm, and even without a leash , will not Artificial Intelligence learn to manipulate us with it, if people are already walking each other, animals are manipulating us in full , some are manipulating us. Well, lets start with the fact that in order to attribute manipulation to someone, you must already give him free will in our sense of the word. Yes, that is to manipulate. Maybe someone who has his own intentions, conscious actions, and so on. And if we talk about the ability of Artificial Intelligence, to mimic , imitate some actions, to be built into our motivation system for managing everyday life. All this has already happened. And for quite a long time, we all deal with this every time. Well, every day it is enough to call, for example, the sberbank hotline and you will communicate with Artificial Intelligence for a long time. And before those until you need to solve some nontrivial problem. It will be a very productive communication, moreover, Artificial Intelligence will not be rude to you. Well at least in case yes will be very patient with you. Well, and so on. That is , he has in this sense, but he knows how to migri. Heres to pretend. Yes, the fact is that he does not pretend, and he does not pretend thats what he is beautiful. Yes, he, in principle, is not able to behave, but in a way different from what was laid down in him, and this is his pluses, this is his minuses to expect from him criminal acts, but actually all the examples that are given in this regard. Yes, machines collect data about us, not a machine collects data, people collect data with the help of machines and indeed and can and certainly victims by others. Yes, they will undoubtedly be abused, because why collect them if they are not used, but this is not at all about Artificial Intelligence. Yes, and to do here is certainly in those in whose hands this tool is located. Yes, of course, of course, this is good denis aleksandrovich, and what scientific breakthroughs in areas of Artificial Intelligence . Are you looking forward to the next 10 years, is everything so rosy and promising, perhaps, right . And firstly, they will continue the victorious march. Such technologies. Like reinforcement learning. This is about robotics, large language models, this is about gpt chat and its evolution, and generative networks. This is the generation of pictures to texts just around the corner, the generation of cartoon films. Eh, that s sad. Ive actually already seen how employees in the press office, and not themselves, write about releases. And they get into this gptchat just keywords, he gives them a great press list. This will come to the point that we will not have literature. It seems to me that they will simply stop writing something beautiful, here is the one, if you take the same chat gt, uh, there is still no creativity in it, it is such a big statistical parrot that, having eaten a lot of information, answers you very well, but there is creativity in Artificial Intelligence is a complex issue. It must be this, it just works and partly, therefore, there was a recent surge of interest in the problem of yes control over Artificial Intelligence, and elon musk, Jeffrey Hinton and other leading researchers in this field, drew attention to the fact that enough is enough, even those language models that are now being created, they are already becoming too strong and have a very strong impact on society if creativity is not an integral part of freedom will, but free will. We say that this is an accident or, in general, this is an opportunity to act. Well, in general, there are no chaotic ones, that s just the point, Artificial Intelligence is not predetermined to act at all. Maybe, to act is not predetermined that it has never been incorporated into any program , you can put a random Number Generator and issue an answer depending on the model by which this generator is used. What, while we define creativity as a non algorithmic decision. This is fundamentally, not a function attributed to Artificial Intelligence, it is in any case algorithmic no matter how difficult it was from this random Number Generator. Its still algorithmic. Yes, he acts, according to that. Uh, that set of rules he was laid either by people, or he learned these rules from the data. Thats just in this aspect. We act people contrary to it, we have a set of rules for life in society. And we do everything, as if not in the rules, not quite people also make decisions based on their past experience from birth to the present moment. And its not in some way in some sense its an analogy for training Neural Networks. All of these also first learn, then they apply, so the practice of some teachings does not help anything. But a. Well, i am so, if in short others directions in which breakthroughs are expected. Well. Personally, i expect that alternative methods for Training Networks will finally be invented. There is such a backpropagation method , and it ruins the whole life in terms of Energy Efficiency in terms of the problem of catastrophic forgetting. This is when we lose knowledge about the first task on the second dataset in terms of continuous learning, and so on. Here, the existing Artificial Intelligence systems are saddled with problems. Its called catastrophic forgetting. Its about that if we learned an Artificial Intelligence system, first on one task, for example, to recognize cats, and then we begin to teach on another task, for example, to recognize these helicopters there, then it will lose the ability to spread cats. Old knowledge will be destroyed. Yes, and this is a consequence of how we teach these networks. It seems to me that nu will inevitably have such teaching methods that are not subject to this problem, which allows us to overcome it. Well and and in my opinion the most important and most significant accomplishment that should if this happens, there must appear an e computing system to perform Artificial Intelligence tasks, meaning hardware, and built on alternative principles, we will break, limiting our von neumann architecture, and there are such areas as quantum computing, photonics, neuromorphic computing. And this is exactly the area in which well hmm we will be able to make a prediction with the help of such solutions. But if we understand, we may not be able to interpret. Uh, then on the basis of which the prediction carried out . Well, this is already more terrible konstantin mikhailovich here. Uh, well, they still calm us down. We will put some algorithms in it, and it will only act on Artificial Intelligence in accordance with the algorithms. In extreme cases , we will put additional ones somewhere. Lets say he wont. And if he himself has free will, without asking us, free will from a technical point of view is a manifestation of the process of the socalled downward causality when the system sets the actions of its subsystems. I wanted to write something my hand took the pen. Neural networks are systems with bottomup causality, like an anthill three reactions, an ant to another ants floferomion, you threw 10,000 ants on a laboratory table, a structure arose. And systems with upward causality cannot explain downward causality, that is , genetics, evolution are processes that descend causality in a person is only 20,000 genes , half as much as in a potato. Our body knows how to read the alphabet of genes to make the right cells. We cut our finger the body knows how to read the dna to repair the skin. Has anyone tried to drive genetic Neural Networks on their home computer . No, i tried for two weeks running the Neural Network cant explain evolution. It instantly fills an ecological niche and adapts to it, but cannot create new species. She cant pull herself out of this niche. This is called a stuck Neural Network. The most famous example of getting stuck was a few years ago by three blacks. We took a selfie photo of ourselves we posted it on the facebook Neural Network signed gorilla. There was a scandal for the whole world, that is, the euroset mixed up blacks and gorillas and facebook said we will fix it and could not fix it in a week, not in a month, or in 3 months, if the euroset is not stuck, so that it cannot pull itself out, it can only be demolished and reeducate and the second scandal was only this year. Thats what was absolutely right, what to say, only to erase it so that she would forget the cats and the cars absolutely this year. On this occasion, the vaunted gpt american chat it turned out that in kenya they are sitting in hundreds of people for a salary of 50 rubles. An hour and they constantly monitor that he doesnt get stuck anywhere literally blacks for a beggarly salary in africa behind the scenes poke him so that god forbid he doesnt get stuck anywhere, because if the skiff gets stuck, that is never our employees at work in this gptchat. Write to yourself about cuts. This is not a civilian salary stuck in the process of writing these press releases. Yes Critical Thinking work on mistakes free will. This is all manifestation of a downward, outgoing cause, and we do not know a single law of biology, a single law physicists who could explain it, that is , hope that we will build a big big Neural Network, and it will gain free will. Its like saying in the middle of the 19th century we are a very large Steam Locomotive and we will go to the moon. No. We need tsiolkovsky who will explain the principles, so that later korolev tsandra and other engineers will bring them to life, then something will work out now. He hopes that Artificial Intelligence itself will gain free will. Its still random. Yes, it will spontaneously arise, all the same that it hopes that in xix century. We will accidentally build Nuclear Reactors or accidentally fly into space. Well, on the other hand. You know , they fired from pushkin luna, but it was worth turning this gun in the opposite direction and actually. That s right for you about from a rocket, but still, yes, in russia no one should guess. Uhhuh and natalya batina, well, lets say, after all, Artificial Intelligence will have freedom. Well, lets say we foresee that thats to insure. How are we supposed to understand this . By what signs we will not understand exactly. Exactly, based on the fact that we are here talking first who can see just see whats allowed. As you say, free will appeared. These are the same developers who used their knowledge of their mathematical apparatus in order to make those very algorithms. Here they will be, and the first front that they suddenly see that something went wrong went out of line outside the algorithm. But in fact, there is one more point, which probably no one has mentioned today in order to reach the free will. Yes, we are living people, yes, and we have a psyche. Here Artificial Intelligence. We need an artificial psyche in the image and likeness of a person. Thats when he will not do any trouble, because again this psyche is for him. A person will program in the image and likeness of the sons of a good person. Okay then i have a question in our guests an optimist from a skeptic. You have already mentioned that today many people are thinking about limiting the possibility of Artificial Intelligence, in your opinion. Which Artificial Intelligence is more dangerous with or without free will well, obviously with free will, because the question is already implies the fact of creating such a decision, but this is true since we have assumed that he has free will. This means he is endowed with consciousness. This means that he has some kind of motivation, a value system and a big question. And what kind of it does it correspond to human, and in general for the good in the name of what this Artificial Intelligence will begin to make a decision, what will be a value for it is not clear. Yes, maybe, yes, maybe not quite a familiar thing, i will say, but in my opinion, there is no simple answer to this question. That is, im not at all sure that a more complicated case, yes, that is, Artificial Intelligence with free will will turn out to be more dangerous than what we have now, because if it comes free will, then along with it, in theory, it should gain the ability to be responsible, and today we have a lot of cases when Artificial Intelligence harms a person and there is absolutely no one to ask, and the most terrible thing in these cases is not their statistics. They are negligible in fact, but the fact that, in contrast to situations where a drunk driver hits a pedestrian. Yes, this is our sad everyday life. Here but nevertheless. We have reason to believe that at least he did not want to, at least. And now he at least feels guilty. And in this sense, how to say. In terms of maintaining the foundations and stability of our society, everything is in place. But if Artificial Intelligence. Yes, there an uber knocks down, yes, an automated one beats a pedestrian, then, in fact. The horror lies in the fact that no one is to blame for asking someone. Maybe maybe he wanted this, but he didnt want to, but its completely incomprehensible to do it. He wont do it again. And the horror here is not in i repeat, not in statistics, but in the fact that this thing does not feel free will, it will not gain, that is, Artificial Intelligence. Yes, yes, and in general, he is not an intellect, that is, he cannot want to kill all people, unfortunately, we have enough people on earth who will set him such a task. Thats the problem, yes, and they will say that he wanted it, but they will simply make you worry about talking. At we have enough people who will set him the task of enslaving all people. Excuse me, they are not even disguised, the entire dovodsky Economic Forum is filled with such people. You thought they funded the development of Artificial Intelligence from the kindness of their souls. Thats it. Well, let s take a short break and think about this advertisement on ntv stars open the summer season, and so you want fire, one of them regularly stands in the pose of a summer resident. Im ready. And who rolled all the beds in concrete . Beds or close thats it. God forbid, more than three days is impossible. Six, what miracles and tools celebrities use of any height, you understand once you put an apple in the middle , who knows what it is and how the stars poison pests more than 500 pieces, i collected the same amount the next day. God is not with a pitchfork. They ordered the scarecrow of the star to be removed and agreed to remove the fence or fortify the shore on the sunday immediately after the show. You are super 60 on ntv you are super 60 new season on sunday 22, when you need something, go to the weekly promotion ozone choose different goods at competitive prices pasta america 4 kg for 199 rubles. Garden trimmer e3x. For 1999 rub. Sberbank presents a profitable loan. The lowest rate over the past 10 years from four percent per annum get a loan from sberbank and get monthly cashback. Obviously beneficial every day the heaviness of the legs. It can be symptoms of varicose veins contributes to the treatment of varicose veins and helps to put you on yulia beautiful free legs and 600 for the freedom of your legs to rent for Outdoor Recreation for purchases from 450 rubles. Pyaterochka helps out, please do not look for meaning in this commercial, better look for cheap tickets for the aesl service to buy cheap air tickets ozon exhibit with love wifi robot vacuum cleaner with carlit from 6. 99 rubles. And diapers panties her kusan for 1. 099 rubles. In the beginning was the word. Then we were the first to launch sms, we were the first to connect to the metro and rose above all so that you communicate not only in words, but also in circles with pictures. And if you need words again, say them looking into your eyes. Which is now available to everyone subscribers. Switch to megafon new fun from russian lotto at any time and you are in the game. Choose your numbers and not even guess a single one. For you in the stoloto app now, when someone decides to steal your food to burger king, he will definitely burn from spicy burgers and spicy snacks. Sharpen anything and try to stay awake at burger king thank you album in which the austrian outfit brand notland gear was born 50 years ago. Thank you for the morning fogs of the boundless sky and the routes that brought us here. Thank you for these peaks the alps. Carat cheesy delight, when you need something, check out the weekly ozone promotion. Choose different products at bargain prices yokosan washing gel for 199 rubles. Outdoor biofireplace maz for 3. 99 rubles. Cash loan on the card immediately deferment of the first payment up to 90 days issue a cash loan and receive money right now. It is the only one with a unique flexible nozzle, just one click for precise hitting and active action in the lesions of the fungus , ready to use hygienically and economically. Order on the site apteka. Ru. Meet the new member deco electric garden trimmer for just rs. Sberbanker. The fan card is a Digital Service for a long and comfortable visit to football matches. The card number is needed to purchase tickets and season tickets for the Russian Premier League games and the russian football cup final. You can apply for registration at Public Services or in the Public Service application. All clubs, 12 cities get a fan card and fight football in the stadium. And the program is on the air science and we are a program about how science will change our lives in the next 10 years. Dear guests , optimists and skeptics, i have a question for you. What is the probability that in 10 years Artificial Intelligence will gain free will . Lets start with you, if we think that human intelligence is endowed with free will. This is the talking head, which, through arbitrary interfaces, is able to send quasilinguistic messages and interpret. Then indeed. Well, not far off. I think its a matter of the near future, when Neural Networks or other systems. It will be quite effective to imitate the presence of a person. Actually all the disputes about whether gpt chat is dangerous for us or not, they actually revolve around this problem. Yes , by and large , it is difficult to recognize text written by gpt chat as written by gpt chat, but as bad meaningless text. Not a difficult problem is that, unfortunately, we ourselves produce most of it