the series of senseless victims did not end there. while butorin was in spain in moscow, fierce wars for spheres of influence continued, this time the orekhovskys clashed, exploded by the izmailov group. lyosha the soldier was assigned to scare the competitors by blowing up a cafe where they often collected a bomb. le explosives with a clockwork, it worked before everything is right. well, like, as the analysis went, there was an explosion in them. well, first of all, if it had been planned, it would have happened when they were there. these people were never there suffers.

what else can i do? i can say the same about kotova, of course, i regret that she got into it. naturally, it would be stupid if i said, yes, i'm glad. of course not, but what can i do? i will say, i have thessaloniki is a pity. actually, because i, well, we were friends with him. after the death of thessaloniki , russian law enforcement agencies increasingly began to receive operational information that sergei butorin was managing the orekhovsky gang from the other world, after a while they managed to find out that he was alive and was in europe

an operation was being prepared to capture and extradite autumn, but the spanish authorities were ahead of their russian colleagues in 2001. butorin and polyansky were arrested at the exit of the brothel, the orekhovskys. they tried to run back, shot back, but unsuccessfully for 7 years. they spent in a spanish prison. before they were sent to their homeland, where the leader of the orekhovskys was charged with organizing a criminal community, banditry and murder , the materials in the case of sergei butorin took almost 70. it turned out that one of the main witnesses for the prosecution is the same businessman stepanov who had already testified against butorin in the mid-nineties, he miraculously

managed to avoid death and now 20 years later. he again spoke on the side of the prosecution against the powerful leader of the orekhovskys. a few days before the verdict in the orekhov gang case , an entrepreneur. vladimir stepanov was shot at the entrance to his house. during the trial, an unpleasant event for the axis occurred, his closest friend marat polyansky made a deal with the investigation in order to achieve a reduction in the term, of course, i expressed my point of view that he had only 11 murders there. here it is the only chance you

get away from capital punishment. well, that is, in his life, the conclusion is just a deal, on the right of the judge, he was forced to make this deal, and he sent me the verdict of the axis and marat listened calmly to polyansky, they gave him 23 years. deprivation of liberty. in a strict regime colony. butorin received a life sentence in a colony. he will no longer leave to move along the corridors will only be under escort and in a half-bent state, and his house will have a tiny cell until his death. now the leader of the orekhovskys lives only with his relatives and family, his brother alexander is alive and well and has long since retired from criminal affairs. axis himself married a second time before his arrest. now he has three sons. he worries about them and does not want the children to repeat his fate. i would like

to address my sons. don't, don't repeat my mistakes. go your own way. direct and honest, good prospects are open to you . you called me with you, so listen to you unpleasant things will happen. the rise of the machines began, you felt that you were communicating with an ordinary living person, and it scared you. yes, it was scary that a person will be replaced by another person who has mastered the tools of artificial intelligence plans to live on this planet, whether or not we wrote our text using artificial intelligence today. no. you know the questions yourself, only

provocative, only debatable now. here is also this artificial intelligence of the neural network, how to control everything and this competition. none at this stage predicts these things. hello this is info shield program. i am vladimir chernyshov who owns the information, he owns the world number, as you know, owns fakes and manipulations with the development of technology , the creator of a fake was generally inflated, well, with the advent of artificial intelligence. in contrast to reality from illusion is already almost impossible, it seems that artificial intelligence takes information wars to a completely different one. level in our program. we will separate facts from fakes so that we can stay on the bright side and not succumb to a false message, reflecting our information shield with a hostile information shield. but

naive people, as they thought freedom of information comes, there is access to it. now everyone has uncensored news in your pocket on your smartphone screen. you're on your own now the editor-in-chief, but it turned out, the stronger the information wave, the wider the sea of ​​information spreads, the easier it is to drown in it and hide the truth in the singing of lies. well, what ’s the only thing? we have learned, at the very least, to calculate fakes and stuffing, like new ones, generation neural networks are ready. again we confuse all the cards of reality. neural networks. now they are actively generating pictures that users can take scandals for pure coin. now they are pouring, as from the horn, abundance. well , remember, here is the footage of the arrest of donald trump or the photo of the pope in a down jacket created by a computer brain. agree, well, even

a human rights organization, amnesty international, got into a real scandal, it promoted an article about violations in colombia in social networks using an input created by artificial intelligence. well, sort of would be a trifle, but the resonance was huge and trust in the organization. now much less, what is the most interesting and frightening neural network. self-learning. she absorbs all the information and analyzes it. well, as a student in the classroom, and as a result, in a moment it becomes smarter and all students and professors of the neural network are already writing coursework essays for students and teachers about this. they don’t even guess, because the text created by artificial intelligence. not just a collection of words and terms, but quite meaningful, he is human now literary blacks. turned out not to be. for the sake of interest, we made a request, in one of these neural networks we were asked to write poems on the topic of the info shield program. here is what the computer bridge gave us. rest is not needed, no lights out. i live online and control everything. so that everything

is in the light. i can find out. i catch viruses. i protect banks. i help people with a couple of failures my radiance. it's like gold. so it's better to never mess with me. well, yes, this is , of course, not pushkin yet, but for now this one, a novice finger will soon learn the technique. since then additions and the onegin stanza will be able to improve the affairs of technology; moreover, we can create a new reality that will become even more real for us by a neural network. maybe to enhance the effect of propaganda pressure on the audience is smarter, a variety of any ear pr man. and here is a series of news, after which well, at least the terminator is in the past, sending the rise of the machines began in australia, the neural network accused the mayor of the city of corruption. he is now suing her. ossetia well , good luck to him, but in colombia the judge created precedent completely entrusting this gpt chat to make a real verdict. and although it was discussed in the case, and the payment of compensation for insurance they are in a criminal case. still, the fact itself

mastorazhivaet and frightening. if justice is administered, neural networks will not turn around, or all this is a real disaster. as they say, how do you like elon mask? artificial intelligence is more dangerous than a defect in the design of an airplane or a car, because artificial intelligence has the potential to someone, it may seem insignificant, but it is not so potential destruction of human civilization. well , i don’t know if it’s possible to trust the killon mask soon, after this interview, he registered, in fact, a company that will deal with artificial intelligence. that is, the mask obviously understands how promising this direction is and the future behind it, the same china is heading for the creation of the most advanced artificial intelligence in the world by the year 2000, the thirtieth year , at least $ 150 billion will be spent on this chinese government, that is, the pecks do not understand how much these technologies are important and the future will largely depend on them understanding this,

of course, and it was not in vain that the biden administration recently convened the heads of the largest ip corporations, including microsoft google and open ai, in washington, dedicated to security issues in the field. indirect intelligence artificial intelligence can help deal with problems like disease and climate change, but it also carries potential risks to our society, our economy, and our national security . we haven't had it for a long time, and so it is clear that all our purchases on travel cards on the internet of reality, our correspondence photos can all be traced. everything can be checked. now here is also this artificial intelligence of the neural network, how to control all this, how will it work out, so that artificial intelligence at some point will want to generally take over a person and human manipulation of the masses will seem like child's play before artificial intelligence manipulates all of us join us vladimir

kobyakov volodya well, how are you personally afraid. this is the development of artificial intelligence. i'm showing this to everyone watching, but vladimir gennadievich can already be seen to find dozens of wary statements of those people. who is involved in the development of artificial intelligence? here we heard elon musk yes, and here is another selection of not the latest people in the field of artificial technology. this is another powerful tool that will be used by some people for really evil purposes if the artificial intelligence gets much smarter than that and it will be very good at manipulating because he will learn it. we have very few examples where more intelligent thing under control less mind people are accustomed to wear sensors that track heart rate activity during sleep in the form of watches or rings companies develop their own neurotechnologies put brain activity sensors into everyday devices. the same corporations can control our personal data and, unfortunately, have access to

what is going on in our heads, and then i am afraid that the right to freedom of thought can significantly interfere with our intimate life. i i worry that this is a future for which we are not ready, but still artificial intelligence. he's different. we are at the household level for now. we mainly play with such text speech developments. here vladimir is otherwise there. 3 minutes ago they tried to write poetry. yes, after all, he did everything for me and admit it, in general, he wrote his text with the help of artificial intelligence today. you know yourself still not yet. i don't trust yet. i think, let him learn a little bit , learn a little about the peculiarities of style. eh, my view of the world. yes, and soon start, yes, used, i think, unfortunately, not far off. yes, time, all these speech and text technologies , first of all, of course, make an impression on the masses there to write an essay to prepare a script for a matinee. well, look. here are the natural

language processing revenue market sizes, but the blue bars are the global market estimates in past years and the forecast for the near future, and the red bars are the russian speech ai market. and after the advent of the gpt bot, almost all world players announced increasing investment in their developments. this is actually such a schrödinger box in which a cat sits and no one. and he knows in what state he is until the box is opened, that is, in what form what quality the speech artificial intelligence services will be, and what it will lead to in general, one of the first investors , the developer of the gpt bot, was , of course, elon musk, then bill gates invested and he said that more billions of dollars are being put into bots. here, according to information , bloombergates will invest $ 10 billion, and its investments and implementation smart botart. google's leadership has already been announced as a competitor's bot is actually threatening, because the search engine may lose its leadership and google's shares may sag . chinese billionaires have already announced that they will soon bring their high-

level speech bots to the market. it is either ba and baidu. largest internet. ah, the concern. china, at the same time, all the leaders of these developments criticize the uncontrolled development of artificial intelligence, and at the same time praise their business and opportunities, and we are damn strong fighting for positive options. you 're using artificial intelligence, let's say we have an algorithm that can automatically detect, for example, breast cancer in a lymph node. and if this algorithm can conduct tests and diagnose the disease faster than a person. and this is a very tedious task - it will save precious time. we may be underestimating the use of artificial intelligence. outside the healthcare system in business, artificial intelligence is ubiquitous and brings incredible benefits for optimizing supply chains for financial analysis and planning.

and here's what i'm interested in in this whole process, this is a multi-stage cooperation between a person and an artificial intelligence person, using a fact-checking tool does this in order to receive data for another machine, which will become more useful again for a person. and i think that over time, if we get this process right, we will be able to solve the impossible problem. well, well, now to all these questions. i hope we we will receive an answer from our guests vladimirovich guests in touch from st. petersburg tatyana vladimirovna chernigovskaya director of the institute of cognitive research, and doctor of biological sciences, and she is responsible for our natural intelligence. and in our studio, alexei masyutin, he is the head of the artificial intelligence center of the higher school of economics, of course, more of an artificial intelligence specialist and immediately a question. please say artificial intelligence. at this stage, it is time for good or evil to establish control over it. or let while everything is free swimming. let's ask a question first. uh, the opportunity to answer

petersburg tatyana vladimirovna do you think control is already needed? yes, of course, it is needed, and quickly. moreover, i’m not sure that we weren’t late, because i haven’t called him lately, including if the developers and western californians are also and they are seriously afraid, if the developers are afraid, then i think , there is every reason to be afraid and just people. and what to be afraid of you can form what as what kind of threat is still easy to formulate? a exactly? that is, simpler than simple, namely, that we lose control over what is happening moreover, uh many people think and i think so that we have already lost, yeah. here, how do you think alexei control? why did the genie come out of the bottle? uh-huh and as you know, he doesn’t

plan to reverse. alexei, how does jin think he got out of the bottle or did he? might as well contain it somehow. you understand that the questions are only provocative, only debatable, but you know it's hard to judge. uh, in that sense , artificial intelligence is a technology. and in fact. e, any technology is a tool, and let's say this is one of the leading researchers of the elikun. uh, on twitter , i think, recently made such a post, where with uh, gave an example. eh, in such a joking style he states that and let's regulate ballpoint pens, because they can be dangerous , they write all sorts of slander or lies and can sign some fake documents there, it’s quite possible, that is, uh in this sense. uh, well, artificial intelligence is some tool that, uh, amplifies our possibilities. where do we direct these opportunities? yes, but it depends directly on the person, because what you say was true five years ago, there and back.

now this is no longer the case. it's not that kind of tool. this is a tool that claims to receive status, ah personality, and it's not a fantasy, it's not a chisel. and this is not azize. well, this is a reality , they talk about it again. checks about a knife with which you can cut off cheese, or you can slaughter a neighbor. this story, of course, is cute, but it is old to the current situation have. well, then vladimir five pieces, rapidly learning. you understand, they learn on their own. moreover, there are signs that they do not learn in the way that we people know about, that is, it may turn out that they are ahead of us not only because they have an infinite memory. they don't sleep. they don't have pain there, their head and so on. because they

went the other way. nobody promised us that today's neural networks, these powerful epithia chats, are their variants. and that they will go in pursuit of the human natural intelligence? nobody promised this. for some reason we decided that. alexei wants to reassure you alexei that he will say an excuse. with a tinkoff vs. subscription, you get a special status, increased income on savings accounts , free maintenance of the tinkoff black card , and much more. connect tinkoff pro subscription and get the first month for free tinkoff is the only one. well, what did he ask? he asked if i knew how to work. we need

people with real experience, we teach those who are hired. find out your credit score for free on the site compare dot ru we will help you choose a loan taking into account your current paths and credit history and check that you have no delinquencies and no one has taken a loan in your name. if the loan, then on the site compare dot ru on vacation. he has a rest on the sea, to put off a vacation, there is nowhere else august 13 is the last chance to win a trip. play dreamed this sunday. dream big with a super prize of 100 million. the sberbank national lottery presents a credit potential service. plan your dreams. know the amount in advance and conditions for all loans available to you. calculate the credit potential in sberbank online right now apply for a loan. at the lowest rate in sberbank from four percent for the first month. summer with tinkoff mobile well, why aren't you with us, i

can see the mountains from my window? and grandfather everything is there, as you like. in general, if you do not come, i will call you every day, bank customers. tinkoff no longer pays per minute. order a tinkoff sim card and get free minutes forever phone. he is so alone we live in one country, but we are so different, the doorbell rang, and we do not notice the distance when we hurry to our loved ones. him into the water.

dear relatives, who wants to wish our young i wish digital technology. but true to tradition order tickets for saturday and pass the money to mom we train multitasking. and ready to fly across the country for the kiss of his beloved. therefore, it is good that there is someone who understands each sberbank for his own, for each his own for 160 million. summer with tinkov is no longer paid per minute. order a tinkoff sim card and get

free minutes, forever millions motorists buy osago or other insurance to compare the car, if the insurance is on the site and in the application. on the site and in stores with tinkoff pay, you can pay online with any phone with a sticker or by attaching the phone to the tinkoff pay terminal , issue any tinkoff card and choose your payment method online with a tinkoff phone sticker, he is the only one. no, the place is a magical adventure, a walk through unknown worlds, inspiring new achievements. after all, in life there is always a place for impressions. one has

only to open them there is no border and in fact, almost all modern modifications of language models are very simple to predict. and of course, but now, if we take the classic transformer models , then their task is to restore words using the so-called masking, which depend on

the surrounding context, when we really feed the model with huge amounts of data, after some time arises. well, there is to understand the word energents, that is, some properties that are initially. eh, it's clearly not there. did not plan to train the system. well , for example, mg5 is an open analogue of the multilingual gpt selection, or, uh, chad g5 by itself. they can have a dialogue with you in several languages ​​at once, because the data arrays contain this data and there are also parallel corpora of texts. and oh, again. and to say that there neural networks have their own goals, in fact. uh, artificial intelligence is some kind of, yes, a big computational black box that imagines a giant form there with billions of different parameters, where there, and divide multiply raise to a power extract the root or apply some function, but at the input it receives numbers and at

the output, it gives numbers further down the list. we know these names, we will call them whatever they don't know. yes? well, i’m such a football player, i’ll say my opinion that, uh, there is an example, too, someone compared that the largest tobacco companies sometimes lobby for bills that introduce barriers for there, well, increased excise taxes and so on. strange. why is it against their interests, sort of somehow, but in fact in this way barriers to entry for smaller companies by their potential competitor. here it is again, no matter how we treat it. here is the company. five, who developed gpti, she kind of claims that it is an open company, but nevertheless , a company specifically american headquarters in america invested 10 billion dollars in it, uh, and they are not disclosing the latest version of the gt4 model. and uh, let's just say, well, it's probably beneficial for them that no

other competitors in china or in russia, well, could not compete sufficiently. let me leave my five. cop. ah, the story is, why are the big players criticizing the development of artificial intelligence? yes, there is already an offer. we have already heard the congress. let's create some kind of international organization like the rich in nuclear power, yes , which will control the development of artificial intelligence, and then the question arises, if we create some kind of world organization under the control of artificial intelligence, which controls the pandemic, the direct intelligence of the organization will dominate. this is the first question. and secondly, will it become a konku tool? well, fight when they say that here is your chinese model. oh, she's harmful. let's get rid of her. your russian is also harmful, and so on. we will introduce only one control tatyana vladimirovna who should control all this? uh-huh, that's exactly what i'm talking about, that the situation is extremely alarming, because that's what a colleague is. uh, my interlocutor

said to uh about, well, as if suppressing competition, everything is absolutely correct. i completely agree with this, but we need this not easier. i'd like to uh hear an example. and here is a specific case, and someone got scared from such comic stories. i heard that some frequency user followed the advice of chad gpt, i think there is some kind of married couple. they filed for divorce. that is, well, roughly speaking, while communicating with the chat bot, and they asked some questions. there, probably, some context of their relationship was set out. and that's what to do there, he develops as a child. we understand that yes, that topic vladimirovna means that it's like the brain neural network develops until they are advised to get a divorce, then they can advise something else. secondly, i'm not talking about jokes. i can't, unfortunately, i just promised not to publicly say these things, but uh recently two weeks ago. i spoke with

one of the developers of this abroad, being. eh, so i talked to this neural network myself. she pointed at me. horror with this. it's something like that. uh, this one is not gps uh, according to uh-huh, but it was a network. uh, morphins, she hasn't gone anywhere yet, maybe they tested her, right? well yes, well, if it can be called that. you are talking, you are a really normal person , asking very complex answers and getting instant answers that are not the result. uh, bulkhead elements. it wasn't you who felt that you were communicating with an ordinary living person. and it scared you. yes, it was scary. well, yes, but this. well, i'm hateful.

so it’s not what i told me there, it rather scared me that it scares the developers who say that firstly it will have to be a huge speed and their training of these networks. and that they are not all understand what is happening. they themselves cease to understand, on the other hand, 25 years ago. imagine that you are walking down the street, and a person is walking towards you and this one says, how to get to the nearest store and puts it in your pocket? it scared you too. and now no. i agree with this. no we actually, my conclusion. from what we are saying, in addition to alanism, so to speak, which is not inherent in my conclusion, this is. we have reached a completely new frontier. this thing could get out of our control, viz. ah, artificial intelligence. we can become more powerful than? naturally, when we

cannot compete, but still, after all, there are some applied moments. we can use and maybe limit these networks to precisely these applied useful moments with algorithms that help us in everyday life. you know how to work with big data. uh, it doesn't matter at all without, uh, this kind of technology, not even technology, but technology. let's listen to alexei what he can offer us as ordinary residents, right? engaged in here is the center, let's say examples i will give, uh, again, here are the language models that they themselves are talking about, this is one of the areas. yes, now it has turned out to be a very large portion of attention. this is us in the media and in public discussion, but uh, for example, a we develop such solutions, and a there is such a thing as video clographs it-trackers

that track eye movement and uh, we learn we build models u based on the recording of the trajectory eye movements of the user, which allows there to classify and identify some kind of condition, that is, a dyslexia problem. this is very important in of the child population, that in order to diagnose it, an hour of active neuropsychological examination is required, which is very laborious and time consuming. and we offer a solution to address the primary source of the problem , namely the reading behavior of the child, where this problem lies to itself and look at the red dot. we recorded the movement of your eyes. while reading these simple sentences. see what it looks like. here is an adult with dyslexia reading this. you see that reading is very different from fluent reading. e child. uh, moves from letter to swell returns rereads we used various machine learning algorithms there is

a reading disorder risk group or e, normotypical group e we achieved a prediction accuracy of 91%. this is one of the straight, second application examples of this technology. show and here artificial intelligence. the fact is that i am not familiar with these devices, i have been working for many years, everything that you say. i know well what artificial intelligence has to do with it, but i will give an example of artificial intelligence here is a set of models that, according to raw data of the eye movement trajectory, perform the classification task , at the end of the conversation i will ask you a question that i always ask and ask and we still plan to plan whether we live on this planet or not. to give everything to these uh beings, this is the question that we must ask ourselves. i understand that neural networks. yes describe symphonies. ah, beethoven is being written. yes, they write the script, all this is understandable, but what will people

do, if you think that people will be freed from hard work, it's time to write strained to play in the summer, then you are mistaken, then how can we artificially slow down. this modernization with artificial intelligence, in order to save yourself or save jobs, is at least necessary. no, well, you need to really look at it. uh, we won’t be able to slow it down, unfortunately, so i would have my answer softer than i said, all this is a transfer to ours, you probably need to be wary , you know, we need to sort of get together and understand that we entered and we won’t enter, a already entered a completely new type of civilization. it really has never been like that. this is what e us competitor. so i want to make a comparison here with us.