[000:00:00;00] we have a tasting room opening in an hour, and then we are planning boat trips on a yacht on the lakes, well, a club, probably until the morning, i feel like a quiet romantic evening is gradually turning into a sabbath, if you ever have the opportunity to be alone , dial, everything is fine, wait, i don’t know your phone number, you’ll take it. yes, gentlemen, it looks like you got your cerebral cortex from an oak tree, are you, mikhailovich, having lunch, or what? yes, no, i was glad, have any of you studied german? i, oh, alfred, you will understand me, the word nach in german means the direction of movement, in russian it is the same, only much more specific, are you, mikhalych, because of a chick, or what ? micah, it so happened historically that i never shook up your unhealthy

liver. but now, it seems, my fatherly love for you is rapidly falling through the cracks, mikhailovich, we understand everything, we’re leaving,

yes, have a pleasant evening, a pleasant evening. artificial intelligence is penetrating deeper into our lives, what revolutionary changes in medicine is it capable of, what areas why it should be used extremely carefully, where to get specialists who will provide the country with technological sovereignty , what place we are in the global it race today, we’ll talk about this with the head of the institute of system programming of the russian academy of sciences, harutyun avitisyan. arkhanovich, the institute of system programming, as far as i understand, has a lot of experience in creating solutions in the field of artificial intelligence, and it has a long history, as far as i understand, this year on december 4 will be 75 years of information technologies in the soviet union, which means in our country, we are at least legal successors, now we have several areas where we use and implement artificial intelligence, we have created a big

data analysis platform, such a basic... platform from which we make specific customers have analytical systems, there are more than fifty different models of artificial intelligence , we are now actively involved in medicine, together with sechinovka we are a world-class center within the framework of this project, for example, we have made, well, probably one of the best ec analysis system, there is pathology , which is based on the analysis of more than 2 million ecs that we received there - from the republic of tatarstan, veliky novgorod, a number of klinnik moscow, this is already an implemented system, right? this is the result we got. now we are at the stage of obtaining the so-called russian device, this is a medical product, next year we will receive it, then the official implementation will begin, because medicine is a very conservative area, and we are essentially a computer power, one of the first in the world, and well, here are the experts in the field of artificial intelligence is constantly warned that it should not be used in critical infrastructure, and they also say that it turns out that artificial intelligence is capable of degrading, how big are such risks?

you see, artificial intelligence in its modern form is actually not very complex, but it is an it technology, yeah, because if we are talking about ordinary software in critical infrastructure, there are established mechanisms for the so-called safe development, yeah, they are also imperfect , but it’s more or less clear how checking the code, from design to maintenance, what to do with it, the field of artificial intelligence is just emerging and not only here on a global scale, here we cannot say that we are behind anyone, it’s just a fact, in this sense, of course, you need to be careful, procedures are needed. we need appropriate standards, state standards, we are actively involved in all this in our country, as far as degradation is concerned, we had an amazing experience when financiers announced some kind of competition, and this was several years ago, even before all these before covid even i think it was about four years ago, suddenly there was supposed to be a competition literally tomorrow, yesterday there was a financial crisis, yeah, not a single model’s accuracy dropped so much that

it was simply inconvenient to show each other. why, because the world has fundamentally changed, the context has changed, that is, you studied on one thing, the world has changed, in this sense , data obsolescence is a very serious problem, but it is also solvable, i don’t think this is... russia after all is in, let's say , a leading position in the world in terms of artificial intelligence, but forgive me my lame question, this is a question of resources, including computing resources, we have everything in order with this, whether we have enough of these resources or not, well, you’re right, that’s a serious question, in fact, i i am confident in our future, we have a very strong school of mathematics, artificial intelligence is still... deeply already, then mathematics is very important there. as for data and knowledge, we are doing well with this, because in the end, our data is our data, this is half the success. as for resources, there are certain difficulties, because our

microelectronics, of course, lags behind a little, and we must honestly admit this, but i think that despite this complexity, we are able to overcome it. we're talking about creating an open repository of trusted solutions, right? and this can create certain competitive advantages for our country, on the other hand, it seems like all the players who want can join it, but in the presence of countries unfriendly to us, you say, why do we need them, good the question is, you know, what’s the matter, because let’s say we conquer some market and even get a billion dollars of money there, yeah, real, yes, for example, any kind of big money, yeah, but in reality, uh, problems in the modern world, competitive long-term development like this is not just about money, we need people, personnel, and we have already moved to the so-called collaborative economy, yeah, what are we talking about, about 70 million people in

github alone, that is, hundreds of millions of people work together creating some technology, github is a huge resource, a huge platform, where there are many, many thousands of open projects that can be used by anyone in the world, these are all our clones. they are the same linux clones as red hat, which takes from the same place, google , which took android, made its own version, a year ago, at the management level it was announced that they were all returning to the vanilla version of the linux kernel, because it turned out to be too long for their own version expensive and not necessary, yeah, it would seem that there were competitive advantages, but no, why, because it is very expensive, that is, joint ownership, joint development is a very important element, well, sometimes we are reproached for the fact that domestically it is... to a certain extent vulnerable, because we are trying something there are billions of lines of code, even if we assume that only 5-10% of them are useful, you don’t need everything else, this is such a volume that no resources will be enough, now within the country, let

’s take it, even take it not artificial intelligence, to make it easier, linux, everyone knows it, we have several dozen companies, the linux operating system, several dozen companies create their own versions , high-quality solutions, of which three or four are at the top there are astra, viola, redsoft, dew, you can talk further, well, i know at least two of them, i even felt it, yes, yes, they take everything from international open source code, add their own codes there, of course, and create a product, and now they also have to , in parallel with each other within the country , ensure the required level of trust, so we, for example, took the linux kernel , turned to companies, we say, let's together ensure the security of the linux kernel, and you will... use it, now we have created a consortium of about 30 companies, i think in the near future there will be 100 hundred, over the past year and a half, 250 security fixes for the linux kernel have been adopted the international community in the last year and a half, please note, this is important, but for all of us, this means that at

the international level they see that russia is a technological country, and yeah, they see that we are open, we are not afraid, because that we are strong, within the country, around the core there are about 60 specialists, not so many yet, but enough. to keep the linux kernel in any state , even if we completely disconnect from the internet and develop it, and of these sixty only literally 20% are inside the institute, even less already, the rest are inside the company, and together we will keep what this means from a practical point of view point of view, if we find a vulnerability in the kernel, it doesn’t matter whether we saw it somewhere that someone found in the west, we literally, according to the regulations, can to say how many days it takes to write a fix for everyone , that is, to improve security, several years ago there was a case in our country when an error of this kind was found, and the company said that we were waiting for the patch to be made at the international level and they will send it and we will give it back, because there was no one to do it, now this problem has been completely removed, we need to do

the same thing, we need to do the same thing for all the main critical libraries, preferably for all software in general, but this is unrealistic right away do it gradually, we are currently holding about 15 projects, in the field of artificial intelligence, this is the pytor flow , the main libraries on which training takes place in our repository, more than 50 critical errors were found there, we say that if we have such a repository , to attract euroses there first, then maybe brix to make it not just russian. and internationally, so that all participants are guaranteed that we will not close either, because we have no goal, to become a center on which everything depends, we have a center build a network structure where we are safe, students from the smallest country, where there is not even it, a small country from any corner of the world, and a student at stanford should be in the same conditions, the digital divide should be eliminated, this is one of the ways, stanford doesn’t agree with you, why agree, if they didn’t agree, they wouldn’t accept, tell me, this is how quickly

, according to your estimates, as a specialist we are switching to russian software, and is this for the better, that is, this is the same principled the task is this, now, you know, to answer this question, you need - a position question, what is for the better, uh, if we want - for the better, what if we want to launch scientifically, if we allow technological sovereignty there, we want, for example , if we want to create such long-term sustainable technological independence, not just shoe a flea once, yeah, it can be done, but in the long term we must have the right model, we need to create an industry, yeah, that’s what i’m talking about if we don’t create an industry , it will be like, here i am i don’t even know what to compare it with, but if instead of creating a nuclear industry, for example, we would have bought 100 bombs there, or even two nuclear power plants, so what would happen next, well, that’s strictly speaking, this is the disease 90, when the technologies that we need, we will buy

them, yes, or we would make one super good product, but an experimental kalashnik, but no, it would be experimental, we could not replicate it, yeah, we need to create this industry, this is very important, but for this we need models, there is practical experience, we share this, including at our conferences, everywhere, it is quite possible, and accordingly, this is import substitution, and i prefer the technological one... it is feasible, but for each individual area it is necessary individual approach, because everything is clear in systems, real-time operating systems, most likely you have to develop them yourself, even if on open solutions, but further develop them yourself, these ideas are already growing into business and people understand that an individual company, even with multi-billion-dollar budgets, they are no longer able to accumulate this in one place, because the colossal problem of people fleeing does not matter. for what reason, too, within the country, we are moving from one company to another, and how it works, if

key people have left, then development is at least seriously hampered, and sometimes you just have to start again, so create a situation where the basic platform does not depend on one of a person, from one will, from one company, it develops regardless of anything, the consumer is more or less the entire country, now, if we get involved in this, if we get involved correctly, we concentrate all this energy, then we will be able to make quite mature decisions, in the future several. years and thus we will close our infrastructure level along with operating systems, with other things , in general i want to say that systems programming, even some kind of analytical review was that in systems programming russia is number two in the world after the americans, so we are quite are able to do this, now society and the economy are transforming into digital ones social economy, and accordingly, it is clear that, probably, for citizens, for business, for the state as well, this creates certain advantages; what, in your

opinion, are the main challenges of this digital transformation? firstly, i want to say that attempts to stop the introduction of technology or transformation , they are doomed, why, because when labor productivity increases, the other day we held a meeting at gem, i was just at a round table on digital digital international relationship, and there i was one of the participants was alexander gasnikov from a scientist in our country, a world-famous mathematician, and he said, here is the gpt chat, we tried to use it when writing articles, yeah, and it significantly increased our productivity, because it many parts of the article were simply generated for us, we then edited them, we saved time and concentrated on real results, you know, what kind of world we are moving into, we are talking about the highest level of science, mathematics, one of the leading

people, he calmly says this, because what he is not afraid, because he... in the same medicine it will appear that using the same signals it is possible to determine diagnostics that were not done at all before, for example, yeah, and gradually the world will move into a state of constant screening, rather than treatment, we are now this has not happened yet, but in 10-15 years all wearable devices will already have an ecg, they will determine a lot of things, and the threats that come from there are serious, but first of all this is what is changing, our consciousness, europeans, for example , followed the path that some parts of artificial intelligence will be prohibited from being used, for example, biometrics in public places to collect social scoring, some other things will be prohibited, and this will happen right away. to the outflow of innovations from europe 18% immediately to a drop in the level of security, mind you, here are philosophical questions that we as a society must discuss internally, but we, what we want, this information flow still affects us, especially if in childhood , the germans, i now know how serious this will be is unclear, but they are discussing what to ban in schools, for example,