Lemmy is a bit of a blind man now . They have already exhausted their possibilities. In my opinion, artificial intelligence is certainly not capable of interpreting the law. You should not believe that such a program will write a lawsuit or respond. The problem is purely engineering. Law, contrary to appearances, is a classical science subject, and if we assume that the system is stochastic, it is random, then in fact there is no law. Yes. This means that on average, 33,000 pages of legal regulations are added to Poland every year . What chance does a judge have at the end to interpret the legislator's intentions? Whether the law becomes a source of chaos is a system of chaos. Is anyone there testing the law? You throw it at millions of people, let them use it, maybe they will survive, maybe they won't. Well, more or less if engineers worked, there would probably be no engineers anymore. System, by non-deterministic system. I saw so much human misery that I came to the conclusion that I should start working on possibly creating a deterministic system. Understand that there is actually no such thing as coherent, uniform established principles of law-making. There are some guidelines of course RCL for example. So if a verdict is passed that violates human dignity, it also leads to a greater drama and catastrophe. Sometimes even greater, but it all happens in the privacy of the courtroom and in the privacy of the courtroom. Every judge in good faith must, because there are no other options, interpret this program, i.e. the law, in his own way. Faculty of law should establish engineering departments, right? For example, 70% of planes go down. What would happen? And in the case of law, well, that's exactly how it is. Here, together with Professor Marek Krokewicz from the Wrocław University of Science and Technology, we decided to propose a new concept of law lectures. We called it the deterministic process of legal interpretation and semantic representation of law. The system is obviously very ambitious. It can be said that it is not so much fundamental as even civilizational. We do not want to create any new law. We want to provide a tool to make the law deterministic and to be able to predict these consequences, and we cordially invite lawyers, computer scientists, and logicians to cooperate with us. Good morning everyone. Hello and welcome to today's podcast. We have a conversation ahead of us that I deeply believe opens an important and necessary chapter in the discussion about the future of the rule of law, technology and the role of humans in an increasingly complex world. In this context, questions about the place of humans in the world of modern technologies, the direction in which the contemporary legal order is heading, and how artificial intelligence affects our understanding of knowledge, responsibility, and the functioning of the state become particularly important. We will consider whether the state in its current form is still able to protect fundamental values, whether it needs a thorough reconstruction, and whether technology can be a tool not for deconstruction but for restoring order. We are facing a conversation that requires calm, courage and intellectual openness. A conversation about boundaries, responsibility and the value that starts it all. about human dignity, which should be the center of every system, including the algorithmic one. Today's guests are developing a new, unique concept on a global scale of deterministic interpretation of law and semantic representation of law using computer science, in particular knowledge engineering. We present to you, Professor of the Nicolaus Copernicus University in Warsaw, Doctor of Laws of the Jagiellonian University in Krakow, Mr. Mariusz Miąsko, representative of the College of Law in Lublin, author of numerous scientific publications on the digitization of law, creator and producer of innovative IT systems used in practice, awarded many times by, among others, IBM, member of the AI ​​working group at the Ministry of Digital Affairs, including the leader of the AI ​​and labor law section, co-author of the official document of the policy for the development of artificial intelligence in Poland for 2025-2030. also the author of groundbreaking legal regulations in force in Poland, including the form of employment, the harvest assistance contract and the rules for posting workers to the European Union, as well as the author of numerous publications, speaker at international conferences, founder of the WIGEN law firm, and for 15 years he has been researching the entropy of chaos in the legal system and is the leader of the scientific component in the project of deterministic and semantic representation and interpretation of law Vera Semantica Juris. On my other side, a professor at the Wrocław University of Science and Technology, PhD in Engineering Marek Krokewicz, head of the semantic processes and systems analysis lab at the Department of Applied Computer Science at the Wrocław University of Science and Technology, has over 30 years of experience in knowledge and information systems modeling , and lectures at Stanford University and the City University of New York, among others. manager of numerous international grants, author of over 50 scientific publications, creator of the Semantic Knowledge Base projects since 2005 and the Association Oriented Database since 2009, founder of the KIK research group. He holds numerous international certificates in project management. Business practitioner, former president and member of the supervisory board of strategic companies, co-founder of Science in Software spółka z o.o. currently the leader of the IT component, deterministic, semantic representation and interpretation of law, which we will talk about today. Gentlemen, the thesis that law can be universally applied by artificial intelligence has been gaining popularity recently. I would like us to discuss this topic a bit, both from the legal and IT perspective. Perhaps we should start from scratch then. Can the law be applied by AI? Let us answer the question of what led you gentlemen to the path of creating another concept of interpretation, since there are already quite a lot of them in force in the world. So, it seems to me that this question needs to be divided into two answers, because the first question is, what is artificial intelligence? I think that the professor will respond to this very well, because I think the answer will be slightly different if we refer only to the issue of LLMs, and different if we refer to other types of artificial intelligence engines. Secondly, we need to consider what the application of law actually means. By applying the law we mean interpreting the law, or by applying the law we mean preparing, for example, certain analyses, models, directions of the defense lines. These are two completely different issues. As for the first issue, whether the law can be applied in the sense of interpreting the law by artificial intelligence, and in particular by films, because this is the view that has been coined, then in my opinion artificial intelligence is certainly not able to interpret the law. We are absolutely convinced of this. We work on the implementation and application of law in its meaning, processing the content of a legal provision through the process of legal interpretation in order to obtain a scope of legal norms. We are working on this project, I personally have been working for 15 years, the professor has been working for a very similar time interval, also over 15 years. I am from the strictly legal side, the professor from the engineering side. And for me it is obvious that artificial intelligence as we understand it today, and especially in the LLM formula, has no competences to apply the law, i.e. to interpret the law. And now the question is, why? Because in my opinion there are two reasons. One of an internal nature, the other of an external nature. As for the reason of such an internal nature, it must be said that artificial intelligence, and LLM in particular, has a whole conglomeration of technological limitations, which the professor will probably outline in more detail, but I would like to briefly address them. This may primarily be a matter of establishing ontology. What is law from the ontological perspective, i.e. what is the scope of law? whether to the legal ontology, i.e. whether the artificial intelligence or the artificial intelligence engine will search exclusively for information y and try to obtain the scope of the norm through information that is formally represented as such among the sources of law, constitutional sources of law . So the Polish Constitution indicates that among our sources of law we have the Constitution itself, but we also have statutes, regulations, lower- level acts, as well as acts of local law and EU regulations. There is a highly controversial issue of the role of directives and the scope of directives. It is often wrongly believed that legal ontology also includes a conglomeration of other apparent sources of law. for example, European Commission guidelines, positions, various types of institutions, offices etc., etc. It is often believed that an anthology of law also includes, for example, a line of case law or a dominant line of case law, or that, for example, the ontology of law includes the achievements of legal science doctrine. Well, if artificial intelligence, if LLM wants to obtain the scope of the norm not only from the content of the provisions, their semantic, semiotic, logical scope, but also wants to obtain this information about the scope of the norm using apparent sources of law, i.e. those I mentioned, including, for example, the guidelines of the European Commission, then it will ultimately obtain a different result, a different result than the one that the artificial intelligence model would obtain from solely using the processing of the content of the provision, for example, a statute, using the principles of legal interpretation into the scope of the norm. Can new technologies and artificial intelligence violate human rights and dignity, or, on the contrary, do they actually protect them? This is an excellent question and it is a fundamental question in general. It must be said that I would reverse the question, whether the current system that we have, the legal system, whether it ensures the average citizen the right to dignity to a sufficient degree. I think that the average Pole is dissatisfied with the functioning of this system. He believes that it is a labile system, that it is non-deterministic. What does non-deterministic mean? This means that it is a system that cannot a priori answer the fundamental question of whether it can make certain decisions based on predictable results. For predictable results. But I would like to address one more issue here, because I would like to finish this previous thread, because it is fundamental. That is, why, in my opinion, LLMs in particular, but also in general, as a rule, artificial intelligence models cannot interpret the law, because we have pointed out that LLMs, wanting to determine the scope of the norm, search too broadly. That is, they search among the collection of books, textbooks, opinions, and judgments. They are not allowed to look in the formal sense, they are not allowed to look for the answer there. yyy aimed at defining the scope of the norm, because even the most perfect scientific studies or even the best court decisions are only a kind of opinion on the scope of the norm. I would like to remind you that they are contradictory and therefore the LLM will not be able to make a decision at all. So this ontology is the first problem that indicates that the application of law by LLM is a dead end. We as citizens very often have this impression, this hope, that the law is deterministic. What does it mean that it is deterministic? This means that we read a certain provision and a priori we know what the decision of the state administration body will be after reading the given provision, what the decision of the court will be. with astonishment, later being defendants or parties in administrative proceedings, we find that there is no, probably almost everyone stated that the decision was very often different from the original idea. There is also this idea, this kind of statement, that where there are two lawyers, there are three opinions. And we should consider on a purely scientific level why this is so. So what is the reason for this situation? And the reason for this situation, ladies and gentlemen, lies in the non-deterministic process of legal interpretation. We all do, because I have to explain something so fundamental here again. In fact, courts and state administration bodies do not make decisions based on the law, but neither do we when concluding contracts. We do not make decisions based on regulations, but on legal norms. This is a fundamental difference. provision, norm of law, these are two completely separate phenomena, one could say issues, and they have fundamentally different scopes. And now, in order to transform a regulation or a group of regulations into a standard, there must be some language that speaks in some deterministic way, i.e. in a repeatable way, to perform this transformation. Well, that's what the set of rules for interpreting the law is for. And now the thing is that we, as citizens, wrongly assume that this set of principles of legal interpretation is deterministic in nature, that there is a predetermined set of first-degree directives and second-degree directives, and we wrongly assume, as it turns out, that there is a closed catalog in law, that it is of such an enumerative nature, a closed catalog of these second- degree directives, we often assume, completely wrongly, that this catalog is completely exhaustive. Meanwhile, we also assume that the relationships between directives, i.e. the order of their application, is predetermined. Well, my next question was supposed to be whether law means order or perhaps anarchy? This is an excellent question. Um, in general, if I were to reduce the word law to one other word, I would say that it is order. And if I were to make it even more clear, I would say that law is not disorder. So there is no such structure in which the state, that is, all of us , created for ourselves a certain artificial entity, because the law is a completely artificial entity, only to introduce disorder. Law is introduced for one sole purpose: to organize our lives, our lives, the functioning of the state and statehood as such, our entire community, that is, our community. And now the point is that if we approach the issue this way, it means that any other formula that is not deterministic is unacceptable. This means that the formula that currently functions is unacceptable, that despite, say, nine legal interpretation concepts in force in Poland for almost 100 years, we do not have a clear and 100% defined catalogue of first-degree directives, and especially second-degree directives, nor is there a defined relationship between them, i.e. the order of application of superiority, subordination, but not only that, because there are many more of these possible solutions and this causes the law to become a source of chaos, a system of chaos. Just imagine if it were like that in mathematics, we all learned the equation with parentheses in primary school and it is valid, each of us knows, everyone who is watching us now knows that first we have to do the interpretation, first we have to do the equation in parentheses and then outside the parentheses, because if we did the exact opposite operation, we would have a result that is exactly different from the one that has been empirically proven to be correct. So every child who has completed the eighth grade of primary school knows this rule. But this rule does not apply in legal sciences. Well, this may be shocking for most citizens, because most citizens, including a significant number of law students, assume that, first of all, all the first- and second-level directives are known, and especially as regards the second-level ones, as I have already mentioned, there is no closed catalogue, an exhaustive catalogue, and even more so, the relationships between them are significantly defined. And what is even worse is the specific relationship between them and the methodology of their application. Therefore, the law is currently a source of chaos. I have been researching the chaos system and the entropy coefficient for 15 years. I am trying to determine the entropy coefficient, i.e. whether this chaos coefficient tends to increase, whether it is constant or decreases. And these conclusions are shocking, because every year in Poland we have an average of 33,000 pages of legal regulations. 33,000 pages, not of regulations, but pages of legal regulations. When we add EU law, we can expect perhaps around 50,000 pages. And each circle, each subsequent recipe on each of these pages creates relationships between other recipes. These relationships are indefinite. This means that we almost don't know what these relationships are like. Of course, we are trying to establish some direction. The thing is that artificial intelligence, such as a classic LLM, in order to have any chance at all of processing the regulations into the scope of the standard, would first have to know the first and second level directives. Later she would have to know the way of relations, the arrangement of relations between these directives, but she has nothing to know, because they do not exist. Because they don't exist. And this is, of course, probably shocking for the average listener and viewer of our channel, but of course, legal theorists are perfectly aware of this and that is why for 100 years they have been trying to create a concept of interpretation that will solve this important, basic, fundamental problem for the Polish state . this state of chaos and the increasing entropy coefficient of this chaos system, this is a chaos system, this is a well-known scientific institution known in physics, generally speaking, in a very simplified way, it comes down to the fact that adding subsequent quantifiers to a given system without modeling them and introducing engineering systems for modeling these quantifiers must lead to collapse, must lead to some kind of collapse of this system inward. This is why we decided, together with Professor Marek Krokewicz from the Wrocław University of Science and Technology, to propose a new concept of law lectures. We called it the deterministic process of legal interpretation and semantic representation of law. But I would like to finish these remaining threads, why we claim that an average LLM and such a classic artificial intelligence model will not be able to handle the application of law in the sense of interpreting the law. Well, we need to know here that in order to transform any information, that is, any information contained in the content of a provision, which is only an editorial unit, some editorial unit, then, apart from this deterministic process of legal interpretation, we must have consistency at the input and output as to what the intention of the legislator was. Because please imagine that the process of creating law is a certain precisely defined sequence of events. Law does not begin in the Sejm, i.e. in any parliamentary group, nor does it begin in a parliamentary subcommittee or committee, and even less so in the Sejm or the Senate. The law starts much earlier. We first have a certain event or a hypothesis of a certain event. For example, someone drives very fast on the highway, an accident occurs too quickly, and undesirable consequences occur. It doesn't have to be this way. this event does not have to happen. In fact, it is enough for us to imagine that this could happen and to want to take certain actions to order, prohibit, reward, etc., etc. Normalizing, regulating, whatever we call it, some kind of action. And at this point we need to imagine the scope of the norm. We have to imagine it. And what do we do next? And then we process this idea of ​​the scope of the norm into the content of the regulations, that is, we select words, conjunctions, commas, punctuation marks, various types of quantifiers, as I would put it, logical function, that reflect the scope of our norm, and only then, when we have processed this idea of this norm, this hypothetical norm, translated it into language, do they then enter the legislative process. they start to go through the legislative process and then they finally become a regulation, from which we will decode it back into the scope of the standard. And now please pay attention to what happens along the way. If we have an idea of ​​the scope of the norm, it is often described in words. Where is it described in words, ladies and gentlemen? It is described by the RCL, the government's legislative center, it is described in the OSR, the assessment of the effects of regulations and there we have the justification and there we have the description that a given regulation is introduced because it is intended to protect someone or protect someone else. Where else do we have such a description? We have, for example, in motifs. As we know, all EU regulations begin with recitals. We don't have before we have the first article, then before this first article we sometimes have several dozen pages of recitals that explain, give us context, give us context, so we have this norm described in words, in many words, in sentences, this hypothetical norm, but pay attention to transforming this hypothetical, this imagined norm, this future formal norm, only when it becomes an actual norm only after the formal concept after the signature of the President of the Republic of Poland, then in order to transform it into the content of the provision, we have to imagine in what context, firstly, and secondly, using what directives of legal interpretation and in what configuration we will identify this scope of the norm with that with the content of this provision. the imperfection of the existing concepts of legal interpretation, because the professor and I draw heavily from them to make it sound good here . We do not deny virtually any of these concepts. We learned from them , built our competences and decided to expand them. We have noticed that, in addition to what is contained in the existing scientific achievements of legal doctrine and legal theory, there is something more that results very simply and very obviously from legal engineering when programming regulations into computer code. That is, we have noticed that the existing concepts of legal interpretation start outside the two basic concepts, de facto start in a sense too late. They begin by trying to decode the provision into the scope of the norm after it has been adopted, but they also, in a way I would say imprecise, refer to the need to determine what the intention of the legislator was, what the legislator wanted to achieve with a given provision. There used to be a rule: Clara non interpretanda, that what is clear does not require interpretation. Nowadays it is believed that omnia is the judgment of the interpreter. Everything requires interpretation, even seemingly clear regulations. This is obvious, because we read one provision, it is clear to us, but it is somehow connected to another provision of another branch of law. Very often, but it has its own connection in terms of scope, and if we do not analyze both of these provisions or a group of provisions, several, sometimes even several dozen provisions, we will not extract a single common norm and at least we will not determine the proper scope of this norm. However, these concepts, in general, apart from not assuming the need to create a deterministic process of legal interpretation, did not take into account the context. We once talked to the professor here about where this context should be placed. However, when we talk about context, it is very important that we distinguish very clearly how we understand context. It's not about the context of the provision in the legal act, in the structure of the legal act, but about the situational context, the external context, and that's something completely different. Now, the professor and I were wondering some time ago where these contexts should be placed so that we could carry out this entire process deterministically, carry out this entire process of interpreting the law deterministically and a priori, so that we could determine the scope of the norm. And the professor rightly pointed out that these contexts should be placed in what we call the notion of a norm before the legislative process, before the content of the provision is created based on this notion, this legal norm. They must be applied there, and only then will the same contexts be taken into account when decoding the provision at the very end, when decoding by the court, by us, by the state administration body, they will be taken into account again, but in this case, not only the contexts will be taken into account, but also the same decoding code will be taken into account at the beginning, at the end and at the end of this path. This may seem difficult, but I will give an example. You all know what an enigma is. Well, everyone knows that Poles, great Polish engineers, mathematicians, and scientific minds, have solved, colloquially speaking, the Enigma, the Enigma encryption machine. In order for the operator who received the information and was sitting at an Enigma machine somewhere in some submarine to be able to obtain this information, he had to have the same set of ciphers as the operator who transmitted this information. If this were not the case, the sequence of events that occurred in the enigma would be purely random, and one could only count on the chance that the information sent coincided with the information received, but this would almost certainly never be the case . Statistically speaking, it is identical in law. This means that if we want to decode information, i.e. determine some scope of the norm at the final stage by the court, then we must have the same code, the same set of quantifiers also at the stage defined primarily at the stage of creation, transformation of this hypothetical norm, the idea of ​​the norm into a regulation. If these circumstances do not occur, we will never develop a deterministic legal system. This means that we will never be able to be sure that we are reading a legal provision and we are absolutely certain that at the end of the year, two, or three, the office, official or judge will decode it exactly the same way. I will give practical examples. This is this is you asked a question about human dignity. How can we talk about human dignity in the current legal system when we have handled thousands of cases in our office. Thousands. Thousands, I don't know how many, but it seems to me that it's something like, when I say thousands, then we're in the region of over 10,000 cases. So some kind of comparative attempt occurred in my life. I believe that the judgments and administrative decisions with which we disagree are in the sense that we believe that the scope of the norm was incorrectly defined, very often inappropriately, and this is not an accusation against either the official or the judge. This is perhaps at the level of 70%, or maybe more. So how can we speak? And let me give you an example, among these events, among these proceedings we have some in which some entrepreneurs received penalties. The record holder received a fine of PLN 500 million . 500 million zlotys. What did this client receive this sanction for? for reading and decoding the recipe in the most obvious way possible, to the best of his school knowledge . I share this point of view, because we read these provisions in the same way when he turned to us after the first stage of the administrative proceedings . What's more, we even deepened the validity of the argument this client presented. So what if, in the end, the Supreme Administrative Court decoded the same provision or, in fact, a group of provisions exactly, oppositely, exactly, inversely, then, ladies and gentlemen , there are thousands, tens of thousands, hundreds of thousands of such cases. I have seen thousands of cases of obvious wrongs, tears, human drama. And for me this is the main source of motivation. I mean, during these 25 years of my legal profession, I have witnessed so much human misery, so many trampled upon people, so many disrespects of human dignity by the system. And in fact, it is not even because of people, not even because of judges and not because of officials, but because of the system. No, not even through lawyers, so that we understand this correctly, but through the system, through the non-deterministic system I saw so much human misery that we came to the conclusion that I started working on creating a deterministic system. The professor and his team worked precisely on the same subject based on specific life experiences, and this is the only direction that gives us the possibility of the functioning of the state as such in the future. In the public space, the concepts of artificial intelligence, language models, data, and knowledge are often mixed up. Please tell me, is LLM Artificial Intelligence? Yes, here indeed a great many ambiguities have arisen, I would even say they are becoming more and more established. Let's start with the fact that in many publications here , I don't want to talk about scientific ones, but it also happens sometimes, the concepts of knowledge, information and data are confused, which is a fundamental issue from the point of view of an IT specialist. Of course, we are talking about artificial intelligence, but the fact that something is something does not mean that it completely fulfills this concept. So it's not like llmy is artificial intelligence, it means that there is nothing except llmy, although at the moment they are simply fashionable and popularly used. Artificial intelligence, on the other hand, has a very long and very extensive history, because, for example, image recognition systems, which are very popular and we know them from mobile phones , there is also artificial intelligence and a lot of other methods, approaches, solutions that are very, very interesting, and they have absolutely nothing to do with linguistic and statistical models, because some of them are, of course, based on statistics. Here, the professor brilliantly explained the shortcomings of both these statistical systems in the context of law. And here we have many hours of discussions and conversations with each other, so we have a great sense of both disciplines, I think. However, I wanted to point out that language models are a wonderful tool that, I think, has a great future ahead of it. However, like any tool, it is suitable for a certain class of tasks, sometimes very large, sometimes narrow. However, I apologize to all LLM creators and fans, but every tool has some limitations and disadvantages, including LLM. Well, the professor actually presented these flaws, so I wo n't even repeat them, but I wanted to draw attention to this determinism and what it's all about. Well, imagine, and I think each of you can imagine, a situation where you are writing a computer program . I know it may sound strange that I said that anyone can imagine it, but I will repeat it , I will risk this statement, because what is a computer program? This is a series of instructions, a sequence of instructions, exactly, where as input we receive this sequence of instructions, some data, and at output we expect some results. Well, now it may be that a theory has long been established about this, that a Turouring machine may be deterministic, not deterministic. In the deterministic case, if we always have the same set of data at the input, we will get the same set of output data at the input. Well, this is a deterministic machine, to put it briefly . And now if we write this program, and we have the data, we would expect the stability, we would say, of this system in the deterministic sense. Okay, what about the law? Well, imagine this and here I will enter the area of ​​law. Probably not very skillful, but I will somehow try to show the dependence, the relationship. namely, imagine that we are such a machine, that we are such a machine, such a computer, only we do not have programs written, for example, in C#, or in C++ or in another language, Python and so on, we only have legal regulations. And what do these legal provisions say? They speak of certain commands, prohibitions, certain possibilities, yes, permissions. and such a dondic modality we would say. So we should in the world, and here you also rightly said about ontology, I will come back to it, in what kind of world do we function? We know this world to a certain extent, of course, which means we are able to get to know the world using our sensors, which means we build an ontology in our mind. This is precisely this vision, this model of the world. What is the difference between large language models (LLMs) for example and ontology-based solutions? That's exactly what I 'm getting at, that's the difference. I will also tell you why I took up this topic. Of course, it is very interesting from a purely IT and legal point of view, of course. However, I also have personal motivations here. Well, it happened to me that I was a defendant, of course. Let me make it clear to you right now that I am not, in principle, a serial killer. The point was that I was the president of the company and some driver committed a traffic offense. And I ended up in court because I didn't indicate who took and used the car at that time . The judge stated, and the high court stated, that there is no regulation that I have to keep such records and that I could not remember, but according to him there should be such a regulation. And it was on this very subject , based on this provision, that he convicted me. And now I thought that I had been wronged, that my dignity had been violated and that it was all just a coincidence, that I was an unhappy person in general. one in a million, but I had the pleasure of meeting the professor, complaining about what happened to me, what a great misfortune, that I must be a terribly unlucky person for something like that to happen to me. And the professor commented that, well, that's typical. And that was the second shock for me. The first one is the high court, he gave me this shock, and the second one is the professor. Well, I found out that this is not an incident, it is a mass, mass accident. Such cases are massive and it got me thinking, since there are so many such cases, the professor told me about many others, of course not thousands, because there was no time, but the ones I heard were shocking. I paid maybe a few hundred zlotys there, but here we were talking about a few hundred million. However, my motivation has increased a lot. It increased a lot because I thought it was a civilizational problem, because if such a large number of people leave the court with a feeling of disrespect for the office, with a sense of lack of dignity, because this happened to me, then it is not a problem. I dealt with it very quickly, but there are people who lose their property, lose their freedom. something needs to be done about this. WC I simply decided that something had to be done. And here, as a result of extremely interesting discussions with the professor, we came to many conclusions, but one of them is fundamental. It's about determinism. Well, there can be no randomness. If we assume that the system is stochastic, it is random, then there is actually no law. Yes? So order ceases to exist. it starts to be random. Um, what happens when there is no law? Well, if there is no law, then by definition there is anarchy. Only, so we go back to what you said at the very beginning. Yes Sir . Exactly. Exactly. And this is this wonderful loop that shows the loop in this sense of looping, shows how we return to the sources and how important they are. Ladies and gentlemen, we need to consider what anarchy is and who anarchists are. I was a legalist. I was such an extreme legalist that I began to fear myself. A legalist yet a statist, a statist on top of that, and not an orthodox one might say. I won't say who I am right now, because it would be quite difficult to define and brutal, but let's say euphemistically, who is an anarchist? Because you probably think that an anarchist is a person who wants to destroy order, right? Ladies and gentlemen, I thought so too. I generally thought that anarchists were the worst evil. No, an anarchist is a person who has stated that the absence of law is better than law, that the harm caused by law is greater than the complete absence of law. That's not there yet, but we're getting there very quickly , because ladies and gentlemen, from my very humble point of view, if there were no law, because I was in court quite rarely, I'll give you a second case, also a very positive one, so that it doesn't seem like there are only negative ones, but if there were no law, I would feel more just . This is tragic. This is for me, this is my worldview has been completely reversed. I am not an anarchist in the sense that I want to destroy the order of law, but I am beginning to see that if the law ceased to function, to exist, people would spontaneously implement justice better than what is happening in the institutions intended for this purpose. And that's for me, for me personally it 's a, uh, failure, you could say, because my whole life has been very focused on legalism, and now if I commit some minor offense, I go home with a great sense of satisfaction. I will give a second case, so that it is, um, um, so that it is not some kind of attack on the courts or on the law as such. Absolutely. Exactly, quite the opposite. Um, I was in court for the second time, because I've only been there twice in total, so I can describe all the cases, i.e. these two. Um, I was in court for the second time . In turn, another institution, a cooperative, I won't mention its name out of pity, sued me because I was supposed to let them use the premises and they posted an announcement downstairs in the stairwell. Well, I didn't see it, because I rented the place and of course they took me to court to pay some kind of fine or, I don't even know the legal form, as, um, I didn't agree with it, it's not about money, I did n't agree with it structurally. Uh, I ended up in court. Judge yy listened to me very patiently, and then quoted the regulation directly stating that the cooperative is obliged to agree on this date with me. Not only that, post it, notify it, but even agree on it. and ended the hearing by dismissing the claim. I only know a few judges personally, and these are truly people, at least I know them, who love their profession very much and try to perform it with utmost diligence, and they often regret that their colleagues, for example, commit certain acts of carelessness, let's put it mildly. So I would be very far from throwing everyone into one bag, because that's not really the case . And that's why we took up this topic with the professor, because it may actually be funny that a computer scientist thinks about human dignity, but a computer scientist is also a human being and he also cares about human dignity. That's why my motivation is enormous. Which approach do you think holds better promise for the rule of law in the long term ? Or is it possible to integrate both approaches? Yes, extremely important questions. Two, of course I don't, uh, I don't feel competent to predict the future, but I will simply use the information that experts around the world are increasingly raising. And these are the opinions that, unfortunately, lm is a bit of a blind man. They have already exhausted their possibilities. Putting in further billions, because here we are talking about billions, or rather tens of billions of zlotys, does not bring much. Moreover, LLMs require enormous computing power. Enormous computing power means an enormous demand for energy, and this is not an enormous demand for energy purely in terms of costs, but a certain German scientist even calculated that if we continue to develop at this rate, in a few to 10 years there will be a shortage of energy on Earth. If only we powered them with nothing else. So this is a bit of a dead end. Besides, they have already touched a certain threshold. Well, what the professor said and what I mentioned, without ontological knowledge, that is, without knowledge of the world, they will not overcome certain issues. And in a situation where we want to address legal issues, ontology is absolutely essential. we may have simply demonstrated this in our earlier discussions . Well, there is neither time nor opportunity here to present certain evidence. However, yyy, this is an absolute, absolute necessity. So, to conclude, there are experts about the future, and experts who have been dealing with artificial intelligence for decades are stating quite often, more and more unequivocally, that systems based on ontological knowledge are the future here. And the connection here, we, the professor, and I had another extremely interesting conversation just yesterday. in the evening at night actually, that combining these two approaches is extremely difficult, as always hybrids are extremely promising and LMY, such a last sentence on this subject are an excellent tool, so that it does not sound like an attempt to simply diminish their achievements. however, they have a limitation. Deterministic systems also have their limitations, but combining them as a sequence is, in my opinion, an extremely valuable approach and extremely promising. And this is precisely the direction in which we are heading, Professor. Yes. I would refer to this statement of the professor referring to the ontological issue, and I would also extend it to include a certain ontological and semantic scope issue. we assume that law is directly related to the world around us . Meanwhile, this law is nothing more than a completely artificial entity. This is, we have to know that this is, we can imagine that this is such an artificial bubble of certain fictitious, in quotation marks I say laws of physics. That is, it is a being in which we create our own internal laws of physics. A lot of people watched the movie The Matrix and there were, when you entered the Matrix, these laws of physics applied, it was an artificial being and artificial laws of physics applied there. This is both a great relief and an opportunity for law-making, but at the same time it is a great problem for law-making. Well, it would seem that we can model every situation and every context by making specific provisions in the regulations in order to create an ideal world. The only problem is that we have to do it in law, that is, in words. In words and in the meanings of these words and in the scopes of these words. The point is that the law is the law is a fiction. The law is, it is some kind of fictitious entity and this law is based on words and unfortunately or fortunately I will not let you judge these words, words are not things, they are not persons, they are not time. Words can at best describe things, people, or time, or the relationships between them. And the accuracy of this description, or rather the sum of the inaccuracies of this description, is the source of interpretation problems. And it's like the professor gave a brilliant example, in my opinion, of a small hypothetical child, a non-existent small child who, let's say, is one month old, already knows all the recipes, but he can't connect them to anything , connect them to anything , because for him it's just a string of specific words. He knows these words, he has them encoded in his memory. We don't know how, but it's a hypothetical child, but it can't relate these words to the entities it sees around it. He sees, watches, observes, sees a table, but does not know that it is a table. sees a chair, but doesn't know it's a chair. He doesn't know that this is what the regulation says. And unfortunately, to put it simply, what I'm saying now is hyperbole, but it's done consciously to better understand the source of this increasing entropy of the legal chaos system leading to the inflation of law that the professor mentioned, i.e. the lack of acceptance of norms, a reduction in sensitivity to the functioning of these norms, these norms. And now let's imagine such a situation and we have to realize the fact that all legal provisions have to be interpreted. And now the interesting thing is that in law it is assumed, in legal doctrine it is assumed, legal theory is assumed , that we have four kinds of language, as if four languages. colloquial language, official, dictionary language, it can be called something else, but I think we all understand what I mean. We can say that it is state-owned. In German there is a word for this: hochhdeut, right? as an officially recognized language. Legal language is the language that is often found in statutes and is often defined and specified by legal scholars and scientists. And legal language. Legal language is that which results from a legal definition, where we have a word and a description in many words of what that one word means. This is a big simplification, but I think it is justified for the sake of better understanding of the podcast. So we have four languages. As many as four languages. And now the thing is that whatever model it is, whether it is a human, whether it is performed by a human, by artificial intelligence, by LLM, by anyone, it has to deal with these issues, that is, assigning a given, a given term, a given phrase, a given, I would say this to, because sometimes it may be a few words, but it constitutes one character to a real entity. And this is the first problem with this law almost in general. This is again hyperbole, because it would offend many people, and seriously scientists, but to put it very simply, such a method aimed at a consistent solution to this problem has hardly ever been attempted by anyone so far. Well, you have to know that such artificial intelligence, such LLM, how would it know which meaning of a word it should assign to a given entity? That is, whether it is a colloquial word, a commonly used dictionary word, a legal word, or a legal term. And now the thing is that, for example, one of the directives of the second level of legal interpretation says that these words from legal language, i.e. those that do not yet have legal definitions described, should be understood in the way that has been generally accepted in a given discipline or a given branch, for example for engineering. Well, but this still doesn't solve the problem at all, because how would we know, if we have opposing opinions, for example, from representatives of the doctrine, there are two professors and they describe the scope of the word differently, then where does llm come from, where does artificial intelligence come from, where does any system or judge come from, how would we know which of them is the leading meaning? This is a broader problem. This is the reason why, for example, any system is not allowed to use the dominant line of case law at all. Well, she's just dominant. It is the result of errors and imperfections in establishing the scope of standards. How would anyone , how would LLM know that this judge was probably right and that one was probably more wrong? How would any system, whether artificial or non-artificial intelligence, know this? And now the thing is that what we proposed in our concept of a deterministic process of legal interpretation and semantic representation of law is the creation of such an intermediate layer, which, as far as I know, no one in the world has worked on so far . Namely, such a metalanguage, it is technically called that, but it is like an artificial language, an artificial language that describes entities in a way that is as indisputable as possible. It 's like literally a separate language. No, I don't mean that it is a language of some kind in the sense of a nation, like, for example, Polish or German or French, but a certain universal code of meaning that will be the format of the communication system. In the format of a communication system. It may turn out that it will not be a language at all in the sense of a classical language, but will be a literary language, but it may turn out that it will be a set of code, a cipher. It will be a cipher. But he has to describe to us all the words that appear in the system. We have about 130,000 legal acts in the Polish system. Let's assume we have 100 acts, 100 regulations per legal act. That's how much I average. I don't know, I've never calculated what the average is. Let's say I had 13 million recipes, but when I break these 13 million recipes down into words, they give us a certain number of words. Some will be repeatable and others will not. But to each, to each repeating group, to each repeating set, there must be assigned one meaning, one specific meaning, one specific range of meanings. If this metalanguage does not arise, and we even know where on this whole axis it must arise, because there is basically only one place where it can arise, then we will never be able to say that law serves human dignity. It will always unconsciously attack human dignity through the system of chaos . You may or may not be aware of this, but to calculate the average overtime pay supplement, there are 18 layers of Polish, EU and international regulations based on Polish international conventions and many more layers. We have both a specialist medicine and a lex generalis, and only when we filter these 18 layers, put them one on top of the other, only then are we able to establish a single scope of the norm. We very often almost say that these are boxes. These are boxes, because please note that law and regulation, contrary to what has often been written in Polish, not only Polish, concepts of legal interpretation and jurisprudence, are not a shape. This is not a shape. Some people talk about the shape of the regulation, that is the scope, but this is something completely different. These are two completely separate entities. Shape and scope. And the thing is that for the law to be coherent and fair to us, it cannot be contradictory. It cannot be internally contradictory. A provision cannot be internally contradictory, nor can it be externally contradictory. That is, even when there are two indisputable boxes, they cannot compete with each other or they must have a clearly defined relationship. If they compete with each other, they must be complete in scope. There are some regulations that state that this does not happen. There are, yes, there are such regulations that do not mean anything, are empty inside, or there are those that are inaccessible to human beings, do not fill any matter or or or even fill matter, but inaccessible to human perception. An example is the term "at the latest after", used for example in Article 12 of Regulation 561 or Article 13 of the international convention Aetr. At the latest after in the logical-mathematical sense means an infinitely short period of time and collapses into the so- called nomen. This thing in which we function is the so-called phenomenon. But if anything, such a collapse is reduced to an infinite nonomenumen. Such a regulation should never have been created because it is beyond human perception. De facto, at the latest means an infinitely short period of time and a closed catalogue, and in a logical sense, an open catalogue is available to human perception. These are two completely different ranges. And now I'm getting to the point that there are many sources of this chaos, as we can see, but we have to, when we talk about whether to use artificial intelligence, what particularly worries me is the belief that we will use artificial intelligence, there are a lot of advertisements, we will use artificial intelligence in law. Very often, new programs are created , new attractive solutions appear, which are presented in a very encouraging marketing way, especially to law firms. We encounter this too, right? So it's something that has clearly taken over the market in some way and is becoming a kind of fashion, an attraction, something that seems possible. You ridicule this term, it becomes a fashion, an attraction, but it is a complete dead end. I have already explained most of the reasons why this is obviously a dead end for me , obviously for you as well . If someone understood how, for example, LLM works, or how not even LLM, but, for example, the rag model works, then programmers would now know what I'm talking about, they would already know that those people who know how these algorithms work inside would already know that it is almost certainly not suitable for applying the law. Unfortunately, very few people are aware of this and fall victim to the illusion that artificial intelligence will prepare the correct appeal or that artificial intelligence will prepare the correct administrative decision or the correct judgment. Evidence of this illusion and the danger of this illusion are the lawsuits against law firms in the United States that fell victim to this illusion and started using artificial intelligence. And she is said to have been hallucinating. She wasn't hallucinating. What does it mean that artificial intelligence hallucinates? This professor will explain it 10 times better than me . But ladies and gentlemen, if there is too little information in the system, too much information, there is poisoned information, there is the right amount of information, but we devote too few kilowatts to processing this information, which means it will search for this information too narrowly, the information will be poorly cut. So we have these so-called changs, because all the information that we upload, if we uploaded the entire codex, we need the system to understand this text beyond these issues of ontological, deterministic nature, etc., etc., it still has to cut and sort this text, and then it has to assign vectors. Vectors are mathematical values, although I do n't want to go into too much detail here, but the point is to be able to put them together logically later. The way the text is cut, to put it simply, determines how the system will respond . So there is a whole mass, a cascade of elements that argue for the need to create a completely new, separate model of legal interpretation and functioning, as well as writing and describing law. And that is why we, Professor Marek Królewicz and I, proposed to create not only the concept of a deterministic process of legal interpretation and semantic representation of law, but also the category of law design engineering in general. Ladies and gentlemen , is this something extraordinary? Well, if you fly on holiday, you think that the plane is flying because some random people at the factory accidentally assembled some components that were lying around. Were they, ladies and gentlemen, each and every screw was designed. Later, the screw-to-element connection was designed and later it was validated, and later larger elements were connected and were designed and validated. And only later was it all tested for years, decades. Billions and billions of dollars were spent to get the new plane into operation. These are, it's at least there, I don't know, a billion, a billion dollars and only then is it released and flies. how is it that the law is precise and gives us a guarantee, a relative guarantee that we will reach the destination, of which every catastrophe is only a confirmation that it is a relative guarantee, but a probability. Probability gives us how it is that the system in which we function, the most fundamental one, because whether we fly a plane for pleasure or not, it is a bit like our good will. We want it, we don't want it, we can take the risk, we can go against them, but in the legal system we are all squeezed. This has nothing to do with our good will. We are simply slaves to the system. We are slaves to order. However, chaos should be properly maintained. Well, the fact is that we spend money on luxuries and we don't build, we don't engineer something fundamental, like what we live and breathe in. Well, we breathe law in the sense that, let's start with the fact that we are considered citurus, we are considered to be alive if we have bubbles, right? That is, if he took in air. Now this concept has changed a bit, but during my studies it was still like this , i.e. it was determined whether the alveoli in the lungs were developed, because if not, it meant that the baby was born dead. So what the air quality is like is also determined by legal regulations. Everything, every element, for this table to be created, for this chair to be created, for every element to be created, certain legal conditions had to be met. I just wanted to address this extremely important issue as an attorney. Ladies and gentlemen, I won't reveal the big secret, er, just jokingly saying that the software is being tested. Yes, that is the software life cycle , how software is created, and I'm not talking about some great software for the, uh, military or for space flights, but every software should be tested. We teach this to students on a daily basis, so to speak. I also happen to teach software engineering, so maybe I can say on a daily basis. For every computer science student it is so obvious that we probably don't even ask about it anymore, because it would be quite offensive for them to even ask about such obvious things, but imagine creating law. Is anyone there testing the law? Imagine a situation where, for example, you invent a law and impose it on millions of people. let them use it, maybe they will survive, maybe not. Well, if engineers worked, there would probably be no engineers anymore, and that's good. Either they themselves would die in their constructs, or their humans would treat them accordingly. And this is what happens in law . For me this is a shock. For me this is a shock. Yes. But Professor, so dynamic, I'll mention right away, please point out that engineering cannot afford disasters because they immediately get media coverage, right? a plane goes down, a construction company goes down, the shares of the construction company or airline immediately fall, and so on. The viaduct collapses, people die immediately from the sound. In law it is like this, in law exactly the same thing happens. So if a judgment is passed that violates human dignity, that is, is extremely flawed, the same thing happens de facto. There is also a bigger drama and catastrophe, sometimes even bigger, but it all happens in the privacy of balance sheets and sheets of paper. They are just so dumb. these pieces of paper with these words with the justification of the verdict and the content of the verdict itself, they are such a silent witness to this, this catastrophe. If we don't see it, or rather, we do n't notice it, but we often do, or maybe we've become a little accustomed to the technology of these disasters, so many of them are happening that we've become a little accustomed to them. This does not in any way release us from the obligation to create a legal engineering system . This is exactly what I wanted to allude to. These gentlemen have been researching the entropy of the chaos of the legal system for 15 years . Now it's time to put this research into practice. So what is legal engineering? So maybe I'll tell you where the issue of engineering, essentially engineering, comes from. What is software engineering for example or knowledge engineering, but software engineering. Well, ladies and gentlemen , there are certain software development methodologies , more or less precisely defined, but we, ladies and gentlemen , as IT specialists, know how to create software. We have, we have rules. I learned from the professor here, as a result of many hours of conversations, that there is actually no such thing as coherent, uniform, established principles of law-making. There are certain guidelines, of course RCL, for example, applies them, but I will not comment on them here. However, generally the process does not end with testing, but the process does not even start properly. So, for example, in the case of computer science engineering, we start with a feasibility study, that is, by determining whether what we want, this idea is feasible at all, whether it makes sense to start working on this. I do n't know if there is an equivalent in law. The professor will probably explain how it is in a moment. Next comes the requirements gathering stage . Well, that's really what we need to do. And what effects do we want this law to have? Yes, this is a very important law in this case. I'm talking about software. Then we create some models, model, design, model and design of the system. Then we implement it, then we test it and finally we deploy it. I have a question. It's like this, it's a process. I'm talking about a very classical process, because there we have quite more sophisticated processes, really iterative and so on. I don't want to get into this, do I? But I'm talking about the simplest waterfall, one that's no longer used, to be honest , because it's the simplest one, but it conveys some meaning. And my question, which I will ask in a moment, I would like to ask, what is it like in terms of this waterfall, in terms of these several stages, what is it like in the law, which elements occur, which ones. And now, to finish this topic on my part, ladies and gentlemen, please imagine that the example that the professor gave, that for example 70% of planes fall down. What would happen? And in the case of law, well, that's exactly what happens: 70% of trials end in something we could call a disaster. But this is the professor who referred to another extremely important aspect, namely iteration. I think that a large proportion of lawyers have trouble understanding the scope of this word. I consciously say lawyers, because if we in the community do not have a common belief as to the scope of defining what iteration is, it means that it is most likely not being implemented. You ca n't build any sustainable system in my experience without iteration. This is the basis. And now, of course, there are rules for creating law. yyy RCL is implementing them. We have regulations there, we even have laws, we have guidelines, but they do not refer to iteration, they refer to a certain shape, a construct and a certain minimum content and certain minimum semantic values, which in my opinion, Professor, constitute perhaps 3% of the need, the demand, to create a deterministic system. So if we don't have iteration at all, well, you could say and or, in other words, and we don't have validation, then we can't have convalidation. Well, for there to be validation, to fix anything, there would first have to be validation and correction. If these two factors are absent, the third factor is absent. Is there no system in our country ? I sent it to the professor tonight at an indecent hour, so I won't say what time, but just ridicule this publication by one of the patrons, well, a tabena of Warsaw. Greetings, Mr. Patron. Besides, I immediately wrote to the patron to congratulate him. Submit this publication about logical values, especially functions. Application of functions. The point is that the attorney, from the logical perspective, took a dozen or so, or less than ten, regulations and began to analyze them at the level of errors in matters as elementary as function and or truth, to put it simply, this study is much richer and the attorney-at-law demonstrates perfectly that for years, some regulations for over 10 years and others for over a dozen years, have had obvious errors that no one is doing anything about. Meaning it does not validate. Why doesn't it stick? Because it doesn't validate differently, or even validates, but there is no iteration that is, I would say, critical, right? I have several registered industrial designs and patents. First I had to design what I wanted to make and patent later, so that it would n't fall apart after the patent was granted. And sometimes we have the impression that we are patenting the system, or giving it all this legalism, i.e. the president signs the votes, but we have not subjected it to the previous stages, i.e. iteration, validation, possibly convalidation and only then putting it into circulation. It seems to me that this is the only inevitable path. This means it is a path to the abyss. If we are publishing tens of thousands of pages a year, pages a year, well, we are, it means that this system is basically dead. He still seems to be there, he still has convulsions, he still seems to be functioning. Well, this is what we are talking about, because we will notice that society is constantly raising this issue, it is constantly being said in the media. This topic is still current, that this amount of bureaucracy is excessive compared to the increasing number of regulations. But does anyone of us know the regulations that apply? Do any of us know the 1%? We are not even aware of the accumulation of the intensity of these regulations, which are introduced in a non-systemic manner. Non-systemic, not in the sense that it is not statutory, but non-systemic. While we're on the subject of systemicity, and we mentioned it here, Professor, you and I once talked about the dominant line of case law, because you were interested in why the dominant line of case law, which is cited by 90%, probably 9% of lawyers in appeals, in general in procedural documents, and so on. The main point of reference, right? The main point of reference apart from the legal basis. Of course, it was the professor who asked why I did not see the dominant line of jurisprudence of common courts. as an ontological element of law, i.e. something on the basis of which the system can make this assessment. Well, if the judge does not have a set of quantifiers with which he can transform the scope of the norm in a reliable manner, within the provisions, into a certain scope of the norm, i.e. such a deterministic result , it means that this result is random. This means that the law is not actually created at the beginning, when there was an intention and it was written down, but it is created at the end, when the judge, in his own way, excuse me for saying this in his own way, does not want to offend any judge, but according to his own criteria. In this sense, in his own way, according to his own criteria, but a bit haphazardly, he randomly defines the scope. A judge who is not a judge cannot avoid falling into this trap. can't in the current system. He has to fall into a trap there because currently in the current legal system there are no good results or bad results. Why? because if this ontological scope is undefined, if the semantic scope is undefined, if the scope of legal interpretation is not defined and if we do not have defined, assigned real entities to the scopes of individual words, it means that there are too many variables based on which the judge can issue a very different scope of the norm. If I may, as an attorney, interject here for a moment, because I would also like to tell our listeners and viewers how this would work in terms of engineering. Imagine we have a computer program and we have a judge on the other side in a different discipline. Yes, we have a program and this is the IT part, and the judge is on the side, the court is on the legal side, and let's clash these two worlds. And we have a computer. A computer is an interpreter. Yes, of course I'm talking about a computer in terms of simplification, because it's obviously a compiler and a whole complex system that decodes these semantics into something that can be operationalized. Okay, so we have such a program and we have different computers and let's imagine that on each computer, every computer, I keep telling myself, the computer is simplifying, the IT system interprets it in its own way, so yes, you write one program or an engineer writes one program and we have 8,000 aircraft and each one behaves according to its own criteria. Well, that's what happens in law. Every judge in good faith must, because there are no other possibilities, interpret this program, i.e. the law, in his own way. If they win in their own way, I will only say that they do it in their own way in the sense that they apply the principles of legal interpretation, but because there is no paradigm, because we are constantly talking about a paradigm, i.e. a strictly defined path of the order of dealing with these variables. What is a paradigm? When a pilot comes in to land, he has a checklist and he has a paradigm of the actions he has to perform, right? randomly perform in a strictly defined order. Then there is the paradigm. However, we will repeat it once again in Polish, only adding here what the professor said so that everyone understands correctly. And what may be shocking is that in the Polish legal system we do not have a paradigm at all. We have directives, we do not even have a complete list of directives based on which we process the provision into the scope of the standard, but the order of their application or discontinuation of their application has never been determined. Well, when writing a program, a computer scientist is absolutely convinced that the machine, and of course the IT system as a whole, will interpret his intentions in the only correct way described in the rules, in the principles of interpretation. grammatical distribution and then, of course, semantics. And ladies and gentlemen, it won't come as a surprise when I say that this is happening. And thanks to this, we fly, drive and generally function in the world of technology relatively safely. However, in the world of law this does not happen, because the legislator, whoever he is, yes, because he is a somewhat elusive being, has certain sincere intentions, let's assume good ones, and at the end of this path there is a judge, a court, an office that interprets. And now please tell me, this is a rhetorical question for you, Professor, because we have already talked so many times. What chance does a judge have at the end to interpret the legislator's intentions? Does he even have a chance? Yes, you do. He would have a chance by using functional interpretation. Functional, of course, in the contemporary sense, not in the sense of when it was created, because the scope is slightly different . This means that when we participate, and we have been participating for years, in the legislative process, many of the regulations in force in Poland are my own, in the sense that, of course, the parliament passed them and the president signed them, but I prepared them, or our team prepared them, so you could say that we are their authors. It is important to know that we justify this direction from a given origin of a given scope, a given legal norm. we justify. So if judges and courts referred to these justifications and started from this interpretation , and not, for example, from a linguistic interpretation, then perhaps the chaos coefficient would be correspondingly smaller. there would still be a chance, but the problem is different, that in regulatory impact assessments, when we write justifications, we have no obligation to provide context. This is what we once talked about, Professor, that if we do not provide contexts at this stage and at the same time we are not obliged to record what values at the level of legal interpretation, we transform this idea of ​​​​a norm before it is even adopted into the content of the regulations. Then we have nothing to decode in the finale. The problem is, I would say, systemically structural. I would say he is an engineer. That is why we talked here in general that law engineering departments should be established at law faculties, because the problem is purely engineering. Law, contrary to appearances, is a classic science subject that requires both humanistic and scientific competences. These are necessary humanities skills, but this is a classic science subject. Someone once asked me when it became a science subject? when the first sentence was created that we considered a law and was written down and we determined whether it was a law or a falsehood law or a falsehood when we gave it a logical value at that moment law became an exact subject, which very many lawyers do not want to hear, the truth is that it is a classic exact subject and it must be treated like any exact science , and not randomly like an exact science. That's why I once wrote a study, a scientific publication, in which I wrote that law, more or less, has such a title as an incompletely mature scientific discipline. Well, it is not a fully mature scientific discipline, since all quantifiers are of a strict nature. I omit axiology. Well, because we do not delve into axiological values . They are very often culturally conditioned. And it is also very important to say that what we create, we do not interfere, we do not create law in this sense, because the professor very rightly said that there is no law. I agree with the professor almost 100% . I say that almost only because, with this subtle nuance, we do not claim that there is no material law, because it has been formally created, developed, it does not exist in the sense that it does not have the ability to obtain a deterministic result, i.e. a result known in advance, by means of a deterministic process, then it does not exist. Well, if there is no determinism, and I read the regulation today and I cannot say how anyone will interpret it in 5 years and that it will be interpreted in the same way as I do today, then there is no determinism and in this sense, the professor is absolutely right. Actually, I didn't say that there is no law, but rather I quoted the professor. Just to clarify, I also give the palm to the first one. Y, however, I fully agree with this statement after the professor's explanations, because for me it was also quite a clear statement at a certain point. Well, but after listening, well, I regret to say that I regret it, because it concerns us, that in fact there is no such law, that is, there is no order. There is no order. But still, but here I must add one more thing. Within our concept of the deterministic process of law and the sematic representation of this law, we have no intention of interfering with substantive law. I mean, that's not our role as scientists. It is not our role to write intentions, to describe intentions in words. So we do not go into the second part, the axiology. This is not our role. That's what philosophers are for, philosophies of law. That's what society is for. It is not up to us, it is not up to us scientists to say what is good or to decide what is good and what is bad. But we want to give a tool, but we give, we create what is turbo unique in this project that I am preparing and is universal on a global scale. This is the middle component, that we provide a tool, an instrument, we provide a code that allows one to process one thing deterministically, that is, in a certain way and in advance, the content, whatever it may be. We do not penetrate into it, into a specific result, whatever it may be, because we do not penetrate into it either. Exactly. Here, I would like to say very clearly that software for law firms, for lawyers, is currently mainly focused on very fast, efficient and indeed it happens extremely quickly searching for, for example, regulations based on, for example, keywords or other texts in general, and this is truly an incredible help. I would love to use such a tool myself and it is absolutely useful. But you shouldn't believe that such a program will write a lawsuit or respond. I used it myself, I won't say which model, because I don't want it to be said that there is some attack on it. four letters, my colleague, the professor knows about it, he wrote a certain chat, he wrote me four procedural letters, extremely simple, because they had four points on one side . Well, then I sat there all day, because I'm not a lawyer, and I turned it into something. He just made up so many things that I just wrote it from scratch, and since I'm not a lawyer, it took me all day, so I wasted my time analyzing it. I could write it myself right away. It just doesn't work that way . However, I wanted to outline very clearly what we are creating. Well, let's imagine a set of information. I say very cautiously, written in a formal, precise language, that is, with unambiguous grammar, but also semantics, interpreted by, as they say nicely, calculating machines, or computers, and which describes the world. This is what we call an ontology, a knowledge graph. sometimes, if it concerns more specific elements, but ontology. All right. Now let's imagine a situation where we have a case at some stage, maybe the initial stage, of some specific, some specific entity. It is also described in detail. Let us also imagine that an event occurs. an event in which this entity, for example, creates a procedural document and it becomes part of the legal system and the question arises, what next? So the question is about the consequences, i.e. what will be the legal status after this event. This is an extremely important question. E, because first of all, most of the clients of the law firm, as I learned from the professor, most of the clients ask about this, that is, they have a certain state and they ask what are the consequences of this or if they do something, what are the consequences, what could be the consequences and what will be the consequences if something happens. And the second aspect is that it is possible to automatically process and generate a longer, admittedly, but finite number of scenarios, and thus to be able to evaluate them according to certain criteria, of course the optimal solution. Imagine how powerful it is for the average person if they could see a list of scenarios for different events. This is something that a law firm cannot do in one day or probably a week in a complex case. It will take a lot of jobs, I mean a lot of man hours it will require and I bet that is all the time, yes, money. I bet, with all due respect to lawyers, that they can't foresee all the possibilities anyway due to the power, let's say computationally , in the case of computers. And this is the essence of our system. As the professor said, we do not want to create any new law. We want to provide a tool to make the law deterministic and to be able to predict these consequences and to make it, as the saying goes, " Czech Republic-friendly," meaning that it's not just the domain of lawyers, but that such a process can also be carried out independently. This is, one might say, the essence of our approach. The system is of course very ambitious, extremely ambitious, one could say it is not so much fundamental as even a civilizational civilizational change. I think you have already noticed this. Uh, we work with some truly amazing minds, and here I bow down to the knowledge of the people I work with. I will not list them here so as not to create such artificial advertising. However, we are also very keen to use, as the professor mentioned, very interesting studies that we can find on the Internet, and here I also bow to the in- depth analysis, for example, of the logical material that the professor mentioned, but we also very warmly invite lawyers, computer scientists, logicians and others to cooperate, because the project is so groundbreaking and so important that I think that participating in it, at least for me, is a great pleasure and honor. So I would like to thank you very much for this extremely valuable conversation today. We touched on fundamental topics ranging from human dignity through the crisis of law to the future of artificial intelligence and knowledge models. I would like to thank our guests for their experience, knowledge and vision of the future of law and artificial intelligence, as well as technology. Thank you very much. Thank you very much for your attention. Speak soon. M.