Cognitive Science and generalized Evolution

Author: Zhao nanyuan 19:32:42

About 40 years ago, at the dawn of computer science, computer experts made three promises: 10 years later Computers can win the world championship of chess. Most psychology theories will take the form of computer programs. This is actually a problem of artificial intelligence in computer science. The Research of artificial intelligence involves cognitive science, and the question of cognition is: how does a human brain (or mind) work? However, traditional artificial intelligence is obviously not able to achieve satisfactory results. This is not a mistake in artificial intelligence technology, but a mistake at the philosophical level. The deep spiritual structure of the Eastern and Western cultures is different. Therefore, it is difficult to make a breakthrough in this Ontology-based approach. Early artificial intelligence started from playing games and proof theorem, hoping to find a general solution to all problems, that is, developing a program for "general problem-solving. However, the complexity of the game varies greatly. The simplest game is a winning game. As long as both parties have mastered this rule, from the very beginning, the game is destined to win first or later. Chess and checkers cannot find a winning method. On the one hand, they encounter difficulties in exponential explosion during computing, and on the other hand, they make people feel bad at thinking habits. When playing chess, people obviously do not consider hundreds of millions of possible chess games. They can only use knowledge to choose from less promising ways. Due to this understanding, in the late 1960s S, artificial intelligence began to research "knowledge", hoping to find a common way to present knowledge in computers and use knowledge to solve problems, an expert system is developed to put expert knowledge in the knowledge base of computers and solve some problems in specialized fields. This research direction also encountered some common difficulties. The first is that the rich knowledge of experts is not completely clear, and the other difficulty is that the expressions of these knowledge are not objective-oriented, that is to say, machines cannot learn new knowledge on their own, but they need a huge amount of knowledge to make practical expert systems. Feng. shortly after nuoman proposed the basic structure of computers, he studied the differences between computers with successive processing and the human brain, and proposed the idea of neural network computers, neural Networks also have difficulties in exponential growth of learning time with the complexity of the problem. Computers can deal with problems with large logic depth, while humans can only deal with problems with small logic depth. Pattern recognition is an important field of artificial intelligence. Its goal is to convert pattern information into symbolic information that is easy for computers to process. The pattern recognition process is to identify the category of a pattern. The premise of pattern recognition is classification. Before recognition, we need to have a category system. If we know what principles the human brain uses to select features to establish a category system, we can find a way to automatically decide how to extract features, instead of relying on case-based techniques to design a pattern recognition system. The inspiration here is that recognition is a process of re-cognition, including how people establish a category system in the brain. There is a reason against evolution: Exponential Explosion. Biological genes contain a large amount of information. For example, the amount of human genetic information can be used in a large-scale library, the time needed to generate such a large ordered information by means of random methods obviously exceeds the lifetime of the universe. However, after all, this shows that the evolutionary mechanism includes algorithms that evade exponential explosion. The West is the creation of God, saying that God is the end of all inquiry, and the West is a theory of God. The East is a pan-God, and God has its own origins. It is a kind of human god. Heaven and the Emperor are not able to copy the human society. The East allows coexistence of views and systems. Modern science originated from the monotheistic theory. The paradigm adopted is the Euclidean paradigm, which has the starting point of metaphysics. The Western knowledge system starts from some premise assumptions without proof and introduces a logical and rigorous knowledge system. The "big fish model" is not only a cosmic model, but also can explain the construction of Western knowledge systems. The main characteristic of this structure based on the "big fish model" is its clear direction, irreversible, and absolute upper-lower relationship. It can only be a large bowl of sea water, not a large bowl of sea water, it can only be a big fish floating on the sea, not a sea water floating on the big fish, upside down, the entire system will crash. China's Knowledge System is an mini-planet system. Many theories are like the floating mini-planet in the universe. There is no obvious support or consistency between them, even if there is a logical conflict between theories, it will not lead to the collapse of this system, because this "minor model" does not mean that there are no conflicting theoretical systems. Metaphysics rejects the analysis, demonstration, and research of any science. The cognitive theory we need is the theory of cognition, both as a tool of cognitive science and as a manifestation of the philosophical hierarchy of cognitive science. Compared with the monotheistic "big fish model" and the multi-God "Minor Planet model", modern science that rejects metaphysics is equivalent to the "Earth model". What is different from the "big fish model" is that, the "earth model" does not need or has no metaphysical support. It only has local and relative upper and lower concepts. It does not have all and absolute upper and lower concepts. It is the place of the Earth, and the following directions are different. There is no absolute starting point or up-and-down relationship between the parts of the scientific earth. theories support each other and are consistent with each other. The method of establishing a general statement based on the collection of special examples and observations is generally called induction and is seen as a scientific indicator. In other words, the use of inductive methods is seen as the demarcation criteria between science and non-science. James raised a question about this method. He pointed out that no matter how large a single observation statement is, they could not logically contain a general statement of no limitation. The question of generalization is also called the question of "the question of the" the ugliness of the House "of philosophy. From the time of the" the question of the question "to the present, this question has been confusing philosophers. Induction is an independent logic principle that cannot be pushed out of experience or other logic principles. Without this principle, there will be no science. Popur provides an acceptable solution to the problem of generalization. He first pointed out the logical asymmetry between the proof and the proof. Seeing that one hundred of the swans are white, the general statement is "all the swans are white", but it is observed that there is a black swan, logically, we can conclude the statement that "Not all the swans are white. Although experience cannot be confirmed, it can be falsified. This means that although scientific laws cannot prove, they can be tested by a series of attempts to refute them. A major contribution of popur is that he clearly pointed out that "truth" is not the only criterion for judging the quality of scientific theories. As a scientific theory, it must be verifiable. If we take the pursuit of truth as the goal of science and the pursuit of truth as the goal of understanding, we will inevitably fall into the theoretical difficulties caused by a large number of absolute problems. As long as we can abandon the concept of truth like rejecting other metaphysical ideas, and change the objective of science and understanding to the pursuit of better theoretical principles and knowledge, then there will be no doubt at all. An Open System Based on "good governance" will never close the door to better theories in the future. As the starting point of the theoretical system of good governance, I propose three basic principles: Existence principle; Economic Principles; Prediction principles. Principle of existence: the existence mentioned here is the existence in the sense of epistemology, not the existence in the sense of ontology, nor the existence in the absolute spirit of idealism. In the sense of epistemology, there is "I think so I am", and "I am" depends on "I think" as the subject of cognition ". The basic principle of existence is stability. It can also be said that "existence is stability". The principle of existence can also be referred to as "stability ". Economic Principles: economic principles can be inferred from the principles of existence and biology. For a creature, survival is the existence of a creature. It is always necessary for a creature to obtain energy from the outside world or substances containing energy through various life activities to overcome the tendency of the Second Law of thermodynamic to maintain a low entropy state, this produces economic principles. If a creature is "non-economic", that is, the total expenditure of energy is greater than the total income, and it is out of loss, it cannot be maintained as a creature. In the process of evolution, creatures with poor economic performance will be eliminated. The economic principle can be expressed as follows: the cognitive processes of all creatures, people who are creatures, people who use nature, and all existence with cognitive ability have a tendency to pursue economy, economy is an important indicator of good performance, which can be converted into efficiency. Prediction Principle: The Prediction Principle comes from the economic principle and the importance of people. An important difference between people and animals is that people can carry out purposeful activities. Increasing survival probability and increasing economy are fundamental goals of human cognition and animal action models. For humans, the more direct objective of cognitive and scientific activities is to Improve foresight capabilities. This is the Prediction Principle and an important principle for evaluating cognitive activities and scientific theories. Popur gave an example of rain. There is a theory that it will rain in the future. Even if the world did not rain in 100, this theory is still correct, because it may rain in 101. However, the weather forecast says it will rain in Haidian District tomorrow, which may be inaccurate but useful. Therefore, the good and the bad of a theory have nothing to do with the true and false of a theory. Instead of looking for truth, they are looking for useful things. The theory is an invention, and the invention is good or bad. It is not directly connected with objective practice and actual things. To give a metaphor: if the natural world is a cow, then the scientific theory is cowhide? Is it a ox bone? None of them. They are the iron ring of the morning bull's nose. They are made on the ox's nose, rather than the ox itself. In this view, there is no longer a problem of authenticity. Ox is a thing of freedom, but it can only be understood if it is useful to me and what I need to understand. Here we talk about the concept of the Cartesian demon. Descartes once wondered whether the things around me really existed or were just an illusion. There is a philosophical significance here: if a person is just born, I will put him in Virtual Reality (Uirtual Reality), just like the movie Matrix, then this person will only think what he feels is true, and from the outside, everything is happening in a box. In fact, the world we recognize is an internal model of the world created in the brain. From a certain section, what we get and output are only some signals, in this case, whether the world actually exists is unknown in principle. Therefore, although the world in the eyes of frogs is totally different from the world in our eyes, we cannot say that frogs are a fallacy, We are truth, maybe in the eyes of more advanced creatures, we are similar to frogs. Another example is that the human eye has three primary colors. If a person has evolved a four-primary-color eye, we are all color blind. From such a problem in epistemology, we will talk about the inverse mirror model. We can see phenomena and find out the theories behind them. So we say this finds the so-called truth. However, there is a conflict in this model. How can we know that the theory behind this phenomenon is the same? Whether this theory exists is doubtful, we have an infinite variety of interpretations, so there is a significant difficulty here. We can refer to the inverse mirror model. There are many explanations in this book, that is, the internal model built in my heart is the world in my heart, but I can only imagine that the world outside is like this. Although it may be wrong, as long as this model can survive, it is a good model. The frog's model is wrong. It cannot see static things, but it can survive by this model, which is enough. The basic meaning of the inverse mirror model is that it is a reflection of the anti-materialism. The reason for saying this is because it has something to do with culture that foreigners cannot think. There are two big models: one is the founding model of God. After God creates everything in heaven and earth, and then creates another person, then people will begin to understand the world. However, this is not the case in China. After chaos is split, the air rises to the sky, and the turbidity falls to the Earth. This idea of China is actually an intellectual problem. It is defined only after being light, heavy, up, and down. The neural trigger in the brain of a fetus is very irregular. It is said that when I was a child, the color was a single block, and the shape could not be understood, and there was no concept of distance. What I didn't understand would become smaller. Therefore, I believe that pangu is fundamentally an intellectual problem. Next, let's talk about the third question: evolution. I like the game theory, but it is not mentioned by me. There are many questions in textbooks, so I will not elaborate on them here. What I created is a soft and hard structure model, which is easy to draw. Later I saw another book titled complexity. What is complexity? One definition is that complexity is at the edge of chaos and order. But what do I do at the edge? Why is it at the edge? I didn't say it in the book, and my soft and hard structure model just solved this problem. The hard structure supports the soft structure, and the soft structure changes the hard structure so that the structure can develop itself. Here is an example. For example, if a tree has a growth point on its buds, and the growth points are split constantly, the tree can grow hard and become a wooden trunk, and the edge of the trunk is active. This is a process of construction. Let's look at another example of Sociology. For example, laws and complete laws all have a provision for changing this law. This is a process of building itself. The law is the foundation of a society and a hard thing. However, the society allows freedom of speech and publishing. thinkers think that there is something wrong with the law and can propose changes to the legislature, hard laws will change. This is also an example of changing the hard structure of a soft structure, but the hard structure also supports the soft structure. Only when freedom of speech is allowed can a thinker spread his or her remarks to influence and lead to changes. Anything that has been invented can be attributed to such a soft and hard structure model. In ancient times, there was a saying that "the benevolent man Leshan, the wise man happy water". "The benevolent man" is the bureaucratic and hard part of the society. It is necessary to work according to regulations to ensure high efficiency and "soft ", the Wise Men will launch inventions and creations to guide the development of society. The fourth part focuses on cognition and thinking. How can we overcome the difficulties encountered in cognitive science and artificial intelligence? As early as 1960s, Japanese scholar du Bian Hui proved a famous theorem called The Ugly Duckling's theorem. The main content of this theorem is: the difference between the ugly duckling and the Swan is as big as the difference between the two swans, And the similarity between all things in the world is the same. From this theorem, we can draw a conclusion that there is no "Pure objective" classification standard, and all the criteria on which people classify are subjective, what criteria is selected for classification is purely subjective evaluation and is a matter of value. To explain this seemingly common sense theorem, let's take an example: According to the biological classification principle, Whales belong to the same category as mammals and even shoot cows, but are different from fish; according to industry classification, whale hunting belongs to the water industry rather than animal husbandry. Whale and fish are in the same category but different from cattle. Pattern recognition is actually a classification problem, and classification requires conscious participation. Why do we need to be conscious and conscious? I can use the soft and hard structure model to explain it. The vast majority of information and signal processing in the human brain is unconscious, which Fred discovered. Before him, people thought that I knew everything I thought, but I did not know anything, but I knew something. The part I knew was the conscious part. Consciousness is clearly related to memory. The Sleepwalking person opens his eyes and sees everything, but does not remember it. Sleepwalking is an unconscious process. Consciousness is related to memory and attention. Where is conscious? Consciousness plays a role in learning. The exponential explosion of knowledge is mentioned above. You cannot learn the same complex things at once, but you can learn things from an early age step by step. For example, if you want to learn a bicycle, you must first learn how to ride it before learning to turn. During this period, you need to have a skilled process. If someone is not good at cycling, this will happen. When they see a waste bin, they thought about not hitting it. Don't hit it, but they hit it directly. This is because when you learn to ride a bike, you first learn to ride in one direction. When you are nervous, you will tune out the basic subprograms you learned before, and you will hit it. Therefore, the role of consciousness is to conduct command learning, to learn this small thing in front of the eyes, and then to accumulate small places into large ones. The so-called consciousness is unconsciously controlled, and consciousness will slowly add something to the basic part, namely, the subconscious. in learning, consciousness plays the role of command accumulation. What is the use of feelings? Only people will laugh. What are the advantages of laughter? Laughter is a learning process. In the words of a computer, it is called interrupt )". When we hear something that we didn't know before, we will laugh. In fact, the role of laughter is to deepen your impression and make you remember things that you don't know, I won't laugh again next time. It is actually a novel detection. A person's feelings are a process of evaluation. The evaluation of whether a thing is useful to himself or not is also evaluated. The evaluation of right or wrong is verified by post-event experience. The most basic evaluation is survival, but other evaluation criteria will also be derived, including prediction. when watching a movie, the results will be known and meaningless, and the results cannot be predicted at all, no. Another problem between the two questions is that he found that he could not derive "what should be" from "what is ". To unify the natural sciences and social sciences, we need to make a small transformation. We believe that natural science is not a question of "what is", or that natural science is not pursuing truth, in this way, the natural sciences and social sciences can be unified. Free will, or the freedom of will, is an important concept closely related to consciousness. For the background of Western culture, the freedom of will is an extremely important basic starting point and a necessary premise hypothesis for establishing multiple theories. Peng Rose said it was wrong to use quantum mechanics to explain the question of free will. Free will is not a random problem, it is not a random process, it cannot be solved by throwing a dice. Minsky has long realized that the mechanical theory about the brain or spiritual world explains the difficult conflict between the free will of rain. In Minsky's words, "the goddess of opportunity is as heartless as the goddess of destiny." If the determination determines what to do, there is no free will. If there is no free will, the "big bowl" should be turned over again, because the basic structure of all morality and ethical society is built on the basis of free will. If you do something wrong, you are willing to be punished, because it is your decision to get rid of it. If it is decided by God, you should not be punished. However, if people are regarded as learning machines, and machine learning can also be done through rewards and punishments. In this way, we can avoid the issue of free will. The morality of society still exists and will not collapse due to the cancellation of free will. Free will is equivalent to the burning element in the theory of fuel, which is incompatible with science. "Free Will" is a religious concept derived from God. In the Bible, people eat the fruits of wisdom, with free will, you must accept rewards and punishments. I got an inspiration from Wang Shuo's book. He accused Peking University Students of saying that they were full of concepts and they couldn't see what they were doing. In learning, the concept must be clear, but why does it become a bad thing? There are many bad concepts in social science. They are masked and hidden unsolved things. Concepts come from philosophy. philosophy comes from theology. Theology was initially designed to mask people like truth and essence, this is something that cannot be proved. Like an apple, Newton discovered that it is the same as the moon, because the essence is quality, and the genetic nature is different, but the importance is different. How to differentiate bad concepts? Concepts in natural science, such as force and quality, are hard to be defined, but there is a definition of operationalism. What force scales out with spring scales is force, quality-Tianping scales out quality. That is to say, concepts are related to something that can be practiced. Without such a relationship, it is a bad concept. This method can be used to clean up the concepts in social science. I think Confucius is the most remarkable. His greatness lies not in what he says, but in what he does not say, and the child does not say "strange force chaos ". He doesn't talk about land farming because he doesn't understand it. There is so little metaphysics in China that it is a very advantageous tradition. Why should I study Ethics? Because of cognitive problems, we need to study what to do and what to do, which is closely related to ethics. Scientific Development has caused ethical issues, while "clone" has caused ethical problems. No! I think this should be said: The development of science has exposed a major problem in ethics. Ethics is neither knowledge nor persuasion. It cannot be justified in science. I think ethics must be rebuilt. What ethics do we need? First of all, it must be effective to answer ethical questions, such as "ABORTION? Ethics should not be disturbed by any religion; otherwise, it is a blind obedience, a belief, or a metaphysics. Second, there must be universality. different cultures have different morals, but there is only one learning to explain them. Ethics has moral risks. If there is an error in ethics, it will cause a huge harm to society. Therefore, I propose the principle of minimizing morality to analyze the economic risk and benefits of each morality one by one. The existence of morality is a kind of constraint on people. It is a kind of anti-freedom, opposite to the natural trend and instinct of people. Morality is useful only when the prisoner's dilemma exists. Morality can be generated spontaneously. Let me talk about the problem in general. If you are interested in the problem, let's discuss it in depth. Sheng Hong: I am very grateful to Professor Zhao. He is talking about a very difficult question. It is an ancient philosophical question, thinking, and reflection. Modern Times are called cognitive science. Professor Zhao just talked about the issues of human understanding of the world, the differences between East and West ways of thinking, and the issues of ethics and science, which all provide us with good questions for discussion. Let us first ask Mr. Li Ming to make a comment. Li Ming: I am very interested in this question today, because my recent research has also started from this perspective. Recently, I also read a book about generalized evolution written by Lazlo, a world-renowned system theory expert. Professor Zhao talked about many things that I saw and thought of in common. Professor Zhao and I both learn science and technology and have many similarities in thinking. Professor Zhao has unique characteristics in terms of human reason and cognition from the perspective of Expert System and AI, I will not repeat Professor Zhao's questions. In my opinion, Professor Zhao's understanding is not enough from the perspective of natural science. As he began to mention the exponential explosion in AI, "Yes" and "yes" have all talked about the problems of understanding and intelligence, it is very difficult to give people a clear and open definition. The arguments from the machine perspective are unclear. This requires a theoretical breakthrough. Lazlo suggested that Darwin's evolution is very difficult to establish in modern times, because the natural choice as the theoretical basis is a pure random process without intelligence and Will, it is an unordered, unexpected collision and choice, which cannot be evolved. Here is an example of an astronomy called boyier, which allows a blind person to twist the Cube once a second when there is no learning process, it may take 127 billion years to twist a plane of the same color. Even from the Big Bang to the present, there is only 15 billion years of human life. However, if you follow certain rules, even a child can draw a face of the same color in 120 seconds. In this sense, it is unreasonable to use a random process to describe evolution. It is possible only when there is a will and a learning process. Darwin's evolution cannot be explained here. The evolution of the whole universe has a common law from beginning to end. Now the general evolution, including Laszlo, is trying to find this rule, but Westerners cannot find it and it is impossible to find it. I think the major breakthrough in the 21st century may only be in the east, not in the West. We should look for it from the ancient oriental culture, like Lao Tzu's words: "daosheng 1, life 2, life 2, LIFE 3, and life 3", "people send to the ground, send to the ground, and send to heaven, tao fa naturally ", all contain simple philosophical principles. The development of AI requires breakthroughs in theory. Mr Zhao's research is very good, but it is only the beginning. To make achievements, we need to study it. I recently studied the history of philosophy and found that Chinese people do not understand logic at all, but there are still some in ancient China, such as those in the Early Qin Dynasty. However, after the Han Dynasty, they had exclusive Confucianism and were all wiped out. So now we must advocate a hundred schools of contention and allow freedom of speech and Thoughts in the system, so that China can have hope. After the Renaissance, the development of Western science and society was promoted by a series of great thinkers. Ding Li: After listening to Professor Zhao's content, I feel very happy because I also believe in the unity of the world, the unity of academics and science, everyone has a similar idea. This is what we call the same things. People of different disciplines and specialties need to communicate like today. However, I am excited and want to ask Professor Zhao a few questions. First, let's talk about the unification of natural science and social science. Professor Zhao believes that the two can be unified by considering natural science as not studying "yes", but also studying "what should be, it is somewhat far-fetched. If we think of science as a kind of activity of mankind, we are actually Symbolic Animals. People use symbols to build models to describe the world. In the era of quantum mechanics, a physicist, wizak, said: The world is in front of people and people are in front of physics. He actually means that we study the world by consciously constructing models to see what the world is, and the question about what the world should be is based on a previous question. For example, the law refers to what the world should be, but the foundation of the law is what the world is ". If you do not know what the world is, but want to change the world and achieve a certain purpose, it is actually reversed at the end. In this sense, it is a misunderstanding that natural science does not study "yes. Professor Zhao proposed that humans are a machine of learning to explain their will. Godeli's theorem says: any complicated mathematical system, if its theorem is recursive judgment, is complex enough to expand the natural number theory in it, then any such mathematical system is incomplete. Because there is a true proposition in this system, and this proposition cannot be proved by this system. In this way, the second gedeli theorem can be introduced: a system cannot prove its own logical consistency. To convert the gordeli theorem into a physical language, we can say that the existence is true, not false. From this, we can see that some of the existence is not explained by strict deterministic theory. That is to say, randomness is an essential feature of the world. This world is not strictly deterministic and is a necessary condition for free will. Next, let me talk about a problem related to my professional game theory. Economics is now defined as a discipline for the study of human behavior, and its position in social science, in fact, it is the same as physics in natural science. The development of modern game theory shares the same things with Teacher Zhao. The current framework and idea of the Game Theory Research is as follows: the game theory explains how a person with a certain degree of rationality chooses his or her choice. The core of game theory is the Nash equilibrium solution, but two different research directions have emerged over the past two decades. One is the direction of Knowledge Theory and the other is the direction of evolution, knowledge theory assumes that the people involved in the game have a high degree of rationality. All the things that the matcher can reason can be rational people. They also have a common knowledge, that is, we all know, we all know that. Balancing in the framework of knowledge theory requires a high degree of rationality. Another research framework at the same time is the evolutionary framework. As long as bloggers have the ability to learn and evolve, they can adjust their actions based on past experiences and rewards and punishments. This shows that learning and evolution can be balanced, but complexity should be introduced in this process, and some knowledge of dynamic systems should be used to solve the problem. Li xiaoning: First of all, we can see the confusions in deterministic science and begin to notice the complexity. This is an improvement. I will discuss several principles. The basic things of Western science are derived from the form-driven thinking system of Aristotle. The definition of a concept comes from the definition of the connotation of a concept, that is, the meaning of a number plus an equal deviation. However, this cannot tell the essence of things, therefore, it is difficult to seek help from external sources. I think Mr. Zhao can use the "big fish model" to describe the Western knowledge system. The use of the "mini-planet model" to demonstrate China's knowledge system and the lack of research into China's thinking. We should start with the language. Westerners start from letters, so they pay attention to logic. Chinese are square characters, and language limits their thinking. To study information theory, they must first learn Chinese, understand Chinese, and learn ancient Chinese. The second is "self-consistency", self is just not stored, self is just an illusion. The third is truth. The scientific process is an error correction process. In addition, the key issue in the evolution theory mentioned above is that selection is not based on the average degree of freedom, but on the key lock points of a certain, very rigid, and stable system, it is relatively closed, stable, and not completely open. Some paths are locked. Let me talk about these points. Cai decheng: in my opinion, in our understanding of nature, "yes" is a core issue in the natural science just introduced by Professor Zhao. Of course, it is necessary to study what models to recognize and what theories to prove. Science is truth. This is a misunderstanding. In fact, science is an endless control process. It is the ability of every generation and generation of human beings to use their limited subjective understanding, understanding the external infinite objective existence is always impossible. It can only be relatively close to it. Future generations will find the mistakes of their predecessors, overturn them, and correct them. Therefore, science develops in the Process of "yes. Beyond the value of science, there is no authority, no conclusion, and no absolute truth. Both natural science and social science are essentially scientific and should be an endless process of understanding, only the objects of research are different. Natural Science Studies the objective existence of the universe. Social science studies the relationship between human society and people. This is the difference. Natural Science can prove false, while social science is difficult to prove false due to frame-based architecture and ideological limitations. I think the essence of science can be expressed as six elements: objective basis, rational doubt, multi-dimensional thinking, equality argument, empirical test, and tolerant encouragement. Natural sciences and social sciences should be unified in the most rational nature. Zhang xiangping: In your soft and hard structure, it is wrong to mention that the Growth Point of the tree is soft structure. The tree is the hardware, the surrounding environment, water, air, minerals should be soft structure, this is a process from disorder to order. The lock point mentioned by Li xiaoning just now is a popular saying. In fact, it is called the "strange machine". It does not seem to be mentioned in your lecture. Zhao: I took it. Zhang: however, without the "Strange convolutionator", we cannot proceed from disorder to order. Zhao: Yes. We will discuss this later. In my understanding and society, as Ding Li explained just now, it is harmful to take the "Godel problem" into consideration. Sheng Hong: time is limited, and the topic is interesting. I have been inspired by it. The soft and hard structure theory proposed by Mr. Zhao has many similarities with the Dissipation Structure Theory and chaos theory. As the definition of "What is life", life is neither fixed nor a brown movement, but between the two. The theory of dissipation structure is also about disorder and disorder. Teacher Zhao just talked about the learning problem in AI. In fact, the learning problem is also involved in many games in game theory, including many small computer models. What is learning, first of all, memory, and then judgment? I think in AI, the learning ability seems to grow. Of course, this is just my guess. What are the following in the "big fish model"? Actually, it is God. What we call rational is actually supported by something beyond rational, the initial principles and concepts cannot be asked, but they are not proved. This is what can be improved. Today's seminar is over. Thank you very much for bringing us a very good speech and a controversial speech.