Artificial Intelligence: The 13th chapter on the debate and prospect of artificial intelligence

The 13th chapter discussion and prospect of artificial intelligence

Teaching Content: This chapter mainly Apple told the debates on theories, methods and technical routes of various academic schools in the development of artificial intelligence, the influence of artificial intelligence on the economy and society and the prospect of artificial intelligence development.

Teaching Focus: enlighten the students to understand the cognition of AI schools, and analyze the development prospect of artificial intelligence.

Teaching Difficulty: the similarities and differences of the three main academic schools of artificial intelligence.

Teaching Methods: The use of classroom teaching and organizational student discussion to guide students to actively predict the future development of artificial intelligence.

Teaching Requirements: A comprehensive understanding of the development of artificial intelligence in the main academic schools of the views and debates, general understanding of artificial intelligence on human society and Economic impact and the development of artificial intelligence, the future of artificial intelligence is full of confidence.

13.1 The controversy of artificial intelligence

Teaching Content: The development of artificial intelligence in the course of various academic schools in various aspects of the controversy.

Teaching Focus: The cognitive view of the three major academic schools of AI.

Teaching Difficulty: the similarities and differences of the three major academic schools.

teaching methods: lecturing mainly

Teaching Requirements: a more in-depth understanding of the development process of artificial intelligence, the main academic schools and debates.

13.1.1 's argument on artificial intelligence theory

1. Symbolism:

Semiotics holds that human cognitive primitives are symbols, and the cognitive process is the process of symbolic manipulation. It thinks that man is a physical symbolic system, and the computer is a physical symbolic system, so we can use computer to simulate human's intelligent behavior, that is, to use the computer's symbolic manipulation to simulate human cognitive process. In other words, people's thinking is operable. It also believes that knowledge is a form of information and is the basis for the formation of intelligence. The core problem of AI is knowledge representation, knowledge inference and knowledge application. Knowledge can be represented by symbols and can be inferred by symbols, so it is possible to establish a unified theoretical system of knowledge-based human intelligence and machine intelligence.

2. Linking doctrine :

It is believed that the human thought primitive is the neuron, not the symbolic processing process. It objected to the hypothesis of the physical symbology, thinking that the human brain was different from the computer, and suggested that the brain work mode of the coupling doctrine was used to replace the computer working mode of symbolic operation.

3. Behavioral doctrine

The behavioral doctrine thinks that intelligence depends on perception and action (so called "behavioral Doctrine"), and proposes the "perceptual-action" mode of intelligent behavior. The behavioral advocates believe that intelligence requires no knowledge, no representation, no reasoning; artificial intelligence can evolve as human intelligence (so called evolutionism), and intelligent behavior can only be demonstrated in the real world interacting with the surrounding environment. The behavior doctrine also holds that: the description of the objective things of the real world and the mode of intelligent behavior of the symbolism (also including the coupling doctrine) is too simplistic abstract, so it can not truly reflect the objective existence.

question: in the first chapter of this book we refer to the three main schools of artificial intelligence, please review the students, which are the three schools, each has what characteristics.

　

13.1.2 's argument on artificial intelligence methods

1. Symbolism:

Symbolism believes that the research method of artificial intelligence should be a function simulation method. By analyzing the function and functions of human cognition system, and then simulating these functions by computer, we can realize artificial intelligence. The symbolism tries to use the mathematical logic method to establish the unified theory system of artificial intelligence, but it encounters many difficulties which can not be solved temporarily, and is rejected by other schools.

2. Linking doctrine :

The link doctrine advocates that AI should focus on structural simulation, that is, simulating human physiology and neural network structure, and considers that function, structure and intelligent behavior are closely related. Different structures show different functions and behaviors. A variety of artificial neural network structures and many learning algorithms have been proposed.

3. Behavioral doctrine:

The behavior Doctrine thinks that the research method of artificial intelligence should adopt behavior simulation method, also think function, structure and intelligent behavior are inseparable. Different behaviors show different functions and different control structures. The research method of behaviorism is also doubted and criticized by other schools, and it is believed that behaviorism can only create intelligent insect behavior, and cannot create human intelligent behavior.

13.1.3 's argument on the technical route of artificial intelligence

1. Special route:

The special route emphasizes the development and development of special intelligent computer, artificial intelligence software, special development tools, artificial intelligence language and other special equipment, such as Lisp Machine, PROLOG Machine, PROLOG, Lisp language, M.I language, OPSS83 language, Expert system development Tools Emycin, expert, INSIGHT2 and guru.

2. General Route:

General-purpose routes believe that the common computer hardware and software can provide effective support for artificial intelligence development, and can solve a wide range of general AI problems. Examples of this include RISC technology based on VLSI, Unix time-sharing operating systems, C language and its improved, sun workstations, and SPARC workstations. The general route emphasizes the development of artificial intelligence application system and artificial intelligence product, should combine with the computer stereoscopic technology and mainstream technology, and regard knowledge engineering as a branch of software engineering.

3. Hardware route:

The hardware Route thinks that the development of artificial intelligence mainly relies on hardware technology, such as VLSI, artificial neural network, brain model, intelligent machine and bot. The route also believes that the development of intelligent machines mainly depends on a variety of intelligent hardware, intelligent tools and curing technology.

4. Software route:

The software Route emphasizes that the development of artificial intelligence relies mainly on software technology, such as enlightening programming, automatic programming expert system, knowledge engineering and other intelligent algorithms. The software route thinks that the development of intelligent machine mainly lies in developing all kinds of intelligent software, tools and application system.

13.2 Effects of artificial intelligence on human beings

Teaching Content: Explain the development of artificial intelligence to human influence.

Teaching Focus: the development of artificial intelligence to human influence degree.

Teaching Difficulties: Why artificial Intelligence development has an impact on human beings.

Teaching Methods: The combination of classroom instruction and organizational student discussion.

Teaching Requirements: A preliminary understanding of the development of artificial intelligence on human economic and social impact.

13.2.1 the influence of artificial intelligence on economy

The development and application of artificial intelligence system has created considerable economic benefits for mankind, and expert system is an example. With the continued decline of computer system price, artificial intelligence technology will be more widely promoted, resulting in greater economic benefits.

Example: 1. Benefits of expert systems 2. AI drives the development of computer technology

　

13.2.2 the influence of artificial intelligence on society

1. Labor and employment problems

Because artificial intelligence can replace human mental work, such as the use of expert systems instead of managers or doctors for decision-making or diagnosis and treatment of patients with disease, so that some people will have to change their jobs, and even cause unemployment. The use of artificial intelligence in technology and engineering can lead some people to lose access to information-processing activities such as planning, diagnosis, understanding and decision-making, and even have to change the way they work.

2. Changes in social structure

On the one hand, it is hoped that artificial intelligence and intelligent machines can replace human beings in various kinds of labor, and on the other hand they fear that their development will cause new social problems. In fact, in the past more than 10 years, the social structure is undergoing a quiet change. Man 椈  Haiyun na called Jiao Mao  War  藯 Intelligent machine 椈  Haiyun na called Jiao Mao gu 〈v Qiao dioxin  Charm Division of Dioxin  quiff  arc H-Wall  bad face lie Shcha Shengzhou gizzard brother Qiao grandmother Curtain  feast descriptively γ⑹ view φ-Xixi , called Jiao

3. Change of thinking mode and concept

The development and popularization of artificial intelligence will affect the way of thinking and traditional ideas of human beings, and make them change. For example, traditional knowledge is generally printed in books, newspapers or magazines, and thus is fixed, and the knowledge of the knowledge base of AI systems can be constantly modified, expanded and updated. Another example is that once users of expert systems begin to believe in the judgment and decision of the system (intelligent machines), they may be reluctant to use their brains, become lazy, and lose their sense of responsibility and sensitivity to many problems and their solution tasks. Over-reliance on computer-based advice, without analysis, will degrade the cognitive abilities of intelligent machine users and increase misunderstandings.

4. Psychological threats

Ai also makes some members of society feel a psychological threat, or a mental threat. It is generally believed that only humans have a sense of mind, and this is the machine. If one day, these people start to believe that machines can also think and create, then they may feel disappointed, and even feel threatened. They fear that, someday, intelligent machines will surpass human intelligence and make humans slaves to intelligent machines and intelligent systems. According to artificial intelligence, it is possible for humans to use machines to plan their future, and even to think of this planning problem as a kind of state space search. While some in society welcome this new concept, others find these new ideas annoying and unacceptable, especially when they run counter to their beloved beliefs and perceptions.

5. Danger of runaway technology

The greatest danger of any new technology is that man loses control of it, or it falls into the hands of those who attempt to exploit new technologies against humanity. Some people worry that the robot will one day anti-guest, slavery their creator will dead Kiss p Catalinas 田ぶ Qiao grandmother Sodium lu liling  broken bark  trampled death sigh mo step ?/p>

discussion: What are the positive and negative effects of artificial intelligence on society?

Example: introduction of the famous American sci-fi writer Asimov (I.asimov) put forward the "Three Principles of robotics" and lead the discussion.

　

The influence of 13.2.3 artificial intelligence on culture

1. Improving human knowledge

In the process of re-elaborating our knowledge of history, philosophers, scientists and AI researchers have had the opportunity to resolve the ambiguity of knowledge and to eliminate the inconsistency of knowledge. The results of such efforts may lead to certain improvements in knowledge so that it is easier to infer new truths of interest.

2. Improving human language

Human language is accompanied by human evolution and development of the emergence and development of human intelligence activities play an extremely important role. According to linguistics, language is the expression and tool of thinking, and the law of thought can be studied by means of linguistics, but the subconscious and subconscious are often "only inexpressible". The use of artificial intelligence technology, the comprehensive application of grammatical, semantic and formal knowledge representation, it is possible to improve the natural language of knowledge representation, while the knowledge is described as the applicable form of artificial intelligence.

3. Improve cultural life

Artificial intelligence technology has opened many new windows for human cultural life. The existing intelligence game consoles will be developed into a more intelligent means of cultural entertainment. Machine Vision Technology can provide a new method of "looking", which may produce a special image transformation corresponding to the image interpretation, which can produce illusion of similar condition, and show the super-realistic representation of receiving effect. The vision system can also reconstruct the scene in a special way, which will have far-reaching implications for the graphic arts, advertising and social education sectors, such as changing the face of television and enabling people to enjoy a more advanced entertainment life in front of the television.

Example: establishing a "knowledge refinery".

　

13.3 The outlook for artificial intelligence

Teaching Content: Looking forward to the development direction of artificial intelligence.

Teaching Focus: the development direction of artificial intelligence.

Teaching Difficulty: Let the students think about the development of artificial intelligence according to their own viewpoint.

Teaching Methods: the main discussion.

Teaching Requirements: to guide students to actively engage in the future development of artificial intelligence discussion, active academic thinking.

discussion: how AI will evolve. Where further research should be carried out.

Exercise: Write a small paper, talk about the future development of artificial intelligence trends and countermeasures.

　

The theoretical framework of 13.3.1 update

There are many problems in AI research:

(1) Macroscopic and microscopic isolation

On the one hand, the intellectual hierarchy of philosophy, cognitive science, thought sciences and psychology is too high, too abstract, on the other hand, the intelligence level of the logic symbol, neural network and behavior of artificial intelligence is too low. These two aspects are too far apart, there are many levels in the middle has not been studied, can not be macroscopic and micro-organic integration and mutual infiltration.

(2) Global and local fragmentation

Human intelligence is the whole effect of brain system, which has rich levels and many sides. However, symbolism only captures the abstract nature of the human brain, and the connection doctrine only imitates the character of the thinking in image, and the behavioral doctrine focuses on the characteristics of human intelligent behavior and its evolutionary process. There are obvious limitations.

(3) Theoretical and practical disconnection

The actual work of the brain, we know a lot on the macro, but the intelligence of the various, unpredictable, complex difficult to get a clear clue. On the microscopic level, we know very little about the working mechanism of the brain, which makes it difficult to find patterns. In this context, the various artificial intelligence theories, only some people's subjective conjecture, can show in some aspects of "intelligence" even if it is quite successful.

The above problems and other problems show that the structure and function of the human brain are much more complex than people think, artificial intelligence research is more difficult than we estimate, the task of AI research is much more difficult than we discussed. At the same time, it also shows that to understand the structure and function of human brain fundamentally, solve the difficult problems and complete the research task of artificial intelligence, we need to find and establish an updated artificial intelligence framework and theoretical system, and lay a theoretical foundation for the further development of artificial intelligence. We need at least a few generations of continuous struggle, multi-disciplinary joint collaborative research, it is possible to basically solve the "intelligence" mystery, so that artificial intelligence theory to a higher level.

13.3.2 Better Technology Integration

Artificial intelligence technology is the integration of intelligent technology and various information processing technology and related disciplines technology. The information technology to be integrated in addition to digital technology, but also includes computer networks, remote communications, databases, computer graphics, voice and hearing, robotics, process control, parallel computing, quantum computing, optical computing and biological information processing technology.

Intelligent systems include not only psychology, linguistics, modern computing, cognitive science and computational linguistics, but also sociology, anthropology, philosophy and system science. Calculation is not only an important part of the intelligent system support structure, but also the vitality of the intelligent system, just as the blood is as important as the human body.

13.3.3 more mature methods of application

The realization of artificial intelligence requires hardware assurance, however, software should be the core technology of artificial intelligence. Many artificial intelligence application problems need to develop complex software systems, which helps to promote the emergence and development of software engineering disciplines. Software engineering can provide a standard procedure for solving certain types of problems, and knowledge software can provide an effective programming method for solving artificial intelligence problems. Because of the complexity and universality of artificial intelligence application problem, the traditional software design method is obviously not enough and not applicable. Artificial intelligence methods must support the development experiment of artificial intelligence system, and allow the system to evolve from a smaller core prototype to a complete application system.

It should be confidently expected that a common and effective approach to artificial intelligence development will be developed. More advanced AI general language, more effective AI-specific language and development environment or tools, and artificial intelligence development-specific machines will be constantly emerging and updated to provide a powerful tool for AI research and development. We are sure that we will eventually be able to study the more sophisticated ways in which artificial intelligence can be successfully applied in more areas.

In the current research of artificial intelligence application methods, there are several noticeable topics, namely, multi-method hybrid technology, multiple expert system technology, machine learning (especially neural network learning and Knowledge discovery), hardware and software integration technology and parallel distributed processing technology. Among them, some people think: the human brain mechanism and distributed artificial intelligence research, established the foundation of a new generation of computers.

With the development of artificial intelligence application, the field of artificial intelligence will be expanded. We can predict that artificial intelligence, intelligent machines and intelligent systems will have a much wider range of applications than current computers. Where there is human activity, it will be applied to AI technology.

13.4 Concluding remarks

This chapter will briefly introduce the argument of AI school, discuss the influence of artificial intelligence on human future, and prospect some problems of AI development. We are optimistic about the future development of artificial intelligence. We believe that artificial intelligence has a better future, though this day may be far off, and that the day comes with hard work and expensive costs, and it takes generations of people to continue to struggle.

　

From:http://netclass.csu.edu.cn/jpkc2003/rengongzhineng/rengongzhineng/jiaoan/chapter13.htm