The Controversy of the Artificial Intelligence Schools

Latest reply: Sep 20, 2019 04:20:32 81 1 2 0

At present, there are three main schools of artificial intelligence: Symbolism, connectionism and behaviorism. 

Symbolism is also called logicism, psychology school orcomputer school. Its principle is mainly the assumption of physical symbol system (i.e. symbolic operating system) and the principle of limited rationality, which is mainly represented by knowledge graph application system. It belongs to the category of modern artificial intelligence, and simulates human intelligence behavior based on the intelligent simulation method of logic reasoning. The essence is to simulate abstract logic thinking of human brain, by studying the function mechanism of human cognition system, to describe human cognition process with some symbol, and to imitate human cognition process by inputting the symbol into the computer that can handle symbols, in a nutshell, “cognition is calculation”. Knowledge graph advocates the use of logical methods to establish a unified theory system of artificial intelligence, but has encountered the obstacle of “common sense” problem.


The representative figures of symbolism are:


Connectionism is also called bionics or physiology school. Its main principle is the neural network and the connection mechanism and learning algorithm among neural networks, which is mainly represented by deep learning, and it is the hot artificial intelligence school at present. The deep learning method greatly improves the efficiency of image recognition and voice recognition as well as plays a role in unmanned driving and smart + industries,. However, the deep learning model architecture itself is a “black box”, it is likely to let you unknowingly lose the opportunity to “find mistakes”. On the other hand, deep learning technology requires a large amount of data as a basis for training, and the results of training are difficult to apply to other issues.


The representative figures of connectionism are:


Behaviorism is also called evolutionism or cybernetics, and its principle is control theory and perceptual-action control system, which is mainly represented by enhanced learning method with reward and punishment control mechanism. Its enhanced learning can be used to accumulate learning performance based on the gain and loss in interaction, which is similar to the real learning mechanism of human beings. However, this method treats the human behavior process too simply and only the simple reward and punishment feedback process is measured in the test. The advocacy of behaviorism is too extreme. It does not study the importance of the internal structure of psychology and the negative consciousness of process, and then opposes consciousness and behavior, limiting the depth development of artificial intelligence.



The representative figures of Behaviorism are:

 150626vrv4w0curp0upff5.png 150626wujue6uucrnfp1qj.jpg 150626qaqd7zu4bquuk0r0.png


From the analysis of the characteristics and shortcomings of the above three schools of artificial intelligence, it is easy to see that people's thinking is difficult to be explained in the existing theoretical framework of artificial intelligence. Then how to do will make it possible to find a way to the bright future of intelligent science, I think that apart from the development of each school, it is necessary to combine each other, and the three major isms will move towards integration. It is believed that the future of AI should be human-machine intelligent integration, making full use of the advantages of people and machines to form a new intelligence.



  • x
  • convention:

MVE Created Sep 20, 2019 04:20:32 Helpful(1) Helpful(1)

good read :)
  • x
  • convention:

Login and enjoy all the member benefits

Login and enjoy all the member benefits