On The Refinement And Perfection Of The Ethical Codes Of Artificial Intelligence

As an action guide for human beings to deal with the ethical challenges of Artificial Intelligence(AI),the number of ethical code of AI has been increasing in recent years.Most of the researches on the ethical codes of AI are limited to one or several texts of the ethical codes,but few review and evaluate the current ethical codes of AI and put forward specific suggestions for the formulation of the ethical codes.Specific Suggestions should be made according to the characteristics of the current A.I.ethics.The purpose of this paper is to make specific suggestions to refine and improve the A.I.ethical codes by studying the characteristics of the current 50 A.I.ethical codes.This paper focuses on the characteristics of AI ethics,and discusses that AI ethics is a special technical ethics that challenges the existing interpersonal ethics paradigm,which is characterized by two seemingly contradictory theoreti cal presuppositions: anthropocentrism and non-anthropocentrism.Based on the c haracteristics of artificial intelligence ethics,this paper sorted out 50 ethical co des of AI and summarized two characteristics of the ethical principles of artifi cial intelligence: 1.There are three types of ethical principles of artificial intell igence: anthropocentric,non-anthropocentric and mixed;2.There is a huge gap between the number of artificial intelligence ethical principles formulated based on anthropocentrism and those formulated based on non-anthropocentrism.In this paper,the concept of "autonomy" to describe human autonomy and autonomy ratio relation of AI,and based on the "level of autonomy plan" to solve two kinds of AI in proportion to their codes of ethics of disparity proble ms,put forward two improved AI ethics advice: first,points type to develop AI of ethics;Second,a complete system of A.I.ethical codes is constructed ac cording to the "autonomy level diagram" to balance the proportion of the two types of A.I.ethical codes on the whole.In this way,the future ethics of AI will be more targeted in the face of different types of AI.