摘要
人工智能技术的飞速发展给人类社会带来变革性和颠覆性影响的同时必然引起学界和业界对其给人类社会所带来伦理挑战的大讨论。以“人工智能基本伦理准则是什么”为研究问题,运用质性研究中实物分析的方法对美国、欧盟、联合国教科文组织等有关人工智能的6份政府政策文件进行编码分析,提炼出人工智能伦理五大准则,并对准则进行讨论分析。研究发现:以人为本、安全可靠、公平正义、透明公开和负责监督是人工智能伦理的五大准则。基于研究发现,提出了以下五个方面的建议:在以人为本准则上,需要协调人机关系;在安全可靠准则上,需要关注数据隐私的安全;在公平正义准则上,需要尽力消除偏见歧视;在透明公开准则上,要加强算法、决策的可解释性;在负责监督准则上,要明晰责任归属以弱化社会矛盾。相关研究对中国当下人工智能伦理准则的研究反思与健康发展具有积极的借鉴意义。
The rapid development of artificial intelligence technology has brought transformative and subversive impacts to the human society.In the meanwhile,it inevitably arouses a great discussion in the academia and industry on the ethical challenges it brings to the human society.With"what is the basic ethical code of artificial intelligence"as main research question,this paper uses the method of document mining analysis in qualitative research to encode and analyze six government documents related to artificial intelligence,such as the United States,the European Union and UNESCO.By adopting Nvivo12plus analysis software,it extracts five ethical codes of artificial intelligence,and discusses and analyzes the codes.The study found that human-oriention,security and reliability,fairness and justice,transparency and openness,and responsible supervision are the five principles of artificial intelligence ethics.In the human-oriented principle,it is necessary to coordinate the man-machine relationship;In terms of security and reliability criteria,we need to pay attention to the security of data privacy;On the criterion of fairness and justice,we need to try our best to eliminate prejudice and discrimination;On the criterion of transparency and openness,we should strengthen the interoperability of algorithms and decisions;On the criterion of responsible supervision,we should clarify the responsibility in order to solve social conflicts.The relevant research could be served as reference for the thinking and development of China's current artificial intelligence ethics.
作者
陈殿兵
朱鑫灿
CHEN Dianbing;ZHU Xincan(Zhejiang Normal University College of Education,Jinhua 321004,China)
出处
《比较教育学报》
CSSCI
北大核心
2024年第1期72-83,共12页
Journal of Comparative Education
关键词
人工智能技术
伦理准则
框架建构
artificial intelligence technology
code of ethics
framework
