人工智能生成内容“可版权性”的法哲学基础——以人工智能哲学理论为视角  

On the Philosophy-of-Law Foundation of the Copyrightability of the AI-generated Work——From the Perspective of Philosophy of Artificial Intelligence

作　　者：刘文献 Liu Wenxian

机构地区：[1]湖南文理学院文史与法学学院

出　　处：《政治与法律》2020年第3期14-26,共13页Political Science and Law

基　　金：湖南省教育厅社科基金项目“后现代视角下知识产权合法性研究”(项目编号:16C096);湖南文理学院博士启动基金项目“后现代视角下知识产权合法性研究”(项目编号:16BSQD38);国家社会科学基金后期资助项目“著作权作者中心主义哲学批判”(项目编号:18FFX033)的阶段性成果

摘　　要：论证人工智能生成内容可版权性领域有结果主义模式与历史主义模式两种典型模式,它们分别与人工智能哲学中的图灵测试和中文屋试验相勾连。这两种模式论证有失偏颇,正确的路径应该是将结果主义、历史主义与创作主体三者融合统一,将图灵测试、中文屋试验与具身人工智能这三种人工智能哲学理论整合起来,作为其合法性基础。当前人工智能生成内容高度依赖于人类的参与,其缺乏可版权性的合法性基础;“道成肉身”的具身人工智能具有“类人”的智能,其生成内容具有可版权性的合法性基础。There are two typical models of demonstrating the copyrightability of AI-generated work:the model of consequentialism and the model of historicism,which respectively correspond to Turing Test and Chinese Room Test in the philosophy of artificial intelligence.Both models are not satisfactory and the right approach is to combine consequentialism,historicism and creative subject,and integrate three theories on the philosophy of artificial intelligence,i.e.Turing Test,Chinese Room Test and embodied AI as the legal foundation.At present,the AI-generated work is highly dependent on the participation of human beings,and thus lacks the legal foundation for copyrightability.The embodied AI with incarnation has the man-like intelligence and thus the generated work has the legal foundation for copyrightability.

关 键 词：可版权性 图灵测试 中文屋试验 具身人工智能 

分 类 号：DF523.1[政治法律—民商法学]