“洛夫莱斯问题”与人工智能表现创造力的计算主义探究  

“Lovelace Problem”and a Computational Exploration of How Artificial Intelligence Appears Creativity

作　　者：束海波 任瀛 SHU Hai-bo;REN Ying(School of Philosophy,Anhui University,Hefei 230039,China)

机构地区：[1]安徽大学哲学学院,合肥230039

出　　处：《自然辩证法研究》2025年第2期74-80,127,共8页Studies in Dialectics of Nature

基　　金：国家社会科学基金青年项目“计算创造力及其哲学问题研究”(23CZX055)。

摘　　要：人工智能要想实现模拟乃至超越人类智能的终极目标,对作为人类智能之标志性特征和核心构成的创造力进行建模,进而开辟专门研究和实现创造力的人工智能分支领域势在必行。然而,由“洛夫莱斯问题”引发的人工智能创造力的可能性难题已成为阻碍当前人工智能进一步发展的瓶颈与桎梏。要化解这一难题,我们既要彻底地批判常识创造观中蕴含的拟人化的心灵观,又要充分地认识到创造力原型实例的多样性,从而为人工智能的创造力建模提供可靠的认识论与方法论。If artificial intelligence wants to achieve the ultimate goal of simulating or even surpassing human intelligence,it is imperative to model creativity,which is the hallmark feature and core component of human intelligence,and then open up a branch of artificial intelligence that specializes in studying and realizing creativity.However,the debate on the possibility of artificial intelligence creativity caused by the“Lovelace Problem”has become a bottleneck and shackle for the further development of artificial intelligence.To resolve this problem,we must thoroughly criticize the anthropomorphic view of the mind contained in the common sense view of creation,and fully recognize the diversity of creativity realization styles,so as to provide a reliable epistemology and methodology for the study of artificial intelligence creativity.

关 键 词：计算 创造力 拟人化 人工智能 “洛夫莱斯问题” 

分 类 号：N031[自然科学总论—科学技术哲学]