大模型的理解力之争与理解观新叙事  

The Debate over the Understanding Ability of the Large Language Model and the New Narrative of the Understanding View

作　　者：肖峰 XIAO Feng

机构地区：[1]上海大学智能哲学与文化研究院 [2]上海大学马克思主义学院

出　　处：《社会科学》2024年第1期41-51,共11页Journal of Social Sciences

基　　金：国家社会科学基金重大项目“负责任的人工智能及其实践的哲学研究”(项目编号:20BZX027)的阶段性成果。

摘　　要：人工智能是否具有理解力的问题在大模型出现后引起了新的关注,目前截然对立的看法之间正在进行难分难解的争议。持否认意见的一方主要从具身性、工作机制等方面论证大模型不可能有人的理解力,持肯定看法的一方则从功能和主观体验等方面来阐释大模型所具有的理解力。从哲学的视野上看,还需要从理解观的更新去解析大模型是否具有理解力的问题,包括整合人本主义与功能主义的理解观,基于动态发展的视野丰富人工理解或机器理解的含义,并走向一种人机互补与合作的新型理解观。The issue of whether artificial intelligence has understanding ability has attracted new attention after the emergence of large language models(LLMs),and there is currently a heated debate between opposing views.The party who holds a negative opinion mainly argues that the LLMs cannot have human understanding from aspects such as embodiment and working mechanism,while the party who holds a positive opinion explains the understanding of the LLMs from aspects such as function and subjective experience.From a philosophical perspective,it is also necessary to:analyze the issue of whether the LLMs has ability from the perspective of updating the understanding view,including integrating the understanding view of humanism and functionalism;enrich the meaning of artificial or machine understanding based on a dynamic development perspective;and move towards a new understanding view of human-machine complementarity and cooperation.

关 键 词：人工智能 大模型 理解 功能 人机互补 

分 类 号：N031[自然科学总论—科学技术哲学] B089[哲学宗教—哲学理论]