特别策划:人文社会科学研究新范式

AI驱动人文社会科学范式升维:从工具依赖到知识共生*

展开
  • (1.青藏高原人文环境数据智能实验室   甘肃兰州   730000)
祝忠明,青藏高原人文环境数据智能实验室研究员;寇蕾蕾,青藏高原人文环境数据智能实验室助理研究员。

收稿日期: 2025-06-11

  网络出版日期: 2025-07-23

基金资助

*本文系青藏高原人文环境数据智能实验室、兰州大学中央高校基本科研业务费专项资金“基于地方志数据挖掘的汉藏民族交融知识图谱构建与循证分析”(项目编号:2025jbkyjd005)研究成果之一。

AI-Driven Paradigm Elevation in the Humanities and Social Sciences: From Tool Dependency to Knowledge Symbiosis

Expand

Received date: 2025-06-11

  Online published: 2025-07-23

摘要

随着数据密集型“第四范式”科学发现与AI驱动“第五范式”科学研究的融合发展,以数据与智能深度协同为核心的数智科学范式正逐步成为当代科研范式演进的重要趋势与关键特征。AI已不再仅仅是提升科研效率的辅助手段,而是日益转变为知识生产中深度参与者和协作者。文章首先剖析了当前AI在人文社会科学中工具性应用的阶段性特征、典型场景及其固有的局限性。继而重点阐释了AI知识共生的理念及其构成要素,强调人机协同的知识生产模式是克服工具依赖瓶颈、激发人文社会科学创新活力的关键。最后进一步论述了构建AI知识共生体系的三大支柱性路径:以语义重构为核心的数据筑基、以开放演化为特征的算法强擎以及以智能体协同为机制的人机共创。在此基础上,探讨了AI知识共生如何赋能具有本土主体性自主知识体系构建,特别是在打破外部话语垄断、提升国际学术叙事能力方面的潜力。

本文引用格式

祝忠明 寇蕾蕾 . AI驱动人文社会科学范式升维:从工具依赖到知识共生*[J]. 图书与情报, 2025 , 45(03) : 13 -25 . DOI: 10.11968/tsyqb.1003-6938.2025029

Abstract

Against the macro-background of the accelerating convergence of the "Fourth Paradigm" (data-intensive science) and the "Fifth Paradigm" (AI-driven science), the data-intelligence scientific paradigm, with the in-depth synergy of data and intelligence at its core, is gradually becoming an important trend and key feature in the evolution of contemporary scientific research paradigms. Artificial Intelligence (AI) has transcended its role as a mere auxiliary tool for enhancing research efficiency and is increasingly becoming an indispensable participant and collaborator in the knowledge production process. This paper begins by analyzing the current stage of AI's instrumental application in the Social Sciences and Humanities (SSH), outlining its phased characteristics, typical scenarios, and inherent limitations. It then elaborates on the concept of AI-powered knowledge symbiosis and its core components, emphasizing that human-machine collaborative model of knowledge production is critical for overcoming the bottleneck of tool dependency and invigorating innovation within SSH. Further, the paper articulates three foundational pathways for constructing an AI knowledge symbiosis ecosystem: building a solid data foundation centered on semantic restructuring; leveraging powerful algorithmic engines characterized by open evolution and fostering human-machine co-creation through mechanisms of agent-based collaboration. Building on this framework, the paper explores how AI-powered knowledge symbiosis can empower the construction of an autonomous and locally-grounded knowledge system, particularly highlighting its potential to dismantle the monopoly of external discourses and enhance the narrative capacity of domestic scholarship on the international stage.
文章导航

/