重庆理工大学学报2026，Vol.40Issue(14)：43-53,11.DOI:10.3969/j.issn.1674-8425(s).2026.07.004
大语言模型驱动下学术研究范式的演化、异化与调适
The evolution,alienation,and adjustment of academic research paradigms driven by large language models:A perspective from DeepSeek-R1
摘要
Abstract
As large language models(LLMs)become deeply embedded in academic research,academic knowledge production is gradually shifting from a model of"tool assistance"toward one of"deep integration".Represented by DeepSeek-R1,LLMs demonstrate notable strengths in reasoning,long-context processing,and open deployment,providing a typical perspective for examining paradigm changes in academic research in the age of artificial intelligence.Drawing on Kuhn's paradigm theory,this paper analyzes the paradigm evolution triggered by the embedding of LLMs in academic research from four dimensions:ontology,epistemology,methodology,and axiology.This study finds that LLMs not only reshape the relationship between research subjects and objects,but also reconfigure paths of knowledge generation,the organization of research processes,academic evaluation standards,and boundaries of responsibility.However,this evolution does not necessarily lead to paradigm optimization.The accompanying risks,including weakened subjectivity,intensified efficiency orientation,outsourcing of knowledge production,and diffusion of responsibility,stem from the coupling of technological logic,academic institutions,and changing modes of knowledge production.Accordingly,this paper proposes a governance framework from four dimensions—subject responsibility,process optimization,rule constraints,and institutional support—to promote the appropriate application and regulated development of artificial intelligence in academic research.关键词
DeepSeek-R1/大语言模型/学术研究范式/异化风险/治理回应Key words
DeepSeek-R1/large language model/academic research paradigm/alienation risk/governance response分类
社会科学引用本文复制引用
张峰,王丁一..大语言模型驱动下学术研究范式的演化、异化与调适[J].重庆理工大学学报,2026,40(14):43-53,11.基金项目
北京市教育科学规划课题重点项目"北京高校青年教师教学学术能力影响机制及提升路径研究"(BAFA25054) （BAFA25054）
