党政研究Issue(4)：27-36,10.
大语言模型时代下的网络政治社会化
Networked Political Socialization in the Era of Large Language Models:Feature Reconfiguration Marked by the Emergence of DeepSeek
摘要
Abstract
The release and application of low-cost,open-source large language models(LLMs),exempli-fied by DeepSeek,are reshaping networked political socialization.BY empowering By empowering online inter-actions,LLMs reconstruct the spatiotemporal logic and intersubjectivity of political cognition,catalyze citizens'incentives for political participation,and facilitate the diffusion of political culture,thereby accelerating the for-mation of civic political attitudes.In the LLM era,networked political socialization is characterised by a duality of actors,a hybridity of content,and a latent bias in orientation.Underpinned by LLM technology,it enables the re-articulation of public agendas,the restructuring of shared social values,and the innovation of social sys-tems.Yet the continued evolution of LLMs also generates new social tensions:the alienation of individuals heightens their desire to reassert subjectivity;the instability of technologically empowered public authority inten-sifies the demand for institutional constraints on power;and clashes of"rationalities"call for a renewed human-istic ethos.关键词
DeepSeek/生成式人工智能/政治社会化/算法大模型Key words
DeepSeek/Generative Artificial Intelligence/Political Socialization/Large Language Models/Al-gorithmic Governance分类
社会科学引用本文复制引用
张爱军,周杨..大语言模型时代下的网络政治社会化[J].党政研究,2025,(4):27-36,10.基金项目
国家社科基金后期资助项目"网络意识形态传播规律研究"(22FZZBO18) （22FZZBO18）
