大模型时代:生成式人工智能的类人属性及社会影响
作者简介:陈茁,南京大学社会学院助理研究员(江苏南京 210023);陈云松,南京大学社会学院教授、南京大学中华文明数智创新实验室首席专家(江苏南京 210023)。
基金项目:
本文系国家社会科学基金重大项目“国家治理视角下基于数智方法的社会风险评估与应对”(24&ZD168)、中国博士后科学基金特别资助“人工智能方法在社会科学中的应用与分析范式研究”(2025T180703)的阶段性成果
摘要: 生成式人工智能已经表现出一系列类人属性,并凭借其广泛的场景适用性、强大的内容生成能力、便捷的交互方式及高度的自适应性等特征,引领着新一轮科技革命。其类人属性使其在特定场景下能够模拟人类的观点、思维和情感特征,并对社会结构和关系模式产生广泛而深远的影响。通过系统梳理近年来数字社会学领域关于生成式人工智能社会影响的主要研究成果与核心观点,以探讨大模型的类人属性为基础,深入剖析其在经济形态与劳动关系重塑、政治权力重构、文化生产与社会关系变革三大关键领域的潜在影响。可发现:生成式人工智能不仅为社会发展带来了前所未有的机遇,同时也揭示出一系列复杂的经济、政治与社会伦理挑战,亟需社会各界共同探讨并开展实证研究。
The Era of Large Models: Human-like Attributes and Social Implications of Generative Artificial Intelligence
Abstract: Generative AI has exhibited a range of human-like attributes.Through its broad applicability,powerful content-generation capabilities,intuitive modes of interaction,and strong adaptability,it is driving a new wave of technological transformation.These characteristics enable generative AI to simulate human perspectives,cognitive processes,and emotional expressions within specific contexts,thereby exerting extensive and profound influences on social structures and patterns of social relations.This article systematically reviews key findings and core perspectives in recent digital sociology research on the social implications of generative AI.Grounded in an examination of the human-like attributes of large models,it provides an in-depth analysis of their potential impacts across three critical domains: reshaping economic paradigms and labor markets,reconfiguring political power,and transforming cultural production and social relations.While generative AI offers unprecedented opportunities for social development,it simultaneously gives rise to complex economic,political,and socio-ethical challenges,underscoring the urgent need for interdisciplinary dialogue and rigorous empirical research.