摘要
生成式人工智能对人类知识生产的渗透正逐渐加深,在这一背景下社会科学如何合理利用前沿技术推动学科发展亟须反思。社会科学自发展初期就存在天然的内部张力,多元哲学基础奠定了其演化轨迹。计算社会科学的兴盛虽部分消解了原有对立,增强了不同范式间的联通性,但也加剧了方法论脱钩于本体论、认识论的局面。人工智能的崛起一方面延续了计算社会科学的固有路径,另一方面则借由其生成能力开辟了硅基样本、复杂因果、社会预报三重新领域。生成式大模型的介入,拓宽了社会科学的本体边界,使认识逻辑从静态性和确定性趋于演化性和情境性,并加强了方法谱系的联构。社会科学家的角色应当从技术借鉴走向深度协同,形成社会问题、技术发展等的跨学科共建。尽管人工智能有望推动社会科学范式进步,但训练偏误、应用不当、制度生态等问题也不容忽视,研究者应持有开放和审慎的态度,合理把握学科发展契机。
Generative artificial intelligence is increasingly permeating human knowledge production.Against this backdrop,how the social sciences can make sound use of cutting-edge technologies to advance disciplinary development requires urgent reflection.Since their early formation,the social sciences have contained inherent internal tensions,and their evolutionary trajectory has been shaped by plural philosophical foundations.Although the rise of computational social science has partially eased earlier oppositions and strengthened connectivity across paradigms,it has also intensified a tendency for methodology to become decoupled from ontology and epistemology.The emergence of AI,on the one hand,continues the established path of computational social science;on the other hand,by virtue of its generative capacity it opens up three new domains:silicon-based samples,complex causality,and social forecasting.The entry of large generative models broadens the ontological boundaries of the social sciences,shifts epistemic logic from static and deterministic orientations toward evolutionary and contextual ones,and reinforces the co-structuring of methodological lineages.The role of social scientists should move beyond mere borrowing of technology toward deep collaboration,forming interdisciplinary co-construction around social problems,technological development,and related agendas.Although AI is expected to propel paradigm progress in the social sciences,issues such as training bias,improper application,and the surrounding institutional ecosystem must not be overlooked.Researchers should therefore maintain an open,yet cautious stance,and appropriately seize opportunities for disciplinary development.
出处
《南开学报(哲学社会科学版)》
北大核心
2026年第1期44-58,共15页
Journal of Nankai University Philosophy,Literature and Social Science Edition
基金
国家社会科学基金重大项目(24&ZD155)。
关键词
人工智能
研究范式
社会科学
硅基样本
因果机制
Artificial Intelligence
Research Paradigms
Social Sciences
Silicon-Based Samples
Causal Mechanisms
