在数字时代，随着大数据和人工智能技术在公共决策和公共服务中的广泛应用，算法科层制这一政府管理体制变革正在中国悄然发生。本文在已有研究基础上构建了一个算法科层制的理论分析框架，基于算法功能和算法应用体制两个关键因素，发展了四种算法科层制形态：部门自动化算法科层、部门预测型算法科层、跨部门自动化算法科层和跨部门预测型算法科层。本文通过四个人工智能算法应用案例，实证探究人工智能算法对科层组织权力关系与激励方式、知识信息处理、组织形态以及服务供给模式的影响及其路径，揭示了四种算法科层制形态的差异性及形成机制。

 In the digital era,with the expanding application of big data and artificial intelligence technologies in public decision-making and public service provision, algorithmic bureaucracy, a revolution in government management system,is quietly taking place in China. This article develops the concept of algorithmic bureaucracy proposed by existing studies, constructs a theoretical analytical framework for algorithmic bureaucracy,and develops four types of algorithmic bureaucracy based on two key factors of algorithmic function and algorithmic application system: departmental automated algorithmic bureaucracy, infra-departmental automated algorithmic bureaucracy, departmental predictive algorithmic bureaucracy, and infra-departmental predictive algorithmic bureaucracy. Through four domestic AI algorithm application cases, the impact of AI algorithms on the power relations and organizational incentives, knowledge and information processing, organizational patterns,and service supply modes and their paths are empirically explored, and the differences between the four algorithmic types and their formation mechanisms are revealed.

欢迎阅读《社会学研究》！您是该文第3269位读者！