数智技术驱动的生产力变革：AI大数据模型与新质生产力发展的理论逻辑

Productivity Transformation Driven by Digital-Intelligent Technology: The Theoretical Logic of AI Big Data Models and the Development of New Quality Productive Forces

摘要: AI大数据模型正在重塑生产力的要素结构，但其作用机制与社会后果尚未得到充分讨论。本文以生产力理论的要素框架为分析起点，并与国际学术界关于技术变革、劳动市场与算法治理的研究展开对话，考察AI大数据模型驱动生产力变革的理论逻辑。研究表明：在微观层面，AI大数据模型通过重塑劳动者的技能结构、推动劳动资料的智能化、使数据成为核心劳动对象，构成生产力变革的要素基础；在中观层面，上述要素变革通过创新循环的加速、要素组合效率的提升与产业组织形态的重组，转化为科技创新驱动、高效能与高质量输出、产业转型升级三重效应，从而将“要素变革”与“效应表现”两个维度联结为层层递进的因果链条。与此同时，这一进程并非纯粹的增益：它还带来就业结构的替代与极化压力、算法决策的公平性风险、劳动过程的数字化重构，以及资本有机构成提高所引致的生产关系张力。本文据此主张，对AI大数据模型的评估应超越单一的效率视角，在技术进步、效率提升与社会成本之间建立均衡的分析框架，并以相应的制度与治理安排回应其社会后果。

Abstract: AI big data models are reshaping the factor structure of productive forces, yet their mechanisms and social consequences remain insufficiently examined. Taking the factor framework of the theory of productive forces as a point of departure, and engaging with international scholarship on technological change, labor markets and algorithmic governance, this paper examines the theoretical logic through which AI big data models drive the transformation of productive forces. The analysis shows that, at the micro level, AI big data models constitute the factor basis of this transformation by reshaping the skill structure of laborers, intellectualizing the instruments of labor, and turning data into a core object of labor; at the meso level, these changes are translated into three effects - innovation-driven growth, high-efficiency and high-quality output, and industrial transformation and upgrading - through the acceleration of innovation cycles, the improvement of factor-combination efficiency, and the reorganization of industrial organization, thereby connecting the two dimensions of “factor change” and “effect manifestation” into a progressive causal chain. This process, however, is not a pure gain: it also generates substitution and polarization pressures in the employment structure, fairness risks in algorithmic decision-making, a digital reconfiguration of the labor process, and tensions in production relations arising from a rising organic composition of capital. The paper argues that the assessment of AI big data models should move beyond a single efficiency perspective and adopt a balanced analytical framework that weighs technological progress, efficiency gains and social costs, supported by institutional and governance arrangements addressing their social consequences.