生成式人工智能的技术风险与人的主体性维护

Technological Risks of Generative Artificial Intelligence and the Preservation of Human Subjectivity

摘要: 生成式人工智能的快速发展，在显著提升生产效率的同时，也引发了技术依赖与主体功能弱化等深层风险。在认知、实践与生存层面，人的认知自主性、实践创造性与价值选择性面临潜在挑战。数字时代的主体功能弱化呈现出内在化、隐蔽化与微观化特征，个体在无意识中倾向于将思考、判断与创造活动让渡于技术系统，从而可能与自身的能动本质发生分离。应对这一风险，需在技术发展与社会进步之间寻求动态平衡，构建包括算法审计、数据伦理审查、人工智能素养教育及“人在回路”式人机协同机制在内的系统性策略，从而在技术赋能与主体维护之间形成良性张力，最终回归以“人”为中心的价值认同。

Abstract: The rapid development of generative artificial intelligence, while significantly enhancing production efficiency, has also triggered deep-seated risks such as technological dependence and the weakening of human functions. At the cognitive, practical, and existential levels, human cognitive autonomy, practical creativity, and value-oriented choice face potential challenges. The weakening of human functions in the digital age exhibits characteristics of internalization, concealment, and micro-level permeation, as individuals unconsciously tend to cede thinking, judgment, and creative activities to technological systems, potentially leading to a separation from their own capacity for agency. To address this risk, it is necessary to seek a dynamic balance between technological development and social progress, and to construct systematic strategies including algorithmic auditing, data ethics review, AI literacy education, and “human-in-the-loop” human-machine collaboration mechanisms. In doing so, a productive tension between technological empowerment and the preservation of human subjectivity can be formed, ultimately returning to a value identity centered on the “human”.