算法权力异化的主体性风险与规制路径

The Subjective Risk and Regulatory Path of Algorithmic Power Alienation

摘要: 智能时代基于大数据和互联网的算法在社会领域应用越来越广泛。具有自主学习和决策功能的智能算法表现出强大的嵌入性，数字空间中算法对于数据信息处理的高效率，使得算法发展脱离了纯粹的工具范畴，这让算法在资源配置中具备了“去工具化”特征和事实上的技术权力。算法权力不断自我迭代和肆意扩张，打破了人机关系中的原有的权力平衡，引发利益失衡、主体降维危机等诸多异化风险，数字空间中人的主体性受到支配和控制。完善数字空间中算法权力的治理体系，需要从人的认知属性、对算法权力赋予理性和道德、法律规制等多方面进行探索，实现人的主体性价值在算法权力中的重塑与复归，完善法律责任的主体归责，构建算法权力运行机制的评估监督，倡导法律规制下的算法权力走向“善治”与“善智”的理想治理建构模式。

Abstract: Algorithms based on big data and the Internet in the intelligent era are more and more widely used in the social field. Intelligent algorithms with autonomous learning and decision-making functions show strong embedding, and the efficient efficiency of algorithms for data information processing in the digital space makes the algorithm development out of the pure tool category, which makes the algorithm have the characteristics of “de-instrumentation” and the de facto technical power in the resource allocation. The continuous self-iteration and wanton expansion of algorithm power has broken the original balance of power in the human-machine relationship, triggering many risks of alienation such as imbalance of interests and the crisis of dimensionality reduction of the main body, and the subjectivity of people in the digital space is dominated and controlled. To improve the governance system of algorithm power in the digital space, it is necessary to explore from many aspects such as people’s cognitive attributes, giving rationality and morality to algorithm power, and legal regulation, so as to realize the reshaping and restoration of human subjective value in algorithm power, improve the subject responsibility of legal responsibility, build the evaluation and supervision of the operation mechanism of algorithm power, and advocate the ideal governance construction model of “good governance” and “good wisdom” under legal regulation.