探索与争鸣 ›› 2021, Vol. 1 ›› Issue (1): 95-102.

• 技术与文明 • 上一篇    下一篇

算法治理与治理算法 | “算法利维坦”的风险及其规制

张爱军   

  • 出版日期:2021-01-24 发布日期:2021-01-24

The Risk of “Algorithmic Leviathan” and Its Regulation

  • Online:2021-01-24 Published:2021-01-24

摘要:

智能算法带来治理效益的同时也会生成新的社会风险即“算法利维坦”。“算法利维坦”是“利维坦”的智能技术形态,即在社会治理过程中算法逐渐成为建构人们生存环境的重要甚至决定性力量,人们逐渐成为算法的附庸,臣服于算法程序监控之下,国家权力运行与社会治理日益依赖智能算法的维持。算法对人们的政治活动与政治行为具有反噬作用。“算法利维坦”在西方政党政治中表现为算法意识形态越位、算法选举越位、算法民调越位和算法政治谣言泛滥。“算法利维坦”权利让渡的半契约逻辑引致技术理性至上的绝对权威逻辑,并最终指向算法技术越位。算法技术越位所造成的后果与其生成逻辑具有内在关联,“算法利维坦”存在公民权利让渡导致算法权力越位的风险,存在算法政治越位导致“人的政治”终结的风险。算法风险规制旨在使智能算法服务于“人的政治”。从平衡逻辑推演,算法风险规制需以算法对抗算法、以舆论监督舆论、以公民权利约束公共权力、以边界剃刀制约边界剃刀等方式防止算法越位。

关键词: 算法利维坦 ,  算法权力 ,  算法政治 ,  算法越位 ,  人的政治

Abstract:

Intelligent algorithms not only bring governance benefits, but also generate new social risks, namely “algorithmic leviathan”.“Algorithm Leviathan”is the intelligent technology form of“Leviathan”, that is, in the process of social governance, algorithm has gradually become an important and even decisive force in the construction of people’s living environment, and people have gradually become a vassal of algorithms. Under the monitoring of the algorithm program, the operation of state power and social governance increasingly rely on the maintenance of intelligent algorithms. The algorithm has a counter-phagocytic effect on people’s political activities and political behavior.“Algorithmic Leviathan”is shown as algorithmic ideology offside, algorithmic election offside, algorithmic poll offside and algorithmic political rumors in western party politics. The semi-contractual logic of the transfer of rights in“Algorithmic Leviathan”leads to the absolute authoritative logic of the supremacy of technical rationality, and finally points to the offside of algorithmic technology. The consequence caused by the offside of algorithm technology is intrinsically related to its generating logic.“Algorithmic Leviathan”has the risk that the transfer of civil rights will lead to the offside of algorithmic power, and there is the risk that algorithmic politics will lead to the end of“human politics”. The purpose of algorithmic risk regulation is to make intelligent algorithms serve“human politics”. From the balanced logic deduction, the algorithm risk regulation needs to prevent the algorithm offside by the algorithm against the algorithm, the supervision of public opinion by public opinion, the restriction of public power by civil rights, the restriction of boundary razor by boundary razor, and so on.

Key words: