摘要
“算法向善”是我国在推进算法治理制度化进程中,对算法技治规则的社会嵌入提出的一种规范性要求。智能算法的高阶发展具备在应用阶段对前置嵌入的道德伦理准则进行迭代的技术可行性,提升了辅助人类引导社会向善的技术治理能力。因此,算法赋能社会治理要正视算法治理的反身性特点,即算法的治理困境也需要算法赋能来解决。然而,作为社会规制对象的算法初始是由资本逻辑与技术逻辑所驱动,在参与社会治理的历史进程中相继演化出具有“自治偏见”的技治主义与监管主义逻辑,引发了压迫式父爱主义、宠溺式母爱主义、监管“失灵”等治理困境,这就要求提倡维系算法有限自治与完全自治之间动态均衡的合作治理逻辑,为具体实现全生命周期算法道德化提供了指导。
The“algorithms for good”is a normative requirement for the social embedding of algorithmic technocracy rules in the process of promoting the institutionalization of algorithm governance in China.It is technically feasible to iterate the pre-embedded moral and ethical standards in the application stage of the developed of high-order intelligent algorithm,which improves the technical governance ability that may assist human beings to guide society to good.Therefore,algorithm-empowered social governance should face up to the reflexive characteristics of algorithm governance,that is,the governance dilemma of algorithm also needs to be solved by algorithm empowerment.However,as the object of social regulation,the algorithm is mainly driven by capital and technical logic.In the historical process of participating in social governance,technocracy and regulatory logic with“autonomy bias”have evolved one after another,which has led to governance dilemmas,such as oppressive paternalism,doting motherly love and regulatory“failure”.This requires the promotion of cooperative governance logic to maintain the dynamic balance between limited autonomy and complete autonomy of the algorithm,which may have implications for realizing the moralization of the full life-cycle algorithm.
作者
俞鼎
YU Ding(School of Marxism,Zhejiang University,Hangzhou 310058,China;School of Social Sciences,Tsinghua University,Beijing 100084,China)
出处
《大连理工大学学报（社会科学版）》
CSSCI
北大核心
2024年第4期10-15,共6页
Journal of Dalian University of Technology(Social Sciences)
基金
浙江省社会科学规划项目“生成式人工智能(AIGC)社会实验的伦理问题及其对策研究”(24NDQN069YB)。
关键词
算法向善
技术治理
治理逻辑
Algorithms for good
Technical governance
Governance logic
