| 2,446 | 24 | 146 |
| 下载次数 | 被引频次 | 阅读次数 |
近年来，在世界范围内涌现出大量的人工智能治理原则。这些文件通常采取“原则主义”的策略，通过提取“最大公约数”的方法对伦理共识进行清单化的列举。这种策略固然有益于回避争议、促进协同，但也难免在具体情境中陷入概念模糊和自相矛盾。为了明晰概念、澄清语境、促进对话，本文搜集了来自全球各地的88份原则文本，在伦理学理论脉络中定位人工智能治理核心议题和相关主张。研究发现治理原则围绕着价值立场、价值目标和正义原则三个核心问题展开，在不同的理论脉络和现实情境中，构成了一个丰富而系统的问题域。推动人工智能的治理实践，需要在静态的对话框架之上进一步构建动态的规制措施，而整合原则文本的伦理基础和治理框架，有助于推动抽象的伦理主张转变为具体的治理实践，以应对复杂现实情境中的道德选择和行动需求。
Abstract：A large number of AI governance principles have emerged in recent years in an attempt to guide governance practices at the ethical and normative levels. However, the ethical framework has problems in the clarity of the concept and the coordination among the principles. With the increasing diversity of governance agents, it is increasingly difficult to coordinate among the proliferating principle texts. In order to establish a unified framework for comparison and dialogue, on which to integrate and clarify the ethical basis of existing AI governance principles, this paper analyzes and summarizes 88 principles texts from around the world. It is found that these principles respond to three interrelated ethical issues: the issue of value standpoint, the issue of value goals and the issue of justice. These core problems derive more specific problems in different theoretical contexts and practical situations, forming a systematic problem domain. Clarifying the theory context of this problem domain will provide a dialogue framework for different AI governance agents.
(1)刘宪权:《人工智能时代机器人行为道德伦理与刑法规制》,《比较法研究》2018年第4期。
(2)彼得·保罗·维贝克:《将技术道德化》,闫宏秀、杨庆峰译,上海交通大学出版社2 016年版,第62页。
(3)Bovens M,"Analysing and assessing accountability:A conceptual framework",European Law Journal,2007,13(4),pp.447~468.
(4)拉·梅特里:《人是机器》,顾寿观译,商务印书馆1999年版,第65页。
(5)(8)Fletcher G,The philosophy of well-being:An introduction,Routledge,2016,p.5,p.95.
(6)Kagan S,"The limits of well-being",Social Philosophy and Policy,1992,9(2),pp.169-189.
(7)Adler M J,"The dignity of man and the 21st century."A Speech Delivered to Members of The Commonwealth Club,1952.
(9)(13)Nozick R,A narchy,state,and utopia,Basic Books,1974,p.1,p.47.
(10)伊曼努尔·康德:《道德形而上学原理》,苗力田译,上海人民出版社1986年版,第8 0页。
(11)Scheffl er S,The rejection of consequentialism, Ox-ford University Press,1994,p.61.
(12)MacAskill W,"Effective altruism:introduction",Essays in Philosophy,2017,18(1),pp.1~5.
(14)Okun Arthur M,Equality and Efficiency:The Big Tradeoff,Washington,The Brookings Institution,p.117.
(15)Marx Karl,Critique of the Gotha Program;reprinted in Marx-Engels-Werke(MEW)vol. 19,Berlin 1978;and in MarxEngels-Gesamtausgabe(MEGA-B),Berlin 1975,vol.Ⅰ25.
(16)(18)约翰·罗尔斯:《正义论》,何怀宏、何包钢、廖申白译,中国社会科学出版社2009年版,第58~69页。
(17)高景柱:《德沃金与运气均等主义理论》,《西南大学学报(社会科学版)》2021年第2期。
(19)于水、查荣林、帖明:《元治理视域下政府治道逻辑与治理能力提升》,《江苏社会科学》2014年第4期。
(20)Gjaltema J,Biesbroek R,Termeer K,"From government to governance…to meta-governance:a systematic literature review",Public Management Review,2020,22(12),pp.1760~1780.
(21)Kooiman J,Jentoft S,"Meta-governance:values,norms and principles,and the making of hard choices",Public administration,2009,87(4),pp.818~836.
基本信息:
DOI：10.14167/j.zjss.2023.01.016
中图分类号:TP18;B82-057
引用信息:
[1]赵瑜,周江伟.人工智能治理原则的伦理基础：价值立场、价值目标和正义原则[J].浙江社会科学,2023,No.317(01):109-118+159-160.DOI:10.14167/j.zjss.2023.01.016.
2023-01-12
2023-01-12