新兴技术视域下的人工智能善治——发展现状、风险挑战与治理思路  

Good Governance of Artificial Intelligence from the Perspective of Emerging Technologies--Development Status,Risks&Challenges and Governance Ideas

作　　者：梁正[1] Liang Zheng

机构地区：[1]清华大学人工智能国际治理研究院

出　　处：《学术前沿》2024年第14期12-22,共11页Frontiers

基　　金：新一代人工智能国家科技重大专项和清华大学自主科研计划的阶段性成果,项目编号分别为:2023ZD0121700、20223080026。

摘　　要：人工智能正在引发一场远超乎于技术场域的时代变革,以大模型为代表的生成式人工智能塑造着未来的主要技术进路。尽管我国已成为一个人工智能大国,但在核心理论创新、基础底座建设、关键软硬件和生态方面还有待提升,迈向人工智能强国还有很长的路要走。在此情况下,如何在保障人工智能有力、有序、有益发展的同时,实现对人工智能的善治是亟需回应的关键问题。因此,平衡好发展与治理之间的关系,践行价值对齐的伦理思路,吸收人工智能国际治理有益经验,探索一条理念开放、主体多元、对象分层、工具灵活的治理新思路,成为探索人工智能善治路径的核心命题。Artificial intelligence is triggering an era of change far beyond the technical field,and the generative artificial intelligence represented by large models is shaping the main technological path in the future.Although China has become an artificial intelligence power,it still needs improvement in core theoretical research,basic infrastructure construction,key hardware-software and ecosystem,and there is still a long way to go for the country to become a strong artificial intelligence power.In this case,the key is how to achieve good governance of artificial intelligence while ensuring its strong,orderly and beneficial development.Therefore,balancing the relationship between innovation and governance,practicing the ethical idea of value alignment,learning from the best practice in international governance of artificial intelligence,and exploring a new governance approach with open ideas,diverse subjects,hierarchical objects,and fexible tools have become the core proposition of exploring the path of good governance of artificial intelligence.

关 键 词：生成式人工智能 新兴技术治理 价值对齐 敏捷治理 

分 类 号：F49[经济管理—产业经济] TP18[自动化与计算机技术—控制理论与控制工程]