The Functions,Issues,and Optimization of General Artificial Intelligence in Administrative Decision-Making
Wu Hao1,2, Zheng Jun3
1. School of Sociology and Public Administration,East China University of Science and Technology,Shanghai 200237,China; 2. New Era National Security Research Center,East China University of Science and Technology,Shanghai 200237,China; 3. School of law,University of International Business and Economics,Beijing 100029,China
Abstract：General artificial intelligence has a significant promoting effect on the modernization of administrative decision-making. During the administrative decision-making process,general artificial intelligence can enhance information collection and analytical efficiency at the input end,capture key problem links at the decision end,and confirm optimized decision-making solutions at the output end,thereby assisting in achieving scientific,democratic,and legal decision-making. However,due to inherent weaknesses in iterative development,natural ethical defects,and practical efficiency limitations of general artificial intelligence,it brings a series of problems across different stages of information operations in administrative decision-making. These issues include algorithmic discrimination difficulties caused by the incomprehensiveness of information collection at the information input stage,technological dependence difficulties due to excessive technology embedding in decision-making at the information processing stage,and algorithmic black box difficulties caused by artificial settings or recognition differentiation at the information output stage. In this regard,it is necessary to optimize the assistance of general artificial intelligence in administrative decision-making,strengthen the management and supervision of information and algorithm deviations to prevent algorithm discrimination and promote algorithm fairness,clarify the auxiliary position and instrumental role of algorithms in the decision-making process to ensure decision-making security,and improve the regulatory formulation procedure and establish an algorithm disclosure system to eliminate the algorithm black box.
[1]余成峰.法律的死亡:人工智能时代的法律功能危机[J].华东政法大学学报,2018(2):5-20. [2]郭雪慧.人工智能在司法审判领域的困境、定位与展望[J].社会科学家,2021(11):129-134. [3]马长山.智能互联网时代的法律变革[J].法学研究,2018(4):20-38. [4]张书玉.论我国行政决策的现代化[J].甘肃社会科学,2017(3):226-230. [5]武俊娴.人工智能技术在道路交通管理中的应用[J].科技与创新,2023(15):115-118. [6]陈美伊.人工智能技术在医疗系统中的应用[J].电子技术,2023,52(9):313-315. [7]秦涛,杜尚恒,常元元,等.ChatGPT的工作原理、关键技术及未来发展趋势[J].西安交通大学学报,2024,58(1):1-12. [8]张宇.人工智能技术对用户决策行为的影响研究[J].无线互联科技,2020,17(11):167-168. [9]赵友华,蔡振华.人工智能时代行政决策体制优化的技术范式与风险因应[J].理论月刊,2020(2):60-68. [10]彭国甫.中国行政管理新探[M].长沙:湖南人民出版社,2008. [11]邓建鹏,朱怿成.ChatGPT模型的法律风险及应对之策[J].新疆师范大学学报(哲学社会科学版),2023,4(5):91-101,2. [12]ALJANABI M,GHAZI M,ALI A H,et al. ChatGPT:Open possibilities[J].Iraqi Journal for Computer Science and Mathematics,2023,4(1):62-64. [13]JIAO W X,WANG W X,HUANG J,et al. Is ChatGPT a good translator？A Preliminary Study[J].arXiv preprint arXiv:2301.08745,2023. [14]孟天广.智能治理:通用人工智能时代的治理命题[J].学海,2023(2):41-47. [15]刘淼.数字鸿沟:定义域的扩展与实践场景的分析[J].湖南师范大学社会科学学报,2023,52(4):135-142. [16]LUGER G R.人工智能复杂问题求解的结构和策略[M].郭茂祖,刘扬,玄萍,等译.北京:机械工业出版社,2017:486. [17]孙伟平.价值哲学视域中的算法歧视与社会公正[J].哲学研究,2023(3):46-55. [18]申海波,韩璞庚.人工智能背景下的治理变迁及其路径选择[J].求索,2018(6):74-81. [19]成龙,王楠.数字异化:表象、根源及其破解途径[J].中国地质大学学报(社会科学版),2022,22(6):1-11. [20]陈峰,王利荣.个人信息“知情同意权”的功能检视与完善进路[J].广西社会科学,2021(8):106-111. [21]宋亚辉,陈荣昌.整体治理:算法风险治理的优化路径[J].学习与实践,2022(7):35-45. [22]石颖.算法歧视的发生逻辑与法律规制[J].理论探索,2022(3):122-128. [23]蔡玮莎.发展智慧养老缩小数字鸿沟[J].中国电信业,2020(12):49-51. [24]颜佳华.提升政府治理算法决策公平性的机制与路径[J].行政论坛,2022(3):34-40. [25]DE LEAT P B. Big data and algorithmic decision-making[J].Acm Sigcas Computers & Society,2017(47):39. [26]昌诚,张毅,王启飞.面向公共价值创造的算法治理与算法规制[J].中国行政管理,2022(10):12-20. [27]唐亦可.论人工智能时代算法的法律规制[J].时代法学,2022(2):99-106.
2025, 
