摘要
数字政府建设的范式革新与技术变迁紧密相关。传统人工智能参与政府治理存在诸多短板,主要体现在数据质量和开放共享有待优化、算法透明度和解释性有待增强、行政决策和执行效率有待提升几个方面。行政领域人工智能通用大模型(GovGPT)为交互性敏捷治理的实现提供了技术支持,通过建构行政领域的人工智能通用大模型,得以克服传统人工智能参与行政治理的诸多难题,从而促进数字政府建设走向交互式与敏捷性的发展方向。在此基础上,首先要树立“数智服务”的行政生态观,其次完善基于GovGPT的交互性敏捷治理的基础机制,最后防范GovGPT大模型可能带来的一系列风险,将人工智能参与数字建设纳入法治化轨道。
The paradigm innovation of digital government construction is closely related to technological changes.There are still many shortcomings in traditional AI participation in government governance,mainly reflected in data quality and open sharing to be optimized,algorithmic transparency and interpretation to be enhanced,and administrative decision-making and implementation efficiency to be improved.The general large mode of artificial intelligence in the administrative domain(GovGPT)provides technological support for achieving interactive agile governance,the construction of general model of artificial intelligence in the administrative domain helps solve various problems associated with traditional AI participation in administrative governance,thereby promoting the development of digital government construction towards interactivity and agility.On this basis,firstly,we should have an administrative ecological perspective of“digital intelligence service”;secondly,we should improve the basic mechanism of interactive agile governance based on GovGPT;and finally,we should prevent a series of risks possibly brought about by the large model of GovGPT,and guide the participation of artificial intelligence in digital construction into the track of the rule of law.
作者
白文静
Bai Wenjing(School of Marxism,Inner Mongolia Normal University,Hohhot 010022,China)
出处
《西北民族大学学报(哲学社会科学版)》
北大核心
2024年第3期91-102,共12页
Journal of Northwest Minzu University(Philosophy and Social Sciences)
关键词
生成式人工智能
数字政府
通用大模型
交互性敏捷治理
generative artificial intelligence
digital government
general large model:interactive agile governance