摘要
以ChatGPT为代表的生成式人工智能促动信息秩序的巨大变局,重新分配社会信息资源和数字权力。数据喂养的偏狭、使用的偏误和伦理的偏颇为生成式人工智能输出的虚假信息、错误信息及恶意信息流入信道创造条件,现有信息秩序面临失序风险。调整信息治理范式与治理方向成为智能传播格局下AI治理的题中应有之义。本文提出实现信息秩序再次和谐的共创进路与治理方向,即以法律为约束,构建软硬法兼施的复合型AI信息治理法律框架,将法治贯彻到生成式人工智能信息治理的全链条多领域,全面提升法律规制效能;以内外兼顾的技术善治实现技术真正服务于社会价值和公共利益的图景,促动技术的自我纠偏与完善;构建生成式人工智能流域化治理模式以直面多元治理难题,将多元数字行动主体纳入差异化规制与协同性治理体系以实现信息秩序的复建。
The monumental upheaval in the information environment propelled by generative artificial intelligence,exemplified by ChatGPT,entails a reconfiguration of societal information resources and digital authority.Biases in data input,errors in usage,and ethical imbalances contribute to the generation of false,erroneous,and malicious information by generative AI,facilitating their ingress into information channels.Consequently,the existing information environment faces the risk of disorder.Adapting the paradigms and trajectories of information governance become a pertinent pursuit within the realm of AI governance amid the context of intelligent communication patterns.This paper posits a collaborative pathway and governance approach to restore information harmony.It proposes leveraging legal constraints to construct a hybrid AI information governance legal framework,integrating both flexible and robust legal measures.This legal framework extends the rule of law across the complete spectrum of generative AI information governance in diverse domains,comprehensively enhancing the efficacy of legal regulations.Concurrently,it underscores a technologically virtuous approach that balances internal and external perspectives,thereby ensuring technology genuinely serves societal values and the common good.This approach fosters technology's self-correction and refinement.Moreover,it advocates for the creation of a compartmentalized governance model tailored for the domain of generative artificial intelligence.This model confronts the challenges of diversified governance by incorporating diverse digital actors into a system of nuanced regulations and collaborative governance.This inclusive approach aims to reconstruct the information environment and realize a coherent and harmonized information landscape.
作者
张文祥
沈天健
孙熙遥
Zhang Wenxiang;Shen Tianjian;Sun Xiyao(NIT University;school of Cyber Science&Technology,Zhejiang University;Institute of Journalism and Communication,Shandong University;Faculty of Innovation and Design,City University of Macao)
出处
《新闻界》
CSSCI
北大核心
2023年第10期41-51,共11页
Journalism and Mass Communication
基金
国家社科基金重大项目“健全互联网领导和管理体制研究”(22ZDA079)。
关键词
生成式人工智能
ChatGPT
信息秩序
智能传播
信息治理
generative artificial intelligence
ChatGPT
information environment
intelligent communication
information governance
