出版科学 ›› 2023, Vol. 31 ›› Issue (3): 94-.

• 多媒体 数字出版 • 上一篇    下一篇

数字出版领域智能语言模型的应用、风险与治理
—基于 ChatGPT 技术特征的分析

  

  1. 中南财经政法大学知识产权研究中心，武汉，430073
  • 出版日期:2023-05-15 发布日期:2023-06-08
  • 作者简介:任安麒，中南财经政法大学知识产权研究中心 2020 级博士生、助理研究员，德国马克斯 • 普朗克创 新与竞争研究所访问学者。
  • 基金资助:
    本文系高等学校学科创新引智计划“新时代科技革命与知识产权学科创新”（B18058）研究成果。

Application, Challenges and Governance of Intelligent Language Models in Digital
Publishing: An Analysis Based on ChatGPT Technology Features

  1. Center for Studies of Intellectual Property Rights， Zhongnan University of Economics and Law，
    Wuhan，430073
  • Online:2023-05-15 Published:2023-06-08

PDF

118

摘要：

以 ChatGPT 为代表的新一代智能语言模型具有算法独创性、能力综合性、应用广泛性、
技术局限性的特点，在出版领域将推动全新智能创作模式、引领智能出版转型升级、助力打造智
慧阅读空间，同时也将引发伦理危机、出版内容同质化、版权保护困境与侵权风险等问题。应探
索应对智能语言模型风险的数字出版治理路径 ：在理念上辩证把握新兴技术的工具属性与价值属
性，在政策立法上探索智能语言模型法制治理模式，在产业中实现数字出版智能升级与复合人才
培养，在技术上完成算法升级并完善相关配套措施。

关键词:

Abstract:

The new generation of intelligent language models represented by ChatGPT is
characterized by algorithmic originality, comprehensive capability, wide application and technical
limitations, which will promote a new intelligent creation model, lead the transformation and
upgrading of intelligent publishing and help build a smart reading space in the publishing fifield,
while also causing problems such as ethical crisis, homogenization of publishing content,
copyright protection dilemma and infringement risk. We should explore the digital publishing
governance paths to deal with the risk of intelligent language model: conceptually, we should
dialectically grasp the tool and value attributes of the emerging technology, explore the legal
governance model of intelligent language model in the policy and legislation, realize the intelligent
upgrading and compound talent training of digital publishing within the industry, complete the
algorithm upgrading and improve the related supporting measures at the same time.

Key words:

Copyright

中图分类号: 

版权所有 © 2019 《出版科学》     鄂ICP备05003330号-1
本系统由北京玛格泰克科技发展有限公司设计开发