摘要
人工智能(AI)技术飞速发展给以创新性学术内容生产为核心的学术出版行业带来前所未有的风险与挑战.本文围绕学术出版中学术内容受AI应用干扰的情况,通过典型案例反映AI的学术贡献争议、学术常识错误、学术诚信缺失3方面的问题,进而分析探讨由AI干扰导致学术出版伦理出现的可控性、可靠性、公平性和安全性4类风险,最后从学术编辑的角色提出应对AI应用干扰进行学术性内容把关的"EDIT"(检测、判别、改善、教导)方案.以此推动AI时代背景下学术编辑职责的升级,直面学术出版内容污染风险,建立人机合作的治理模式.
Abstract
The rapidly developing artificial intelligence(AI)technology has brought unprecedented risks and challenges to academic publishing industry,which is centered on the production of innovative academic contents.This article focuses on the interfered academic contents by AI applications in academic publishing,and reflects the arising problems on contribution controversy,academic common sense error,and lack of academic integrity caused by AI through typical cases.Furthermore,we analyzes and discusses the risks in controllability,reliability,fairness,and safety arising from AI interference in academic publishing ethics.Finally,we proposes the"EDIT"(i.e.Examine,Discriminate,Improve,Teach)strategy for guarding academic contents from the perspective of academic editors in face of the interference of AI applications,aiming to promote the upgrade of responsibilities of academic editors in the AI era,and to establish a governance model for human-machine cooperation to address the risks of academic content pollution in academic publishing.
万方数据