生成式AI时代机器翻译译后编辑中的政治意识培养?  

Enhancing Political Discernment in the Age of Generative Al:Implications for Machine Translation Post-Editing

作　　者：高玉霞[1] GAO Yuxia(Ocean University of China)

机构地区：[1]中国海洋大学

出　　处：《翻译界》2024年第2期18-30,共13页Translation Horizons

基　　金：“中央高校基本科研业务费专项”(202413011);广东省人文社会科学重点研究基地广东外语外贸大学翻译学研究中心招标课题“面向大语言模型应用的国家翻译实践能力建设路径研究”(CTS202313)阶段性成果。

摘　　要：在生成式人工智能兴起的当下,ChatGPT、DeepL等机器翻译工具正在重塑语言服务行业的格局。这些工具不仅带来效率的飞跃,也引发了人们对翻译技术应用和伦理问题的广泛关注。机器翻译结果可能含有不易察觉的政治偏向,这为使用者带来了不容忽视的政治风险。翻译从业者以及翻译教育领域的学生对于机器翻译的政治倾向存在认知偏差。本研究旨在探讨如何在机器翻译译后编辑中培养政治意识,以识别和规避这些风险。文章建议通过增强翻译技术教育、提升风险意识、构建具有政治敏锐性的师资队伍等方式,来强化翻译人员的政治意识。In the burgeoning era of generative artificial intelligence,machine translation tools such as ChatGPT and DeepL are reshaping the language industry.These innovations offer leaps in efficiency and have simultaneously ignited widespread concern over the application of translation technology and its ethical implications.Machine translation outputs may harbor subtle political biases,posing significant political risks that cannot be overlooked.This risk is particularly pertinent to translation practitioners and students in the field of translation education,who may harbor cognitive biases regarding the political leanings of machine translation.This study seeks to explore the cultivation of political awareness in post-editing machine translation to identify and mitigate these risks.The paper proposes enhancing translation technology education and raising awareness of potential risks,as well as building a politically sensitive faculty,as means to strengthen the political acumen of translation professionals.

关 键 词：生成式AI ChatGPT 机器翻译 译后编辑 政治意识 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] H085[自动化与计算机技术—控制科学与工程] H315.9[语言文字—语言学]