国内外生成式人工智能新闻文本汉英翻译质量对比研究——以ChatGPT与Kimi为例
A Comparative Study on the Quality of Chinese-English News Text Translation by Domestic and Foreign Generative Artificial Intelligence—A Case Study of ChatGPT and Kimi

当前人工智能技术快速发展，给翻译领域带来前所未有的机遇与挑战。国外ChatGPT在通用文本翻译中表现出色，国内语言模型Kimi在处理大语言数据和实现多语种翻译方面也展现了强大能力，但二者在新闻文本等垂直领域的表现仍需探究。本研究选取中国日报发布的双语新闻文本作为语料，对比ChatGPT和Kimi生成译本与官方译本，通过文本分析探讨两者在新闻文本翻译中的质量。结果表明，ChatGPT在语言地道性和流畅性方面表现出色，而Kimi在处理中国特色文化元素方面表现更佳。基于此，本研究进一步提出人类译者在人机协作这一模式下需充分发挥其核心素养，以提升新闻翻译的整体质量。
The rapid development of artificial intelligence technology has brought unprecedented opportunities and challenges to the field of translation. While ChatGPT, an international model, demonstrates outstanding performance in general text translation, the domestic language model Kimi exhibits strong capabilities in processing large-scale language data and achieving multilingual translation. However, their performance in vertical domains such as news texts still requires further investigation. This study selects bilingual news texts published by China Daily as a corpus, conducting comparative analysis between translations generated by ChatGPT and Kimi against official reference texts to evaluate their quality in news text translation. The results indicate that ChatGPT excels in linguistic authenticity and fluency, whereas Kimi shows superior performance in handling Chinese-specific cultural elements. Based on these findings, this study proposes that human translators must fully leverage their core competencies within human-AI collaboration frameworks to enhance the overall quality of news translation.