Özet:

Artificial intelligence applications, which are gaining currency day by day due to their new features, in parallel with the development of technology, have attracted attention with the development of neural machine translation models, such as ChatGPT in the field of translation, especially in recent times, and have been regarded as a valuable phase and promising development in the translation world. With the development of ChatGPT, which is a neural machine translation model, “translation proofreading” has gained popularity in the translation industry and some changes have been made in translators’ job descriptions. Also, the question that has been a topic of discussion in the translation industry since the first years of the development of machine translation models has gained currency again: “Can machine translations replace translators?” The purpose of this study is to describe the quality of Arabic-Turkish translation outputs and the distance covered by artificial intelligence in the context of translation as a result of these developments. Artificial intelligence-based machine translation models, their development stages, and their most recent stage were included in the study, and then the ChatGPT Arabic-Turkish translation outputs were analyzed comparatively. In this study, the language of the source text has been determined to be Arabic and the language of the target text to be Turkish, and comparative analyses were carried out within the framework of ChatGPT translation outputs and human translations. Plain text, literary text, and technical texts, which were translated from Arabic into Turkish, were re-translated from Arabic into Turkish using ChatGPT, and the ChatGPT translation outputs were examined in the context of semantic and linguistic equivalence in general. The success rate achieved in the translation of different text types in the Arabic-Turkish language pair were evaluated.

Anahtar Kelimeler: