Speaker
Description
Artificial intelligent (AI) applications and their fast pace of development have had tremendous impacts on almost every field in life, with no exception for education. People in general and language learners in particular have enjoyed enormous benefits from the popularity explosion of ChatGPT as an AI-powered tool. However, it is an undeniable fact that ChatGPT with its great leaps forwards is not always error-free in translating texts, resulting in occasionally questionable accuracy and reliability of its outputs. This paper presents initial findings of an investigation into different types of translation errors produced by ChatGPT in translating texts from English into Vietnamese and vice versa. Based on the theory of text types developed by Reiss, K. (1971) and Halliday, M.A.C (1994) and the framework of translation error typologies proposed by House, J. (1997) and Vinay, J. and Darbelnet, J. (1958), various text types in both English and Vietnamese are selected for translation, and an analysis of ChatGPT-generated translations is conducted to identify and categorize errors. By examining frequency and types of errors, the findings provide insights into limitations of ChatGPT as a translation tool. This study is intended to help Vietnamese learners, particularly translation students, in recognizing ChatGPT’s translation errors, enabling them to critically evaluate and refine AI-generated outputs in their own translation practices.
Biography
Hong Nguyen is a lecturer of English at Foreign Languages Department, Vinh University, with 20 years of professional teaching experience. She holds a master’s degree in English Applied Linguistics and has great research interests in Applied Linguistics, Translation, and Testing and Assessment.
