Perception of Generative AI Use for Japanese Speaking among Indonesian Workers in Japan

Abstract

This study aims to determine the perception of Indonesian workers in Japan towards the use of generative Artificial Intelligence (AI) as a medium for practicing speaking Japanese. The background of this research is from the development of Society 5.0 technology in Japan which emphasizes the integration between the physical and digital worlds, including in the field of language learning. Generative AI, such as ChatGPT and similar applications, offers great potential in supporting interactive, adaptive, and contextual speaking exercises without time or place limitations. This study used a survey method with the distribution of online questionnaires to a number of Indonesian workers in various sectors in Japan. The data obtained were analyzed descriptively quantitatively to identify the level of acceptance, perceived benefits, and obstacles faced in the use of AI as a learning medium. The results of the study based on PLS-SEM showed that knowledge and use of generative AI had the most dominant effect on the use of generative AI (β = 0.698; t = 10.234), difficulty speaking Japanese (β = 0.329; t = 3.652), confidence, exposure to speech and demographics was not significant, and the predictive ability of the model was moderately strong (R² = 0.635). The use of generative AI as a medium for Japanese speaking practice in Indonesian workers in Japan is determined by AI literacy and the level of difficulty in speaking, not psychological or demographic factors. This research is useful for identifying the advantages and disadvantages of using AI so that it helps improve the Japanese language communication skills of Indonesian workers in the workplace.

Ähnliche Arbeiten