Generative AI technology and language learning : global language learners’ responses to ChatGPT videos in social media

Abstract

Generative AI technology has great potential for language learning and teaching. Since the launch of ChatGPT at the end of 2022, its digital affordance on language learners remains to be determined. This sequential mixed-method study aimed to detect the emotions in language users’ self-reflection of their learning experiences with ChatGPT. This study analysed commentary data (n = 1088) from 19 most widely discussed YouTube videos on using ChatGPT for language learning between January 2023 and April 2023. A coding system was developed based on the visualisation tool WordCloud and human analysis results to identify the patterns in these comments. The key findings of this study highlight five crucial digital affordances of ChatGPT that enhance language learning: (1) opportunities to practice language, (2) individualised feedback, (3) engagement and interaction, (4) authentic input of language use, and (5) real-time assistance on linguistic knowledge. Surprisingly, with the assistance of the sentiment analysis tool and hierarchical linear regression, results identified two significant predictors of less positive emotions in language learning experience with ChatGPT: (1) Opportunities to practice language and (2) Individualised feedback. The implications of digital affordances in ChatGPT assisted language learning are discussed.