Exploring the potential of generative AI and open source tools in the creation of a language learning game

Abstract

This is a project-based thesis that explores the potential of generative AI in the context of language learning. The result is a proof of concept of a language learning video game which allows the player to practice their conversation skills with an AI character via either text or speech. I begin the thesis by familiarizing myself with Stephen Krashen’s theory on language acquisition in order to understand the shortcomings of traditional language learning approaches. I then compare his theory with systematic research done on gamified solutions such as language learning games and apps. While there is very little published research when it comes to the use of generative AI in language learning, results from preliminary research are promising, especially in the context of promoting learning autonomy and reducing anxiety. In the result section of this thesis, I discuss the development of the video game. As the game is designed to help the player practice their Japanese conversation skills within a cozy anime environment, I resorted to using various open-source solutions in order to increase immersion as well as to supplement the player with various learning aids. I managed to complete all the milestones I had set out for this project and based on my research, I believe generative AI has the capability to address some of the shortcomings of traditional or gamified language learning solutions.