Surpass chatbots and human teachers? Explore the effectiveness of conversational AI in language learning

Abstract

Conversational Artificial Intelligence (ConvAI) has evolved from early rule-based systems and simple chatbots (e.g., Siri) to complex learning tools enabled by large language models, opening up new possibilities for language education. ConvAI is a type of AI that enables computers to understand, process, and respond to human language in a natural and human-like way. The introduction of ConvAI to language learning is especially timely and meaningful, as it overcomes the limitations of traditional learning methods, such as limited practical scenarios, mechanical or unengaging feedback, and high human costs. ConvAI systems, sometimes represented as avatar, can simulate real-life conversations, provide immediate correction, and offer adaptive learning options as well as context-aware feedback. This study evaluated the effectiveness of ConvAI in language education from two ConvAI-enabled language learning tools: Loora and Talkpal. Detailed differences between these two platforms are summarized. Moreover, the overall analysis reveals that ConvAI-enabled learning tools can break temporal and spatial barriers for language learning, offer highly personalized and adaptive learning experiences, provide vivid, immediate, and granular feedback to language learners, democratize learning costs, and ultimately enhance students’ grammatical accuracy and overall English proficiency. Our study also highlights key challenges in ConvAI-enabled language learning tools, such as limited depth, lack of human interaction, difficulties in fully replicating cultural and contextual nuances, and data privacy risks. Finally, we compare learning practices between human teachers and ConvAI avatar teachers. The comparison results arouse educators’ attention on effective Human-ConvAI collaboration in language learning.