AI-Assisted Oral English Teaching: A Review of Generative Large Language Models
TL;DR
It is found that GenAI has seen broad applications in other ways before, during, and after class, and its users speak fluently and readily to themselves, but still suffer from language bias, content accuracy, and learning bias.
Abstract
As generative large language models like ChatGPT are developing, their applications have become increasingly widespread and there are new technical directions for traditional English oral language teaching. However, studies have focused on writing and grammar and their systematic review of oral English teaching is still lacking. This article discusses generative large language (LLMs) models in English oral linguistic teaching, with a particular emphasis on their usage in different teaching situations and their impact on learning results. The article also compares development directions and differences between research in countries and outside the world. It is found that GenAI has seen broad applications in other ways before, during, and after class, and its users speak fluently and readily to themselves, but still suffer from language bias, content accuracy, and learning bias. Based on this, the paper suggests future development directions from the perspective of technological optimization and teaching integration in order to provide a systematic guideline for empowering English oral Language teaching with artificial intelligence.