Abstract: Artificial intelligence (AI) is evolving in a way that makes the distinctions between certain fields of application fuzzier and increases its capacity to be applied in a variety of contexts. A significant step in this direction has been made with the public's availability of ChatGPT, a large languages model (LLM)-powered generative AI chatbot. Professionals expect this technology will impact education and the job of teachers as a result. Nevertheless, how instructors might really utilize the technology & the nature of its connection with teachers remain understudied, given some presumptions regarding its influence on education. In this study, ChatGPT's interaction with instructors was investigated with an emphasis on understanding the complimentary functions that each play in education. Over the course of two weeks, ChatGPT was requested to be used by eleven language instructors. They subsequently shared interaction logs created during their usage of the device and took part in individual interviews to discuss their experiences. Four ChatGPT roles—interlocutor, content provider, teaching assistant, and evaluator—and three teacher roles—orchestrating various resources with quality pedagogical decisions, empowering students as active investigators, and fostering AI ethical awareness—were identified through qualitative analysis of the data. The potential application of LLM-powered chatbots for educational purposes is also discussed, along with its implications.

Keywords: ChatGPT, Large Language Model, Chatbot, Artificial Intelligence, AIEd, Human-Computer Interaction, Large Language Models-Powered Chatbot

| DOI: 10.17148/IARJSET.2023.10807