Application Risks and Coping Strategies of Major Language Models Such as Chatgpt

With the development of deep learning technology, large language models represented by ChatGPT have been rapidly applied in the field of education. Its penetration rate in artificial intelligence teaching scenarios is leading. While it empowers teaching through intelligent Q&A, homework correction and other functions, it also raises issues such as technical reliability, teaching essence alienation, and blurred ethical boundaries. However, existing research lacks systematic exploration of potential risks in the field of artificial intelligence teaching. This study adopts literature analysis and case study methods, and finds that there are risks of technical limitations, teaching effect alienation, and ethical and moral misconduct in the use of big language models in artificial intelligence teaching. The study proposes to construct a three-dimensional response framework of “technology teaching ethics”, including optimizing model technical performance, innovating teaching mode design, and establishing an ethical governance system, providing theoretical and practical references for promoting the safe and effective application of big language models in artificial intelligence teaching.