Prompt engineering in the practice of a higher education teacher: Pedagogical potential and limitations

Abstract

In the context of the rapid introduction of generative language models into education, prompt engineering is of particular importance as a system of methods for formulating text instructions for AI in order to solve pedagogical problems. The purpose of the study presented in the article is to analyze the pedagogical potential and limitations of prompt engineering in the context of higher education, as well as to determine the directions of its effective use in the professional activity of a teacher. Key functional areas where prompt engineering can be effectively applied are identified: learning and teaching, methodological, research, project, educational, expert and analytical, communicative. Based on the analysis of theoretical sources and the results of empirical research, a classification of prompts has been developed based on the principles of their construction, with reference to didactic tasks. It is shown that prompts can be used not only to generate educational content, but also to organize reflection, develop critical thinking, and model professional situations. The barriers and risks limiting the implementation of prompt engineering are described: methodological, technological, personnel and organizational. It is noted that the successful integration of technology requires the formation of methodologically sound strategies for teachers, the development of digital competencies and legal regulation of interaction with AI. The presented conclusions were confirmed during the implementation of the program of additional professional education “Artificial intelligence tools and services in the work of a higher education teacher”. The conclusion is made about the high importance of prompt engineering as a component of digital pedagogy and the importance of developing institutional strategies for its implementation.