Abstract:

This paper examines the integration of generative artificial intelligence (AI) in educational practices, with a specific focus on the optimization of prompt formulation for large language models to enhance the creation of teaching materials, streamline lesson preparation, and ease administrative tasks. The study is based on a professional development workshop designed for mathematics and science teachers in Slovenia, where eight key strategies for effective prompt engineering were introduced. The workshop involved 17 participants who generated prompts before and after the training, allowing for a comparative analysis to gauge improvements. The findings showed significant improvements in the precision, specificity, and audience targeting of the prompts, indicating that educators became more adept at crafting focused and relevant inputs for AI systems. However, certain challenges persisted, particularly in the areas of employing negative prompts and guiding the models to request necessary data for generating high-quality responses. The study offers valuable insights into how educators can leverage prompt engineering to improve lesson planning and content creation, contributing to a broader understanding of AI’s role in modern education. By highlighting both the successes and areas for improvement, this research reinforces the importance of equipping teachers with the skills needed to navigate AI-driven learning environments, ensuring more effective and adaptive teaching practices in the future.

Keywords:

Generative artificial intelligence, prompt engineering, teaching mathematics, educational technology.