Transforming Teaching: Prompting Techniques and Assessment Strategy for Learning with AI

This talk explores how generative AI tools, particularly LLMs, can be scaffolded to foster critical thinking rather than replace it, especially when supplemented with Wolfram Language. Using case studies in physics problem solving, we compare student-led versus model-led prompting strategies and analyze their impact on reasoning depth. We propose a framework aligning prompting strategies like chain-of-thought and ReAct with inquiry-based pedagogies and suggest assessment rubrics that center student–AI interaction as a site of learning.