How to Model Effective AI Use in Classrooms

Source: Mike Kentz Substack
Author: Mike Kentz
Original source: https://mikekentz.substack.com/p/how-to-model-effective-ai-use-in

Private backup: the full article text is archived in the private repository at archives/articles/mikekentz-substack-com-how-to-model-effective-ai-use-in.source.md. It is not published on the public Quartz site.

Summary

Mike Kentz argues that prompt-engineering acronyms and starter frameworks are insufficient because effective AI use is context-specific, iterative, and language-based. He proposes teaching AI interaction through writing pedagogy: students compare stronger and weaker chat transcripts, annotate them, discuss what makes one better, and co-create criteria for better use. This treats AI chats as a new genre of text rather than a technical trick. Kentz also notes unresolved questions about transcript design, facilitation, transfer beyond humanities, multimodal systems, and agentic AI.

Big ideas

Claims

Key evidence and examples

  • Kentz compares prompt acronyms to sentence stems: useful starts, but not foundations for skilled communication.
  • He adapts writing-class routines—comparison, annotation, discussion, voting, and rubric-building—to AI chat transcripts.
  • Students identify stronger AI use by noticing specificity, context, nuance, iteration, and the quality of follow-up moves.
  • The article reports versions tested across middle school, high school, college, and a Grade 12 classroom pilot.

Education relevance

Very relevant for AI literacy, writing pedagogy, and teacher professional learning because it gives educators a classroom routine for moving beyond prompt sheets toward visible, discussable AI-use practice.

My notes