How Grading the Chats Makes Learning Visible
Source: Mike Kentz Substack
Author: Mike Kentz
Original source: https://mikekentz.substack.com/p/how-grading-the-chats-makes-learning
Published: 2024-11-18
Source type: essay
Private backup: the full article text is archived in the private repository at archives/articles/mikekentz-substack-com-how-grading-the-chats-makes-learning.source.md. It is not published on the public Quartz site.
Summary
Mike Kentz presents “Grading the Chats” as the AI-era equivalent of “show your work.” Because AI can generate polished final products, he argues educators should assess the student’s process by examining how they prompt, question, evaluate, iterate, and refine during AI interactions. The article recommends designing large, ambiguous tasks, sharing model transcripts, giving process feedback, starting small, and having teachers first grade their own AI chats. Its central contribution is a process-based assessment model that treats AI chat transcripts as evidence of student thinking and AI literacy.
Pull quotes
Process over final product
“When we evaluate the process rather than the final product, we gain valuable insights into how students think, question, and solve problems.”
Show your work for AI
“By evaluating chat transcripts between students and AI, educators can observe how students interact with these tools, making their thinking visible in a way that wasn’t possible before.”
Big ideas
- AI literacy should help people notice how AI changes what counts as knowing
- Learning still needs some struggle, even when AI can make things easier
- AI literacy has to be taught inside real subjects
- Students need to bring the purpose; AI should not supply it for them
Claims
- AI chat transcripts can make student thinking visible
- In an AI world, assessment should focus on watching students think
- Prompting AI is a literacy practice, not just a technical trick
- AI literacy should teach students what to do with AI, not just what to think about it
Key evidence and examples
- The calculator analogy frames chat transcripts as the “show your work” equivalent for AI-assisted learning.
- Kentz names process categories such as prompt writing, breaking down problems, response analysis, iteration, and refinement.
- He recommends big, ambiguous tasks that AI cannot trivially complete and model transcripts that make successful interaction visible.
- Teachers are encouraged to first grade their own AI interactions before grading students’ chats.
Education relevance
Very relevant for assessment redesign, AI literacy, writing instruction, project-based learning, academic integrity, and teacher professional development.
Durability note
Kentz’s specific “Grading the Chats” framing is tied to current generative-AI classroom practice, but the durable idea is broader: when final products become easier to outsource, assessment has to recover evidence of process, judgment, and metacognition.