Teaching the Machine to Teach: Toward a Human-Centered, AI-Forward University
A talk on the pedagogical harness around AI—knowledge graphs, win conditions, teaching rules, and feedback loops—and what it frees humans to do.
A talk delivered at the Marshall University Cabinet Retreat, July 13–14, 2026, as part of a leadership conversation about the university’s 2030 vision. It brings together two threads from my writing—Designing for the Machine Learner and The Win Condition Problem—into a working demonstration of how you actually build an AI that teaches. The through-line: the model is just the engine, and everything around it is the pedagogical harness that helps it learn.
The frameworks
The Board — Representing a knowledge domain as a machine-readable graph that maps concepts, skills, and how they relate, so an AI can reason about where a learner is and where they’re going.
The Win Conditions — Making success criteria explicit through AI-readable rubrics that capture what mastery looks like at each node. This is the talk version of The Win Condition Problem.
The Rules of the Game — Encoding teaching principles into the prompt, like answering with questions before answers, so the agent behaves like an instructor rather than an answer key.
The Feedback Loop — Continuous assessment across performance, behavioral, situational, contextual, and affective signals, which is what makes precision personalization possible.
Well-designed systems like these don’t replace the teacher. They free humans for the work only humans can do: demonstrating embodied skill, sharing lived experience, and breaking a learner’s frame when it needs breaking. The talk closes with a live prototype that puts the four pieces together with knowledge graphs, AI agents, and synthetic learners.
Author’s note: This essay was written with the help of generative AI, used as a thinking partner to explore framings, surface assumptions, and refine language. AI-generated outputs were treated as provisional material, not authoritative conclusions; all judgment and final decisions remain my own.