Connected Classroom · Workshop
Socratic Dialogue with AI: Restoring Critical Thinking in the Classroom
AI processes fast. Your students know things it never will. The Dialogic Learning Model is a workshop for designing lessons that put those two facts to work together.
Book This WorkshopThe Inversion Crisis
When Anthropic studied how university students actually use its AI, it found an inverted pattern of Bloom's taxonomy. The machine was doing the higher-order work. Creating and analyzing, the two operations that sit at the top of the hierarchy, made up most of what students handed over. Remembering and understanding, the foundation everything above depends on, were rare.
Source: Anthropic Education Report, How University Students Use Claude (2025).
This is the inversion. Students are handing the top of the pyramid to a machine before they have built the base for themselves. They are trying to put a roof on a house that has no foundation. The critical thinking a lesson is meant to build is the exact thing the tool is quietly removing.
The response is not to ban the tool. A ban only moves the behavior out of view. The response is to change the task, so that finishing it requires the student to think in a way the AI cannot do for them.
The Dialogic Learning Model
The Dialogic Learning Model is a way of designing lessons around what only the student can bring. Their lived experience, their local context, their own reading of a problem. An AI has none of it, so the task cannot be completed by prompt alone. The student has to do the thinking, because the thinking is the part that is theirs.
The move is small and repeatable. Take a task the AI can finish, and redesign it so the AI's output becomes the raw material the student has to interrogate, not the answer they submit.
| Outsourced to the AI | Returned to the student |
|---|---|
| Summarize the themes of the novel. | Test the AI's summary against a moment from your own life it could not know about, and mark exactly where it falls short. |
| Explain the causes of the event. | Name what the AI's account leaves out about your community, and argue why that omission changes the conclusion. |
| Generate three solutions to the problem. | Interrogate the AI's three solutions, then defend a fourth that only someone with your context would propose. |
The workshop teaches this through a protocol teachers can run the next morning: See, Think, Wonder, applied to AI output. Students see what the model produced, think about what it assumes and misses, and wonder about what it could not have known. The tool stays in the room. The thinking comes back to the student.
Preview the Workshop
This is the full workshop deck, the same one presented to hundreds of educators. It is free to view and free to share.
How Creative AI Dialogue Can Restore Critical Thinking, by Connected Classroom
Inside the Workshop
The deck shows the argument. The workshop is where a teaching team builds it into their own lessons. Sessions run in person or online, for a department, a whole staff, or a conference audience, and they are worked, not watched. Teachers leave with lessons they can teach that week.
- The Dialogic Workbook. The implementation guide that walks a team through the Inversion Crisis data and the full Dialogic Learning Model, step by step. Worked through together during the session and provided to participating teams.
- The Dialogic Thinking Prompts. A ready-to-use set for interrogating AI output and forcing students to surface what only they know. Built for immediate classroom use.
- Live task redesign. Facilitated See, Think, Wonder on the real assignments your teachers already use, so the model leaves the room as lessons rather than theory.
Bring This Workshop to Your School
Tell me your context, your grade bands, and what you want teachers to walk away able to do. I will tailor the session to it. Available for schools, districts, and conferences, in person or online.
Book a WorkshopGo Deeper
The workshop sits on top of a larger body of work on AI, cognition, and the classroom.
- Protecting the Developing Mind The research pillar on cognitive privacy, and what surveillance and AI do to the conditions children learn inside.
- The Cognitive Offloading Crisis Why students stop thinking for themselves when AI does the work, and what that costs a mind still under construction.
- Making Human Capability Visible in AI Assessment Designing assessment that shows what a student can do that an AI cannot.
- AI Memory and the Barnum Effect How personalized AI can flatter a student with generic output they mistake for real insight.
- Connected Classroom The full body of work on AI, cognition, and human capability in the classroom.