DROPS
20 July 2026
FORMAT
Asynchronous
POSITION
05 / 05
SERIES
AI + Practice
WORKSHOP 05
AI and Discussion Design
Use AI to design discussion structures that ask every student in the room to think.
The seminar where the same five kids talked while the rest watched. By the end of the workshop, you have discussion structures that ask every student in the room to do the thinking, not just the five who like to talk.
01
THE CURRENT
What discussion usually looks like by week six.
You designed a great Socratic seminar. The five strong students sat at the inner circle, the rest of the class watched, and the conversation was thoughtful. You felt good about it, until you graded the participation rubric and realized you had assessed the same five kids that always get assessed for discussion. The other twenty performed their attendance, and you did not know what any of them had thought.
By week six, "discussion" in your classroom has quietly become the rotation of voices that were always going to volunteer.
The work this workshop addresses is designing discussion structures that ask every student to do the thinking, not just the five who like to talk.
REF · HATTIE
Classroom discussion ranks among the higher-effect-size practices in his meta-analysis. The variance in effect lives in design.
REF · COE
Learning happens when people have to think hard. The structure of the discussion decides who has to think hard.
ASIDE
Performance of participation is its own genre of literacy.
02
THE REFRAME
Discussion as the structure that makes the thinking happen.
When discussion works, every student in the room is doing cognitive work, even the ones not speaking. The deliverable is a discussion architecture where the structure of the conversation requires thinking from everyone, not just airtime from a few. The teacher's pedagogy decides what kind of thinking the discussion is building toward.
AI's default failure on discussion is generating discussion questions. Fluent prompts, plausible, often well-phrased, and almost always structured for the same five kids to answer. The questions are not the architecture.
What AI can do with your pedagogy underneath is help design the architecture itself, the structure of who thinks about what, in what order, with what evidence, in response to what. The design happens with your knowledge of your students in the lead. AI handles the structural calculus that would otherwise eat the planning period.
DEF · DISCUSSION ARCHITECTURE
The structural design of who is doing what kind of thinking in response to what, independent of who speaks aloud. The architecture decides whether the discussion is building thinking or performing it.
REF · SPECTRUM
Eight analytical lenses students learn to apply across texts. In discussion, the lens choice routes who is thinking about what, so the same passage opens up eight ways at once.
REF · ATWELL
Discussion as conferring, one-to-one or small. The architecture is not always whole-class.
SEE ALSO
Fig. 5, lower-right. The cell this workshop is for.
03
BY AUGUST
What you walk into the year with.
You have a discussion architecture that asks every student in the room to think. You know what kinds of cognitive work each discussion structure is doing, what variation across discussion types fits your texts, and how the pattern of discussion design across the year is teaching students to think in conversation with other readers.
You also have an AI that knows your discussion pedagogy. Not a generic question generator. An assistant that knows what kind of thinking you want the discussion to build, what your texts ask of student readers, and what structural variation fits your classroom. The articulation work you did in the workshop becomes context the AI carries every time you sit down to plan a discussion.
NOTE
The architecture is yours. The AI is a tool that holds it.
ASIDE
The students who do not speak in discussion are often the ones doing the most thinking. The architecture's job is to honor that without leaving them invisible.
04
INSIDE
The design work and what supports it.
You get audio that works through the pedagogy underneath discussion, where most discussion systems break down, and the reframe the rest of the workshop builds on. A structured prompt set walks you through the design work, with each prompt scaffolding a stage of the system you are building. Walkthroughs show how the pieces come together so you can see the design moves in action. You also build tools along the way that hold the architecture of your design and let you see whether the system is doing the teaching work you designed it to do.
The workshop does not stop at the drop. As you start implementing and questions come up, you send them my way, and I answer them in audio form on a private podcast feed that goes only to workshop buyers. The feed stays in your library alongside the rest of the kit.
Part of the workshop is teaching you to get AI output you actually want to use. That happens by helping your AI understand what you value about discussion, including your pedagogy, your standards, and the way you think about this practice. The AI then functions as a specialized assistant for this work rather than a generic one. The same articulation work transfers to any other practice you want to train AI for after this workshop.
You will work across Gemini, NotebookLM, and Claude or ChatGPT, with a curated research base underneath the design work.
SHAPE
Audio. Prompt set. Walkthroughs. Tools you build. The articulation work for your AI. A curated research base underneath. Private podcast Q&A after the drop.
PLATFORMS
Gemini · NotebookLM · Claude or ChatGPT.
PREREQ
None. Comfort with a few AI platforms is enough. The workshop teaches the rest.
05
FOR MEMBERS
Take the discussion work further into the year.
Inside Teaching for Today, the discussion work extends past the workshop into your year-long workflow. Membership infrastructure tracks what kinds of thinking each student has been asked to do in discussion across the year, where the architecture surfaces some voices and not others, and where the design might need adjustment. The same infrastructure runs the student-facing communication that helps each student see the kind of thinker they are becoming in discussion.
SCOPE
The workshop is one-time. The membership is the running layer underneath: the tracking, the communication, the longitudinal view.
NOTE
Current members get the bundle of all five workshops at the price of a single one.
06
TAKE IT
Take this workshop, or take it inside a bundle.
This workshop runs $49 on its own. If you are thinking about more than one, the bundle options on the series page give a per-workshop discount that grows the further you go. The System Pair runs $89. The Responsive Trio runs $129. The full five-workshop series runs $199.
PRICING TABLE
1 · $49
2 · $89 (System Pair)
3 · $129 (Responsive Trio)
5 · $199 (Full Series)
DELIVERY
One-time purchase. Lifetime access. Asynchronous from drop day onward.
07
WHO IT IS FOR
This workshop was built for and by English teachers. The discussion design assumes the work students do involves discussing literature, argument, evidence, interpretation, and that the thinking you want them doing in discussion is the kind of thinking that grows readers.
A serious second-year English teacher belongs here. A veteran biology teacher does not.
COLOPHON
AI + Practice · Workshop 05 of 5. Built by an English teacher reacting to discussion pedagogy that confuses airtime with thinking.
- Customer
- Payment
Contact information
Payment information



AI and Pedagogy Series - Discussion Design$49
- Total payment
- 1xAI and Pedagogy Series - Discussion Design$49-+
- Discount
- Shipping$0
- Sales tax$0
- Total
- Today's payment
- AI and Pedagogy Series - Discussion Design$0
- Discount
- Shipping$0
- Sales tax$0
- Total
- Future payments
- AI and Pedagogy Series - Discussion Design$49
- Discount$0
- Shipping$0
- Sales tax$0
- Future amount
$0 - Today's payment:
$0 - Today's payment
-
$0
All prices in USD
This site is protected by hCaptcha and its Privacy Policy and Terms of Service apply.