AI can draft a full lesson in seconds. That speed is real, and it is also the trap: the faster the draft arrives, the easier it is to stop thinking. The goal of an AI-facilitated lesson is not to hand the work to a model — it is to spend your time on the decisions only a teacher can make, and let the model handle the parts that are mechanical.
Here is the framework we teach in the Instructional Design Bootcamp.
Start with the outcome, not the prompt
Before you open any AI tool, write the single sentence that describes what a student should be able to do by the end of the lesson. If you cannot state the outcome in one sentence, no amount of generated content will save the lesson.
The model is excellent at expanding a clear outcome into objectives, activities, and checks for understanding. It is terrible at deciding what the outcome should be — that requires knowing your students, your standards, and your constraints.
Let AI draft the scaffolding
Once the outcome is fixed, this is where AI earns its keep. Ask it to generate:
- three or four measurable objectives aligned to the outcome
- a warm-up that surfaces prior knowledge
- a worked example and a guided-practice set
- two formative checks at different difficulty levels
Treat all of it as a first draft from an eager but inexperienced teaching assistant. It is raw material, not a plan.
Keep the judgment human
This is the step that separates AI-facilitated teaching from AI-replaced teaching. Read the draft and make the calls the model cannot:
- Which example will actually land with this class?
- Where will students predictably get stuck, and is there a scaffold there?
- What did the model make sound rigorous that is really just busywork?
Cut, reorder, and rewrite. The draft should leave the lesson, on average, better than blank — but never untouched.
Close the loop after you teach
The best lesson-planning input you have is what happened last time. After the lesson, jot two lines: what worked, what flopped. Feed those notes back in next time so each iteration compounds. Over a semester this is what turns a generic generated plan into something genuinely yours.
AI facilitated. Human led. That order is the whole point.