Ask AI to test each lesson against two questions: does it have one clear learning outcome, and can a student apply it independently? If a lesson fails both tests, it needs to be split or merged — AI can tell you which.
The Over-Stuffed Lesson Problem
Educators who know their subject deeply often build lessons that cover too much. Not because they’re disorganized — because they can see how everything connects. The problem is students can’t. When a lesson tries to accomplish three things at once, students leave having absorbed one of them, half-understood the second, and forgotten the third entirely.
The flip side also happens: lessons that are so granular they don’t stand alone. A five-minute explanation of one concept that only makes sense in context of the previous lesson isn’t really a lesson — it’s a subsection. Combining it with what comes before or after would actually make it stronger.
How AI Evaluates Your Lesson Structure
Paste your lesson titles and brief descriptions into Claude and ask: “Which of these lessons are trying to do too much and should be split? Which are too thin to stand alone and should be merged with an adjacent lesson? Give me a recommendation for each.” Claude will work through your list and flag the ones that need attention.
You can also give AI the actual lesson content and ask it to identify the core learning outcome. If it struggles to name one clear outcome, the lesson is probably over-stuffed. If the outcome it names is too narrow to justify a standalone lesson, it’s a candidate for merging. This test takes about two minutes per lesson and consistently surfaces structural problems that would take students weeks to complain about.
What This Means for Educators
Well-sized lessons matter more in online teaching than in any other format. In a classroom, you can read the room and adjust on the fly. In an online course, once the lesson is recorded or the session is planned, you’re committed. Getting the lesson size right before you build saves you re-recording, restructuring, and the uncomfortable realization that your course has a pace problem halfway through a cohort run.
The Simple Rule
One lesson, one outcome, one application. If you can’t name what a student will be able to do after this specific lesson — not after the module, after this lesson — it’s probably not sized right. Use AI to audit your lesson list before you build, not after. Five minutes of structural review with Claude before you record will save you five hours of restructuring later.