In this video, I connect three worlds:
How chunking supports student learning
How we teach it through Python variables and string formatting
And how AI models rely on chunking to function
You probably talk a lot about AI in the classroom: how to prompt it, how to use it, how it doesn’t think, but in actuality we don’t really talk about how it DOES mirrors human thinking. This video is about chunking: not just as a teaching strategy, but as a bridge between how humans learn and how AI models process information.
In cognitive science, chunking is how we manage complexity—breaking things into smaller parts so we don’t overwhelm our working memory. It’s how we remember phone numbers, take notes, and scaffold learning for students.
Now here’s the interesting part—AI models, do something similar. They process language in tokens and use parameters like stream=False to control how output is delivered. When that setting is off, the AI waits until it’s “done” to give you everything at once. No chunks, no partial thoughts—just a final, uninterrupted stream.
Sounds helpful, right? Sometimes. But just like students, AI doesn’t always do its best work when forced to hold too much at once.
There’s a connection here worth exploring: humans chunk to learn, and machines chunk to reason. When we teach students to notice that parallel, we are helping them understand not just how to prompt but how to think, structure, and communicate more clearly.
Share this post