The classroom is a product lab

Teaching gives me a very direct view of where an idea stops being usable.

I can write an assignment that feels perfectly clear at my desk. Then twenty people try to use it. The same question appears three different ways. Everyone pauses at the same step. Half the class produces something technically correct that still misses the point.

At that scale, repeated confusion is not an individual student problem. It is evidence about the system.

The classroom is full of this kind of evidence. Instructions have information architecture. Rubrics have hierarchy. Demonstrations have pacing. Feedback has an interface, even when the interface is a conversation. Every one of those things can make the next decision easier or harder.

That does not mean every difficult moment should be removed. Learning needs friction. The useful distinction is between productive difficulty and accidental difficulty.

Productive difficulty asks someone to compare two type choices, defend a hierarchy, or notice why an interaction feels wrong. Accidental difficulty asks them to decode what the assignment is requesting, remember where a file was posted, or translate a vague comment into an actionable revision.

I want to preserve the first kind and design away the second.

This is where many of my tools begin. Not with a feature idea, but with a sentence I have said too many times. If I keep explaining why a motion curve feels cheap, perhaps the explanation should become something people can manipulate. If critique keeps ending as a wall of advice, perhaps the feedback needs a clearer structure and a ranked place to start.

The best teaching artifacts externalize part of the instructor's thinking. They let someone compare, name, adjust, and try again without waiting for me to stand next to their screen.

That changes the test for whether the artifact works. A clever demo is not enough. I am looking for independence. Does the student ask a more specific question afterward? Can they explain why they changed something? Do they have a method they can reuse on the next project, when the example and the instructor are gone?

Building from the classroom also keeps product ideas honest. Students will ignore anything that takes too long to understand. They will find every missing instruction. They will use a tool in ways its tidy demo never anticipated. That can be uncomfortable, but it is extraordinarily useful.

Then the loop goes the other way. A tool built from repeated classroom friction returns to the classroom as a new way to practice. It becomes easier to see which part teaches and which part merely performs.

I do not think of teaching as separate from product work. It is where I get to watch language, systems, and interfaces meet real people every week. The classroom does not just produce assignments. Sometimes it reveals the product hiding inside the explanation.