Are We Designing Better Lessons or AI-Proof Assignments?

As teachers, we understand what makes a good lesson.

  1. It’s engaging for students so they want to learn about it and do the work associated with it.

  2. It is something the students can understand. It’s not too hard or too easy.

  3. It is valuable to the students giving them a reason to be engaged.

  4. The instructions are clear so students aren’t confused about what to do.

We should be considering these things every day. But maybe our own motivation for doing these things has changed. Maybe it’s no longer about creating a good, engaging lesson, maybe now it’s more about preventing students from using AI to do their work.

I get that teachers are overwhelmed and overworked. Sometimes you just need a quick lesson. Sometimes you don’t have the time or the bandwidth to make the lesson as exciting as you might have. From that perspective, it’s a system problem. Classes are too full and teachers don’t have enough help to manage them well.

To put it bluntly, the crap rolls downhill. If the teacher is overwhelmed and is taking short cuts in their planning because classroom survival requires that, and we have a world where students can have AI do their homework for them, something is going to give. This isn’t a new problem. Students will find ways not to do their homework. AI just simplifies the problem of figuring out how to hand something in without doing much. Students have always been on the losing end of teachers trading off engagement for survival, long before AI. But AI makes the trade increasingly more of a gamble than simply the consequence of a system that overwhelms its teachers.

Teachers have to address the motivation issues, the instructional design problems, the somewhat under-planned, less than ideal lesson because they ran out of time, and also wrestle with the idea that if they don’t do all of that and students mentally check out, they are more likely to get an AI completed assignment.

Many of the solutions that I’ve offered in the past have intentionally changed the structure of teaching to move assignments back into the classroom to be done by hand on paper. Why? Because it doesn’t add any additional work. It doesn’t make the teacher try to do more in the same amount of time. It allows the teacher to control the environment, so that even if the lesson is not ideal, the student doesn’t get access to AI so they have no choice but to figure it out or not do it.

Let’s be honest, that’s not ideal either. The ideal solution is to have enough time and resources to give every student the attention and support they need. Even that doesn’t completely solve the problem, because even with all of that present, some students will still choose not to do the work.

We were never able to completely prevent cheating. All the rules in the world did not prevent some students from trying to cheat. I believe AI misuse will fall into that same pattern. It will never be a fully solvable problem. But can we do better than we are now?

If our only motivation for building better lessons is to prevent AI misuse, I think AI exposed something we didn’t want to admit: Teaching those better lessons is a lot of work and a lot of time that most teachers simply don’t have. That’s not an individual problem. That’s a system problem that we’ve been paying lip service to for a long time.

Yes, there have been plenty of laws limiting or creating targeted class sizes. Here in Manitoba, the target is 20 kids in Kindergarten to Grade 3 classrooms and 25 kids in grades 4 to 8, with no defined limits in high school.

While the school divisions are doing relatively well after having looked at their class size trackers*, the number of kids in the room is not the only determining factor for a teacher. How many needs are there in that same room? You can have only 5 students in a room, if just one of those students is bouncing off the walls, that room can be difficult to manage alone. We’re not talking about classrooms of 5 students either, we’re talking about 18, 20, or more students in a room, with potentially multiple students bouncing off the walls making it increasingly difficult to manage the classroom alone.

Are teachers the ones responsible for solving the AI cop-out problem through better planning?

No, of course not because in the current structure there is no way to consistently give them enough time to plan and give every student the needed level of individual attention they require. To fix the system purely from a ratio driven standpoint of teachers to students, the ratio would have to be considerably lower than 1:20. Even a ratio of 1:5 can be challenging depending on who is in the room, which makes fixing ratios alone an incomplete answer.

But I think there is a better question. Are we suddenly more concerned about the quality of lessons because students aren't learning, or because AI has made it harder to determine whether they're learning or not?

AI misuse brought lesson planning back into the discussion. That discussion seems to be coming from a couple of different directions. The first is driven by the question of whether to get technology out of classrooms or not. The second is driven largely by planning better in an attempt to gain more buy-in from the students themselves.

The first option of simply removing technology from the classroom and going back to a more old-school approach to teaching seems like the simplest option. It would be easy for many teachers to revert back to physical textbooks, pencils, and paper. Students hand write everything. They go to the library and research in physical books instead of online. The only technology in the school is a calculator. But did that environment do a good job of teaching students to think? Did that create an ideal world of intelligent thinkers? Would that actually fix the problem?

The second option requires teachers to solve a problem they didn’t fully create. The problem is largely systemic. So what’s the answer?

I believe we need to stop treating technology, education, thinking, and employment as competing goals. They all affect each other. It’s hard to do one without needing or affecting one or more of the other ones.

If they aren’t competing priorities and they all affect each other, then how do we balance them?

I think that’s the question that remains unanswered. A lot of the discussion moves in the extremes of needing to remove one to focus on another. I don’t believe that’s the right approach.

Teaching students to think and preparing them for the larger world means teaching technology and that technology is increasingly including generative AI.

How do we manage generative AI and still assess learning?

If I’m teaching ecosystems as a means of teaching the foundational skill of scientific inquiry, then I can’t simply test how much a student learned about ecosystems because that doesn’t tell me whether or not they understand scientific inquiry. AI will generate an ecosystem essay. It can tell me everything I ever wanted to know about ecosystems for a grade 7 assignment. That’s the thing we’re trying to avoid.

As the teacher, how do I assess the student's ability to conduct scientific inquiry rather than simply assess the product of that inquiry? This brings us back to a very simple question that teachers ask every time they plan anything. What method best teaches or assesses a given skill?

Sometimes the answer is pencil and paper. If we’re assessing scientific inquiry through ecosystems in a grade 7 classroom, then maybe having the students do the assessment with a pencil and paper is the best way to assess that learning. But if the path to that assessment is entirely driven by the content knowledge of the ecosystem and the student generates all of that work with AI, then we can’t be surprised when they can’t do an assessment that asks them for the scientific inquiry the assignments never actually required.

The foundational skill we’re trying to teach determines the method. Not all skills are best taught on paper anymore because technology and AI are needed sometimes. But not all skills are served by technology and AI either. Printing is a skill that kids need to learn with pencils and paper. We learn to multiply without a calculator. But then by high school, there are places where a calculator serves the objective better. AI and technology are no different.

But unlike the calculator, I don’t believe that students are best served by creating Kindergarten to grade 8 classrooms that are entirely technology free. I do believe in balance, but I also believe in the slow integration and increasing use of technology as we progress through the grades.

Technology, to me, as we move into the later grades, is like pen and paper. It is a tool, not a subject, and therefore it should be used as a tool. In Kindergarten to grade 3 classrooms, it is more of a separate intentional subject because most of the skills being taught are so foundational, those objectives are not necessarily best served on the computer. As we move into grades 4 to 6, technology is slowly becoming less of a separate subject and more of an intentional daily tool. By grade 8, technology is simply another tool for teachers to use.

All of this requires one thing: teacher training. Systemic issues such as the ones I’ve addressed, are not resolved by individual teachers alone. They are resolved by the system intentionally training its teachers to integrate technology and use AI well. That takes training and practice. It takes time.

The immediate fix is not to ban AI or remove technology from the classroom, it is to begin the process of learning how to teach in a technology rich, AI driven world. The foundational skill of teaching has not changed. What has changed is the introduction of a new method, which is technology based learning. That means deliberately designing instruction around the capabilities technology provides, not just simply using technology to perform familiar classroom tasks. Technology has changed what knowledge is important, what skills are needed, and how we teach all of it.

Until we get beyond the argument of all or none, I don’t think we’ll ever fully address the underlying issue or answer the question. It’s not a question of whether we teach technology or not. It’s a question of how and when we teach it. We answer that question by first understanding what we’re actually trying to teach and then figure out what method best serves that outcome. The question doesn’t have a black and white answer. There are many ways to teach an individual skill.

We need to stop debating whether teaching and learning have changed and instead start figuring out what we need to do differently. It starts with teacher training and support, access to technology, and a willingness to learn a new way of doing things.

*Class Size Trackers

Winnipeg One School Division

Louis Riel School Division

Pembina Trails School Division

River East Transcona School Division

St. James-Assiniboia School Division

Seven Oaks School Division

*Originally posted on Substack.


AI is used in my writing process to explore ideas, challenge assumptions, and assist with research and editing. The opinions, arguments, and conclusions are my own, and I take full responsibility for the final content.

Discuss...