Small Design Changes That Shift Everything

In my last post, I explored a question that I think higher education needs to wrestle with more directly: What counts as evidence of learning when AI can generate increasingly sophisticated outputs?

Once we start asking that question, another one quickly follows: What do we actually need to change?

For many faculty, that question can feel overwhelming. Discussions about AI often sound like they require complete course redesigns, entirely new assessment systems, and a fundamental rethinking of teaching itself. I don't think that's true.

While AI is certainly creating pressure for change, meaningful improvement often starts much smaller than we think. In many cases, a few intentional design decisions can significantly improve how students engage with learning in an AI-rich environment.

Part of the reason is that AI didn't create many of the challenges we're talking about today. It simply exposed them more quickly. Long before generative AI existed, students were already looking for efficient ways to complete coursework. Many assignments could be finished through memorization, formula-following, or surface-level completion. Students have always balanced competing priorities and searched for the most efficient path to success.

AI didn't invent those behaviors. It just made them more visible. That's why I believe the goal shouldn't be to create "AI-proof" assignments.

I'm not convinced those really exist.

Instead, we should focus on designing learning experiences where student thinking becomes more visible. Sometimes that means asking students to explain why they chose a particular solution instead of only submitting an answer. Sometimes it means asking them to connect course concepts to a class discussion, lab activity, workplace experience, or current event. Sometimes it means including a brief reflection on what they found difficult, what changed during their process, or what they would do differently next time.

Other times it means shifting from generating answers to evaluating them. Students might compare competing solutions, critique recommendations, identify weaknesses in an argument, troubleshoot a problem, or justify a decision based on available evidence. These may seem like small adjustments, but they change where the learning happens.

The focus moves away from producing an answer and toward making sense of information. And that distinction matters. As AI becomes increasingly capable of generating content, the value of education is less about producing information and more about interpretation, judgment, reasoning, decision-making, and reflection.

In my own courses, some of the most effective learning moments occur when students are diagnosing problems, troubleshooting unexpected results, making decisions with incomplete information, collaborating with others, or explaining their reasoning in real time.

Those activities aren't necessarily immune to AI assistance. Students can still use AI as a tool within those experiences. But the learning doesn't reside solely in the final product.

It resides in the thinking. And when we design activities that make that thinking visible, we gain a much clearer picture of what students actually understand.

But there is another benefit to making thinking visible that I don't think we talk about enough. It's not just about helping instructors see student learning. It's also about helping students see their own learning.

Many students move through educational experiences without spending much time examining how they arrived at an answer, why they made a particular decision, or what strategies helped them succeed. They focus on completing the task, receiving the grade, and moving on to the next assignment.

AI can make that tendency even stronger.

When powerful tools can generate explanations, summarize information, suggest solutions, and produce polished work in seconds, students can easily focus on the outcome while paying less attention to the thinking that produced it.

That's why I believe some of the most important instructional design changes we can make are the ones that encourage students to look at their own thinking.

·      What did they find difficult?

·      What assumptions did they make?

·      Where did they get stuck?

·      How did they decide which information to trust?

·      What changed in their understanding during the process?

Questions like these help students develop metacognitive skills, the ability to think about their own thinking.

And those skills may become even more valuable as AI becomes more capable.

Because the students who benefit most from AI won't necessarily be the ones who can generate the most content. They'll be the ones who can evaluate their own understanding, recognize gaps in their knowledge, question their assumptions, and make thoughtful decisions about when and how to use these tools.

In other words, making thinking visible isn't just an assessment strategy. It's a learning strategy.

The more students understand how they learn, how they reason, and how they make decisions, the better prepared they will be to work alongside increasingly powerful AI systems.

That's why I think one of the healthiest approaches for faculty right now is to resist the urge to redesign everything at once. Instead, start small.

·      Pick one assignment.

·      Pick one activity.

·      Pick one discussion.

Then ask yourself: "What small change could make student thinking more visible here?"

·      Maybe it's a reflection question.

·      Maybe it's a justification prompt.

·      Maybe it's a process checkpoint.

·      Maybe it's an opportunity for students to explain not just what they did, but why they did it.

Those changes may seem minor. But small design decisions often have a much larger impact than sweeping policy statements or attempts to control technology use.

AI is creating pressure for change in higher education. But meaningful course evolution rarely starts with a complete redesign. More often, it starts with one thoughtful adjustment that helps keep thinking at the center of learning.