What Yellowstone, Rwanda and AI Taught Me About Change

Over the past two years, I’ve spent countless hours talking with teachers, parents, students and school leaders about artificial intelligence. The questions they have are remarkably consistent: How should schools respond? When should students use AI? How do we preserve authentic learning in a world where answers have become nearly instantaneous?

They’re important questions, but I’ve noticed something interesting about them. Our conversations rarely stay focused on the technology itself. Before long, they drift somewhere deeper. They become conversations about thinking. About curiosity. About judgment. About what learning is actually for when a machine can generate an essay, solve a problem, or summarize a chapter in seconds.

Oddly enough, the most helpful answers I’ve found didn’t come from AI research. They came from Rwanda and Yellowstone National Park.

For years, those sites occupied completely different parts of my life. One was a trip that changed me. The other was an ecological story I couldn’t stop thinking about. It took me much longer to realize they were helping me answer the very questions I was now asking about artificial intelligence in schools. 

That realization began on a dusty road, on an otherwise ordinary day in Rwanda.

I went to Rwanda as part of a group helping schools transition from French to English in the classroom. That day, I happened to be carrying a yellow jerrycan of water from the watering hole back to the village alongside people who make that miles-long journey every day. Before long, my hands ached and my shoulders tightened. The hard plastic handle dug into my fingers as I wondered how much further we had to walk. The people beside me never seemed to notice the weight. For me, carrying water was extraordinary. For them, it was simply another walk home.

That contrast stayed with me long after I returned home. The walk answered very little. Instead, it left me with a question that has followed me ever since: How do we help students understand realities they have never lived?

That question would resurface years later in a sixth-grade classroom. Our teaching team designed an interdisciplinary unit around the novel “A Long Walk to Water,” hoping literature, conversations with a friend of mine from Sudan, and local field experiences would help students move beyond learning about water insecurity and begin to care about the people living it.

By most measures, the lesson was a success. Then we sat down to debrief.

“I don’t think it worked,” one teacher said.

Almost immediately, heads began nodding around the table.

“I was thinking the same thing.”

“Me too.”

No one defended the lesson. No one blamed the students. Instead, we realized we had been measuring engagement when what we really cared about was understanding.

Looking back, I think that meeting changed our teaching far more than the lesson itself. Years of working together had built enough trust for one honest observation to give everyone else permission to be honest too. We stopped trying to protect the lesson and started asking a different question.

Not, How do we make this activity better? But, How does understanding actually grow?

We’ve been refining that project ever since, and that conversation forced us to confront a much larger question: If meaningful change doesn’t happen because of a single lesson, how does it happen?

To understand how meaningful change happens, consider what happened in Yellowstone National Park.

When scientists reintroduced gray wolves to Yellowstone National Park in 1995, they expected the elk population to change. They didn’t expect rivers to change as well. As elk altered where they grazed, willow and aspen trees recovered. Beavers returned. Wetlands expanded. Vegetation stabilized riverbanks. An entire ecosystem began reorganizing itself in ways few people had anticipated.

The wolves didn’t directly reshape Yellowstone. They changed the relationships within the ecosystem. And everything else followed.

Learning about Yellowstone changed the way I thought about change itself. It made me realize that the most meaningful transformations rarely begin by fixing one thing. They begin by changing the relationships within a system. Once I saw that pattern, I couldn’t stop seeing it in schools.

We often assume meaningful change happens by improving the thing right in front of us. We redesign a lesson. We adopt a new curriculum. We write a new AI policy. We expect better outcomes because we’ve improved one part of the system.

The wolves didn’t change the rivers directly. They changed how elk moved through the landscape. That altered where trees could grow, which brought back beavers, reshaped wetlands, stabilized riverbanks, and eventually changed the rivers themselves. One dramatic intervention didn’t drive the transformation. It emerged because relationships within the ecosystem had changed.

I’m beginning to think schools work the same way.

Looking back, our sixth-grade project didn’t improve because we discovered a better activity. It improved because the relationships around the project changed. Teachers trusted one another enough to challenge assumptions. Students were invited to wrestle with ideas instead of simply completing assignments. Learning became less about delivering an experience and more about cultivating understanding.

Once I began seeing schools that way, I couldn’t think about artificial intelligence in quite the same way anymore.

AI isn’t simply introducing another tool into classrooms. It’s changing relationships: The relationship between students and thinking. Between effort and accomplishment. Between teachers and expertise. Between knowing something and finding it instantly.

Those may prove to be the relationships that shape the future of learning far more than any particular AI model or school policy.

That’s why my conversations about AI are almost never about the technology itself. They’re about what we choose to value.

Parents wonder whether students will gradually stop wrestling with difficult ideas if a chatbot can do the work for them. Teachers ask how to preserve productive struggle when every answer is just a prompt away. School leaders wrestle with an even deeper question: If AI can do more of the thinking, what kinds of thinking should remain unmistakably human?

I don’t think there’s a simple answer. What I do know is that answering those questions has required me to revisit experiences that had nothing to do with artificial intelligence — at least not at first.

I keep finding myself returning to that dusty road in Rwanda.

Carrying a jerrycan didn’t help me understand what it means to live with water insecurity. It wasn’t meant to. What it changed was my relationship to the people carrying it beside me. 

Our sixth-grade project taught me something similar. We eventually realized our goal wasn’t to create a memorable activity. It was to create the conditions where students might begin asking deeper questions about lives very different from their own.

Increasingly, I think AI presents schools with that same choice. We can use it to remove every obstacle from learning, making every task faster and easier. Or we can become even more intentional about preserving the kinds of thinking that develop judgment, curiosity, empathy and discernment.

Looking back, I don’t think Rwanda, Yellowstone, and AI were ever separate stories.

Rwanda changed my relationship to people whose lives I couldn’t fully understand. Our faculty meeting changed the relationships among teachers. AI is changing the relationship students have with thinking itself.

Rwanda taught me that understanding begins with humility. Our sixth-grade team taught me that meaningful change begins with the courage to question our own assumptions. And Yellowstone taught me that lasting change happens by changing relationships within a system.

Technology will continue to change. Our responsibility as educators won’t.

We still have to create experiences that ask students to think, question, struggle, and understand in ways no technology can do for them. Those were the questions before AI. I suspect they’ll still be the questions long after today’s technology has been replaced by something none of us can yet imagine.