Students as Partners:
My Journey with AI in Math Education
How an open-ended question, student skepticism, and a willingness to learn alongside my students transformed my teaching practice.
The Beginning: A Question Worth Asking
When I started the Cohort 21 Action Plan journey, I had a clear vision: “How might we use AI as a learning partner in math?”
I imagined my Grade 7 students turning to Google Gemini for their math questions, getting feedback on their thinking, reviewing concepts, and tackling assignments with an always-available AI tutor at their fingertips. It seemed logical, innovative, and exactly what 21st-century math learning should look like.
I was confident. I had a plan. What I didn’t have was what my students would actually tell me.
When Students Push Back
The first hint that my grand vision needed revision came when I listened to my students. They had concerns—real ones:
- AI is bad for the environment. The computational footprint bothered them.
- AI gets things wrong. Why trust something unreliable?
- We shouldn’t use it. A blanket skepticism that made me pause.
These weren’t the objections of luddites. These were thoughtful young people wrestling with legitimate questions about AI’s role in their lives and their learning. Instead of dismissing their concerns, I realized they were teaching me something essential: before diving into innovation for innovation’s sake, we needed to address trust.
That’s when the pivot happened.
The Pivot That Changed Everything
Our school approved NotebookLM for Grade 7 use—a tool that felt safer, more contained, and importantly, designed specifically for learning. Around the same time, I discovered something unexpected: Gemini wasn’t just useful for students. It transformed my administrative work, helping me manage tasks and think through complex planning challenges.
But the real game-changer was deciding to create a safe sandbox environment within NotebookLM. Working closely with our tech team, we built a space where students could explore AI-powered learning features without the open-ended overwhelm of a tool like Gemini. We knew where boundaries were. We understood what data was being used. And critically, we could control the experience.
What Students Actually Discovered
Once inside NotebookLM, something remarkable happened. Students weren’t just passively using a tool—they were discovering what it could do.
They found flashcards. They generated videos. They built quizzes and created podcasts. They explored the chat features and dug into sources. The interface was familiar—these were learning strategies they already knew—but powered by technology that made creation effortless.
Here’s what they said about it:
“I discovered I can create new quizzes based on what I wanted to learn about. I discovered there are multiple different podcasts. I discovered I can look at the sources the AI uses.”
“You can go back and review your quiz after you’re done. The videos are really helpful and understandable. For the quiz if you put the wrong answer it will explain why it’s wrong and explain why the right answer is right.”
This wasn’t forced adoption. This was genuine engagement born from autonomy and choice.
The Math Learning: Real Concepts, Real Understanding
But engagement without learning is just entertainment. What made this real was what students actually learned.
Using NotebookLM, students deepened their understanding of integer concepts, worked with the zero property, explored counter representations, practiced operation rules, and solidified their definition of integers. The tool provided interactive quizzes and explanations that adapted to their responses. When they got something wrong, they weren’t just told the answer—the AI explained the reasoning.
This is scaffolding done right: not hand-holding, but strategic support that builds understanding.
The Numbers That Matter Most 
100% of students said they would use this tool again.
Let that sink in. Not 80%. Not “most.” All of them.
This came from a cohort that started skeptical. That questioned whether AI had a place in their education. That worried about environmental impact and reliability. And through actual experience, through discovery rather than mandate, through choice rather than prescription, they became enthusiasts.
Why Familiarity Beats Open-Endedness
Here’s a critical insight I learned: I started too broad with Gemini. Open-ended, generative, unlimited possibility—sounds great to adults building a vision. For Grade 7 students deciding whether to trust an unfamiliar tool? It’s paralyzing.
NotebookLM worked because it met students where they were. They already knew quizzes. They were comfortable with flashcards. Videos made sense. Podcasts were a format they understood. The scaffolding wasn’t about simplifying—it was about building on their existing competencies while introducing the power of AI.
Students didn’t need to reinvent learning. They needed to see how AI could make the learning strategies they already valued faster, more interactive, and more responsive to their individual needs.
The FuturePrize Experience: Learning With and From Students
The story gets deeper through my participation in FuturePrize—a design sprint that pushed my thinking further.
I worked with three students. I received professional development on AI and human-centered design. Our team came away as runners-up out of 24 teams, but the real prize was something else: a fundamental shift in how I understand my role.
Through FuturePrize, I learned:
- About students’ genuine feelings toward AI—not my assumptions, but their actual hopes and concerns
- About how to scaffold learning so students build confidence gradually
- About how students can use AI to build on their own ideas—as a partner in thinking, not a replacement for it
- About the power of small groups—where real dialogue happens and students shape the direction
Most importantly, I learned that learning alongside students as partners is more powerful than positioning myself as the expert.
The Big Takeaway: Partnership Over Prescription
Here’s what I’m taking away from this entire journey:
Teaching in the age of AI isn’t about mastering the technology and deploying it to students. It’s about learning alongside them. It’s about giving them autonomy to discover what works for their learning. It’s about understanding how they learn best and creating the conditions for them to make informed choices.
The shift from “here’s a tool I think will help you” to “let’s explore what this tool can do, and you tell me what’s useful” changed everything. Students weren’t recipients of innovation. They were partners in shaping it.
When we position students this way—as collaborators, as thinkers, as people whose perspectives and concerns matter—they step up. They ask harder questions. They push back on things that don’t make sense. They become invested in the outcome.
That’s when real learning happens.
Lingering Questions: The Work Ahead
This journey isn’t over. In fact, it’s raised questions that will shape my practice going forward:
- How can I use AI to help with assessment? Not just as a quiz generator, but as a tool that helps students see their own growth and understand what they still need to work on.
- How can I use it to communicate better with parents and students? How do I close the feedback loop so that the insights from AI-powered learning actually translate into conversation that helps students improve?
- How can we continue learning together? Building on the partnership model, how do we deepen the feedback cycle—not information flowing down, but understanding flowing in both directions?
These questions will shape what comes next. And I’m excited to explore them alongside my students.
Connecting to Our Moment
Susan’s journey—and I say this as someone reflecting on my own practice—is a microcosm of where education needs to go in 2026 and beyond.
Cohort 21 exists because we know the future demands more than content delivery. It demands partnership, design thinking, agency, and the ability to learn alongside rapidly changing tools and contexts. It demands that we see students not as passive recipients of our expertise but as collaborators in shaping what learning looks like.
AI in the classroom isn’t about replacing teachers or streamlining instruction. It’s an opportunity to fundamentally shift who gets to be a designer of learning. When we let students in on that design process—when we listen to their concerns, honor their questions, and learn from their discoveries—we unlock something powerful.
We unlock their learning. And we transform our teaching.
For Educators Starting Their Own Journey
If you’re wondering whether AI belongs in your classroom, my advice is this:
- Start by listening. What are your students’ concerns? What do they already value about learning? Build from there, not from a predetermined vision.
- Choose your tool thoughtfully. Not every tool works for every context. A contained, purposeful tool beats open-ended access when you’re building trust.
- Create a sandbox. A safe space where students can experiment without stakes lets them discover value on their own terms.
- Step back and observe. What do students actually find useful? What surprises you? Those discoveries are data.
- Position yourself as a learning partner. You don’t need to have all the answers about AI. You need to be genuinely curious about how your students experience it and what they think.
- Give them autonomy. Students making choices about how they use tools leads to better engagement, better learning, and a deeper sense of ownership.
The future of education isn’t about teachers disappearing. It’s about shifting from experts dispensing knowledge to partners discovering it together.
What’s Next
This is just the beginning. I’m heading into next year with new questions, a deeper understanding of how my students think, and a commitment to continuing this partnership.
I’m exploring how to use AI for more meaningful assessment. I’m thinking about how to involve parents in the feedback loop. I’m designing spaces where students have more voice in how learning happens.
And I’m doing it alongside them—not because they need me to be the expert on AI, but because real learning happens when we’re all figuring it out together.
That’s the lesson from my Cohort 21 journey. And it’s one I’m carrying forward.
Resources
- NotebookLM — Google’s tool for interactive learning, research, and content generation. Designed with educational use in mind, it offers a bounded environment for AI-powered learning activities.
- Cohort 21 — An initiative focused on 21st-century learning, design thinking, and partnership-based education. Learn more at cohort21.org.
- FuturePrize — A design sprint program that brings educators and students together to explore how technology can serve learning. Applications open annually.
- Something Big Is Happening – Why should we think about using AI? Let’s understand it. Use it. Adapt it.
- AI for Students — Podcast: Resources on helping young people think critically about AI’s environmental impact, reliability, and appropriate uses in education.
About the Author: Susan Matthews is a Grade 6 & 7 mathematics teacher committed to partnership-based learning and thoughtful technology integration. Through Cohort 21, she’s exploring how AI can empower students as designers of their own learning. Her work bridges skepticism and innovation, grounded in listening to what students actually need and value.
