What Happens When You Trust High School Students with a Hard Question
I want to start with a Mentimeter response.
20+ students from four Iowa City high schools had just walked into a room together for the first time. Most of them had no idea what they’d signed up for. I put a question on the screen: What is one way AI has already changed your life or the life of someone you know?
The word cloud that appeared was dominated, predictably, by efficiency. Easy. Learning. Faster.
And then, in smaller text scattered around the edges: slop. bad tiktoks. lazy. dirtying their water.
Dirtying their water.
I didn’t know it yet, but that phrase was going to echo through the entire day. Because the same student who wrote it or another student thinking the same thing also responded to my second question, “When you hear the word ‘AI,’ what’s your honest first reaction?” with this:
“It’s destroying ecosystems and dirtying peoples water to the point its unusable.”
And another student wrote:
“I hate AI. I truly don’t think it’s useful; if anything it is ruining our world faster.”
And another:
“I think of slop, and then I think of people abusing it and losing autonomy.”
And one more, quietly devastating in its specificity:
“I’m getting a 0.”
This was not a room of AI enthusiasts. This was not a curated group of tech-forward students who’d been pre-selected for interest in artificial intelligence. This was twenty-one teenagers from West High, City High, Liberty High, and Southeast Junior-Senior High who had by their own responses largely skeptical, worried, or actively hostile feelings about the technology I was about to ask them to spend eight hours building with.
That tension is the entire story of yesterday. And it is the most hopeful thing I have experienced in a long time.
What They Were Actually Thinking
Before I could do anything else, I needed to understand the room. The third Mentimeter question was: What’s one question about AI that nobody has given you a real answer to yet?
I’ve asked versions of this question to adult educators, administrators, and policymakers across Iowa. I’ve never gotten answers like these.
Multiple students asked, in various forms, about water. Not abstractly, but specifically.
“How does it waste water? Why do people say it uses up water? Like I still don’t understand how it takes up/dirties water. Is anyone making an effort to make AI environmentally sustainable?”
These students had encountered something about AI’s environmental footprint and the data center water consumption issue is real and documented and nobody had explained it to them in a way that made sense. That gap between knowing something is true and not being able to explain why is a specific kind of intellectual frustration, and they were sitting in it.
Other questions came from places just as sophisticated:
“Does it result in long-term neural pathway change due to dependence? Will it disrupt our human reward circuitry?”
“Does it hinder the creative process of humans by solving problems for us?”
“Why do people tend to put less effort into learning something when using AI?”
The first is a neuroscience-adjacent question from a high schooler. The second is a philosophy of mind question. The third is metacognition, a student watching themselves and their peers and asking why.
And then, more quietly, this one:
“I don’t use AI almost ever... I suppose I’ll learn a lot today.”
I think about that student a lot today. What it took to show up with that combination of honest self-disclosure and genuine openness. What that kind of intellectual humility looks like in a seventeen-year-old who walked into a room full of strangers and admitted they were starting from zero.
The Frame That Changed Everything
The questions told me what I needed to know about how to open the day. This was a room full of smart, skeptical, genuinely curious young people who had real concerns and unanswered questions and who did not trust the technology they were being asked to work with.
So I didn’t ask them to trust it. I asked them to design with it.
Today you are not students. You are designers.
Not AI enthusiasts. Not ambassadors for technology. Designers solving a real problem for a real community, the community they actually live in, Iowa City, right now.
I have used versions of this frame before, but I have never seen it land so quickly and so completely. Within twenty minutes of that moment, I watched students who had written “bad” and “cheating” and “I’m getting a 0” in response to the word “AI” lean forward and start arguing about what problem they wanted to solve. The skeptics became the sharpest problem-framers. The student who understood that AI was “a system created to recognise patterns in human speech and writing”, a technically accurate description that most adults can’t produce, helped their team write a problem statement that named a specific user and a specific barrier with a precision that took my breath away.
When you stop asking young people to have an opinion about technology and start asking them to use it to solve something they care about, something shifts. The question stops being “what do you think about AI?”, which produces performed attitudes and becomes “what problem do you see in your community that nobody has addressed yet?” And that question, it turns out, these students have been thinking about for a long time.
Eighteen Minutes
Here is the truth about yesterday that I am still processing: we ran behind. Go figure. I got long winded. Jumped on my soap box on AI. Asked them more and more questions as they were so upfront and brutally honest....
The Mentimeter opening conversation was rich enough that I let it breathe longer than I should have. The questions students asked, the water questions, the cognitive impact questions, the “will we even have jobs” questions, deserved more than a quick acknowledgment before moving on. So I gave them more. And then we were behind. And as a huge fan of time boxing agendas and creating authentic check ins for learning I knew we had to get moving and fast. 10:30 was approaching where they had their first pitches to adults they have never met or seen.
Which meant that from the moment teams formed around their shared interests to the moment I expected a working first prototype on screen was approximately eighteen to twenty minutes.
I want to be honest about what I thought when I looked at the clock. I thought: we might not have anything to show the adults at 10:30. And this could ruin the day as students will shut down because that is not motivating to keep going AND I already knew this would be a bit tough as the first prototype always sucks and rarely works.
I was wrong. Completely wrong. And the reason I was wrong is the Gem I had built, a Google Gemini Gem designed specifically to walk teams through a structured design sprint sequence without allowing them to give generic answers. It pushed back on vague user portraits. It refused to accept “students who are confused about AI” as a meaningful problem statement. It compelled specificity at every step and produced, at the end, a prompt students could paste directly into Canva Code to generate a first prototype. Before they could use the gem they had to hold conversation and identify the true problem and map out the details in their design packet. This discourse could have easily been 30 minutes, but we had to move!
In eighteen minutes, teams went from forming to having something interactive on a screen. It was rough. It was imperfect. It was exactly what a first prototype is supposed to be, a hypothesis about what might help, waiting to be tested by a real person.
But it existed. And at 10:30, real adults sat down and used it.
The Magic of the Loop
The structure of yesterday had one design decision that I believe made everything else possible: adults gave feedback at 10:30, came back to the same teams at 2:15, and then watched the final pitches at 3:30 to the entire room in a large space.
Before/after. Before/after. Before/after.
I have done design thinking workshops where adults come in once, give feedback, and leave. That structure produces gratitude. This structure produced transformation. Because when the same adult who said “I don’t know if this is for me or for my kid” at 10:30 came back at 2:15 and said “Oh. I get it now”, every student in the room felt what design actually means. Not the concept of user-centered design. The lived experience of changing something and watching a real person’s confusion become clarity.
The adults were blown away. Several of them told me afterward that they had expected to give feedback to half-formed ideas and instead found themselves looking at prototypes that had genuinely changed between the morning and the afternoon. Some teams had made over ninety revisions to their Canva projects over the course of the day. Ninety. That’s not iteration as a concept. That’s iteration as a practice. That’s students who cared enough to keep going even when they were beyond frustrated with limitations of AI and things not working as they planned. Like any good design or PBL the power of ownership of learning, real audience, and real impact keeps the learning moving when it is difficult.
What They Actually Built
I want to be careful here. These are student prototypes built in a single day. They are not polished products. They have rough edges and incomplete logic and UX choices that a professional designer would revise. That’s not the point. The point is what they reveal about what these students understand and care about.
One team built FutureVote, a civic engagement tool focused on the intersection of AI and youth political participation. Another built an AI literacy tool aimed specifically at making AI concepts accessible to people who feel left behind by the conversation by helping teachers create personalized interventions for students in their classrooms. A third built a college matching tool, Real Talk, they called it, that was honest about what AI can and cannot tell you about your future. Another built Galactic Quest, using space exploration as a frame to create a K-12 game based learning system for kids who get overwhelmed with too much text.
These were not prompts I gave them. These were not topics I suggested. I gave them a question, How might we prepare our community for an AI-shaped future?, and space to figure out what that question meant to them. And this is what they came back with.
I did not know what topics they would choose. I want to repeat that. I deliberately gave them the widest possible runway to brainstorm and cluster and follow genuine interest. I had no idea going in whether they’d focus on jobs or school policy or mental health or environmental impact. I had no framework for predicting what 20+ students from four different schools would care about when given real freedom to decide.
What they chose reminded me that when we talk about students being lazy or disengaged or not caring, we are almost always describing what students look like when they’re asked to care about something they didn’t choose.
Give them the choice, give them the runway, give them a real constraint and a real audience, and watch what happens.
When Students and Educators Are Thinking the Same Thing
There’s a detail about the day I haven’t mentioned yet that I keep returning to.
Before the ICCSD AI Summit, a separate teacher workshop generated a set of strategies for AI in education as part of the teacher workshop designed by the amazing CodeJoy team. Educators and leaders put their heads together and identified what they believed would matter most. On the day itself, Kelsey shared those strategies after their pitch showcase for educators to see real time with students and asked them to vote: Which of these do you feel will be most beneficial?
Here is how they ranked them:
Teach students to use AI, use it thoughtfully, and with nuance
Design learning tasks that emphasize creativity — reward process over product
Prioritize human connection, community, and keeping the “human in the loop”
Keep questioning everything — listen well — slow down our decision-making
Try everything yourself — be an active participant — be transparent
Read that list again alongside the Mentimeter responses from the morning. The students who wrote “I hate AI” and “I’m getting a 0” voted most strongly for teaching thoughtful, nuanced AI use. The students who worried about losing autonomy voted for keeping the human in the loop. The students who spent the entire day making ninety-plus revisions to their prototypes voted for rewarding process over product.
They weren’t anti-AI. They were anti-thoughtless AI. There’s an enormous difference and the education system has spent a lot of energy responding to the wrong one.
We’ve built policy frameworks around preventing students from using AI carelessly. Meanwhile students are sitting in classrooms thinking: Nobody has taught us how to use this well. Nobody has slowed down long enough to help us think about what it’s actually doing to us. Nobody has given us a creative task worth doing.
The educators who designed those five categories were, in some sense, already listening. They had diagnosed the same problem from the adult side that students were describing from the student side. They just hadn’t been in the same room yet.
Yesterday they were. And the vote confirmed what both groups already knew: the path forward isn’t restriction or uncritical adoption. It’s nuance, creativity, human connection, and the willingness to keep asking hard questions.
That convergence is not nothing. In an education system that often feels like it’s having two separate conversations about AI, one in the staffroom and one in the student’s head, a moment where both conversations point at the same thing is worth paying attention to.
The Unanswered Questions
There is something that has stayed with me since I drove home last night and pondered and reflected in the rain and over coffee this morning.
Those Mentimeter responses, the ones about water, about autonomy, about neural pathways, about jobs, were not answered yesterday. We didn’t have time to answer them, and more importantly, answering them wasn’t the day’s purpose. The day’s purpose was to do something in the presence of those questions rather than wait for someone to provide answers.
But those questions are real. And the fact that these students are carrying them and that they know AI data centers use staggering amounts of water, that they’re worried about what dependence does to human cognition, that they’re watching themselves use AI and wondering what it’s costing them, that is a kind of literacy that most of the adults in their lives do not have. Including, honestly, a lot of educators.
We have spent so much of the AI-in-education conversation worrying about whether students will use AI to cheat, whether AI will make them lazy, whether we can detect AI in their essays. Yesterday showed me a different conversation happening inside students, largely unheard:
We have concerns. Real ones. Sophisticated ones. We’re watching what this technology does to us and to our world. We just haven’t been given a space to do something with those concerns except feel them.
Design gave them that space. For one day, the concern about water became a design constraint. The worry about cognitive dependence became a problem statement. The frustration about schools treating AI as purely an academic integrity issue became a prototype that tried to imagine a different relationship between students and AI tools.
What I’m Taking With Me
I have been doing this work of professional development, AI education, supporting teachers and districts navigating this moment for a long time. Some days I remember clearly why I’m still in it. Other days are harder.
Yesterday I needed the reminder. And these students gave it to me, freely and without knowing they were doing it.
I left the pitch showcase feeling like a proud parent. That’s the only honest description. I didn’t know any of them before 8:30 AM. By 4:00 PM, I had watched them form teams, develop deep empathy for specific users, build things that didn’t exist eight hours earlier, fail and iterate and fail better, and stand up in front of real judges and defend every design decision they made with real evidence from real observation.
I had also watched a student who wrote “I’m getting a 0” on a Mentimeter card spend eight hours building something he cared about with a tool he distrusted, and pitch it with genuine conviction.
That’s not a small thing.
The student who said “I don’t use AI almost ever... I suppose I’ll learn a lot today” did, I think, learn a lot. But not what I would have taught them in a lecture. Not what any policy document or curriculum framework would have prescribed. They learned what happens when you stop being skeptical in the abstract and start being specific and when distrust becomes a design constraint instead of a reason to disengage.
That is the thing I keep coming back to. The skeptics made the best designers. The students who were most critical of AI in that opening Mentimeter were often the ones who pushed hardest in the observation block, who asked the sharpest questions of the adults using their prototypes, who were most unwilling to accept a vague problem statement because they knew from personal experience what a vague answer to a hard question feels like.
Their distrust was a resource. We just had to give it somewhere to go.
And it turns out the educators in the room had already named that somewhere. Teach it thoughtfully. Reward the process. Keep the human in the loop. Keep asking.
The students confirmed it. They were waiting for exactly that.
The driving question for yesterday was:
“How might we prepare our community for an AI-shaped future?”
I don’t think any of us answered it completely. I don’t think it can be answered completely, certainly not in a day, and maybe not ever. But with students from four high schools spent eight hours taking it seriously with their hands and their code and their observations and their genuine concern for the community they live in.
That’s preparation. That’s more preparation than most adults are doing.
That’s why I’m still in education.
And yes, I have this entire AI Design Thinking Workshop completely laid out, captured, documented with protocols and all the things if you are interested in learning more.