NC's AI Solve-a-Thon Proved Students Are Ready. Are We?

Education
Written By Lydia Kumar

Season 3, Episode 33 of Kinwise Conversations · Hit play or read the transcript

Episode Summary: NC's AI Solve-a-Thon Proved Students Are Ready. Are We?

Most debates about AI in schools ask the wrong question. Instead of debating whether students are ready, North Carolina's Department of Public Instruction decided to find out, by giving them real problems to solve. In this first episode of Kinwise's special series on the NC AI Solve-a-Thon, host Lydia Kumar sits down with the two architects of the competition: Vera Cubero, NC's AI and education leader and one of the 2025 Top 100 Leading Women in AI, and Matthew Mayo, Director of Digital Technology and Learning at NC DPI.

Together, they share how the Solve-a-Thon came to life, what they learned from watching students tackle homelessness, DMV navigation, and rural resource access, and why one student said he learned more in those few months than in his entire school career.

Key Takeaways for Superintendents, K-12 Leaders & Mission-Driven Educators

  • AI literacy is the floor, not the ceiling. Students need AI fluency, the ability to apply human judgment, maintain agency, and use tools to benefit their communities. The Solve-a-Thon showed what becomes possible when we stop debating and start building.

  • Students are already using AI. The question is whether schools will guide that use thoughtfully or let students figure it out on their own. North Carolina chose to create a structured environment where curiosity could lead.

  • Real challenges unlock durable skills. Empathy, collaboration, critical thinking, and a learner's mindset don't develop naturally in traditional classrooms. Project-based, community-connected work is where those skills come alive.

  • Coaches are the unsung infrastructure. By equipping and supporting adult coaches across the state, NC DPI distributed leadership and created unintentional professional development for educators right alongside their students.

  • Instructional redesign is not optional. If the end product can be AI-generated, schools need to shift their focus to the process, learning portfolios, metacognitive reflection, and assignments that AI can't simply do for students.

How the NC AI Solve-a-Thon Came to Life

Lydia: I want to begin by giving you all a chance to share your journey. What led you to the Solve-a-Thon and how did it come into being?

Matthew Mayo: I think one of the things Vera has done is an amazing job at the state level creating our AI guidelines. That really set the framework for AI and education. In ed tech, anytime there's a new tool, we're kind of like, nope, let's ban it. Vera and the team worked really hard on those guidelines and implementing them across the state.

When we think about students engaging with technology, we really wanted to move them from just consuming the technology to actually using it to make things better. We wanted to create something that felt different from a traditional competition. Instead of asking students to build the most technically impressive project, we asked them to identify real problems in their communities and explore how AI could help address them. The result was incredibly powerful. Students weren't just experimenting with technology, they were thinking deeply about people, systems, and the challenges their communities face every day. The curiosity, the empathy, and the innovation just came together.

Vera Cubero: There's still a lot of fear and uncertainty, and a lot of it is well founded. But we both felt that AI is here, and it's something our students need to start grappling with and learning to work with and co-create with. We tossed around the idea last year but felt we weren't yet in a place where we'd have adequate participation. So we waited until this year. We were hoping to get 50 teams, and we did with every region of the state represented.

The things the students built really showcased what students can do when given intellectual freedom. Those durable skills everyone talks about, empathy, critical thinking, collaboration, a learner's mindset, those are not naturally developed in a traditional classroom. The Solve-a-Thon was a way to really emphasize those skills while preparing kids for the future. And the students exceeded our expectations. They really did.

What Students Built

Lydia: That must have been so inspiring to see. I think for people listening who are thinking about how you find someone to support students through that process: what did that look like? Were students on their own or did they work with an adult?

Vera Cubero: We required that they have an adult in the school. We were a little worried about it at first, but we got our 50 teams and close to 50 coaches. A lot of the educators were really excited. We shared it out in all of our statewide communications and they came together. We did have some people reach out afterward to say they hadn't heard about it and would have liked to participate, so I think as Matthew said, next year we'll probably get even more.

Lydia: What were some of the things students built?

Vera Cubero: We had students solving for issues like homelessness, lack of access to resources in rural communities, and improving the DMV. Those were our three top winners. But we also had a group working to solve the problem of white squirrels being killed on roadways. One was about invasive species. One was a translation tool for ASL users. There was a mental health chatbot. A lot of really creative things that showed students were thinking and connecting to their communities.

We didn't want to call it a prompt-a-thon. A lot of those bring students together, give them a problem, and they have a few hours to put something together. We wanted to focus on the solving: they were really solving real-world issues and showcasing that AI can be used as a force for good. Students started in September, and we had our first virtual competition in October to narrow down to our ten state finalists. So it was a multi-month competition. The things they built were real and practical, and a lot of them are ready for production if they get sponsors.

"We wanted to focus on the solving part, that they were really solving these real world issues and showcasing that AI can be used as a force for good."

The Coaches: Unintentional, Intentional PD

Lydia: You created this competition for students, but at the same time you were really upskilling the adults. The coaches. So you created a community of adults who had excellent training with Play Lab and came together to learn how to lead students in thinking this way. You're distributing leadership. You can't work with every single student, but you can work with these adults who then work with the students.

Matthew Mayo: It's like unintentional, intentional PD for everyone. Many educators walked away realizing something really important, that students are capable of far more than we sometimes give them credit for. When students are given something meaningful and have the tools to explore solutions, they bring such fresh perspectives that some of us adults might not immediately see. The coaches saw how quickly students can adapt to emerging technologies when the learning environment encourages curiosity rather than fear.

AI Literacy Is the Floor. Fluency Is the Goal.

Lydia: What do you think this revealed about how AI literacy should look in schools?

Vera Cubero: I think it revealed that we've got to move more to application. While a lot of people are still debating whether to introduce any AI in their schools, we have kids who are already building things that solve major issues with AI. AI literacy is just the floor. Students need AI fluency. The students that are going to succeed in the future are going to be the ones who are competent in working with tools, but who also have the ability to implement human judgment and agency throughout the whole process. It's time to stop debating whether our students are going to use AI. They already are. The question is: are we going to guide them in the right way?

Matthew Mayo: I think about a video called "Ban Pencils," a mockery of how teachers responded when pencils first came out. The literacy piece is the component we've got to get right. COVID was the great accelerator for technology in the classroom, and we rushed so fast that we didn't take time to teach kids how to use things effectively. We've got to go back a little bit. It's about developing students' ability to ask good questions, evaluate information critically, and understand how technology influences the world around us. All of us need to learn when to question AI, when to verify its outputs, and how to use it responsibly.

Lydia: Vera, you said AI literacy is the floor and we need to move toward fluency. What's the difference?

Vera Cubero: A big part of AI fluency is that discernment, that human agency piece. Students don't just need to know how to use chatbots as input-output machines, because the technology is already moving toward autonomous agent teams. Basic AI literacy is what is it, how does it work and all of that is shifting beneath our feet. But I consider fluency to be what every one of these students who did this project achieved. They went through the process of really guiding and evaluating AI while keeping human judgment central, and ensuring the outcome benefits humanity. That ethical piece rises to the top when we're really talking about fluency. It's data literacy, discernment, knowing if you should even use the tool. Anthropic has an AI fluency framework, the four Ds, and discernment is a big part of it.

"AI literacy is just the floor. The students that are going to succeed in the future are going to be the ones who are competent in working with the tools and who have the human judgment and agency to oversee them."

Instructional Redesign Is a Must

Vera Cubero: We have to really face the fact that instructional redesign is an absolute must. In a time when these tools can give us the answer to any traditional assignment, we have to start rethinking those assignments. If we just continue to give students worksheets and five-paragraph essays and then battle with them over whether they cheated with AI, we're totally missing the point.

Our division is doing a pilot on micro PBL, AI-assisted, purely standards-aligned. Teachers can take a standard, put it in, add parameters like their location to bring in that local connection and empathy piece, and it will co-create a three-to-five-day micro PBL unit. What makes it different is that teachers have to be engaged throughout the whole process. It reminds them often that they're the teacher, that they bring a lot to it and need to have oversight.

And learning portfolios. If you can't tell whether the end product was created by AI, it's time to stop evaluating only the end product. Process over product. If you're going to do that, you have to collect it, something as simple as a Google Site capturing the whole messy, beautiful process of learning. John Hattie's research shows metacognitive reflection has an effect size close to 0.7, nearly eight months of growth in a school year if done well.

I quote John Dewey often: if we teach today's students as we taught yesterday's, we rob them of tomorrow. That has never felt more relevant.

What Students Proved When We Trusted Them

Lydia: What do you think the Solve-a-Thon revealed about what students are capable of when we trust them with meaningful challenges?

Matthew Mayo: It opens our eyes to what kids are capable of when we give them a blank canvas and some guardrails. We've got to do more of that. How do we get kids to truly design? How do we get them to think critically and collaborate? The soft skills that were developed … teams had to communicate with each other, build on each other's ideas. We hear so many times that kids are always on their phones. Well, they had a lot of face time together for this. And it goes back to creating a culture where we allow kids to fail forward. The system has always been pass-fail, but in my own career some of my greatest failures have actually been my greatest successes because I learned what not to do next time.

Vera Cubero: I had one coach tell me a couple of weeks ago that one of the young men on one of the winning teams came to her and said he had learned more in the last four or five months on that project than he had in his entire school career. And then there was a school with a very high poverty rate that has historically struggled, and these kids created an amazing project and won at a state level. The pride it brought them, and then the pride it brought their whole school community. That was really something.

"He had learned more in the last four or five months on that project than he had in his entire school career."

Guest Bios

Vera Cubero is an AI and education leader who spearheaded North Carolina's statewide AI guidelines and was named one of the 2025 Top 100 Leading Women in AI. A key collaborator with NC DPI's Digital Learning team, Vera is a national voice on AI fluency, ethical decision-making in education, and what it looks like to move students from AI consumers to AI creators.

Matthew Mayo is the Director of Digital Technology and Learning at the North Carolina Department of Public Instruction, where he leads statewide initiatives at the intersection of technology, equity, and student readiness. A longtime advocate for project-based learning and fail-forward culture, Matthew was a co-architect of the NC AI Solve-a-Thon.

Connect with Vera Cubero and Matthew Mayo

Resources Mentioned & Related Concepts

→ See how this connects to Kinwise Coach

Lydia Kumar
Next

Schools are Asking the Wrong Question About AI