Students Cannot Just Be Consumers. They Need to Build.

My Students Were Gamers

When I was teaching Computer Science and Robotics Pathways at Beaumont High School, most of my students were gamers. This is not unique to Beaumont. Esports and gaming as a whole are gaining traction year over year. Student attention is easily captured by video games.

I do not think it is wise to overconsume video games. But the idea of building your own video game is an attention catcher that bleeds into real educational territory: game development, programming, design thinking, AI literacy. When a student who spends 4 hours a day playing games learns that they could build one, the conversation shifts. They stop being passive consumers and start thinking like creators.

That shift from consumer to creator is what drives everything I have built since leaving the classroom.

Project-Based Learning from Day One

I did not have a sudden revelation about project-based learning. From the start of my teaching career, I focused on having students build things. Every year, students were required to complete a number of projects. At the heart of it, they needed a website that would house those different projects, forming their portfolio.

That portfolio was not just a school assignment. It had real value for work experiences later on. A student who graduates with a website showcasing the applications, games, and tools they built has something tangible to show an employer or a college admissions office. That was always the intention: make the work matter beyond the classroom.

At Beaumont High School, I watched students go from not knowing what HTML stood for to building functional web applications in a single school year. Not because I lectured them into understanding. Because they had something they wanted to build and I gave them the framework to do it.

Why I Built Clever Games

The first pilot for Clever Games launched in the 2025-26 school year for students in grades K through 5th. The timing was not random.

The mass adoption of AI and the rapid development of agentic software created a unique window. For the first time, it became possible for young students to build real, playable games without needing to write code themselves. AI handles the technical generation. The student handles the creative direction: what the game looks like, how it plays, what the characters do, what happens when you win or lose.

The Department of Labor AI Literacy Framework also contributed to the decision. There is a growing recognition at the federal level that students need to understand AI, not just as users, but as builders. Clever Games is built around that idea. Students are not consuming AI content. They are directing AI to create something that is theirs.

We do create accounts for students on the platform, but no actual student information is visible to the public. Only the representing and participating schools are shown. Student privacy was a design constraint from the beginning, not something we added later.

What We Learned from the K-5 Pilot

We just finished our pilot program for grades K through 5th. The results are informing everything we plan to launch in the 2026-27 school year.

The most surprising thing was not what students built. It was the diversity of what they wanted to build. One student wanted a football game. Another wanted a Barbie dress-up game. Then another one wanted a Windows 95 computer that shoots objects out of the sky. These are kindergarteners through fifth graders. Their imaginations are not constrained by what they think is "possible" in software. They just describe what they want.

That range of ideas confirmed something I suspected but had not seen at scale: when you remove the technical barrier, creativity is not the bottleneck. Every kid has a game in their head. Most of them have never had a tool that could bring it to life.

We are now refining the platform based on insights from the pilot. That includes improving the visual and performing arts curriculum integration, building out live portfolio capabilities, and addressing the student privacy architecture more thoroughly.

The Hard Parts Nobody Warns You About

Developing an AI-driven software application while simultaneously building other solutions and products is proving difficult. The variable token costs alone create planning challenges that traditional software does not have. Every game a student generates consumes API tokens. Scaling that across hundreds of students means the cost model is fundamentally different from a standard SaaS product.

But the harder challenge is not technical. It is educational.

While it is exciting to say "look what I built with AI," the real value is getting students to understand the interconnected educational pieces and branching subject areas. A student who builds a platformer game has touched game design, visual art, storytelling, logic, and basic AI interaction. But do they know that? Can they articulate it? Can they connect what they did in Clever Games to what they are learning in math class or science class?

That is the bridge we are still building. The technology works. The educational framework around it is what needs the most refinement.

What I Tell Educators About AI

Anyone not using AI right now is falling behind. I use AI constantly: to build software, to ship products and content, to organize, and to communicate. AI is allowing everyone to build, and build faster. That presents an incredible opportunity for students who can imagine and think big.

When I introduce Clever Games to teachers and administrators, most of them are not fully in tune with the current capabilities of AI. They are always interested when they first see the platform. My approach is simple: I tell them we can set them up with an account and have them try it themselves. No pitch deck. No slideshow. Just use it. That has been more effective than any presentation I have given.

The conversation I have most often with educators is not about whether AI should be in schools. That question has already been answered by every student who is using ChatGPT on their phone. The conversation is about what role we want AI to play.

My position is straightforward: students cannot just be consumers. They need to build. They need to create things, not just interact with things other people created. If AI can lower the barrier to creation so that a first grader can build a video game and a fifth grader can build a portfolio of creative projects, then AI is doing exactly what technology should do in education.

It is not replacing teachers. It is giving students the ability to act on their ideas instead of just having them.

If you want to see what Clever Games looks like in practice or explore a pilot for your school, email me: hello@manuelzavala.com. I will walk you through the platform and show you what students are actually building.