91����

A Maine teacher may have an idea for an interactive science demonstration, a classroom game or a multimedia lesson that could help students understand a difficult concept. 

Turning that idea into a working digital tool, however, can require programming experience that many educators do not have.

“Teachers have lots of great ideas for the technology they and their students could use in the classroom,” said Justin Dimmel, associate dean for academics and student engagement in the University of Maine College of Education and Human Development. “But teachers are very busy. They don’t usually have the time, capacity or expertise to build their own tools.”

Dimmel is leading a 91���� team exploring whether artificial intelligence can help remove those barriers while keeping educators in control of the creative process.

The U.S. National Science Foundation selected the team’s project, The Forge, for a planning award through FINDERS FOUNDRY, a national initiative focused on challenges in K-12 education. 

The 91���� team will bring together educators, researchers, technologists, parents and students to explore a teacher-centered environment for creating interactive learning materials with AI-assisted coding tools.

Putting teachers first

The Forge is not a finished platform being introduced into Maine schools.

The team will first study what teachers need, how AI could support learning and how schools can address concerns including student privacy, data security, inaccurate information, bias and appropriate classroom use.

Those questions matter to parents and teachers alike: Who decides how AI is used in the classroom? How will student information be protected? And can the technology support learning without replacing a teacher’s judgment and creativity?

The team includes co-principal investigators Zihan Wu of the Maine College of Engineering and Computing and Kamal Chawla of the College of Education and Human Development. Five teachers from different Maine schools and two parents also will participate, with students expected to contribute as teachers develop and test early ideas.

The project grew from a challenge submitted by a Maine teacher who wanted educators and students to have a greater role in creating educational content. 

Teachers often imagine games, demonstrations and other interactive experiences that could benefit their students, Dimmel said, but many ideas are never developed because the process is technically demanding and time-consuming.

The researchers describe that unrealized potential as an “innovation reserve” within schools. The Forge will explore whether AI-assisted design — a practice commonly referred to as “vibe coding” — can help educators tap into it.

From classroom idea to working tool

A chemistry teacher, for example, might want to create an interactive demonstration. Another educator might want to produce a video for a flipped classroom or build a digital learning game app. AI-assisted coding tools could make it easier for teachers without programming experience to turn those concepts into working prototypes.

The goal is not to replace teachers, and the proposed environment would differ from a generic chatbot, automated lesson-plan generator or learning management system. Instead of beginning with an empty prompt box, educators could receive guidance, workflows and specialized tools for different classroom projects.

The details have not yet been determined. Dimmel said the planning process will allow educators and families to help shape the project rather than have researchers build a product based only on their own assumptions.

“We want to try to put teachers’ creativity at the center of what we’re going to design,” Dimmel said. “We see this as an accelerator or an incubator, not a replacement.”

While much of the public conversation about generative AI has focused on what the technology might automate, The Forge asks how it could expand what teachers and students are able to create.

“We’re trying to think about putting teachers and students in the roles of being content creators, rather than consumers,” Dimmel said.

Working directly with the technology also could help teachers and students understand what it can do, where it can fall short and how it can be used responsibly.

Testing what teachers need

During a concentrated design process in September and October, participating teachers will use existing AI-assisted coding tools to build early versions of their instructional ideas. Researchers will study the obstacles they encounter and what support could make the process more accessible.

Dimmel said the work also could show that existing general-purpose AI platforms are sufficient if educators receive appropriate training. The team is not assuming that a separate platform will necessarily be the best answer.

For Dimmel, success would not depend solely on whether the work leads to a new platform. Participating teachers could become a core group of educators prepared to help their schools and students approach AI with greater knowledge, creativity and confidence.

“We inspire teachers to consider themselves and their own creative agency,” Dimmel said.

Expanding who gets to create

Dimmel compared AI-assisted coding to the printing press, which lowered barriers to producing and sharing information. Until recently, creating an application or interactive digital resource often required significant programming knowledge, time and other resources. AI tools could make that type of creation more accessible to people who have ideas but do not consider themselves software developers.

“We want to make sure that teachers are participating in that, that they have the skills, experience and knowledge that they feel like their agency is valued, and they can see a way to get their ideas out into the world,” Dimmel said.

Contact: Casey Kelly, casey.kelly@maine.edu