Preparing science educators to use and teach AI in the classroom

An interdisciplinary team of researchers funded by the U.S. National Science Foundation has created a teacher-driven curriculum called Shark AI. This optional online program introduces Florida middle school teachers and students to artificial intelligence through the science of paleontology. With Shark AI, science teachers learn how to incorporate AI and machine learning concepts into their classrooms to help their students build essential skills and prepare for 21st-century STEM careers.

"Most K-12 AI learning occurs with computer science standards and learning goals," said Bruce MacFadden, University of Florida (UF) distinguished professor and principal investigator (PI) on the project. "Shark AI takes a novel approach by using AI to teach concepts that are connected to biology, paleontology and the nature of science, while at the same time connecting to computer science education goals as well," he said.

Developed with and for teachers, Shark AI is a free, optional online curriculum with five flexible modules — all aligned with middle school science standards — that teachers can choose to integrate into their instruction plans. The lessons aim to demystify AI by teaching students about data collection and object classification using fossil shark teeth, showing students how to train and evaluate machine learning models while recognizing a model's limitations, and supporting students as they create unique AI models.

But to best prepare students to use AI, teachers must also know how to use the technology. One critical component to using Shark AI is having access to real fossils.

Credit: Christine Wusylko, University of Florida

"With Shark AI, students have access to high-quality, real fossils that they can touch and feel," MacFadden said. "The kids love having that, and so do the teachers."

The other key piece is the support provided through professional development for Florida middle school educators who are voluntarily implementing Shark AI in their classrooms.

"If teachers aren't given opportunities to learn, they can never bring that knowledge back to their schools," said Seema Naik, an engineer-turned-sixth and eighth grade educator, and this year's Career Technical Education Rookie Teacher of the Year for her school in Broward County. "I never imagined that you could actually make a connection between fossils and AI, that you can train a model and do this. That was quite fascinating," she said.

The Shark AI team hosts a weeklong summer workshop to support teacher cohorts. The educators learn from various experts, including other middle school 'teacher leaders,' who use the curriculum.

Credit: Erika Nortemann/Florida Museum

"One thing the teacher cohorts enjoy most about the professional development is working with other teachers — their peers — who facilitate a lot of these AI learning experiences," said Pavlo "Pasha" Antonenko, UF professor of educational technology and co-PI on the project. "From this project, we have learned that teachers can use AI. Teachers are using AI. We just need to help them do it more effectively," Antonenko said.

Importantly, this peer-to-peer engagement helped teachers like Naik unpack some of the common misconceptions she and others held about AI.

"I had several questions and concerns about the technology before I went through the training, but now I am comfortable using AI in my class," she said. "Learning how AI could be used positively, that was pretty powerful."

Credit: UF College of Education, University of Florida

The community teachers develop while participating in the Shark AI training is also key as they navigate the new curriculum. For example, Naik mentioned that some Florida schools use Microsoft applications while others use Google. This caused a few technical hurdles for some of the Shark AI teachers, but they worked together to troubleshoot and ultimately share their best practices.

And to help teachers make clear connections for their students between the science classroom and the STEM workforce, the project offers "role model visits," where a research team member visits science classes to engage directly with the students using Shark AI.

"Going to a class and saying: 'I am a real scientist, and if you want, you can be too. You are classifying fossils using machine learning tools, and that is exactly what I do as a scientist.' That is very powerful for teachers, not only for the content support, but also for the science identity that the students might come away with," MacFadden said.

For Naik, Shark AI has not only given her a better understanding and confidence to use AI in her science classes, but she recognizes how the curriculum is giving her students "ownership of their learning," something she says is key as they move into high school and become more self-sufficient.

"This project really champions teachers," emphasized Christine Wusylko, UF postdoctoral researcher on the team. "Shark AI teachers feel special, and they feel part of a club. The curriculum is flexible, and it's led by teachers. And as a former science teacher myself, that is invaluable."