Middle School Data Science Workshop (Jefferson Lab photo/Lindsay Cunningham)
Jefferson Lab hosts workshop to help middle school students learn about AI
NEWPORT NEWS, VA - Middle schoolers are learning that training Artificial Intelligence (AI) models can be a serious business, but sometimes it can be as fun as engaging in a viral game.
The 2024 Middle School Data Science Workshop, organized and hosted by the Science Education team at the U.S. Department of Energy’s Thomas Jefferson National Accelerator Facility, used a viral mobile game to help engage a group of middle schoolers in learning about AI.
Jefferson Lab’s Science Education team is committed to encouraging and promoting a diverse and talented future STEM workforce. This new workshop focuses on building skills and confidence in STEM for this young demographic. By focusing on middle school students, the workshop aims to spark curiosity in STEM early by providing opportunities for hands-on learning.
“These students aren’t researchers working on our projects,” said David Lawrence, a Jefferson Lab software scientist and member of the lab’s Experimental Physics Software and Computing Infrastructure group who volunteered to lead the workshop. “These are the future scientists we want to encourage.”
Lawrence’s passion for working with middle school students began in 2011 when he coached the Hines Middle School robotics team.
“I’ve always enjoyed connecting with middle schoolers,” Lawrence said. “Sixth graders get attention because they’re new and eighth graders receive recognition as they prepare for high school, but seventh graders often don’t have as many programs focused on them.”
The lab chose to focus the program on seventh graders. Lawrence’s experience with middle school students made him the perfect person to lead this program.
Jefferson Lab’s Science Education team invited 30 students from three middle schools in the Newport News Public Schools district to participate in the workshop that was held in October. Teachers were encouraged to select students with emerging potential in STEM classes to help create a more inclusive and diverse group.
“We looked for students who don’t believe they can succeed in STEM,” Lawrence said. “These students have more potential than they realize.”
During the workshop, students received a tour of the Jefferson Lab Data Acquisition Center. They also used Google Colab to learn how to train AI through the popular game Flappy Bird.
Students began by playing the side-scrolling arcade game on their laptops. After each game, they watched a database plot their scores in real-time on a histogram. Next, they trained AI to play the game, observing and fine-tuning how the AI improved over time. As they did this, the histogram displayed live updates, comparing their scores to the AI’s. These live visuals helped them understand data and how to read graphs.
“AI training is just adding, subtracting and multiplying numbers on a large scale—skills they already know,” Lawrence said. “I wanted to show them that this is something they’re capable of doing.”
Lawrence also wanted students to understand they can influence modern technology.
“I don’t want students to think, ‘This is for smarter people in bigger places,’” Lawrence said. “I want them to know the world is accessible and that all of this is within their reach.”
Although this is the first year Jefferson Lab has hosted the workshop, Lawrence believes it will continue annually and lead to more opportunities for students to explore science.
“I’ve seen the benefits of inclusion in science, and this is something everyone needs to address,” Lawrence said. “We should recognize that not everyone has the same opportunities and it’s up to us to create them.”
Further Reading
Jefferson Lab's Science Education Programs
PIER Plans
Bridging the Gap Between Physics and Computing
By Xavier Gardner
Contact: Michelle Alvarez, Jefferson Lab Communications Office, malvarez@jlab.org