Two Penn GSE Doctoral Students Selected as Penn AI Fellows

Artificial intelligence is reshaping research across disciplines, and at Penn GSE, two doctoral students are helping lead that exploration. Luis Morales-Navarro and Shruti Mehta, both PhD candidates in the Learning Sciences and Technologies (LST) program, were recently selected as members of the Penn AI Fellows program, a new interdisciplinary initiative connecting researchers working with AI across the University.

The fellowship brings together doctoral students and postdoctoral researchers from nine Schools at Penn who are applying AI in their work. Through regular meetings, collaborative events, and research exchanges, the fellows will examine how AI methods can be translated across fields and how scholars from different disciplines can learn from one another.

For Morales-Navarro and Mehta, the opportunity connects their education research to a broader network of scientists, engineers, and social scientists exploring the future of AI.

Morales-Navarro added that the fellowship offers a chance to extend his work into new interdisciplinary spaces, bringing his focus on young people’s interactions with AI into conversation with researchers across fields. As he emphasized, “young people have important insights into how AI systems affect their lives,” and the program provides an opportunity to elevate those perspectives within broader discussions about the future of AI.

“This program grew out of our Data Science Fellows initiative,” said Colin Twomey, executive director of the Data Driven Discovery Initiative at Penn’s School of Arts and Sciences. “As AI became increasingly relevant across fields, it made sense to expand beyond a single school and bring together researchers who are using and developing these methods in very different contexts.”

That expansion became possible through a partnership with PennAI, a new initiative launched by the Office of the Vice Provost for Research. The new Penn AI Fellows program represents the first University-wide cohort. Fellows participate in weekly lunches with faculty researchers, collaborative discussions, and peer-led events focused on emerging AI tools and research methods. The program’s central goal, Twomey said, is to break down disciplinary silos that often shape academic research. 

“You spend most of your time working within your own field,” he said. “The fellowship creates opportunities for longer-distance conversations across disciplines. That’s where surprising new ideas and collaborations can emerge.”

Twomey also emphasized that education researchers bring a particularly important perspective to the AI conversation.

“When people think about AI, they often focus on the technical side,” he said. “But questions about learning, teaching, and human interaction with technology are just as important. Having scholars from education involved in these conversations helps broaden how we think about the impact of AI.”

For Morales-Navarro, the fellowship builds on years of research examining how young people understand and interact with AI.

Born in Costa Rica and educated at New York University (NYU) Abu Dhabi, Morales-Navarro first became interested in computing education while teaching introductory programming courses as a recitation instructor at NYU Shanghai. Watching students struggle with coding helped him recognize how differently learners respond to challenges in technical fields.

“Some students saw debugging as a puzzle to solve,” he said. “For others, it felt like a reflection of their own abilities.”

Those experiences led Morales-Navarro to pursue graduate study at Penn GSE, where he first completed an LST master’s degree before continuing into the PhD. Working with advisor Yasmin Kafai, the Lori and Michael Milken President’s Distinguished Professor at GSE, he developed research examining how young people engage with emerging technologies. His work focuses on helping children move beyond being passive users of AI and instead becoming active participants in their design and evaluation.

“I’m interested in children first,” Morales-Navarro said. “I study how young people understand AI systems in their everyday lives and how we can involve them as partners in designing and evaluating those systems.”

Through participatory design research, Morales-Navarro works with students to examine how technologies, such as generative AI tools, function and influence everyday digital experiences. His dissertation research explores teens’ mental models and understandings of generative language models and how these may change when engaging in model design and auditing activities. He spent last summer teaching high school students how to build their own small generative language models—or “babyGPTs”—at the Franklin Institute.

“Many conversations about AI literacy focus on teaching students how to use AI tools,” Morales-Navarro said. “But it’s also important for learners to understand how these systems work and to recognize that they are designed by people.”

Mehta’s research focuses on how technology can support learning in large-scale and asynchronous educational environments.

Originally from New Delhi, India, Mehta studied mathematics and education as an undergraduate at the University of Delhi, then pursued a master’s degree in society and culture at the Indian Institute of Technology Gandhinagar before coming to Penn GSE for her doctoral studies.

Her research explores how AI can help identify effective instructional strategies in environments where teachers cannot provide direct guidance, such as massive open online courses (MOOCs) or other asynchronous learning platforms.

“When you don’t have a teacher physically present, instruction looks very different,” she said. “My work asks what kinds of strategies actually support learning in those environments.”

Using large datasets from online courses and AI-based analytical tools, Mehta studies how different instructional approaches influence learner engagement and success. She is particularly interested in understanding how AI can help support students in under-resourced contexts where teachers face large class sizes or limited time.

Growing up in large classrooms and later teaching students in India, she saw firsthand how asynchronous resources, such as online videos and open courses, could help learners access knowledge beyond traditional classroom structures.

“Sometimes teachers simply don’t have enough time or resources to support every student,” Mehta said. “Technology can help fill some of those gaps.”

At the same time, she emphasizes that AI should complement teaching rather than replace it.

“I see AI as a tool for experimentation,” she said. “The key question is how those tools actually work when they reach real classrooms.”

“The idea is not to replace teachers,” said Associate Professor Seiji Isotani, Mehta’s advisor and faculty director of Penn GSE’s Learning Analytics and AI program. “It is to augment their capacity to support students.”

For Isotani, the Penn AI Fellows program represents an important opportunity for education researchers to contribute to the broader conversation about artificial intelligence.

“AI is a huge field,” he said. “If you only approach it from the technical perspective, you miss insights from areas like education, psychology, and policy. When students from different disciplines interact regularly, they begin to see new possibilities for their research. Those interactions often lead to new ideas and innovations.”

“Artificial intelligence is not just a technical problem,” said Twomey. “It has social, educational, and ethical dimensions. Bringing together researchers who study those different aspects is incredibly valuable.”