AI research discovery event highlights interdisciplinary strength at Tufts University

The excitement for AI research at Tufts University was evident at the recent Human-AI Interaction Research Discovery Event. Hosted by the Human-AI Interaction Center (HAII), part of Tufts Institute for Artificial Intelligence (TIAI), the evening prompted rich exchanges among AI-focused researchers at Tufts. Dozens of faculty, staff, and students gathered in Joyce Cummings Center to share their current projects and explore future research possibilities in AI.

Expertise from many disciplines strengthens AI work

The attendees did not fit neatly into a single academic discipline. Their backgrounds, spanning community health, philosophy, computer science, and politics, to name a few, were as varied as their current work. The evening began with a poster session where students shared their current interests and recent projects. The posters on display wove together expertise from students in the School of Engineering, School of Arts and Sciences, Tufts University School of Medicine, and Tufts University School of Dental Medicine, with AI as the common thread.

Tufts’ interdisciplinary ethos drew Shepard Rodgers, A27, to the university, where he is pursuing an undergraduate degree in computer science. “Having a computer science department that’s intentionally interdisciplinary, hands-on, and active was very appealing to me,” he shared. Over the summer, he was part of a group that worked at the intersection of AI and social contexts to create a community platform for residents in the Talbot-Norfolk Triangle neighborhood in Boston. The AI-driven platform integrates residents’ own perspectives and input with official data. Rodgers and fellow researchers hope the platform will become a central hub for neighbors to share and seek important information.

Julia Santaniello earned her undergraduate degree in neuroscience and transitioned to a master’s in computer science through the School of Engineering’s post-baccalaureate in computer science program. Now, she’s a second-year Ph.D. student working with Associate Professor Jivko Sinapov on projects that combine her interests in neuroscience and computer science. One example is a framework that uses brain signals to help train AI agents. Their work paves the way for future brain-driven reinforcement learning human feedback systems. She credits the collaborative spirit among Tufts researchers with helping her find her niche. “Here, it’s really collaborative. If someone has a gap they need to fill it’s easy to find someone that can fill that need.”

AI research shines a light on collaboration at Tufts

From granular improvements in AI development to more sweeping applications, AI research brings together unlikely collaborators. One such faculty pair: Peter Levine, associate dean for academic affairs and Lincoln Filene Professor of Citizenship & Public Affairs in the department of political science and James Intriligator, professor of the practice in the department of mechanical engineering. The pair are working together on an AI-enabled chatbot that could help organize democratic community groups. The team presented during the faculty lightning talks, which were divided into four categories: Medicine, Health and Education, Autonomy Learning and Shared Space, LLMs and Challenges of Competence, and Civic Society and AI’s Purposes.

Nearly 20 faculty presented ideas including how to leverage wisdom from wheelchair users to improve robot navigation (Clare Booth Luce Assistant Professor Elaine Short), how AI can support engineering education (Director of the Center for Engineering Education and Outreach [CEEO] Merredith Portsmore and Robotics Engineer Milan Dahal, CEEO), and how to build large language models that say no rather than hallucinating (Professor Reuth Mirsky). Professor of the Practice Mat Rappaport of the Department of Film and Media Studies discussed his collaboration with data analytics graduate student Ishmam Khan to create media art installations that explore how AI visually represents descriptions of ecological environments from the early 1900s using contemporary generative AI.

Using AI to enhance medical training and practice

Across the board, Tufts researchers are committed to creating a better future through AI. Michael Song, A23, started at Tufts as an undergraduate student in biology and community health. Now, he is at the forefront of integrating AI into dental training and practice. With over 200 students, Tufts University School of Dental Medicine boasts one of the largest dental student populations in the United States. The school recently taught a mandatory course in AI and dentistry for third-year dental students—one of the first of its kind in the United States. Song evaluated student feedback data about the usefulness of the course and came up with recommendations to fine-tune the curriculum moving forward. “For a profession as hands-on as dentistry, continuing education is really important,” he said. “We need to prepare tomorrow’s clinicians for tomorrow’s problems.”

Medical students Shreya Asher, M27, and Vajipayajula Dhruv, M28, are similarly focused on how AI can help train the next generation of healthcare workers. They devised an AI-driven simulation model for medical students to practice wound care treatment. Beyond training for medical professionals, faculty members like Wittich Family Assistant Professor Mike Hughes of the Department of Computer Science and Associate Professor Benjamin Wessler of the School of Medicine are improving medical devices for better health outcomes. Their upgraded stethoscope uses deep learning algorithms to better detect heart disease in primary care screenings.

Building connections in AI research

Director of the Human-AI Interaction Center and Karol Family Applied Technology Professor Matthias Scheutz organized the evening with the help of TIAI staff members Thomas Arnold and Meghan Rodriguez. During the event he demonstrated AVA, a social robot that welcomed people at the door and alerted participants when it was time to move to the faculty talks. Scheutz expanded on this demonstration during his faculty talk, where he discussed the possibility of integrating social robots into coffee shops or dining halls on the Tufts campus. Scheutz hopes the evening strengthened connections and sparked new ideas for attendees. Put simply, “the point is to grow the Tufts AI community,” he said.

Tufts Institute for Artificial Intelligence serves as the central hub for AI research, education, and collaboration at Tufts University. By fostering interdisciplinary partnerships across departments and research domains, TIAI brings together faculty, industry experts, and students to ensure that AI makes a positive, meaningful, and ethically responsible impact on the world.

Safety and ethical use are top of mind for Tufts AI researchers like Human-Robot Interaction Ph.D. student Katie Kowalyshyn. She examined different methods that developers use to safeguard AI. Combining her passion for computational linguistics and philosophy, her work explores the critical question: what happens if AI goes rogue? Ph.D. student Sveta Paster agrees that there is a lot of room for growth in AI development. “I think a lot of people are afraid of AI or think that large language models (LLMs) can do everything. There’s a lot of interesting problems that still need to be solved. The creative part is still our responsibility,” she remarked. That creativity was on full display at the research discovery event and continues to grow as Tufts researchers work together to shape artificial intelligence in useful and responsible ways.

See the booklet of abstracts and one slide presentations.