University of ConnecticutCollaborative research initiative
Reimagining AI interfaces.
The PAL (Pluralistic AI Lab) is a collaborative research initiative at the University of Connecticut, an interdisciplinary effort among the Connecticut Advanced Computing Center (CACC), the Department of Digital Media and Design (DMD), and the UConn Humanities Institute (UCHI). It is led by a cross-disciplinary team of artists, designers, technologists, researchers and grassroot organization leaders. The lab’s mission focuses on reimagining AI interfaces and architectures to move beyond “monolithic” single-person chatbots.
Researching multi-persona and multi-agent systems that present users with diverse, even conflicting, perspectives to reorient AI interactions from single, authoritative, or “polite” consensus to preserving “honest friction.”
Creating culturally calibrated AI agents that enable marginalized communities to participate in policy simulations and democratic deliberation rooted in their lived experiences.
Combating the “three faces of harm” in standard chatbot designs: cognitive offloading (erosion of critical thinking), ideological narrowing (manufactured consensus and bias), and epistemic compression (reduction of cultural and value diversity).
Our work spans theoretical frameworks, system design, and real-world pilots with underrepresented communities.
The Co·Agents project—a collaboration between researchers at UConn and Tayo/FYLPRO—develops a multi-agent AI system for cross-cultural cooperation in democratic deliberation. By utilizing culturally calibrated AI agents, Co·Agents enable underrepresented communities to run policy simulations at a fraction of the cost of traditional “mini-publics.”
The project leverages Tayo’s role as a trusted community information hub to ensure AI-driven governance is rooted in the lived experiences of the global Filipino diaspora, and implements the Kierans framework to measure “honest friction” and misalignment.
AbstractThe current paradigm of artificial intelligence (AI) design elevates conversational chatbot interfaces, making them the default mode through which hundreds of millions of people interact with AI systems globally. This narrative review identifies three related harms produced by the single-voice chatbot interface, drawing on insights from literature in cognitive science, political studies, AI alignment, game design and human computer interaction. Firstly, cognitive offloading of user thinking, secondly, ideological narrowing of user beliefs, and thirdly, epistemic compression of cultural and value diversity. We then propose multi-persona interfaces, in which multiple personas are presented to the user simultaneously, as a cost-effective intervention that addresses all three harms at once. We ground this intervention in traditions the AI literature has largely overlooked, drawing on techniques from game design and participatory simulation, including the magic circle, role differentiation, and persona embodiment. We propose a taxonomy for the future design of multi-persona interfaces, and call for a research agenda that closes the gap between technical pluralistic alignment methods and empirical studies of human-AI interactions with pluralistic interfaces.
Researchers and technologists spanning AI, art, design, cognitive science, political theory, and engineering.
Artist, Designer, and Assistant Professor in the Department of Digital Media and Design at UConn. Researching software simulation as a mode of collective un-looping and re-making. Her computational art has been curated by the New Museum, Art Basel, Ars Electronica, and Electra, and she has spoken at ELO, CAA, New York Creative Tech Week, and ISCP.
Lead Technical AI Policy Researcher at Hugging Face and Research Affiliate at UConn. He leads the EvalEval Coalition, a multi-stakeholder initiative for rigor in AI evaluations. His work spans algorithmic bias, agent autonomy, and vulnerability disclosure, and has been covered in The New York Times, Forbes, The Guardian, and Wired. His research has shaped AI regulation including in New York City.
Assistant Professor in the School of Data and Information Sciences at UNC Chapel Hill and Affiliate Researcher to the Center for Information, Technology and Public Life. His work is on technological bias against historically marginalized populations often overlooked in traditional fairness research.
MPhil candidate in Human-Inspired AI at the University of Cambridge, where she focuses on developing methods for evaluating and mitigating bias in LLM-powered social robots. She previously studied computer science and law at the University of Washington, founded a digital literacy program for marginalized learners, and has worked to bridge design, engineering and governance in robotics and AI.
MPhil candidate in Human-Inspired AI at the University of Cambridge. Previously a Senior Product Manager at Atlassian building AI systems used by hundreds of thousands globally, and co-leading their responsible AI governance program. Her research investigates how AI shapes individual beliefs and how those beliefs transmit to naive third parties, with implications for epistemic autonomy and democratic resilience.
Designer and technologist whose work sits at the intersection of architecture, urbanism, and computer science. Holds a Master of Science from MIT with experience in full-stack development, AI-driven urban platforms, shared mobility simulation for autonomous vehicles, and award-winning IoT devices for climate monitoring and beehive optimization.
Technical engineer and socio-technical policy researcher whose work centers on differential privacy and ethical AI governance. He has conducted research in AI alignment and pluralism at UConn and algorithmic fairness at UMich. As Director of Tufts AI Safety, he established the AI Policy Fellowship and Hackathon, and serves on the board of the Auster Center Civic Innovation Program.
UX/Product Designer and MFA candidate at UConn specializing in UX/UI, motion, and interactive media design. With a professional background in product development and strategy, she creates thoughtful, user-centered digital experiences that connect research, design, and emerging technology.
Filmmaker and visual storyteller, and an MFA candidate in Digital Media & Design at UConn. His work explores storytelling across forms and media, including narrative film and AI-assisted virtual reality projects. At PAL, his research examines the foundations of perspectival plurality and what is lost when many viewpoints are collapsed into a single authoritative voice.
This project has received the generous support of the University of Connecticut Humanities Institute, the School of Fine Arts Research Grant, and the Office of the Vice President for Research in helping to make this project possible.
We also work alongside community organizations to ground our research in real needs and lived experiences, and are grateful to our community partner, Tayo / FYLPRO.