Empowering communities through responsible AI systems
The Human-Centered Responsible AI (HRAI) Lab is dedicated to designing, developing, deploying, and studying interactive artificial intelligence (AI) systems to enhance human-AI collaboration, augment human intelligence, and democratize access to AI empowerment technologies in a responsible way.
HRAI adopts a system-centric approach to creating AI-enabled socio-technical systems that augment human capabilities in various practical applications that tackle society’s “wicked problems”.
The mission of the lab is to advance fundamental research in the science of human-AI interaction with a clear lens on societal challenges. The lab will leverage the existing strengths of the Institute by utilizing and amplifying collaboration with the industry community, incorporating state-of-the-art machine learning and AI methods, and engaging with the local community for broader impacts.
Since its establishment in 2024, the Lab has been directed by Toby Jia-Jun Li, assistant professor of computer science and engineering. Li is also a Lucy Family Institute fellow.
Focus
Key characteristics of HRAI’s focus include:
- System-focused methods: RHAS focuses on the design, building, deployment, and analysis of practical systems as a method of inquiry that contributes generalizable scientific knowledge in three areas:
- Developing new theoretical frameworks, design methodologies, and evaluation metrics for human-AI interactive systems in service of society
- Assessing the ecological validity of AI techniques and models to identify gaps and provide insights that guide foundational AI research.
- Validating, characterizing, and further evolving the theories and guidelines in social and behavioral sciences
- Multidisciplinary efforts: HRAI champions an inherently interdisciplinary approach, grounding and informing its system design and study methodologies in social, behavioral, and liberal arts disciplines, including psychology, cognitive science, sociology, economics, design, education, ethics, and others. It aims not only to apply these insights to computing research but also to contribute back to theories, empirical findings, and implications in these areas.
- Human-centered approaches: At the core of HRAI’s research philosophy are the diverse needs, preferences, and abilities of individuals, organizations, and communities. It specifically targets to democratize AI technologies to underserved stakeholders, aiming to bridge the digital divide and mitigate disparities in AI access and impact.
- Comprehensive metrics: HRAI employs a multifaceted process-centric approach to planning and assessing the impact of the systems it creates. This approach goes beyond measuring the quality and efficiency of outputs, focusing on human-centric factors such as trust, reliance, agency, transparency, knowledge acquisition, behavioral changes, sustainability, ethical implications, and long-term community impacts.
