Fall Public Lecture: Alignment Beyond Correctness – ARIA

An evening lecture on what it means for AI to align with human values in domains without clear right answers.

Areas of Study

Trustworthy Multimodal Interaction

Studying how people perceive and interact with AI across speech, vision, text, and gesture to build trustworthy systems.
Reasoning in Humans and AI

Exploring how humans and AI systems reason differently and what this means for collaboration and decision-making.
Alignment in Subjective Domains

Investigating how AI can align with diverse human values, preferences, and cultural contexts in areas without clear-cut answers.
Participatory Design

Centering the people most affected by AI in the design process, especially communities traditionally excluded from tech development.