Human-AI
Interaction
Design
@ Cornell

This is the home of the DesignAI group at Cornell Computing and Information Science, a boutique research group focused on advancing human-centered tech design methods for the age of Artificial Intelligence. We have three superpowers: We create first-of-their-kind AI applications that address messy real-world problems. We conduct qualitative user research and evaluation in the wild. And from that work, we derive novel design thinking, design methods, and design tools.

News Our Dagstuhl Seminar β€˜NLP Meets HCI’ was accepted πŸŽ‰ Join us at Schloss Dagstuhl, Wadern, Germany, May 17–20, 2027! Two papers accepted to the AAAI/ACM conference on AI, Ethics, and Society (AIES) πŸŽ‰ See you in MalmΓΆ, Sweden! Talia Wise successfully passed her Ph.D. thesis proposal (A exam) πŸŽ‰ Welcome Yuewen Yang, our new Ph.D. student, to the group! πŸ™Œ πŸ“’ Qian Yang is going on a talk tour! Join her at Cornell IS Colloquium (9/9), Harvard CHARM group (9/24), Stanford HCI Seminar (10/9), UC Berkeley (10/13), Carnegie Mellon HCII Seminar (10/23), and U Washington DUB Seminar (date coming soon)

HAI Design Research

  1. Profiling AI as a Material for Software Design

    When and how do existing human-centered technology design methods fall short in translating AI's technical advances into benefits to users and society?

    Featured work
  2. Methods and Tools for Designing GenAI as Thinking Partners

    How can we design generative AI applications that influence users' thinking at the right time, in the right way, and to the right extent?

    Featured work
    • CoAuthor πŸ… Understanding the cognitive impact of writing with LLMs. Paper Β· Dataset & platform
    • PAIL πŸ… Designing vibe coding tools that augment software design across multiple levels of abstraction. Paper
    • Truman Agents Crafting interactive, multi-agent social media simulations to cultivate bystander intervention against cyberbullying. Paper Β· Platform
  3. Towards Aligning Infrastructural AI platforms with Public Interest

    From AI-mediated social media platforms to ChatGPT and Claude, many of the most impactful AI applications are for-profit societal infrastructure: they mediate critical societal functions (e.g., civic discourse, public health education), yet in ways shaped by commercial incentives that do not always align with the public good. Existing research calls for regulation or pursues activism to address this misalignment. What can we, as designers, do to help, given our distinctive expertise in building AI things that address ill-defined, messy societal problems?

    Featured work

HAI Design Courses

HAI Design Events