Autumn 2025
Understanding Online Community Norms to Build an AI Agent
Co-Directed by PhD student Soobin Cho, Dr. Mark Zachry and Dr. David McDonald
Every online community has norms and policies that shape member behavior. This DRG focuses on Wikipedia, one of the largest examples of online collaboration, where massive collaborative efforts create encyclopedic articles supported by complex community policies. Wikipedia policies vary from relatively simple to intricate. Some policies are complex and multi-faceted making them difficult for humans to understand, presenting an opportunity for AI assistance.
In this DRG, students will gain foundational knowledge about online collaboration with a focus on Wikipedia and AI training methodologies. The core work involves collecting and conducting qualitative analysis of real cases where Wikipedia policies are applied, contributing to a dataset that will be used to train norm-aware AI agents.
We are looking for a small group of DRG participants with:
- Interest in online collaboration and online community dynamics
- Prior experience with qualitative analysis
This is a 2-credit DRG offered to undergraduate (HCDE 496) and graduate (HCDE 596) students. Everyone participating in the DRG must be registered for either HCDE 496 or 596. Students participating in this DRG will be meeting weekly on Wednesdays from 3:00 PM to 4:30 PM. Additionally, students are expected to spend 3 to 5 additional hours per week outside of class.
Deadline: Apply for this DRG by September 19th by filling out this Google Form. We expect to select and notify participants by September 22. Please reach out to Soobin Cho through email (soobin30@uw.edu) or HCDE Slack with any questions.
Autumn 2025
Designing the Future of Human-Agent Interaction
Led by Kevin Feng and David McDonald
AI agents—compound AI systems that can tackle open-ended tasks over an extended period of time—may usher in the next wave of AI and UX innovation after today’s chatbots. AI agents that learn from our preferences and act on our behalf can amplify our abilities and empower us to accomplish more in our personal and professional lives. However, they also bring consequential risks, from acting against human instructions to causing mass job loss due to labor automation. Carefully designing human-agent interactions is critical to realizing the potential of AI agents and mitigating their risks.
In this DRG, you’ll explore what future human-agent interactions and interfaces may look like. You’ll be challenged to think creatively beyond the bounds of existing interaction paradigms to design new ones tailored for AI agents. Concretely, you can expect to learn about:
- Design challenges for human-centered AI systems, and in particular, AI agents
- Key design ideas proposed by academic papers and industry products
- Gaps and opportunities for design innovation in today’s landscape of agentic systems
The benefits for you are:
- Developing useful skills and knowledge for being a designer in the AI agents era
- Creating a series of designs which you can showcase in your portfolio
- Work done as part of this DRG may be integrated into a larger project that has a direct impact on industry stakeholders (i.e, tech companies building agents).
- Obtaining DRG credits for fall quarter
This DRG will have very limited capacity. If you’re interested, please apply by filling out our short application form. The strongest applicants may be:
- Curious about or have prior experience designing AI-powered applications
- Experienced with UX design tools (e.g., Figma)
- Not afraid to dig into technical concepts
If you are accepted to this DRG, you will be notified by Friday Sept. 19, 2025. This DRG will meet weekly for 90 minutes at a time to be determined once the quarter starts. Students taking this DRG must enroll in a minimum of 2 credits.
Find the application form here. We look forward to receiving your submission!