Stanford CS547 HCI Seminar | Winter 2026 | Computational Ecosystems

Key Concepts

• Computational Ecosystems: Integrated systems where practices, social structures, and technologies are designed as a coherent whole rather than piecemeal.
• Critical Technical Practice: A methodology involving deep reflection on how technology is designed and the underlying assumptions of the practice it supports.
• Socio-technical Configuration: The combination of human social structures and technical tools that define how an activity is performed.
• Regulation Skills: Cognitive, metacognitive, motivational, and emotional skills required for self-directed work (e.g., planning, help-seeking, risk diagnosis).
• Dialectical Activities: Human activities that are intrinsically valuable and cannot be fully captured or optimized by output-oriented technology.
• Opportunistic Collective Experiences (OCEs): Systems that facilitate remote connection by identifying shared environmental affordances rather than relying on passive social media scrolling.

1. The Philosophy of Computational Ecosystems

The speaker argues that HCI (Human-Computer Interaction) often focuses too heavily on "progressing" through incremental technological improvements without questioning the underlying "way of doing things." The core argument is that many human problems cannot be solved by better technology alone; they require a redesign of the entire socio-technical ecosystem.

• Systemic Design: Instead of layering technology over existing, broken practices, designers should conceive of practices, social structures, and tools simultaneously.
• The "Atul Gawande" Perspective: Echoing the medical field, the speaker emphasizes that we are obsessed with components (best drugs, best tools) but neglect how they integrate into a functional system.

2. Real-World Applications and Case Studies

The speaker presented five specific domains where this ecosystem approach was applied:

• Community-Informed Planning (KAI/CSCW Conferences): Replaced manual, opaque scheduling with a "mixed-initiative" system. It engaged 1,500+ community members to provide preferences, allowing organizers to resolve conflicts while retaining human oversight.
• Flexible Coordination (Hit or Wait): A system for mobilizing crowds for local tasks (e.g., lost and found, accessibility data). It uses decision theory to ping users opportunistically when they are already "on the go," minimizing effort while maximizing system coverage.
• Opportunistic Collective Experiences (OCEs): A shift from passive social media scrolling to active, shared experiences at a distance. The system identifies when users are in similar physical contexts (e.g., watching a sunset) to facilitate meaningful connection.
• Readily Available Learning Experiences (Isopleth): A tool for novice developers that transforms professional websites into learning resources. It visualizes event-driven relationships and data flow, moving away from simplified tutorials toward "participation in expert practice."
• Agile Research Studios (ARS): A socio-technical model for research training. It replaces the "master-apprentice" bottleneck with a community-based model where students learn regulation skills (planning, help-seeking) through peer-to-peer interaction and structured reflection.

3. Methodologies: The "Regulation-Informed" Approach

A significant portion of the talk focused on the Capstone Notes framework used in the DTR (Design Technology and Research) program:

1. Identify Patterns: Instead of troubleshooting individual work outputs (e.g., a broken feature), mentors identify underlying "regulation gaps" (e.g., a student’s tendency to prioritize delivery over understanding).
2. Practice Objects: Computational representations that track how a student’s practice evolves, linking work issues to specific regulation behaviors.
3. Holding Space: Weekly studio meetings (45–60 minutes) dedicated to reflection, sharing, and acceptance, treating the research process as a vehicle for personal growth rather than just paper production.

4. Key Arguments and Perspectives

• Beyond Consequentialism: The speaker critiques HCI for being a "consequentialist enterprise" that values only what is produced. They argue for designing systems that support "intrinsically valuable" human activities.
• The "Unflattened" Internet: Current platforms (like Yelp) flatten human experience into a single metric (e.g., "cozy"). The speaker proposes "experiential computing" (e.g., the Differ platform) that accounts for diverse contexts, accessibility, and individual needs.
• The Role of AI: The speaker notes that their current state-of-the-art solutions do not rely on Generative AI. They view GenAI as a "strong component" but caution against using it to merely optimize existing, potentially flawed, socio-technical configurations.

5. Notable Quotes

• "What’s broken is not only our technology that supports our current ways of doing things, but our way of doing things in and of itself."
• "The real question for all of us in HCI isn’t just what can our socio-technical systems produce, but also what kind of lives do they invite us to live?"
• "I can ask for help, that everyone asks for help, and it doesn’t make them stupid to need help." (Student testimonial on the ARS model).

6. Synthesis and Conclusion

The main takeaway is a call to action for designers and researchers to move beyond "technological values"—the tendency to align our lives with what technology provides—and instead build systems that sustain our human values. By acknowledging the limits of technology and focusing on the "larger ecology" of human activity, we can create systems that foster personal growth, deep engagement, and meaningful connection. The speaker emphasizes that this work is slow, difficult, and requires a commitment to "dancing with not knowing," but ultimately leads to a more fulfilling and authentic way of living and working.