Stanford CS547 HCI Seminar | Winter 2026 | Computational Ecosystems
TL;DR
The speaker argues that to solve persistent human problems in HCI, designers must move beyond building better tools and instead critically reimagine entire socio-technical ecosystems. Through examples in event planning, crowdsourcing, social connection, and education, he demonstrates how redesigning human practices—what he terms "critical technical practice"—can unlock values that pure technological advancement has failed to address.
💭 Critical Technical Practice 3 insights
Fundamental limitations of current approaches
Despite decades of computing advances, persistent human problems like effective student mentoring remain unsolved because the socio-technical practices themselves are broken, not just the supporting tools.
Redesign ecosystems, not just artifacts
True innovation requires reimagining entire socio-technical configurations and workflows rather than simply building better optimizers or interfaces.
Reflection as methodology
Drawing on Philip Agre's concept of "critical technical practice," designers must engage in ongoing reflection on their values and the inherent limitations of how they currently work.
🤝 Collective Coordination 3 insights
Community-informed event planning
By engaging over 1,500 community members to source preferences and constraints, a new mixed-initiative approach reduced academic conference planning time by more than an order of magnitude while resolving hundreds of hidden conflicts.
Flexible crowd coordination
The "Hit or Wait" system applies decision theory to ping volunteers opportunistically based on their location and convenience, achieving near-optimal outcomes for local services like lost-and-found without rigid directives or chaotic volunteerism.
Listening for opportunities
Effective coordination requires listening for when contributions align with both individual convenience and collective goals rather than enforcing predetermined workflows.
🌱 Connection and Learning 3 insights
Opportunistic Collective Experiences
OCEs replace passive social media scrolling with context-aware shared activities, proving as connecting as direct messaging but with lower initiation barriers and enabling novel experiences like synchronized sunset watching across time zones.
Learning from professional code
The Isopleth tool transforms complex professional websites into interactive learning resources, allowing novice developers to learn directly from production code rather than simplified tutorials.
Technology as expression of care
These redesigned ecosystems empower users to express care—whether shaping conferences, helping neighbors, or maintaining distant friendships—by creating new human practices rather than merely adding features.
Bottom Line
Advance human values by critically reimagining entire socio-technical ecosystems and human practices rather than incrementally improving existing technologies.
More from Stanford Online
View all
Stanford CS336 Language Modeling from Scratch | Spring 2026 | Lecture 16: Post-Training - RLVR
This lecture explains why RLHF hits overoptimization limits with learned reward models, and how RLVR (Reinforcement Learning from Verifiable Rewards) enables unlimited compute scaling on verifiable tasks like math and coding through simpler algorithms like GRPO.
Stanford CS336 Language Modeling from Scratch | Spring 2026 | Lecture 15: Mid/Post-Training
This lecture explains how post-training transforms raw pre-trained models like GPT-3 into instruction-following systems like ChatGPT through supervised fine-tuning and reinforcement learning, emphasizing that high-quality data curation matters more than algorithmic sophistication.
Stanford CS336 Language Modeling from Scratch | Spring 2026 | Lecture 14: Data
This lecture details the pre-training data pipeline, covering the transformation of raw HTML and PDFs into linear text and classifier-based filtering strategies to curate domain-specific datasets, while emphasizing the strategic trade-off between data quality and training duration.
Stanford MS&E435 Economics of the AI Supercycle | Spring 2026 | Infrastructure, Capstone Case
Sachin Katti, OpenAI's head of industrial compute, details the infrastructure economics driving the AI supercycle, explaining how the company plans to scale to 30 gigawatts by 2030 while navigating the shift from training to inference-heavy agentic workloads and managing massive energy and supply chain constraints.