Stanford Webinar - Human-Centered AI: Designing Systems People Trust
TL;DR
Stanford professor James Landay argues that truly human-centered AI requires expanding beyond traditional user-centered design to address community and societal-level impacts, while tackling embedded Western cultural biases and the global rise of sovereign AI initiatives.
🎯 The Human-Centered AI Framework 2 insights
Expand Beyond Traditional User-Centered Design
Human-centered AI must address impacts on non-users such as healthcare patients, resource allocation recipients, and data labelers who are affected by but do not directly interact with systems.
Design Across Three Critical Levels
Effective AI development integrates the user level, community level (indirectly affected groups), and societal level (broad cultural impacts like social media's influence on mental health or news consumption).
🌍 Cultural Bias and Sovereign AI 3 insights
Western Cultural Ontologies Dominate AI Models
Research reveals large models default to Western perspectives, such as viewing trees as isolated objects rather than interconnected living systems, thereby limiting representation of Indigenous and non-Western philosophical frameworks.
Sovereign AI Addresses Cultural Control
Countries are pursuing sovereign AI across infrastructure, data, models, applications, and talent to ensure national security, economic independence, and cultural representation independent of dominant Western technologies.
Training Data Reflects Geographic Cultural Bias
Large models train predominantly on North American and Western European internet data, creating systems that may not reflect the values, languages, or ontologies of communities in Iran, Singapore, or Indigenous populations.
🏥 Real-World Health Applications 2 insights
AI as Personalized Health Coach
The Bloom application uses motivational interviewing techniques to co-create fitness plans with users, featuring a visual garden interface that grows based on weekly activity levels to encourage sustained engagement.
Ambient Sensing for Aging in Place
Researchers are developing privacy-preserving ambient sensing AI to help older adults live independently at home longer while carefully considering impacts on care workers and family members who interact with the system.
Bottom Line
AI developers must expand design processes to include community and societal stakeholders beyond end-users while actively addressing Western cultural biases embedded in training data to create technologies that genuinely reflect global values and earn widespread trust.
More from Stanford Online
View all
Stanford CS221 | Autumn 2025 | Lecture 20: Fireside Chat, Conclusion
Percy Liang reflects on AI's transformation from academic curiosity to global infrastructure, debunking sci-fi misconceptions about capabilities while arguing that academia's role in long-term research and critical evaluation remains essential as the job market shifts away from traditional entry-level software engineering.
Stanford CS221 | Autumn 2025 | Lecture 19: AI Supply Chains
This lecture examines AI's economic impact through the lens of supply chains and organizational strategy, demonstrating why understanding compute monopolies, labor market shifts, and corporate decision-making is as critical as tracking algorithmic capabilities.
Stanford CS221 | Autumn 2025 | Lecture 18: AI & Society
This lecture argues that AI developers bear unique ethical responsibility for societal outcomes, framing AI as a dual-use technology that requires active steering toward beneficial applications while preventing misuse and accidental harms through rigorous auditing and an ecosystem-aware approach.
Stanford CS221 | Autumn 2025 | Lecture 17: Language Models
This lecture introduces modern language models as industrial-scale systems requiring millions of dollars and trillions of tokens to train, explaining their fundamental operation as auto-regressive next-token predictors that encode language structure through massive statistical modeling.