Stanford CS547 HCI Seminar | Winter 2026 | Creation, Evolution, and Formalization of Notations
TL;DR
This seminar challenges the traditional linear model of 'informal-to-formal' notation development, arguing that humans dynamically create new notations through collaborative practice while current AI systems are limited to 'instant formalization' into existing structures. The speaker presents a three-stage historical model of notation evolution—from culturally situated invention through community dispersion to institutional sanctification—to guide future HCI system design.
🤖 AI's Blind Spot: Instant Formalization vs. Creation 3 insights
AI locks users into dominant languages
While AI enables 'instant formalization' from vague prompts to runnable code like React or Python, this capability incentivizes users to abandon preferred languages for AI-optimized ones, suppressing notational diversity.
The creation dimension is missing
Current AI systems can translate informal ideas into existing formal representations but cannot dynamically co-create new abstractions with users to ground emerging shared understanding during collaborative work.
Risk of notational monoculture
As engineers default to languages like Python solely for better AI assistance, the feedback loop threatens to homogenize computational expression and erase context-specific notation innovations.
📊 Rethinking Formalization: Beyond the Spectrum 3 insights
The linear spectrum is insufficient
Traditional HCI theory (Shipman and McCall) models formalization as a simple horizontal progression from informal text to formal semantic nets, failing to capture how humans actually iterate on whiteboards.
Ad-hoc notation invention is constant
In practice, humans continuously invent and extend notations mid-think—such as modifying Venn diagrams with squares for AI and circles for humans to represent human-AI common ground.
Formalization requires two dimensions
Notation work involves both horizontal translation toward existing formalisms and vertical creation of entirely new representations, with AI currently supporting only the former while ignoring the latter.
📜 Historical Evolution of Notations 3 insights
Stage 1: Invention through metaphor
Notations originate as culturally situated inventions employing linking and grounding metaphors to accurately describe specific empirical phenomena within their birth communities.
Stage 2: Dispersion and divergence
As notations spread to broader communities, they naturally diverge and adapt to local needs unless active social and material infrastructure manages the dispersion process.
Stage 3: Institutional sanctification
Maturation culminates in standardization committees establishing canonical versions, effectively 'sanctifying' specific formal structures and often freezing further evolutionary adaptation.
Bottom Line
To advance human-AI collaboration, we must design systems that support the dynamic creation and evolution of new notations alongside users, rather than merely optimizing for instant formalization into existing dominant languages.
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.