Concept·Concept Design·Automation·Developing·CON-083
Interaction Pattern Suggestion
Value hypothesis
Designers consider more pattern precedents during the concept stage, letting them make choices grounded in explicit trade-offs, while not missing underused options that may be better solutions.
Quality · Efficiency
The designer hits a specific interaction problem (a state transition, an empty state, an onboarding sequence, an error recovery path, a complex data input flow) and asks an LLM for established patterns that address it, drawn from documented pattern libraries and platform conventions. Candidate patterns come back with trade-off notes on complexity, learnability, accessibility implications, and platform precedent; the designer weighs each against the product's existing design system, the user population's expectations, and the specific edge cases the scenario introduces.
Risks in application
Homogenization
The patterns the LLM returns most readily are the ones its training corpus over-represents: iOS Human Interface Guidelines, Material Design, a handful of high-profile SaaS interfaces. Drafts built from these suggestions converge on a global interaction vocabulary that reads as competent, while the product stops developing the distinctive interaction choices that would differentiate it from the interfaces the corpus is saturated with.
Constraint Blindness
Suggested patterns may match the interaction problem in the abstract but conflict with constraints invisible to the LLM: existing component library affordances, platform-specific accessibility requirements, or upstream design system decisions that bound the available solution space.
Expertise that differentiates
Interaction Design
Pattern libraries document what works in general; the designer decides what works here, given this product's existing vocabulary, this user population's literacy with conventions the LLM assumes are universal, and the accessibility constraints the scenario description never captured. A well-known pattern applied without that fit check produces interactions that look industry-standard and feel foreign, because the standard was set by products serving different users under different constraints.
AI Fluency that assures
Task Delegation
CON-083-s1: 'asks an LLM for established patterns that address it, drawn from documented pattern libraries and platform conventions'.
Process Description
The designer hits a specific interaction problem (a state transition, an empty state, an onboarding sequence, an error recovery path, a complex data input flow) and asks an LLM for established patterns that address it, drawn from documented pattern libraries and platform conventions.
Pattern libraries document what works in general; the designer decides what works here, given this product's existing vocabulary, this user population's literacy with conventions the LLM assumes are universal, and the accessibility constraints the scenario description never captured.
The designer weighs each against the product's existing design system, the user population's expectations, and the specific edge cases the scenario introduces.
Possible Indicators
Expert assessment delta
Difference in pattern appropriateness between AI-assisted and memory-alone selection, rated by experienced interaction designers against the product's existing design system, the user population, and accessibility constraints.
Effort reduction
Hours spent researching and evaluating interaction patterns per concept cycle, versus baseline for manual pattern library scanning and documentation review of equivalent scenario scope.