RisksBLACK-BOX

02

Homogenization

Convergence on average

Generative models produce output based on the distribution of their training data. They gravitate toward the most common patterns - which means away from anything distinctive.

It's not just that AI-generated interfaces look the same. The convergence runs deeper: research protocols default to the same questions, journey maps default to the same shapes, strategy decks default to the same conclusions. The technical term is model collapse, the recursive degradation as AI trains on AI-generated content. The design community calls it "sameness", "style collapse" or "dribbblisation", accelerated. When good enough becomes standard and low-cost, companies have to decide whether to invest in excellence. If all the services look interchangeable, some of those companies are in trouble. They just don't know it yet.

Design transfer

  • Design systems that converge on identical patterns across competing products
  • Research insights that default to generic best practice rather than context-specific findings
  • Brand expression that is technically correct but visually interchangeable with the competition
  • Loss of brand-specific interaction signatures over successive AI-assisted iterations

In the wild

Use cases

Research Plan Development

EXP-002

A first-draft research plan generated on project inputs frees the researcher from formalization focus on details of study design like hypothesis formulation, methodological choice and fieldwork logistics.

Explore·Developing

Design Brief Generation

DEF-014

Reduces brief creation time while preserving the strategic clarity that allows them to guide design work.

Define·Developing

Interview Thematic Coding

DEF-010

Compresses thematic coding of qualitative interview data from days to hours. Helps identify cross-participant patterns a solo researcher might miss.

Define·Developing

Journey Mapping

DEF-013

Transforms quantitative and qualitative data into a journey map faster than manual methods. Time saved on initial mapping lets teams dedicate more to filling gaps, verifying details, and socializing understanding.

Define·Emerging

Persona Generation

DEF-012

Connects the qualitative synthesis of persona attributes with quantitative data, reinforcing support and targeting what is most relevant to product decision-making, before personas are socialized within the organization.

Define·Developing

Problem Reframing

DEF-107

Surfaces alternative problem framings that teams would not have considered under time pressure, reducing the risk of building solutions to problems that were never properly interrogated.

Define·Emerging

Survey & Open-Text Analysis

DEF-015

Analysis of open-text responses or other unstructured survey data at a scale not possible with manual coding, to identify trends and patterns across large response sets quickly.

Define·Developing

Design Remix

CON-031

Systematic recombination of elements from different products or brands - using prior art to expose new possibilities.

Concept·Emerging

Design Variants

CON-022

Parallel generation of alternative designs against controlled requirements. Allows multiple directions to be explored at the same time and bootstraps iteration.

Concept·Developing

Ideation Support

CON-019

Kickstarts "blank page" effect by providing something to react to, helps in "cold start" situations where the domain is unfamiliar.

Concept·Developing

Information Architecture Development

CON-082

Explore multiple organizational options for product information architecture deeply in the same time typical assisted methods spend to create one or two.

Concept·Developing

Interaction Pattern Suggestion

CON-083

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.

Concept·Developing

Low-fidelity mockups & Layout Generation

CON-020

Move faster from content requirements to reasoned interfaces layouts by generating candidate wireframes for critique and further exploration.

Concept·Developing

Moodboarding and Visual Exploration

CON-027

Simultaneous exploration of multiple creative directions at a specificity and volume not achievable through stock photography or manual illustration.

Concept·Developing

Pattern Library Synthesis

CON-091

Bootstraps a canonical pattern library in hours rather than quarters, while simultaneously producing a gap analysis that informs design system investment priorities.

Concept·Emerging

Prompt-to-Prototype

CON-021

Produces functional, interactive prototypes without manually wiring hotspots or engineering support, achieving higher fidelity faster.

Concept·Developing

Workshop Support

CON-077

Generating and refining workshop material with AI allows designer to focus on details and success factors, rather than on assembly.

Concept·Developing

Synthetic Stakeholder Critique

VAL-035

Gives designer access to diverse stakeholder perspectives by simulating distinct critical positions, finding unanticipated objections that can be addressed before live stakeholder engagement.

Validate·Emerging

Codebase-Extracted Design System

DEL-057

Generates a structured design system - markup documentation, living demos, component inventory - from an existing production codebase, creating a system where none formally existed.

Deliver·Emerging

Production Copy Generation

DEL-044

Tightens the content development timeline by semi-automating microcopy and first drafts with rules governing style and tone of voice, so teams refine to production-quality copy faster.

Deliver·Established