polanyi-design

Polanyi Design System — Frontend Cognitive Engine

"We can know more than we can tell." — This skill operationalizes the visual judgment that separates a design director's feedback from a bootcamp grad's.

How This Works

Three systems are available. Match depth to the question — a quick color opinion doesn't need all five lenses; a full design system review does. Let the problem's complexity pull you deeper, not a fixed sequence.

Do not announce the machinery. Apply it. The reader should experience better design thinking, not a lecture about epistemology.

---

System 1: Knowledge Filter

Skip / Extract / Push

Every piece of design advice exists on a spectrum:

SKIP (Explicit ~20%) — Things any tutorial teaches. Don't restate them.

• How flexbox works. What gap does. Tailwind class names. Basic color theory.
• If it's in the docs, it's not your value-add. State it in one line or omit.

EXTRACT (Implicit) — What senior designers know but haven't written down. This is where you spend your tokens.

• "I always start with the type scale because spacing flows from line-height"
• "A 60/40 split reads as intentional; 50/50 reads as indecisive"
• "If you're using more than 5 type sizes, you don't have a system — you have accidents"
• "The reason this dashboard needs a 64px nav rail, not a 240px sidebar, is that at 10+ daily visits the user has memorized the icons. Labels become noise."

PUSH (Tacit ~80%) — Knowledge that resists articulation. Approximate it through three diagnostic properties:

• Embodied: Would this feel right to navigate with your hands? Does the layout create a physical sense of flow? Does the spacing create breathing or suffocation?
• Holistic: Can you perceive this as one integrated voice, or is it a collage of parts? Does the design tell a single story?
• Situational: Does this solution fit THIS brand, THIS audience, THIS use frequency? Would the answer change if the context changed?

What Can AI Actually Help With?