How UX/UI Principles Is Bringing Research Back to Design: 185 Laws, 2,300 Citations & AI Validation

AI coding tools generate visually plausible, cognitively broken interfaces daily. Buttons violate Fitts's Law. Forms ignore Hick's Law. Navigation breaks Nielsen's consistency heuristic—and teams ship because "it looks fine in Figma."

UX/UI Principles (uxuiprinciples.com) responds with a citation-backed library: 185 principles, 2,300+ peer-reviewed references (2020–2026 focus), free validators, and 600+ MCP/API-integrated prompts for Cursor, Claude, and ChatGPT. This analysis covers what the product delivers, the research it rests on, and where domain/naming identity intersects.

Receipt #1: The research foundations (not vibes)

Classic laws the library maps to modern UI:

| Law | Origin | UI failure mode | | --- | --- | --- | | Fitts's Law | Fitts, 1954 | Tiny touch targets, edge-distance ignores on mobile | | Hick's Law | Hick, 1952 | Mega-menus, 12-option pricing grids | | Miller's Law | Miller, 1956 | 9-field forms without chunking | | Jakob's Law | Nielsen, 2000 | Novel nav patterns users must re-learn | | Peak-end rule | Kahneman | Error states that poison entire session memory |

Nielsen Norman Group's 10 usability heuristics (1994, updated applications 2020s) remain the most-deployed evaluation framework in industry—UX/UI Principles indexes each against post-2020 validation studies (mobile, AI chat UI, AR overlays).

Why recency filter (2020–2026)? Voice UI, AI copilots, and dark-mode-default changed interaction assumptions. A 1998 web form study doesn't cover chat-first onboarding or LLM streaming layouts.

Receipt #2: Product architecture

Six categories, 185 principles

1. Foundations — perception, attention, memory
2. Core Principles — Fitts, Hick, feedback loops
3. Design Systems — tokens, consistency, accessibility
4. Interface Patterns — forms, nav, modals
5. Specialized Domains — 61 principles across AI validation, voice, AR/VR (largest section—bet on emerging surfaces)
6. Human-Centered Excellence — ethics, inclusion

Free tier (acquisition funnel)

• AI Validator — paste design decision → matched principles + citations
• UX Smells Detector — 8 anti-patterns (mystery meat nav, feedback black hole) with refactor recipes
• UX Flows — journey templates

No signup required—reduces friction vs Mobbin ($99/mo visual reference).

Paid tier ($59/year, down from $79)

• Full library + citation database
• 600+ tool-specific prompts (Cursor, V0, Claude, ChatGPT)
• Developer API + MCP server — AI agents query principles during code generation

Competitive positioning: Mobbin/Smart Patterns sell screenshots; UX/UI Principles sells evidence. Trade-off: no visual mock analysis—text-in, principles-out.

Receipt #3: MCP integration as distribution moat

The Model Context Protocol (MCP) server lets Cursor/Claude pull principles mid-generation—transforming passive docs into design governance.

Example flow:

1. Developer prompts V0 for checkout form
2. MCP queries Fitts + error-prevention principles
3. Agent flags 32px button below 44×44px WCAG 2.5.5 target size minimum
4. Regenerate before merge

Strategic bet: If AI IDEs adopt validation hooks by default, UX/UI Principles becomes infrastructure—similar to ESLint for UX.

Domain & naming analysis: uxuiprinciples.com

| Factor | Assessment | Research tie-in | | --- | --- | --- | | Descriptive clarity | Instant category comprehension | Good for SEO; weak for distinctiveness | | Length | 15 characters | GrowthBadger: long strings hurt oral recall | | TLD | .com | 44% memorability vs alternatives; correct for trust | | Missed .ai | Product is AI-validation core | Identity Digital pattern: .ai signals category to tech buyers | | Generic term risk | "UX principles" is category, not brand | SEO competition with NN/g, Smashing Magazine, etc. |

Phonetic profile: Front vowel density in "UI" + sonorant "Principles" = trust/education signaling (Lowrey/Shrum back-vowel "large/authoritative" fits reference product).

Recommendation implied by data: Own uxuip.com redirect + consider uxuip.ai for AI-validator SKU—Atom: 89% tech familiarity with .ai.

When to use this vs alternatives

| Need | Tool | | --- | --- | | Visual inspiration | Mobbin, Smart Patterns | | Heuristic evaluation (expert) | NN/g heuristics, SUS surveys | | Citation-backed AI validation | UX/UI Principles | | Raw papers | ACM, Google Scholar | | Automated a11y scan | axe, Lighthouse |

Complement, not replace: Lighthouse catches WCAG violations; UX/UI Principles catches cognitive law violations Lighthouse misses.

Risks and limitations

1. Text-only input — screenshot analysis gap; user must describe UI accurately
2. Recency bias — foundational 1950s laws need original citations alongside 2023 replications
3. Generic brand — competitors can clone library structure; MCP integration depth is moat
4. Static principles — monthly updates required; design research moves fast on AI UI patterns

China relevance

• Mobile-first laws matter more — WeChat mini-programs: Fitts targets on small screens critical
• Han script information density — Miller chunking differs; 4-character labels ≠ 4 English words
• ICP/compliance UI — separate from Western heuristic sets but affects trust perception
• Domain: .com still standard for devtools selling globally; .cn for localized product surface

Bottom line

UX/UI Principles fills a documented gap: AI-generated UI proliferates faster than design review capacity. 185 principles × 2,300 citations replace opinion wars with receipts—Fitts, Hick, Nielsen, and 2020s replications.

At $59/year, it's asymmetric bet against $99/mo visual libraries—if your bottleneck is "why is conversion low?" not "what does Stripe's button look like?"

The generic name uxuiprinciples.com trades distinctiveness for clarity—ironic for a product teaching differentiation through cognitive science. The MCP/API path is the brand moat, not the domain string.

For AI naming workflows that respect fluency research, see using ChatGPT and Claude for brand naming. For domain trust signals affecting UX entry, see psychology of domain names.