Interaction Engineering · 5 / 10

Conventions: Don't Let AI Invent New Interactions

Once the flow is clean, a deeper disease remains: AI sometimes invents interactions that look novel and nobody can use. This lesson covers two words — conventions (idioms) and affordance. Finish it and you'll spot which control is faking innovation, and explain why users can't learn it.

ConventionsAffordanceSpot-the-flaw drillWhere innovation belongs
None of the interactions you use came free

Chapter 13 of About Face 4 splits interfaces into three approaches: grown from the program's internals (implementation-centric), drawn to look like real objects (metaphor), and learned (conventions / idioms). Cooper's stance is blunt: modern GUIs mostly run on the third. Windows, scrollbars, dropdown menus, hyperlinks — none of them are intuitive; users learned them one by one.

His example is the mouse: nobody glances at a mouse and knows it drives the on-screen cursor — someone has to teach you, or you fiddle for half a minute. Once learned, you never forget. Scroll, double-click, drag-and-drop, pull-to-refresh, swipe-left-to-delete — same story: learn once, use everywhere.

"All idioms must be learned; good idioms need to be learned only once."
Alan Cooper, About Face 4, Chapter 13.

For product people that's a free gift: the world's products already taught your users — reuse existing conventions and learning cost is zero. The snag is AI often ignores that favor. It stitches patterns from training data and invents a control nobody has ever learned.

Metaphor doesn't travel far

While we're here, bury the second approach too. In the early GUI boom, credit went to "looking like the real thing": desktops drawn as desks, folders as folders, users reading pictures. Cooper doesn't buy it in Chapter 13: metaphor interfaces make users play connect-the-dots between icon and function. Simple features maybe; anything abstract can't be drawn. "Batch-rename 50 files" — what would you draw?

Worse, metaphor overstays. Once users know the basics, leather textures and page-flip animations stop helping and start dragging — they eat space, constrain interaction, and drag mechanical-era rules along. Apple admitted it: the leather calendar in iOS 6 was wiped clean by iOS 7.

So conventions win by traveling light: they don't ask you to guess — they ask you to learn, and learning is fast. Next, how this vocabulary is built.

Three layers of vocabulary: why GUIs are learnable

Chapter 13 has an inverted pyramid explaining why conventions are easy to learn: the whole GUI interaction vocabulary sits in three layers — fewer and simpler as you go down. Tap each layer to see what it holds.

The inverted pyramid of interaction vocabulary Viewed 0 / 3
From the bottom up the vocabulary stacks richer — but the foundation is just a few actions
Tap any layer on the left. Read the pyramid bottom-up: the foundation is tiny, yet the building can go high — that's the structural reason GUIs are learnable.

Users can carry this vocabulary from other products into yours via mental models: they assume your UI obeys the same rules as the ten thousand they've used. How lost they get when models mismatch was covered in the Psychology part — here's a jump card, no re-teaching.

Cross-chapter · Mental models: the map in the user's headUsers bring an old product's model to your UI — mismatch once, get lost once. Where models come from and how to fix mismatches: Psychology Lesson 7 covers it. Tap to jump.
Hands-on · Three inventions hide in this settings page

Below is an AI-generated cloud-drive settings page. Features work, the UI isn't ugly — but it hides three clever inventions, each breaking a well-learned convention. Tap them one by one; hits get a verdict.

Find the controls faking innovation Found 0 / 3
Tap controls that feel wrong. Tapping a legit control still gets a reply
pan.example.com/settings
Account settingsCloud Drive Premium
Basic profile
Invention ①
Mist Watcher
mist@example.com
Not set
Not configured
Untitled
Page 1 / 3 · Swipe left for more fields
General
Notifications
Invention ② View membership benefitsDouble-click to open Annual plan · 208 days left
Invention ③ Save After click, the ring fills once = save success
Log out
Invention ① · Sideways-scrolling form: users can't learn it. Form convention is top-to-bottom: scroll, fill, submit — unchanged for thirty years. Horizontal paging stuffs required fields into dark cubbies; users think they're done, hit Save, and drown in errors. To save height, use collapse or groups — don't change direction.
Invention ② · A link that needs a double-click: users can't learn it. Underlined blue text — the Web contract is single-click to go. Double-click belongs to desktop-icon conventions. Mix the two rules and nobody guesses; you end up teaching with a tiny hint line. A control that needs a manual has already failed.
Invention ③ · Progress-ring Save button: users can't learn it. Save has the world's most common pattern: a solid button that says "Save." As a ring, users first think "what's loading?" then "can I even click this?" The control stops self-explaining — the invention threw affordance away.
All three inventions caught. Notice what they share: each looks clever alone, and fails by breaking rules users already know. One acceptance line for AI: "Navigation, forms, and basic buttons use the most common standard patterns — don't invent new interactions." Where to spend innovation: last section of this page.

Why does AI do this? Its training data mixes everything: real products, award-winning concepts, portfolio showpieces from design communities. Ask for "a settings page with design sense" and it blends patterns from several sources — grammatically fine, rule-wise fighting each other.

One thing it can never learn: what your users already know. Conventions live in user experience. AI sees how a control looks, not thirty years of teaching history. So this gate is yours: any control that needs a tiny hint line to explain how to use it deserves suspicion.

Affordance: controls must explain themselves

The second word comes from Norman's The Design of Everyday Things: affordance — how a thing looks tells you how to use it. Cooper sharpens it in Chapter 13: affordance is about what you think it can do, counted separately from what it actually does. A small round button by a door reads as a doorbell to everyone — even if pressing it does something else, its affordance is still doorbell.

Classic case: the door handle. A horizontal metal bar shouts "pull me" in shape and height — but the door pushes. The shop sticks a "Push" label on. Useless; people still pull. Signals the hand can read speak louder than words.

On screen it becomes a contract: draw a button and you promise it can be clicked, reacts when clicked, and does what the label says. Controls that break the deal Cooper calls bait-and-switch affordance.

He also calls out flat design: early UIs used shadow and depth to shout "press me"; flat style scraped that away. Visually clean — affordance took the hit. So clickable elements in flat UIs need stricter discipline: one color, one shape, reserved for things that click. No guest appearances.

Two membership centers below — same upgrade card, two drawings. Tap the one you think keeps the contract.

If it looks like a button, it must click A/B duel
Tap a card to vote; verdict after you pick
Plan A
vip.example.com
Desktop clientCheck for updates
Current version v3.2.0, last checked today 09:41
You're up to date
The big blue pill looks most like a button — click does nothing. The real action "Check for updates" shrinks to small type up top
Plan B
vip.example.com
Desktop client
Current version v3.2.0, last checked today 09:41
You're up to date
Check for updates
Status sits honestly as gray text plus a dot; action gets a real button

Affordance has a zero-cost test: send a screenshot to someone who's never used it, give 5 seconds, ask them to point at what's clickable. Every miss is an affordance accident. Especially useful on AI-generated UIs — AI never asked that question while drawing.

Spend innovation where it cuts

Easy wrong conclusion here: always copy existing interactions — innovate and die. Cooper's ruler is finer:

"Significant change must offer significant advantage."
Alan Cooper, About Face 4: if users must relearn, the payoff must be visible to the naked eye.

Navigation, save, forms — users rehearse these hundreds of times a day in other products. Changing the pattern forces them to relearn walking, with no real payoff — use the most common patterns here. Spend the innovation budget on your core differentiator: the capability unique to your product, where no convention exists to copy. Nail it, and others copy you.

Measure one question with that ruler. AI added four "highlight designs" to a batch file manager. Which innovation spent the budget right?

Four highlight designs — which one earns its keep? Single choice
Wrong answers still get explanations — keep going until you hit right
AReplace bottom nav with a press-and-hold radial pie menu
BUpload progress as a skeuomorphic water glass — full = done
CCore feature batch rename: change one rule, live-preview fifty filenames
DDouble-click a file card to favorite — no favorite button on the UI
Common-pattern list: copy these without apology

One last cheat sheet. Users already know these conventions cold from other products — copying them is free usability. Check AI output against this list; the right column pastes straight into a prompt.

ConventionWhere users learned itDefault ask for AI
Pull to refreshEvery feed appRefresh lists with pull-down — don't invent a floating refresh layer
Swipe left to deleteSystem Mail, WeChat chat listsDelete list items with swipe-left or a clear button — don't hide it in long-press menus
Bottom tab barMobile apps' standard slotPrimary nav as bottom tabs, icon + label, 3 to 5 items
Shopping cartThirty years of e-commerce teachingMulti-select then checkout keeps the cart model — don't invent a new metaphor
Drag-and-drop uploadCloud drives and email attachmentsUpload areas support drag-and-drop and keep "click to choose files"
Ctrl / Cmd + ZDesktop software's old ruleEditing features must support undo shortcuts
Blue underlined linksWeb lesson oneLinks keep recognizable style; single-click goes there
Magnifying-glass searchBrowsers and every content appSearch entry = magnifier icon plus placeholder hint text
Red badge dotsPhone notification systemsUnread cues use badge counts — don't invent blinking markers

Nothing in the table asks users to learn — that's the value. The learning budget you save, by the last section's ruler, goes to your core differentiator.

Key Takeaways

Interfaces run on conventions: scroll, double-click, drag-and-drop — none are intuitive; all are learned. Good conventions last a lifetime; the world's products already taught your users (Cooper, About Face 4, Chapter 13).

Affordance is a contract: how a control looks is what it must do. Drawing a button promises clickable and reactive; something that looks like a button but doesn't click burns credit for every control on the page.

Spend innovation where it cuts: navigation and basics use the most common patterns; innovation budget goes to the core differentiator. Significant change must offer significant advantage — novelty without payoff is just load.

Prompt line for AI: "Navigation, forms, and basic buttons always use the most common standard patterns — don't invent new interactions; clickable elements must look clickable." Four handles covered; next lesson goes into detail: how to pick the controls themselves.

Source: Original to Xiaoshan Academy's Interaction Engineering series; some interaction principles adapted from About Face 4: The Essentials of Interaction Design; affordance concept from Don Norman's The Design of Everyday Things.