Back to database

Elizabeth Lin's AI use case

Independent design educator at Design is a Party

Uses Cursor as a creative partner for design exploration: generating visual styles, building interactive prototypes like a working piano with sound, and refining rough AI-generated interfaces into more polished designs.

The problem

What was broken before AI

Designers often hit friction at the beginning and middle of a project. A blank page can make it hard to choose an aesthetic direction. Traditional prototyping tools are great for screens, but harder for interactions with sound, motion, or real code behavior. And when AI generates ugly interfaces, the designer still needs a way to guide it toward better taste without rewriting everything manually.

What changed

What the use case made possible

Elizabeth uses Cursor to make design exploration more immediate. She can ask for aesthetic directions, combine styles like Brutalist and Y2K, restore checkpoints, and run the same prompt again to get a different variation. She can also build interactive prototypes, such as a digital piano with sound, from a simple prompt. When the output is ugly, she uses specific feedback, references, and design principles to push it toward something more intentional.

Why this matters

Why this use case is worth studying

Elizabeth’s workflow is valuable because it gives designers a more playful relationship with AI. The tool is not just producing finished screens; it is helping her ask better design questions. What happens if this feels more Y2K? What if it sounded like an old Mac? What if the dashboard borrowed from Robinhood, Cash App, Stripe, or Edward Tufte? The work becomes a conversation with prototypes instead of a long wait between idea and test.

Use this when

When this pattern applies

Use this pattern when a design idea needs to be felt, clicked, heard, or tested before it can be judged. It works especially well when static mockups are not enough, when you want to explore several visual directions quickly, or when the interaction itself is part of the idea.

Exponential Builder analysis

01

Start wider than the screen — Elizabeth uses AI to surface aesthetics, movements, and reference directions before committing to one.

That makes the blank-page phase more concrete because the designer reacts to options instead of trying to invent a direction from scratch.

02

Prototype the behavior, not only the look — A working piano with sound reveals more than a polished static mockup ever could.

Code-based AI tools are especially useful when the design question depends on interaction, motion, audio, data, or some other behavior that needs to be felt.

03

Your taste is the steering system — The AI can generate variations quickly, but the quality rises when Elizabeth gives specific critique: what to remove, what aesthetic to move toward, and which products or principles should guide the next pass.

The builder skill here is learning to translate taste into operational feedback.

Who this is for

Best fit

Product designers

Brand and visual designers

Design educators

Founders prototyping early ideas

PMs exploring product concepts

Designers who want to experiment with code without becoming full-time engineers

Teams trying to make rough ideas easier to react to

What to avoid

Mistakes and warnings

Where this pattern can go wrong if you copy it too literally.

Do not expect AI to have taste without guidance from you.

Avoid vague feedback like “make it better” when specific critique would help more.

Do not confuse a surprising aesthetic with a usable interface.

Use checkpoints before experimenting so you can recover good directions.

Keep your design judgment active; the AI can generate options, but you still decide what works.

Public workflow preview

The shape of the workflow

A high-level look at how the use case works, with the reusable pattern made clear.

01

Start with creative direction, not a final answer

Elizabeth asks Cursor what styles, movements, and aesthetics it can work with before choosing a direction.

02

Combine references to spark variation

She gives the AI specific styles or product references so it can generate a surprising first pass.

03

Use checkpoints to explore safely

If a direction feels wrong, she can restore the previous version and try again without losing momentum.

04

Prototype interactions, not just screens

Cursor can build working examples, like a digital piano with sound, that would be hard to prototype quickly in a static design tool.

05

Teach the AI through design feedback

Instead of saying “make it better,” Elizabeth gives concrete critiques, references, and principles that move the interface closer to good taste.

Copy the pattern

The reusable idea

Pattern in one sentence

Use AI coding tools to turn design ideas into small interactive prototypes, then use your taste to steer the output instead of expecting the first version to be right.

Reusable idea

Elizabeth’s approach works because she uses AI to widen the design process before narrowing it. A designer does not need to start with a perfect prompt or a polished concept. Start by asking for directions, references, and rough variations. Then use your taste to react: keep what has energy, restore what does not, and give the model sharper language each round. The useful skill is not prompting once; it is learning how to steer.

Steal this workflow

Use Cursor as a design exploration loop:

1

Pick one small surface: a homepage, dashboard, widget, or tiny interactive idea.

2

Ask Cursor what styles, aesthetics, and movements it can implement, and have it describe them before generating anything.

3

Choose two directions to combine, such as “Brutalist and Y2K,” then ask for a redesign.

4

Treat the first result as material for reaction. Save a checkpoint before changing direction.

5

If the design feels wrong, give feedback in three parts: remove what is bothering you, name the feeling you want, and cite product or design references.

6

For interaction-heavy ideas, ask for a working prototype instead of another static screen.

7

Restore, rerun, or constrain the prompt until the prototype gives you a direction worth developing.

Suggested prompt

Suggested pattern based on Elizabeth Lin’s Cursor workflow: “I’m working on a small prototype for [describe page or interaction]. First, list visual styles, design movements, and aesthetics you are comfortable implementing for this kind of interface, with a short description of each. Then redesign the prototype by combining [style/reference 1] and [style/reference 2]. Keep it usable, make the interaction actually work, and after the first pass, wait for my feedback before refining.”

Field notes

Get new AI use cases in your inbox

A short weekly note on how real people are using AI to save time, make money, build tools, and run their lives.

No spam. Just useful AI use cases.

Related use cases

Keep exploring nearby systems.

Browse all