The key point is this: writing good prompts is not a trick. It is a communication and design skill that transforms implicit knowledge into clear instructions for machines. Just like writing a requirement or defining a hypothesis, well-crafted prompts describe intent, context, expected behavior, and constraints.

1. Prompts as a Tool for Clear Communication

One of the biggest wins for designers using AI is communication. Drafting messages, adapting tone for different audiences, and translating technical language into simple explanations all become faster and more consistent.

Examples

• Rewrite a difficult email to sound clear, professional, and collaborative.
• Turn vague stakeholder feedback into concrete design guidance.
• Explain a technical concept to a non-technical audience.

Why this matters: clear communication reduces rework and helps cross-functional teams move forward with fewer misunderstandings.

2. Product Thinking Supported by AI

Beyond writing, AI can support strategic thinking. Designers can use it to explore alternative perspectives, simulate user feedback with personas, or review decisions based on usability principles.

Examples

• Review this interface using common UX heuristics.
• Imagine you are this type of user. How would you react to this screen?
• Act as a UX mentor and suggest approaches for this leadership challenge.

This works like a second brain for quick reflection, critique, and ideation.

3. AI in Visual Production and Prototyping with Figma Make

ChatGPT and tools like Figma Make work well together. While ChatGPT helps structure language, logic, and reasoning, Figma Make interprets prompts to generate visual components, layouts, and interactive prototypes.

The main difference is the type of prompt and the level of detail required.

Based on real designer workflows:

1. PRD-based prompts
   Use detailed descriptions of what to build, how it should behave, and which constraints apply. This leads to more consistent and usable prototypes.
2. Assisted prompt generation
   Start with a rough idea and let the system structure it into a complete prompt. This is useful for quick exploration.
3. The TOKEN framework
   Task, Output, Key elements, Expected behavior, Notable constraints.
   This framework helps ensure your prompt contains all critical elements before sending it.

4. An Integrated Flow from Idea to Prototype

Step 1.
Explore with ChatGPT

Use structured prompts to think about the problem, challenge assumptions, clarify requirements, and turn vague ideas into clear instructions.

Step 2.
Translate into a Figma Make prompt

Convert these requirements into a format that defines task, output, UI elements, expected behavior, and constraints.

Step 3.
Iterate between human and AI

Generate early prototypes, refine them in Figma, and adjust your prompts based on what you learn from each output.

5. Best Practices for Writing Effective Prompts

To get real value from AI in design, keep these principles in mind:

A. Provide context and purpose
Without context, AI will make generic assumptions. Always explain where this screen or feature fits and why it exists.

B. Describe expected behavior
Do not just say “create a dashboard.” Explain how elements should react, update, or interact.

C. Add constraints and style rules
Mention platform, accessibility needs, design system rules, or brand guidelines when relevant.

D. Reduce ambiguity
A good prompt does not hide uncertainty. It exposes it and resolves it.