Beyond the Button Podcast: AI Strategies for UX

Episode 2: EDEN—When Design Thinking Meets Prompt Engineering


Listen Later

Most people approach AI prompting backwards. They throw random one-liner prompts at it and hope for good results. But UX designers have a massive secret weapon that’s been hiding in plain sight. It’s the same systematic, human-centred approach you already use to solve complex design problems, but applied to prompt engineering. When you combine design thinking principles with AI prompting, something remarkable happens: you stop getting random outputs and start creating predictable, high-quality results that actually solve real human problems. The best part? You can build prompting systems that your entire team can use, turning AI from a personal productivity hack into a strategic business advantage. (Something we are teaching a small group of select individuals in our AI-Powered UX Design Masterclass.)

In episode 2 of our podcast, we discussed the Eden framework—a four-step design thinking approach that changes chaotic AI interactions into systematic, human-centred solutions.

Empathize: Start with human context, not AI capabilities

Before you write a single prompt, step back and understand the human problem you’re solving. Most people jump straight to “what can AI do for me?” But design thinking teaches us to start with “what do humans actually need?”

Ask yourself:

* Whose problem am I solving?

* What are the real stakes here?

* Who are the end users, and what are their constraints?

If you’re creating a prompt for user research analysis, don’t just think about what the AI can analyze;  think about what your stakeholders need to make decisions, what format they prefer information in, and what level of detail serves their goals. This empathy phase lets your AI outputs serve real human needs instead of just demonstrating technical capabilities.

Define: Frame the problem clearly before you start prompting

The biggest mistake in prompt engineering isn’t bad syntax — it’s solving the wrong problem. In the define phase, get clarity about the specific challenge you’re addressing. Instead of “I need help with user feedback,” define it as “I need to identify the top 5 usability issues from 15 user interviews to present actionable recommendations to the product team by Friday.”

This precision transforms vague AI outputs into targeted solutions. Think of this like writing a design brief ;  the clearer your problem definition, the more focused and valuable your AI solution becomes.

Explore: Use ensemble methods to iterate beyond your first solution

Don’t settle for your first prompt or first output. This is where UX designers have a huge advantage — you’re already trained to iterate and explore multiple solutions. Apply ensemble methods to your prompting: try role-based prompting where you ask AI to respond as different stakeholders (UX writer, software engineer, product manager) and compare their perspectives.

Use model ensemble by running the same prompt through Claude, ChatGPT, and Gemini to see different approaches. Test prompt ensemble by writing multiple versions of your request and seeing which produces better results. Then combine the best elements from each approach to create a superior final solution.

Navigate: Test your prompts with real humans in real contexts

This is where most people fail . They optimize prompts for themselves but never test them with their actual users. Move beyond solitary prompting to collaborative prompt development. If you’re building AI systems for your team, test them with colleagues who have different knowledge levels and use cases. Try to break your prompts intentionally:  what happens when someone uses them incorrectly or asks edge case questions? Build error handling into your prompts using natural language: “If the user asks about [specific scenario], respond with [desired guidance].” Create prompt templates that your team can use and polish together, treating them like shared design systems.

Transform individual productivity into team advantage

The real power of the Eden framework isn’t just better personal prompts — it’s building AI systems that elevate your entire organization. Phase one of AI mastery is optimizing your individual workflow. Phase two is creating systems that make your whole team more effective. Document successful prompts in a shared knowledge base where colleagues can test, refine, and improve them collaboratively. This transforms prompting from a solitary skill into a team capability that compounds over time.

Systematic prompting gives you a strategic advantage

You already understand iterative design, user testing, and human-centred problem solving. These skills directly translate to prompt engineering excellence. While others are focused on technical tricks, you’re approaching AI with systematic methodology. You know how to empathize with users, define problems clearly, explore multiple solutions, and test with real people. This design thinking foundation makes you naturally better at creating AI solutions that actually work in practice.

Entry-level AI usage is becoming commoditized. Anyone can ask ChatGPT to write an email. But strategic AI application requires the exact skills you’ve been developing as a UX professional. When your manager asks “how can we optimize our workflow with AI?,” you’ll have concrete systems and frameworks ready to implement. You’ll be the person who can build reliable AI tools, not just use them occasionally.

Stop thinking of yourself as someone who uses AI and start thinking of yourself as someone who designs AI interactions. Just like you design user interfaces, you can design AI experiences that serve specific human needs. The EDEN framework gives you a repeatable process for creating AI solutions that work reliably for real people in real contexts.

AI isn’t replacing UX designers at all. It’s elevating the ones who learn to design with it systematically. Master the EDEN framework, and you’ll be building the AI-powered design solutions that others can only dream about.

Support Beyond the Button: AI Strategies for UX by joining for a FREE subscription.

The UX design industry is splitting into two groups: those who embrace AI and thrive, and those who resist it and get left behind. The choice is yours, but the window to make this transition is closing fast. And we’re offering you an invaluable Masterclass to easily get ahead.



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit uxandai.substack.com
...more
View all episodesView all episodes
Download on the App Store

Beyond the Button Podcast: AI Strategies for UXBy Lennart Nacke & Reza Mogavi