63 of 63 cards · Click any card to flip · Double-click to view details
The camera didn't replace the painter — it created entirely new forms of visual communication. Photography freed painters to explore abstraction, emotion, and conceptual art. AI won't replace your craft; it will free you to explore what only humans can do.
The designers who will thrive aren't the ones who know AI best — they're the ones who ask the best questions OF AI, FOR customers. You already know how to ask great questions. Now you're learning to ask them at scale.
Write down your biggest fear about AI in your design practice. Now, as a team, reframe that fear as an opportunity. What new capability does this fear point to?
AI is not a replacement for thinking — it's an amplifier for thinking. A microphone doesn't sing for you. A telescope doesn't see for you. AI doesn't design for you. It amplifies the questions you ask and the patterns you notice.
Use this prompt to see your own role through a new lens.
"I'm a [your role] with [X] years of experience. Describe 5 ways AI could amplify my existing expertise rather than replace it. For each, give a specific example of how my human judgment remains essential."
Imagine it's 2030. AI tools are as common as Figma is today. Write a 3-sentence "day in the life" of a designer in your craft. What does your role look like? Share with the group.
AI excels at: Pattern recognition, volume processing, alternative perspectives. Humans excel at: Judgment, empathy, contextual nuance, ethical decision-making. TOGETHER: Deeper customer understanding than either alone.
List 10 tasks you do in a typical week. Mark each as: (H) Only humans can do this well, (A) AI could do this faster, (B) Best done together. Discuss: What surprised you?
Use AI to find historical parallels to the current moment.
"What are 5 historical examples of creative professionals who feared a new technology would replace them, but instead it expanded their field? For each, explain what new roles or capabilities emerged."
For the next 5 minutes, argue the OPPOSITE of your actual belief about AI in design. If you're skeptical, argue it's transformative. If you're enthusiastic, argue it's overhyped. What did you learn?
When photography emerged, Bauhaus designers feared illustration was dead. Instead, they invented graphic design, photomontage, and visual communication as we know it. Every tool that threatens the old creates the new.
Identify ONE specific task you'll do next Monday. Plan how you'd do it without AI, then with AI. What's the difference in time? Quality? Insight? Commit to trying the AI version.
Surface patterns that human analysts typically miss.
"Analyze this [research type] from a [UX/Service/Visual/Content/Research] perspective. What patterns, edge cases, or unspoken needs might a human analyst miss? What questions should we be asking that we haven't?"
Go beyond what users say to what they feel.
"Analyze the emotional journey in this customer data. Map the emotional highs and lows. Where is there anxiety, frustration, delight, or confusion? What emotions are present but not explicitly stated?"
Take the same customer research artifact. Spend 4 minutes analyzing it manually — identify top 3 pain points and 2 opportunities. Then spend 4 minutes using AI prompts. Compare: What did AI surface that you didn't? What did YOU catch that AI missed?
Create rich personas from raw research data.
"Based on this [interview transcript/survey data/support tickets], create 3 distinct user personas. For each, include: demographics, goals, frustrations, unspoken needs, and a direct quote that captures their essence."
Find where users say one thing but mean another.
"Review this research data and identify contradictions — places where what users say doesn't match what they do, or where different users have conflicting needs. What design tensions does this reveal?"
Exchange your research artifact with another team. You have 3 minutes to analyze THEIR data using AI. Fresh eyes + AI = unexpected insights. Report back what you found.
Find the users everyone else forgets about.
"Based on this research, who are the edge-case users we might be overlooking? Consider: accessibility needs, cultural differences, technical literacy levels, emotional states, and situational contexts that could change the experience."
AI can process 100 interview transcripts in minutes. But it takes a human to notice that the participant's voice cracked when talking about their mother. Use AI for volume. Use yourself for depth.
Transform raw data into a structured journey map.
"From this customer data, create a detailed journey map with these stages: Awareness, Consideration, Decision, Onboarding, Usage, Advocacy. For each stage, identify: actions, thoughts, emotions, pain points, and opportunities."
Your team has 3 different research artifacts. Use AI to find the common themes across ALL of them. Create a single "insight statement" that captures the most important finding. Present it in one sentence.
Expose the hidden assumptions in your research.
"Review this research methodology and findings. What assumptions are baked into how this research was conducted? What biases might be present in the questions asked? What populations or perspectives are missing?"
Set a timer for 60 seconds. Each team member writes ONE insight from the research on a sticky note. No AI allowed. Then use AI to generate 10 more insights in 60 seconds. Which set is more surprising?
Challenge the assumptions hiding in your problem statement.
"What assumptions am I making about [user group/problem/context]? List at least 10 assumptions, then for each one, suggest what would change if that assumption were wrong."
Get diverse expert perspectives on your challenge.
"What would a [behavioral scientist / accessibility expert / anthropologist / child psychologist / economist] say about this problem: [describe problem]? Give me each expert's unique perspective and the questions they would ask."
Push past safe, predictable ideas.
"Generate 10 ideas for [challenge], but make 5 of them feel uncomfortably different from what we'd normally do. Push into territory that feels risky, weird, or counterintuitive. Explain why each "uncomfortable" idea might actually work."
See your problem through the eyes of innovative companies.
"What would [IDEO / Apple / Duolingo / IKEA / Nintendo] do if they were solving this problem: [describe problem]? For each company, describe their likely approach based on their known design philosophy."
Build a "Question Map" in three rounds. Round 1 (3 min): Ask AI to help you understand the problem differently. Round 2 (3 min): Ask AI to help you ideate unconventionally. Round 3 (4 min): Ask AI to stress-test your best idea. Map the trail of questions that led to unexpected insights.
Go deeper than surface-level problems.
"For this problem: [describe problem], perform a "Five Whys" analysis but at each level, also ask "Who else is affected?" and "What would happen if we did nothing?" Give me the full chain of reasoning."
A 6-year-old asks "Why?" about everything. Use AI to generate 10 questions a curious child would ask about your product or service. Which of these "naive" questions reveals a genuine design opportunity?
Flip the problem on its head.
"Instead of solving [problem], what if we asked: "How could we make this problem WORSE?" Generate 10 ways to make it worse, then invert each one into a potential solution."
AI is best used not to find answers, but to find better questions. The quality of your AI output is directly proportional to the quality of your input. Great designers don't need AI to think for them — they need AI to help them think bigger.
Add creative constraints to break out of conventional thinking.
"For this design challenge: [describe challenge], generate 5 creative constraints that would force us to think differently. For example: "What if it had to work without any text?" or "What if the entire experience lasted only 30 seconds?""
As a team, generate as many questions as possible about your challenge in 5 minutes using AI. Then vote on the top 3 most surprising questions. These go into the shared Question Library for the community of practice.
Draw TWO cards from different categories. Use AI to find a connection between them and apply it to your current challenge. The more unlikely the connection, the more creative the result.
Kill your idea before it kills your project.
"Take this idea: [insert idea]. Imagine it's 6 months from now and this idea has completely failed. What went wrong? Give me 10 specific reasons for failure, ordered from most to least likely."
Get the harshest critique before your stakeholders do.
"Act as a cynical, risk-averse executive who has seen many projects fail. Review this idea: [insert idea]. List all the reasons this will fail, what could go wrong, and who might this exclude or harm."
Check who your design might leave behind.
"Review this design approach: [describe approach]. Who might this exclude? Consider: age, ability, literacy, language, culture, economic status, technology access, and emotional state. For each excluded group, suggest an alternative."
Your current best idea is now FORBIDDEN. You cannot use it. Draw three Prompt Cards from any category and use them to generate a completely new approach in 5 minutes.
Examine the ethical implications of your design.
"Analyze this design decision through an ethical lens: [describe decision]. What are the potential unintended consequences? How could this be misused? What data privacy concerns exist? What power dynamics does this create or reinforce?"
Take your current solution and design the EXACT OPPOSITE. If your solution is simple, make it complex. If it's digital, make it physical. If it's fast, make it slow. What insights emerge from the opposite?
The best critique isn't "I don't like it." The best critique is "What would happen if...?" AI can generate 50 "what ifs" in seconds. Your job is to judge which ones matter.
Stress-test your idea at different scales.
"Take this solution: [describe solution]. What happens when 10 people use it? 1,000? 1,000,000? What breaks at each scale? What new problems emerge? What would need to change?"
Swap your challenge with another team. You have 5 minutes to understand their problem and offer a completely new AI-powered approach. Fresh eyes see what familiar eyes miss.
Strip your idea down to its essence.
"Explain this design solution in one sentence a 10-year-old would understand: [describe solution]. If you can't, what complexity needs to be removed? Now suggest 3 ways to simplify without losing the core value."
Take your current idea. Use AI to generate 5 variations in 3 minutes. For each variation, identify one thing that's better than the original. Combine the best elements into a new, stronger version.
When you're stuck, change the rules. Don't try harder — try differently. Ask AI: "Give me a completely different way to frame this problem that has nothing to do with my current approach."
Check your user flow for barriers.
"Analyze this user flow from an accessibility perspective. Identify potential barriers for users with: visual impairments, motor disabilities, cognitive differences, and situational limitations (one-handed use, bright sunlight, noisy environment)."
Identify unnecessary friction in your experience.
"Review this checkout/onboarding/signup flow: [describe flow]. Identify every point of friction. For each, classify as: necessary friction (security, legal), reducible friction (can be simplified), or unnecessary friction (should be eliminated)."
How might we reduce checkout abandonment for elderly customers? Use AI to understand the specific barriers elderly users face, then generate 5 solutions that would work for them WITHOUT making the experience worse for younger users.
"What specific UX barriers do elderly users (70+) face in digital checkout processes? Consider: vision, motor skills, cognitive load, trust, technology familiarity. For each barrier, suggest a solution that benefits ALL users."
Map the emotional landscape of your service.
"For this service journey: [describe service], map the emotional experience at each touchpoint. Where does anxiety peak? Where is there an opportunity for delight? Where do customers feel lost or abandoned? Suggest interventions for the 3 worst emotional moments."
How might we make mortgage applications feel less intimidating? Use AI to understand the psychology of financial anxiety, then design 3 service interventions that reduce fear without oversimplifying the process.
"What psychological factors make mortgage applications intimidating? Consider: complexity, jargon, financial risk, time pressure, power imbalance. For each factor, suggest a service design intervention."
Design the invisible parts of the service.
"For this customer-facing service: [describe service], what happens backstage that the customer never sees? Map the internal processes, handoffs, and potential failure points. Where could backstage improvements dramatically improve the frontstage experience?"
Communicate trust through visual design.
"How can we visually communicate trust and security for [product/feature]? Analyze what visual elements (color, typography, spacing, imagery, iconography) signal trustworthiness. Provide 3 distinct visual approaches with specific design recommendations."
Explore unexpected visual directions.
"Generate 3 different mood board concepts for this brand/product: [describe]. Each must be inspired by a different source: one from nature, one from architecture, and one from a historical art movement. Describe the colors, textures, typography, and imagery for each."
How might we communicate trust in our new security features? Use AI to explore how different visual languages convey safety, then create a visual design brief that balances security with approachability.
Make complex information accessible.
"Rewrite this technical explanation for a 5th-grade reading level: [paste text]. Use an analogy to explain the core concept. Then provide 3 alternative versions: one formal, one conversational, and one that uses a story format."
How might we explain data privacy to non-technical users? Use AI to find the right metaphors and language that make privacy feel understandable without being patronizing.
"Explain data privacy to a non-technical user using 3 different approaches: a metaphor, a story, and a Q&A format. Each should cover: what data we collect, why, how it's protected, and what control the user has. Avoid jargon entirely."
Find the right voice for the moment.
"For this user scenario: [describe scenario], write the same message in 5 different tones: reassuring, direct, playful, empathetic, and authoritative. Which tone best serves the user's emotional state at this moment? Why?"
Discover what you don't know you don't know.
"For this research question: [describe question], what are we NOT asking that we should be? What adjacent fields or disciplines might have relevant insights? What methodologies from other domains could we borrow?"
What don't we know about why customers leave? Use AI to generate hypotheses about customer churn that go beyond the obvious. Then design a research plan to test the top 3 most surprising hypotheses.
"Generate 15 hypotheses for why customers might leave [product/service], ranging from obvious to highly unconventional. For the 5 most unconventional, suggest a research method to test each hypothesis."
Audit your research for hidden biases.
"Review this research plan: [describe plan]. Identify potential biases in: participant selection, question framing, data interpretation, and researcher assumptions. For each bias, suggest a specific mitigation strategy."