Unit 4: Ideation to Visualization - Generative AI as co-creator - Subjective Questions

Co-imagination refers to the collaborative creative process where humans and Generative AI systems jointly ideate, explore, and generate design solutions, treating the AI not merely as a tool but as an active creative partner.

Key aspects:

• Shared authorship: Both the human designer and the AI contribute ideas that build upon each other.
• Augmented creativity: AI expands the solution space by producing variations, unexpected combinations, and rapid iterations.
• Dialogue-driven: The designer refines prompts while the AI responds, creating an iterative feedback loop.

Why it is a paradigm shift:

• It moves design from human-only ideation to human-AI collaboration.
• It accelerates the divergent phase of Design Thinking by generating hundreds of possibilities.
• It raises new questions about originality, ownership, and ethics that traditional design did not face.

In essence, co-imagination reframes the designer's role from sole creator to curator and orchestrator of machine-augmented creativity.

Plagiarism in the context of Generative AI arises because these models are trained on vast datasets of existing human-created work, and their outputs may closely resemble or directly reproduce copyrighted material without attribution.

Why it is a concern:

• Uncredited reproduction: AI may regenerate portions of training data, unknowingly copying an artist's style or content.
• Blurred authorship: It becomes unclear whether the output is original or derived.
• Legal risk: Using such outputs commercially may infringe intellectual property rights.

Mitigation strategies:

• Verification: Cross-check AI outputs against existing works using reverse-image or plagiarism-detection tools.
• Transparency: Disclose the use of AI in the creative process.
• Substantial transformation: Use AI outputs as a starting point and add significant human creativity.
• Licensing awareness: Use models trained on licensed or public-domain data.
• Attribution: Credit sources where identifiable.

Ethical use requires treating AI outputs as drafts requiring human judgment, not final deliverables.

Creativity and originality are related but distinct concepts:

Can AI-generated work be truly original?

• Argument against: AI recombines patterns from training data, so outputs are fundamentally derivative rather than truly novel.
• Argument for: Novel combinations not present in the training set can be considered a form of computational creativity.
• Balanced view: AI exhibits combinatorial creativity but lacks intentionality and lived experience. True originality still depends on the human's conceptual framing and purpose.

Conclusion: AI can enhance creativity, but originality in the deepest sense remains tied to human intent and meaning-making.

The transition "From Ideas to Action" bridges the gap between abstract ideation and tangible implementation, ensuring that generated ideas do not remain theoretical.

Key steps involved:

1. Idea Filtering & Selection:

   • Evaluate ideas against feasibility, viability, and desirability.
   • Use tools like impact-effort matrices.

2. Prioritization:

   • Rank ideas based on user value and business goals.

3. Prototyping:

   • Build low-fidelity models to make ideas tangible.

4. Testing & Feedback:

   • Validate assumptions with real users.

5. Iteration:

   • Refine based on feedback loops.

6. Action Planning:

   • Define tasks, resources, timelines, and responsibilities.

7. Implementation:

   • Execute the refined solution in the real world.

Role of Generative AI:

• AI can rapidly prototype visuals, generate action roadmaps, and simulate outcomes, accelerating the leap from ideas to action.

Key principle: Ideas gain value only when they are executed, tested, and iterated into real solutions.

Framing is the process of defining the problem in a specific way, establishing boundaries, assumptions, and perspectives. Reframing is the deliberate act of viewing the same problem from a different angle to unlock new solution possibilities.

Framing the problem space:

• Defines what problem is being solved.
• Sets the scope and constraints.
• Uses techniques like "How Might We" questions.

Reframing the problem space:

• Challenges existing assumptions.
• Shifts perspective (e.g., from user to stakeholder, from feature to need).
• Asks: "What if we looked at this differently?"

Why reframing is critical for innovation:

• Avoids solving the wrong problem: A poorly framed problem leads to irrelevant solutions.
• Opens new solution spaces: New frames reveal opportunities invisible under the original frame.
• Breaks fixation: Prevents anchoring on obvious but suboptimal ideas.
• Enables breakthrough thinking: Many innovations come from redefining the problem itself.

Example: Instead of "How do we build a faster elevator?" (technical), reframing to "How do we make waiting feel shorter?" (experiential) led to installing mirrors — a cheaper, effective solution.

Ideation tools help teams generate a large quantity and diversity of ideas during the divergent phase of Design Thinking.

1. Brainstorming:

• Free-flowing generation of ideas without judgment.
• Rules: defer criticism, encourage wild ideas, build on others' ideas, aim for quantity.

2. Mind Mapping:

• A visual technique branching out from a central concept.
• Helps explore connections and associations between ideas.

3. SCAMPER:

• A checklist prompting: Substitute, Combine, Adapt, Modify, Put to other use, Eliminate, Reverse.
• Useful for improving existing products.

4. Brainwriting (6-3-5 Method):

• 6 participants write 3 ideas in 5 minutes, then pass sheets to build on others' ideas.
• Reduces dominance by vocal members.

5. "How Might We" (HMW) Questions:

• Reframe problems into open-ended opportunity statements.

6. Worst Possible Idea:

• Generate bad ideas deliberately to reduce fear and spark unexpected good ones.

Role of Generative AI: AI tools (like image and text generators) can supplement these methods by producing rapid idea variations and unexpected combinations.

Idea Selection is the convergent phase of ideation where the large pool of generated ideas is narrowed down to the most promising ones for prototyping and development.

Common methods and tools:

1. Dot Voting:

   • Team members place dots on their favorite ideas.
   • Quick, democratic prioritization.

2. Impact-Effort Matrix:

   • Plots ideas on two axes: impact (value) vs. effort (cost).
   • Prioritizes high-impact, low-effort "quick wins."

3. Now-How-Wow Matrix:

   • Categorizes ideas by feasibility and originality:
     • Now: Easy and normal ideas.
     • How: Original but hard to implement.
     • Wow: Original and feasible (ideal).

4. Feasibility-Viability-Desirability (FVD) Framework:

   • Desirability: Do users want it?
   • Feasibility: Can we build it?
   • Viability: Is it sustainable/profitable?
   • Best ideas sit at the intersection.

5. Weighted Scoring Matrix:

   • Rate ideas against weighted criteria and total the scores.

Purpose: Idea selection ensures resources are invested in ideas that balance user needs, technical feasibility, and business value.

Visualization is the practice of representing ideas, data, and concepts in visual form (sketches, diagrams, models, maps) to communicate and explore them more effectively.

Importance in Design Thinking:

• Makes abstract ideas tangible: Converts vague concepts into concrete, discussable artifacts.
• Enhances communication: Visuals transcend language barriers and align diverse stakeholders.
• Reveals gaps and insights: Patterns and problems become visible that text alone may hide.
• Supports empathy: Journey maps and personas help visualize user experiences.
• Accelerates feedback: Stakeholders can react to visuals faster than to written descriptions.
• Aids decision-making: Comparing visual options clarifies trade-offs.

How it enhances the process:

• Empathize phase: Empathy maps, personas.
• Define phase: Problem statement visuals, affinity diagrams.
• Ideate phase: Mind maps, sketches.
• Prototype phase: Wireframes, storyboards, mockups.
• Test phase: Annotated visuals capturing feedback.

With Generative AI: AI can rapidly produce concept visuals, mockups, and illustrations, dramatically speeding up the visualization stage.

Conclusion: Visualization is the shared language of Design Thinking that bridges thinking and making.

Design Thinking employs multiple visualization methods to represent ideas, users, and processes.

1. Sketching:

• Quick hand-drawn representations of ideas.
• Application: Early concept exploration.

2. Storyboarding:

• A sequence of frames depicting a user's interaction over time.
• Application: Visualizing user experiences and scenarios.

3. Journey Mapping:

• Visual timeline of a user's experience across touchpoints, capturing emotions and pain points.
• Application: Identifying opportunities to improve experience.

4. Mind Mapping:

• Branching diagrams showing relationships between concepts.
• Application: Ideation and organizing thoughts.

5. Wireframing & Mockups:

• Low- to high-fidelity layouts of digital interfaces.
• Application: Designing apps and websites.

6. Personas:

• Visual profiles of representative users.
• Application: Keeping the team user-centered.

7. Affinity Diagrams:

• Grouping sticky notes/ideas into themes.
• Application: Synthesizing research data.

8. Infographics & Data Visualization:

• Charts and diagrams presenting data clearly.
• Application: Communicating findings.

Generative AI contribution: AI tools can auto-generate mockups, illustrations, and diagrams from text prompts, enhancing all these methods.

Comparison of Traditional vs. AI-Assisted Ideation:

Advantages of Traditional Ideation:

• Rich in context and empathy.
• Ideas grounded in real human experience.
• Strong team ownership and collaboration.

Limitations of Traditional Ideation:

• Slower and prone to groupthink and fixation.
• Limited by participants' knowledge.

Advantages of AI-Assisted Ideation:

• Rapid generation of many variations.
• Breaks fixation by offering unexpected options.
• Explores broad solution spaces.

Limitations of AI-Assisted Ideation:

• May produce generic or derivative ideas.
• Raises ethical/originality concerns.
• Lacks genuine understanding and intent.

Best practice: Combine both — use AI for breadth and speed, and human judgment for depth, empathy, and ethical curation.

Using Generative AI as a co-creator introduces significant ethical responsibilities. A robust ethical framework ensures responsible co-imagination.

1. Transparency & Disclosure:

• Always disclose when AI has contributed to a work.
• Example: Labeling AI-generated images in a portfolio.

2. Attribution & Intellectual Property:

• Respect the rights of original creators whose work trained the AI.
• Avoid passing off AI-derived work as fully original.
• Example: Not replicating a living artist's signature style commercially.

3. Avoiding Plagiarism:

• Verify outputs are not near-copies of existing works.
• Add substantial human transformation.

4. Fairness & Bias Mitigation:

• Recognize AI reproduces biases in training data.
• Example: Ensuring generated personas represent diverse users, not stereotypes.

5. Accountability:

• The human designer remains responsible for the final output.
• AI is a tool, not a scapegoat.

6. Authenticity & Originality:

• Preserve genuine human creativity and intent.
• Use AI to augment, not replace, human imagination.

7. Consent & Privacy:

• Do not use personal data or likenesses without consent.

8. Sustainability:

• Consider the environmental cost of large AI models.

Applying the framework:

• Before: Choose ethically trained models.
• During: Prompt responsibly, avoiding harmful or infringing requests.
• After: Review, transform, verify, and disclose.

Conclusion: Ethical co-creation balances the power of AI with the designer's duty of honesty, respect, fairness, and accountability, ensuring technology enhances rather than undermines creative integrity.

The journey from Ideation to Visualization with Generative AI as a co-creator involves several integrated stages.

Stage 1: Framing & Reframing the Problem Space

• Define the problem clearly using "How Might We" statements.
• Reframe from multiple perspectives to widen the opportunity space.
• AI role: Generate alternative problem framings and reveal overlooked angles.

Stage 2: Ideation (Divergent Phase)

• Use tools like brainstorming, SCAMPER, mind mapping.
• AI role: Rapidly produce large volumes of idea variations, unexpected combinations, and stimulus prompts to break fixation.

Stage 3: Idea Selection (Convergent Phase)

• Apply methods like Impact-Effort Matrix, Now-How-Wow, and FVD framework.
• AI role: Cluster similar ideas, summarize themes, and even score ideas against criteria.

Stage 4: From Ideas to Action

• Convert selected ideas into concrete plans and prototypes.
• Prioritize, plan resources, and define next steps.
• AI role: Generate action roadmaps and simulate potential outcomes.

Stage 5: Visualization

• Represent ideas visually via sketches, wireframes, storyboards, and mockups.
• AI role: Instantly generate concept visuals, illustrations, and prototypes from text prompts.

Stage 6: Ethical Review

• Check outputs for plagiarism, bias, and originality.
• Add human refinement and disclose AI involvement.

Integration & Feedback Loop:

• The process is iterative — visualization often feeds back into reframing and re-ideation.

Conclusion: Generative AI accelerates each stage, but the human designer orchestrates the process, ensuring the flow from vague ideas to concrete, ethical, and user-centered visual solutions.

"How Might We" (HMW) questions are open-ended, opportunity-focused statements that reframe problems and insights into actionable challenges to guide ideation.

Structure: How Might We [action] for [user] so that [outcome]?

Role in framing the problem space:

• Optimistic framing: "Might" suggests possibility without guaranteeing solutions.
• Solution-neutral: They invite many answers rather than prescribing one.
• Right scope: Not too broad (overwhelming) nor too narrow (limiting).
• Bridge: Connect the Define phase to the Ideate phase.

Examples of Good HMW statements:

• "How might we make waiting for the elevator more enjoyable?"
• "How might we help students collaborate remotely without feeling isolated?"

Examples of Poor HMW statements:

• Too broad: "How might we redesign education?" (unfocused)
• Too narrow: "How might we add a blue button to the app?" (prescribes solution)
• Not user-centered: "How might we increase profit?" (business-only)

Best practice: Generate multiple HMW questions from a single insight, then select the ones with the best scope for productive ideation.

The Impact-Effort Matrix is a prioritization tool that helps teams decide which ideas to pursue by evaluating them on two dimensions.

Construction:

• Draw a 2x2 grid.
• X-axis: Effort (low to high) — resources, time, cost required.
• Y-axis: Impact (low to high) — value or benefit delivered.
• Plot each idea in the appropriate quadrant.

The Four Quadrants:

Interpretation:

• Quick Wins are prioritized for immediate action.
• Major Projects are worthwhile but need strategic planning.
• Fill-ins are low priority.
• Thankless Tasks should generally be dropped.

Benefits:

• Simple, visual, and collaborative.
• Focuses resources on the highest-value opportunities.

Limitation: Estimates of impact and effort can be subjective, so team discussion is essential.

Combinatorial creativity is the process of generating new ideas by combining existing concepts, elements, or knowledge in novel ways. It is one of the most common forms of creativity.

How Generative AI exhibits it:

• AI models are trained on massive datasets and learn patterns across domains.
• They recombine learned elements to produce outputs that did not exist explicitly in the training data.
• Example: Generating an image of "a Victorian house in the style of cyberpunk" merges two distinct concepts.

Characteristics of AI's combinatorial creativity:

• High volume: Produces countless combinations quickly.
• Cross-domain: Blends unrelated fields effortlessly.
• Pattern-based: Relies on statistical associations, not understanding.

Does it diminish human creativity?

Arguments it might:

• Over-reliance may weaken original human thinking.
• Homogenization if everyone uses similar tools.

Arguments it does not:

• AI handles combinatorial breadth, freeing humans for conceptual depth and intent.
• Humans provide meaning, context, and judgment AI lacks.
• It augments creativity by offering stimuli that spark human ideas.

Conclusion: Rather than diminishing creativity, AI's combinatorial power can amplify human creativity when used as a collaborative springboard, provided humans retain critical and intentional roles.

The Now-How-Wow Matrix is an idea-selection tool that categorizes ideas based on two dimensions: originality/innovation and ease of implementation (feasibility).

Structure:

• X-axis: Feasibility / ease of implementation (easy to difficult).
• Y-axis: Originality / innovation (normal to original).

The Three Categories:

1. Now (Blue):

   • Normal ideas that are easy to implement.
   • Ready-to-use, low-risk solutions.
   • Best for: Immediate implementation.

2. How (Yellow):

   • Original ideas that are currently hard to implement.
   • Innovative but need future technology or resources.
   • Best for: Long-term vision / research.

3. Wow (Green):

   • Original ideas that are easy to implement.
   • The sweet spot — innovative and feasible.
   • Best for: Prioritized action.

How it balances feasibility and originality:

• It prevents teams from choosing only safe, ordinary ideas (Now) or unrealistic dreams (How).
• It steers focus toward Wow ideas that maximize both innovation and practicality.

Benefit: Encourages a mix of quick wins and breakthrough thinking while remaining realistic about resources.

Storyboarding is a visualization technique that uses a sequence of drawings or frames to depict how a user interacts with a product, service, or experience over time — much like a comic strip.

Structure of a storyboard:

• A series of panels/frames.
• Each frame shows a scene, action, or step in the user's journey.
• Often includes captions describing context, emotions, or dialogue.

Role in Design Thinking:

• Narrative visualization: Turns abstract ideas into a concrete story.
• User-centered: Focuses on real scenarios and touchpoints.
• Sequential clarity: Shows the flow of an experience step by step.

How it helps communicate design ideas:

• Empathy: Places the user at the center, revealing emotions and pain points.
• Shared understanding: Aligns team members and stakeholders around a common vision.
• Identifies gaps: Exposes missing steps or friction in the experience.
• Low cost: Quick and inexpensive to create and revise.
• Persuasion: A compelling story communicates value better than specifications.

With Generative AI: AI can quickly generate storyboard illustrations from text descriptions, speeding up the visualization of scenarios.

Conclusion: Storyboarding transforms ideas into relatable, human-centered narratives that make design concepts tangible and communicable.

The Double Diamond model describes design as alternating between divergent (expanding options) and convergent (narrowing options) thinking across two diamonds: Discover-Define and Develop-Deliver. Generative AI supports both modes.

Divergent Phase (Expanding possibilities):

• Goal: Generate as many ideas as possible.
• AI contributions:
  • Produces large volumes of idea variations rapidly.
  • Suggests unexpected, cross-domain combinations.
  • Breaks cognitive fixation by offering novel stimuli.
  • Generates multiple problem framings and "How Might We" questions.

Convergent Phase (Narrowing choices):

• Goal: Select and refine the best ideas.
• AI contributions:
  • Clusters and summarizes similar ideas.
  • Scores ideas against defined criteria.
  • Highlights patterns and themes in large idea sets.
  • Helps refine selected concepts into polished prototypes.

Double Diamond mapping:

• First diamond (problem): AI aids in exploring the problem space (diverge) and defining it clearly (converge).
• Second diamond (solution): AI aids in generating solutions (diverge) and refining the final deliverable (converge).

Caution: In the convergent phase, human judgment must lead, since AI lacks genuine understanding of user context and business value.

Conclusion: Generative AI is a powerful accelerator across both diamonds, but effective use requires balancing AI's breadth with human critical evaluation.

The Feasibility-Viability-Desirability (FVD) framework, popularized by IDEO, is a lens for evaluating and selecting ideas by examining them from three critical perspectives.

The Three Lenses:

1. Desirability (Human lens):

   • Question: Do people actually want or need this?
   • Focuses on user needs, emotions, and experiences.

2. Feasibility (Technical lens):

   • Question: Can we build/deliver it with available technology and capabilities?
   • Focuses on technical and operational capability.

3. Viability (Business lens):

   • Question: Is it financially sustainable and profitable?
   • Focuses on business model and long-term sustainability.

Why an ideal solution must satisfy all three:

• Desirable but not feasible: A great idea users love but cannot be built — remains a fantasy.
• Feasible but not desirable: A buildable product no one wants — commercial failure.
• Viable but not desirable: Profitable on paper but no user demand — fails in market.
• Sweet spot: The intersection of all three is where successful, sustainable innovation lives.

Visualization: Three overlapping circles (Venn diagram); the ideal idea sits at the center overlap.

Conclusion: The FVD framework ensures balanced decision-making, preventing teams from pursuing ideas strong in one area but weak overall.

The rise of Generative AI as a co-creator challenges long-held notions of authorship and originality in design.

Impact on Authorship:

• Blurred ownership: When AI generates significant portions of a work, it is unclear who the author is — the designer, the AI developer, or the creators of the training data.
• Legal ambiguity: Many jurisdictions do not grant copyright to purely AI-generated works, requiring meaningful human input.
• Collaborative authorship: Emerges as a new model where humans and AI share the creative process.

Impact on Originality:

• Derivative concern: AI recombines existing works, raising questions about whether outputs are truly new.
• Homogenization risk: Wide use of similar models may produce visually similar outputs, reducing diversity.
• Redefinition: Originality shifts from the artifact to the conceptual framing and curation by the human.

Safeguards to preserve human creative identity:

1. Meaningful human input: Ensure substantial human decision-making, editing, and intent in the final work.
2. Curation over generation: Position AI as a source of raw material that humans selectively refine.
3. Transparency: Disclose AI involvement to maintain honesty.
4. Personal voice: Infuse work with unique human perspective, values, and style.
5. Ethical prompting: Avoid mimicking specific living artists.
6. Verification: Check outputs for unintended plagiarism.
7. Documentation: Record the creative process to demonstrate human authorship.
8. Skill development: Continue nurturing human creative skills rather than outsourcing them entirely.

Conclusion: Generative AI does not have to erode human creative identity. When designers act as intentional curators and directors — bringing meaning, context, and ethical judgment — they preserve authorship and elevate originality to the conceptual level, ensuring the human remains at the heart of creativity.