AI for Product Designers
AI for product designers.
Where AI helps—and where human judgment still matters.
AI can accelerate research synthesis, exploration, documentation, and repetitive design tasks. But it cannot replace the judgment required to understand people, navigate ambiguity, prioritize product decisions, or determine whether a solution is right.
AI is most valuable when it improves the work—not when it replaces the thinking.
The goal is not to add AI to every step of the design process. The goal is to use it intentionally where it can create more time for research, collaboration, iteration, and better decision-making.
I use AI as a collaborative tool for organizing complex information, generating alternatives, testing different ways of communicating, and accelerating early exploration.
I override or reject AI output when it lacks context, invents evidence, oversimplifies users, introduces bias, weakens the product strategy, or produces a solution that does not reflect real workflows.
AI can accelerate the process. Designers remain responsible for the outcome.
Every AI-assisted activity still requires review, interpretation, validation, and accountability from the product designer.
Speed, structure, exploration, and operational support.
- Organizing interview notes and identifying possible themes
- Summarizing large amounts of product or stakeholder information
- Generating early questions, hypotheses, and alternative directions
- Exploring multiple content structures and interaction concepts
- Creating first drafts of documentation and design specifications
- Comparing patterns, requirements, and research findings
- Reducing repetitive production and administrative tasks
Context, empathy, ethics, prioritization, and accountability.
- Determining whether research findings are accurate and meaningful
- Understanding motivations, emotions, constraints, and workarounds
- Recognizing organizational, operational, and cultural context
- Balancing user needs with business and technical considerations
- Identifying bias, unsupported assumptions, and misleading output
- Deciding which problem should be solved and why it matters
- Accepting responsibility for the product experience and outcome
Human direction at every stage.
AI supports specific activities within my process, but research evidence, product context, and human review determine what moves forward.
Define the problem before involving AI.
I first clarify the product problem, audience, business goal, known constraints, and unanswered questions. Without this foundation, AI tends to generate polished but generic output.
- Organize existing context
- Identify missing questions
- Structure discovery plans
- What problem matters
- Who needs to be involved
- What evidence is required
Use AI to support analysis—not replace user contact.
Interviews, observation, analytics, usability testing, and stakeholder collaboration remain the foundation. AI can help organize the evidence after it has been gathered.
- Summarize interview notes
- Suggest emerging themes
- Compare findings across sessions
- Whether a pattern is real
- What context changes its meaning
- Which findings require more research
Accelerate organization while protecting the evidence.
AI can quickly group information and propose themes, but I review every conclusion against the original research to avoid invented relationships or oversimplified insights.
- Group similar observations
- Create draft research summaries
- Surface repeated language
- Which insights are supported
- How findings connect to behavior
- What should influence the product
Generate more possibilities before narrowing the direction.
AI is useful for quickly exploring alternative content models, workflow ideas, edge cases, and early concepts. These outputs are starting points rather than finished solutions.
- Generate alternative approaches
- Identify possible edge cases
- Challenge an early concept
- Which direction fits the research
- Which tradeoffs are acceptable
- What deserves prototyping
Transform ideas into intentional product experiences.
I use research, interaction principles, accessibility, information architecture, and design-system standards to turn promising directions into usable product experiences.
- Draft interface content
- Explore information hierarchies
- Generate placeholder scenarios
- How the workflow should behave
- What information receives priority
- Whether the experience is accessible
Test with people instead of trusting generated confidence.
AI can simulate questions or critique a concept, but it cannot replace validation with real users working through meaningful tasks in the correct context.
- Draft testing questions
- Identify possible usability risks
- Organize test observations
- What needs to be validated
- How behavior should be interpreted
- Which changes should be made
Use AI to reduce documentation work without losing intent.
AI can accelerate first drafts of specifications, summaries, and release documentation. I review and revise the output so engineering teams receive accurate, actionable direction.
- Draft specifications
- Summarize design decisions
- Create documentation outlines
- Whether details are accurate
- How interactions should be implemented
- What must remain consistent
Fast output is not the same as a good product decision.
AI output must be challenged whenever it conflicts with research, lacks context, creates risk, or reduces the quality of the user experience.
The output invents evidence.
AI may present assumptions as facts or create unsupported user needs. I return to the original research and remove conclusions that cannot be verified.
The solution is generic.
Generated recommendations often reflect common design patterns without accounting for the specific workflow, environment, or business requirements.
The user is oversimplified.
AI-generated personas and journeys can flatten complex behavior into stereotypes. Real research must define the people and situations being designed for.
The recommendation creates risk.
I reject output that could weaken accessibility, privacy, security, transparency, user control, or confidence in the product.
The interaction ignores context.
A pattern that works in one product may fail in an operational, mobile, regulated, or high-pressure environment.
The answer sounds certain without validation.
AI confidence is not proof. Important product decisions still require evidence, stakeholder alignment, usability testing, and measurable outcomes.
Where AI creates meaningful value in product design.
The strongest use cases reduce low-value effort while keeping product decisions grounded in real users, clear goals, and human review.
Organizing complex qualitative information.
AI can help structure interview notes, compare sessions, identify repeated language, and create an initial summary for review.
Human review requiredExpanding the number of directions considered.
AI can generate alternative flows, content approaches, edge cases, and questions that help challenge the first solution.
Research determines directionCreating and comparing interface language.
AI can draft labels, explanations, error messages, and onboarding content that designers can refine for clarity and context.
Designer protects meaningSupporting documentation and pattern consistency.
AI can help draft component guidance, compare specifications, and identify inconsistencies across documentation.
Teams define standardsPreparing stronger evaluation scenarios.
AI can help draft tasks, identify possible failure points, and organize observations after testing sessions.
Real users validate designsMaking complex decisions easier to explain.
AI can help turn detailed research and design rationale into structured summaries for product, engineering, and leadership.
Designer owns the recommendationThe standards I use when bringing AI into product design.
AI-assisted work should remain transparent, evidence-based, reviewable, and accountable.
Start with a real problem.
Do not use AI simply because the technology is available.
Protect sensitive information.
Do not place confidential user or business data into tools without the appropriate protections and approval.
Verify the source and output.
AI-generated summaries and recommendations must be checked against the original evidence.
Design for transparency and control.
Users should understand when AI is involved and retain appropriate control over important decisions.
Evaluate bias and exclusion.
AI output should be reviewed for assumptions that could create unfair, inaccessible, or incomplete experiences.
Keep accountability human.
The product team remains responsible for every experience, recommendation, and outcome.
AI should create more time for the work that requires human understanding.
Organizing notes, drafting documentation, repeating production work, and starting from an empty page.
Listening to users, exploring meaningful problems, collaborating with teams, testing ideas, and improving product outcomes.
Better tools should support better product decisions.
I use AI to accelerate the design process while keeping research, product judgment, accessibility, ethics, and user needs at the center of every decision.
