How I think about AI’s impact on design
Leadership | 7 minute read
OVERVIEWAI excels at production (done right), helps with exploration (right solution), but struggles with problem framing (right problem). Design leaders need to invest in strategic capability, not just automation.
A framework for exploring AI in design
Since the first ChatGPT models arrived, I’ve wrestled with AI in two ways: first, bringing an AI assistant to market at NetDocuments; second, understanding how AI reshapes design itself.
Every design leader I know is asking how to use AI. I think the better question is what parts of design work AI actually changes and what that means for how we lead. To explore this question, I’ll lean on a framing from a former Fidelity Investments colleague, Jen Cardello, who broke the design research process in three parts:
Right Problem: Are we solving the right problem (intent, framing, business alignment)?
Right Solution: Are we building the right solution (exploration, validation)?
Done Right: Are we executing well? (production, quality, delivery)?
My experience has shown me that AI impacts differs dramatically at each of these phases. I’ve learned that AI excels during “done right”, helps with “right solution”, but struggles with “right problem”. Understanding these differences has rewritten my thinking on how to structure teams, where I need to invest time developing designers, and which capabilities to build vs. automate.
Right Problem
Problem framing, Business alignment, Strategic contextAI impact: Lags
Right Solution
Exploration, Option Generation, ValidationAI impact: Helps
Done Right
Production, Execution, Quality, DeliveryAI impact: Excels
When most design organizations are still debating how to incorporate AI as a production tool, I think the real opportunity is understanding where AI best helps and how to avoid creating new risks. The right problem, right solution, done right framework is perfect for design leaders to think this through.
Right problem: Where AI lags
Spoiler alert: humans are more necessary as ever. The ‘right problem’ stage is the most difficult and least familiar to many organizations. These are the question to ask at this stage:
Are we solving the right problem for the business?
How do we balance user needs with business constraints?
Which problems move the needle of revenue, retention, or other strategic imperatives?
What are we choosing not to solve?
This is where design thinking intersects business strategy which requires a careful consideration of business context, stakeholder needs, technical constraints, and market dynamics.
Let’s remember that AI is trained on existing solutions to known problems. This focus on what’s known means AI struggles to intuit what could be. McKinsey worded it plainly in recent research that AI is best at interpreting data, not generating new signals. In other words, AI can help you explore known problems but can’t tell you which problems to explore.
While at NetDocuments, I used ChatGPT to analyze a long list of customer ideas. It was fantastic at converging, categorizing, and ranking ideas by votes, but utterly failed when I asked if these were the most impactful issues to solve (though it’s instructions tried to convince me otherwise). ChatGPT lacked context. It didn’t know that we had a strategic prospect hammering on a nonexistent feature they wanted before signing, or that a past architectural decision made executing an idea extremely difficult, or that a competitor was generating tremendous buzz with a new capability we lacked. I had this context because I was spending time with the business and this is where designers, and design leaders, need to be directing their attention.
As we’ll see, AI supercharges production, so design needs to invest more in this first phase to ensure the right things get produced. This means more time in strategy discussions with product, more time understanding what success looks like for sales, more time asking “what business outcome can will impact by solving this problem” across your organization. Design leader must make time for every designer to do this. When new designers start on my teams, I direct them to spend the first few weeks developing this critical context before jumping in to solutioning.
Right solution: Where AI helps and hurts
Once you’ve defined the right problem to solve, you need to find the right solution. This includes generating and evaluating/testing multiple options, eventually narrowing possibilities down to a few promising solutions. Traditionally, this phase can consume a few days to a few weeks depending on the complexity of the problem to be solved. AI has compressed this cycle down to hours and days by accelerating generation when you need volume and variation.
While at NetDocuments, we had a complex UI for uploading files. Critically, it lacked any sort of queuing function. Using the AI-assisted design platform, Uizard, I was able to generate 30 variations within an hour of back and forth prompting. Better yet, some of the variations were ideas I wouldn’t have likely considered. If you’re already embracing a fail fast, learn fast model this leap in generative design can be incredible powerful. Generate options quickly, evaluate against user needs and business goals, choose the best options to move forward with.
Despite this generative boost, there can be a hidden costs. Over convergence of ideas, where solutions begin to gravitate toward a common norm, can be problematic. With the models training on the same data and designers using the same tools, solutions start looking similar. In one Wharton study, teams using AI generated far fewer unique ideas, just 6% as original as human-only groups. Volume itself can be a trap. When it’s trivial to create dozens upon dozens of options, the burden of discerning promising ideas requires even more effort.
When reflecting back on the 30 solutions Uizard produced, there were only 3 clear themes, with small variations on the edges within each theme. Without business context and design discretion, teams may ship fast but deliver what everyone else is shipping. This underscores the first point - when exploring the wrong problems with AI, you risk converging on the wrong solutions faster. Our role as design leaders is to create systems that harness AI’s speed without sacrificing originality.
Done Right: Where AI shines
Once you’ve pursued the right problem and solution, you need to execute. This is production work: detailed design, accessibility conformance, UI copywriting, design system use, documenting. This is where AI deliverers on the productivity gains that garner the headlines.
McKinsey reports 70% faster cycle times for production tasks similar to this phase. What took designers days can drop to hours when using AI to detailing UIs. In my teams work at NetDocuments, we found that AI-assisted design using tools like Lovable and Bolt got us to roughly 60% good very fast with the designers delivering the critical 40% of detail and refinement in Figma.
[measuring]
With production work compressing, strategic work can expand. This shift has influenced designer goals and expectations with less time spent on design details and more time spent on influencing strategy. Teams still need designers with taste, craft, and technical abilities but I now weigh business thinking and strategic judgment heavier when hiring. Designers can spend more time building relationships and influence than chasing tweaks to detailed designs. Design leaders must redefine what success looks like for designers else risk leaving them without a clear sense of value.
How this changes design leadership
Where designers spend their time
Thinking about AI through this framework has evolved how I lead designers. My teams are spending more time aligning our activities with business outcomes, framing problems with stakeholders, and formalizing relationship and strategic influencing as cornerstones of their professional development along side craft. Designers are asking “is this the right problem?” more and less “is this the optimal shade of blue?” For many designers, this is very uncomfortable as they were trained and rewarded for craft. So, I strive to create psychological safety by providing lots of opportunities to practice and by developing skills that help them discern which problems are best to apply craft to.
Hiring and developing designers
When hiring, I’m increasingly favoring business acumen. Portfolio case studies need to clearly demonstrate a strategic approach, including how they determined the problem to be solved. When developing designers, I’m building their skills in fostering relationships, framing problems, group facilitation, and the language of business. When onboarding, I’m creating space for designers to understand the business context first before using Figma in anger.
A fork in the road
Applying this framework to AI reflects a choice about what kind of design organization you’re building. One path is to use AI to reduce headcount, automate executions, treat design as a production function. The other path is to use AI to amplify human judgment by expanding strategic influences, developing judgement, and position design as a business function. IDEO’s research supports this: companies using AI for innovation, not cost-cutting, see stronger long-term growth. AI won’t replace designers, but it will reveal which designers think like leaders.
Where designer's spend their time has changedTraditional -> AI-Assisted
20% Framing problems 40%
30% Exploring options 35%
50% Detailing a solution 20%
Why this matters now
The decisions organizations make now will shape their design teams for the foreseeable future. Team structure, process, capability investment, and what to automate will be hard to reverse as AI continues to accelerate. I’ve witnessed teams make quick efficiency gains only to struggle months later when they realize they’ve addressed the wrong problems. Thinking through how you want to apply AI through the right problem, right solution, done right framework can help organizations weigh how to optimize.
If you’re hiring design leadership, ask where they see AI helping and where they see risks across these three phases. Ask how they’re reimagining teams and developing talent. Ask whether they advocate investing in strategic capability or optimizing for execution efficiency. Their answer will give you insight as to if they see AI as a strategic shift or an exercise in selecting the newest tools. I think the difference matters.
TakeawayI’ve spent over 15 years building design teams that drive positive business outcome and much of my success can be attributed to learning early on that design activity must be connected to those outcomes. AI represents a tremendous opportunity for entire design teams to operate in a more business savvy manner, not just design leadership. Organizations that recognize this now, using AI as a way to expand design’s business impact, will see durable success.
If you’re building that kind of organization and looking for design leadership that thinks this way, I’d welcome a conversation.