The Theatre Lights Are Coming On
In his recent piece on Product Coaching and AI, Marty Cagan makes a pointed observation. AI is replacing a significant portion of what passes for product management today: aggregating requests, generating roadmaps, and writing PRDs. An engine, or your engineer, could just as easily do the mechanics.
Design has its own version of this theatre. The pixel-pushing, the endless variations, the "making things look nice" that was always more execution than thinking. AI is already doing the work faster and it will only get better.
"AI isn't the immediate threat you should be worried about... it's the person who knows how to leverage it effectively that should concern you." — Matthew Stephens, The Future of Design Leadership in the Era of AI
The uncomfortable truth is a significant portion of what passes for "design work" was always execution, not strategy. AI doesn't threaten design. It threatens the theater of design.
What Design Leadership Actually Is
Julie Zhuo, former VP of Product Design at Facebook, has written extensively about the difference between design levels. In her framework, junior designers solve isolated problems. Design leaders frame what senior designers need to solve by working at increasing levels of abstraction, eventually identifying the biggest problems themselves and designing solutions.
"The more senior the designer, the more abstract the problem they should be solving." — Julie Zhuo, The Looking Glass
This matters more now than ever. At the execution level, the facets, the layers, the accents, the components, variations, AI is a capable collaborator. But at the strategic level, understanding which problems matter, why users behave as they do, what the business actually needs, the work remains deeply human.
Zhuo goes far, offering a definition of design that transcends tools entirely:
"Design is the act of intentionally trying to influence an outcome... Design is a sword against chaos. Design is the pixie dust for innovation." — Julie Zhuo
Design leadership, then, isn't about proficiency in Figma. It's about having the judgment to identify the right outcomes and the skill to influence them.
The Ownership Shift
Marty Cagan argues that product people need to develop shared sense, the judgment to make good outcomes in ambiguous situations. Design leadership requires something parallel: the ability to own the problem space, not just the solution space.
Jared Spool has been making this case for years. His distinction between UX outcomes and business outcomes cuts to the heart of it:
"User experience outcomes are leading indicators of business outcomes, with profit as the lagging indicator." — Jared Spool
Design leaders who understand this relationship, who can articulate how improving the user experience drives retention, reduction, or whatever the business cares about, become strategic partners. Design leaders who can only talk about craft become order-takers. The key word is ownership. There's a world of difference between:
- Owning deliverables: "I shipped 47 screens this quarter"
- Owning outcomes: "We increased conversion by 23% by simplifying the onboarding flow"
The first needs activity. The second measures impact. AI can help produce more deliverables faster. It can't help you take accountability for outcomes that requires judgment, conviction, and a willingness to be wrong.
Design Leadership is Change Management
Peter Merholz and Jesse James Garrett, in their podcast Finding Our Way, have spent years interviewing senior design executives. A recurring theme emerges that reframes what the job actually is:
"Design leadership is change management." — Peter Merholz
Katrina Alcorn, GM of Design at IBM, put it this way:
"I think we're change agents because all of us doing this are still part of a movement to change how businesses work, how they run." — Katrina Alcorn
This reframing matters for the AI conversation. If you think design leadership is about maintaining a team that produces design artifacts, AI is an existential threat. If you think design leaders are actually changing how an organisation thinks and operates, AI is a tool that accelerates your mission.
Karen Hawson, who led design at Intuit and other major companies, articulates what this looks like in practice:
"You have to get into the operations of the company... What are the expectations for designers? What are the expectations for product managers? How are we bringing people on this stuff? Those are all the operating mechanisms you have to infiltrate." — Karen Hawson
This is the work AI can't do. Not yet. Perhaps not ever. The politics, navigation, the relationship building, the slow work of embedding design thinking into how a company operates, this requires human judgment, empathy, and persistence.
AI as Design Coach
Cagan's framing about "product-led coach" applies directly to design. An aspiring designer can now get 24/7 feedback on their work. They can learn principles on demand, get critique at 2 AM, and accelerate their development dramatically.
"Any aspiring product creator... now has 7x24 access to the advice and assistance of an experienced product coach, representing the aggregated learnings of some of the best minds in product." — Marty Cagan
For design, this is transformative. The traditional apprenticeship model, learning from a senior designer over years of close collaboration, was always scarce. Most designers never had access to the caliber of mentorship that the best designer's received. AI changes the equation fundamentally.
But there's a catch. The hardest parts of design leadership, the relationship building, the organisational politics, the culture-shaping, still require humans. AI can coach craft. It can teach process. It can critique compositions and suggest improvements. It cannot (yet) navigate the CEO who thinks "make it pop" is a brief, or build trust with an engineering lead who sees design as decoration.
The Craft Question
There's a version of this argument that deserves acknowledgement. Design has always valued craft. The careful attention to detail, the pixel-perfect obsession, the obsessive refinement, these are part of design's identity. Right?
AI challenges this in ways that feel uncomfortable. If a model can generate a consistent interface in seconds, what's the value of spending hours refining one manually?
The answer isn't to abandon craft. It's to understand when craft is the point.
"Great designers are strong at 'product thinking.'" — Julie Zhuo
Craft in service of outcomes remains valuable. The designer who can discern why one solution works better than another, who can make judgment calls about tradeoffs, who can see what the model misses, that designer becomes more valuable, not less. Craft as performance, beautiful work that doesn't move metrics, becomes harder to justify. This was always true. AI just makes it undeniable.
The Opportunity
For design leaders willing to step into true ownership, strategic thinkers who can articulate why design decisions matter to business outcomes, AI is a force multiplier. Consider what becomes possible:
- Move faster: Generate and evaluate more concepts in less time
- Test more: Use AI to prototype variations and run experiments
- Think more: Spend less time on production, more on decision-making that matters
Greg Petroff, former CDO at GE, describes what effective design leadership looks like:
"I'm all about co-defining outcomes. I think that's a missing gap in software development. A lot of product teams start without actually having a lot of clarity about what they're trying to accomplish." — Greg Petroff
Design leaders who excel at co-defining outcomes, who can sit with product, engineering, and business stakeholders to align on what success looks like, become indispensable. AI doesn't threaten this skill. It makes it more important.
A New Operating Model
What might design leadership look like in this new era? Some principles emerge:
- Lead with outcomes, not outputs. Measure success by impact, not activity. The question isn't "how much did we ship" but "what changed because we shipped it?"
- Develop strategic fluency. Understand the business metrics that matter. Learn to translate between design language and business language.
- Build for influence, not control. The new model focuses on embedding design thinking across the organisation, using AI to scale influence without scaling headcount.
- Embrace AI as a thinking partner. Use models to stress-test ideas, generate alternatives, and accelerate learning.
- Double down on judgment. A model can generate options. Humans must choose between them. The ability to make good calls becomes the core competency.
The Path Forward
Cagan has a piece with a note of optimism. Despite early worries that AI would block entry to the profession, he now believes AI coaching can actually democratize access to product skills:
"I didn't envision that the models would be able to get good enough, fast enough, that they could help to dramatically accelerate the learning curve for aspiring product creators and product leaders." — Marty Cagan
The same applies to design. A motivated designer anywhere in the world, San Francisco or São Paulo, London to Lagos, now has access to coaching that was previously reserved for those lucky enough to land at top companies with strong design cultures.
The bar for design leaders will rise. Those who've been coasting on craft alone will struggle. Those who understand that design leadership was always about influencing outcomes, not producing artifacts, will thrive.
This piece is a response to Marty Cagan's "Product Coaching and AI" and draws on writing from Julie Zhuo, Jared Spool, Peter Merholz, and conversations from the Finding Our Way podcast.