AI Has Changed What Makes a Great Designer

For most of my nearly three decades teaching design at Parsons School of Design, there was a relatively straightforward answer to the question of what made a young designer valuable: technical proficiency. Students needed to know how to draw, construct, render, prototype, photograph, present, and eventually master whatever digital tools were transforming the profession. The tools changed constantly. The underlying expectation did not.

Then AI arrived.

Today, a designer can generate polished images, presentations, variations, and even entire visual directions in minutes. The ability to produce is becoming abundant—and when production becomes abundant, something else becomes scarce.

Judgment.

This shift has consequences far beyond design schools. It should change how organizations hire designers, evaluate creative work, cultivate junior talent, and structure the creative process itself. The organizations that adapt will not be those that simply give designers better AI tools. They will be the ones that understand that the value of a designer is increasingly determined before anything gets made.

The competitive advantage has moved upstream

Five years ago, I might have told a design student that mastering Adobe Creative Suite, CLO3D, Rhino, or another emerging tool would help determine their future. Technical proficiency was one of the clearest ways to distinguish yourself. Today, that advice is outdated.

AI can generate countless design variations almost instantly. Beautiful images are no longer scarce, and therefore they are no longer much of a competitive advantage. What matters more is knowing which image is worth making in the first place.

This is not simply a design-industry phenomenon. Across organizations, AI is making it cheaper and easier to produce content, analyze information, brainstorm alternatives, and complete tasks that previously required substantial human labor. But design provides an especially useful lens because designers have always worked at the intersection of production and judgment.

A designer does not simply make a chair. A designer decides whether a chair is needed, who will use it, what problem it should solve, how it should feel, what it communicates, and what assumptions about human behavior are embedded in it. AI can increasingly help with the making, but it cannot take responsibility for the decision. That distinction should fundamentally change how organizations identify, evaluate, and develop creative talent.

Stop hiring designers for what AI can do

Many organizations still evaluate designers through portfolios that emphasize slick, polished outcomes: the finished campaign, the rendered product, the completed identity system, the beautiful interface. Those things still matter. But they tell us less than they once did. A polished image demonstrates that someone can produce a polished image. Increasingly, AI can do that too.

The more revealing questions are different:

Why this problem? Why this audience? Why this solution? What alternatives were rejected? What assumptions informed the decision? What did the designer notice that others didn't?

These are questions of judgment. That does not mean organizations should stop hiring for technical ability. It means technical proficiency should increasingly be treated as a foundation rather than the destination.

When I was a student, a polished portfolio could distinguish one designer from another. Today, the more revealing portfolio may be the one that allows us to see the thinking behind the work: the failed experiments, contradictory directions, discarded concepts, research, observations, and questions that eventually led somewhere. The messy work matters because it reveals how someone thinks. It highlights how they feel. It captures something about who they are. It conveys the way they see the world. That is precisely what organizations need to see more clearly.

AI can accelerate expertise—and complicate how expertise develops

There is an important complication to this argument. AI does not simply make inexperienced workers less valuable. In some circumstances, it can make them substantially more productive. A 2025 study of 5,172 customer-support agents found that access to a generative-AI assistant increased productivity by 15% on average, with particularly large gains among less-experienced and lower-skilled workers. Those workers improved both the speed and quality of their work.

That is good news for organizations. But it raises a tricky question:

If AI allows less-experienced employees to perform tasks before they have accumulated the experience traditionally required to perform them, how do organizations make sure they are still developing judgment?

For decades, junior designers learned by doing: making hundreds of bad drawings, receiving criticism, observing experienced designers, testing materials, defending ideas, watching projects fail, and trying again. Those experiences were inefficient. They were also educational. The organizational temptation will be to eliminate this inefficiency. But some inefficiency is the very mechanism through which expertise develops.

More AI does not necessarily mean more creativity

There is another reason organizations should be careful about simply inserting AI at the beginning of every creative process: more generated options do not automatically produce more original thinking. AI can even weaken a designer's creative muscle.

In a 2024 experiment involving 60 participants, researchers examined the effect of AI-generated images on visual ideation. Participants who used an AI image generator showed greater design fixation and produced fewer ideas, with less variety and lower originality, than participants in the baseline condition. The researchers also found that the effectiveness of AI-assisted ideation depended partly on how participants approached prompting and responded to AI suggestions.

This matters because organizations often equate more options with more creativity. But if AI-generated possibilities cause teams to converge too quickly on what is already visible, the technology may accelerate the creative process while narrowing it.

The solution is not to keep designers away from AI. It is to think more carefully about when AI enters the process.

Redesign the creative process—not just the toolset

The question for leaders, then, is not:

How can we get designers to use AI faster?

It is:

Where in the creative process should AI enter—and where should humans remain deliberately responsible?

Recent workplace research suggests that this distinction matters. A 2025 field experiment involving 250 employees at a technology consulting firm found that generative AI increased creativity, with significantly stronger effects among employees with stronger metacognitive strategies—people who actively monitored and regulated their thinking, evaluated their progress, and revised their strategies. In other words, AI was not a plug-and-play creativity solution. Its value depended partly on the user's ability to think about how they were thinking.

That finding points toward a different approach to AI adoption. Rather than treating prompting as the primary new creative skill, organizations should teach employees how to evaluate, challenge, redirect, and sometimes reject what AI produces.

I would suggest four changes.

1. Evaluate the thinking, not just the output

When reviewing creative work, ask designers to explain the problem they chose to solve, the alternatives they considered, and why they rejected them. A final image should be the beginning of the conversation, not the end.

Organizations that reward only polished outcomes will increasingly reward capabilities AI can replicate. Organizations that examine reasoning will have a better chance of identifying genuine creative judgment and uncovering opportunities for innovation.

2. Give designers problems before giving them prompts

One of the dangers of generative AI is that it can make the solution arrive before the problem has been adequately understood.

Instead, give teams time to observe customers, examine context, conduct research, and frame the problem themselves before introducing AI into the process. This is particularly important because the most consequential design decisions often happen before ideation begins.

A hospital, for example, may ask a design team to improve signage. A conventional process might begin by generating alternative visual systems. A better process might begin by asking why patients are getting lost in the first place, which patients struggle most, how anxiety affects their ability to navigate, and what information they need at each moment.

The AI-generated signage may ultimately be useful. But the more important design work happened before the prompt was written.

3. Protect productive struggle

Organizations are understandably eager to eliminate wasted time.

But not all inefficiency is waste. Sketching, experimenting, arguing, making physical prototypes, pursuing an apparently bad idea, and starting over can look inefficient from a productivity dashboard. Yet these activities can develop intuition, material understanding, confidence, and judgment.

My experience teaching design has made me particularly cautious about treating this kind of struggle as waste. Students often learn most when an idea fails and they have to determine why. AI can help them move past that failure faster—but moving past it faster is not necessarily the same as learning more from it.

Leaders should therefore distinguish between friction that prevents progress and friction that produces expertise. The first should be removed. The second should sometimes be protected.

4. Hire for questions, not just answers

As AI becomes better at generating answers, organizations should pay greater attention to who can formulate consequential questions.

In interviews, instead of asking only candidates to solve a design problem, ask them:

· What would you need to know before solving it?

· Whose perspective is missing?

· What assumptions might we be making?

· What would make you reject your first solution?

· What problem might we actually be solving?

These questions reveal something a polished portfolio often conceals: the quality of a person's thinking. And this may become one of the most important distinctions between designers who merely know how to use AI and designers who know how to direct it. Most importantly, it helps organizations build teams that turn better questions into meaningful design innovation.

The designer's role is becoming more about decisions—and less about production

Recent research with professional designers suggests that this shift is already occurring. In a 2026 study based on interviews with 23 designers, researchers found that designers were assuming responsibility for determining whether and how AI should be incorporated into their professional practice. The important question was not simply whether AI could perform a task, but how designers negotiated its use within professional practice.

That is a subtle but important distinction. The future designer will not necessarily be the person who knows every new AI tool. It will increasingly be the person who knows when to use one, what to ask of it, what to distrust, what to change, and when not to use it at all.

This is why I increasingly think of the future designer as an architect of meaning. The phrase may sound grandiose, but the responsibility is practical. Designers determine what people see, what they notice, what they understand, how they navigate environments, what they trust, and sometimes how they feel about themselves.

An apparel collection can help someone reclaim confidence after chemotherapy. A chair can restore dignity to someone whose body has changed with age or disability. A hospital's design can calm an anxious patient before they ever meet a doctor. In each case, the object matters. But what matters even more is what the object does to the human experience.

AI can generate thousands of possible products. It cannot decide which human problem deserves attention. That remains a human decision—and increasingly, a leadership question.

Organizations need to rethink what they mean by "creative"

The most important implication for senior leaders is not that designers should use AI more carefully. It is that organizations need to redefine what they consider creative performance.

If a company continues to reward speed, volume, polish, and technical fluency above all else, it will increasingly find itself rewarding capabilities that AI makes abundant. If it instead rewards problem framing, discernment, experimentation, interpretation, and meaningful human outcomes, it creates a different kind of competitive advantage.

This does not mean abandoning efficiency. It means recognizing that efficiency and value are not synonymous. The designer who produces 100 concepts in an hour may be more productive. The designer who recognizes that 99 of them are solving the wrong problem may be more valuable. And the research discussed above suggests that this distinction is not merely philosophical. AI can raise productivity, improve creative performance, and help less-experienced workers—but the gains depend on how people engage with the technology, what they already know, and how thoughtfully they evaluate its output.

That is why organizations should stop asking only how much creative work AI can automate. They should ask a harder question:

What human capabilities become more valuable when AI can perform the production?

The answer is not simply creativity. It is judgment about creativity: knowing which problem matters, which idea deserves attention, which output should be rejected, whose perspective is missing, and what a design should ultimately make possible for another human being.

The future of design will not belong to those who can make the most things. It will belong to those who can see what is worth making—and to the organizations wise enough to recognize the difference.

----

About the author:

Steven Faerm is Professor of Fashion at Parsons School of Design, where he has been teaching for nearly thirty years. His fourth book, The Future of Design Education: Theories, Speculations, and Emergent Practices, is set to be published in 2027 (Routledge). www.stevenfaerm.com

Statement on the Use of Generative AI. During the development of this article, the author used ChatGPT 5.5 as an editorial tool to support idea generation and exploratory thinking, and to improve the clarity, organization, and language of the writing. ChatGPT was not used to generate the article's ideas, arguments, analyses, or conclusions. All research, citations, editorial decisions, and final content were reviewed and approved by the author.