UX Design with AI ┃ 1.4 How AI Fits Into the Design Workflow

Module 1. AI Fundamentals for the UX Workflow

In this lesson, we'll look at how AI is finding its way into the design workflow — not by replacing designers, but by changing where they spend their attention. We'll start with a real project from Nike and use it to pull out a few ideas you can apply in your own work.

What Actually Changes When AI Joins the Process

The shift is bigger than people expect — but it's not the shift the headlines describe. AI isn't "replacing designers." What's really happening is a change in the division of labor: AI takes on the repetitive, time-consuming work, and designers move up to the decisions that need human judgment — defining the problem, setting direction, deciding priorities, and reviewing quality.

Concretely, AI can help you:

• Explore more, faster. Generate many design options so you can widen your search before committing to one.
• Personalize. Tailor experiences using data about how people behave and what they prefer.
• Cut the busywork. Reduce repetitive cleanup and organizing so the team's workflow runs leaner.

That's the upside. But there's a catch worth taking seriously.

So How Should You Use AI Here?

Treat AI as an assistant that speeds up UX analysis — not an authority that settles it. Rather than applying an AI suggestion as-is, check four things:

• the reasoning behind it,
• how accurate the tool actually is,
• how it was validated, and
• whether it connects to real user data.

The higher the stakes — again, anything touching conversion and revenue — the more scrutiny each suggestion deserves.

And before you even hand a task to AI, a quick gut-check helps. Ask yourself:

• Is this task repetitive and time-consuming?
• Are the criteria for comparing and choosing between outcomes reasonably clear?
• Can a human review and correct what AI produces?
• Will automating or assisting this free me up to focus on more important judgment calls?

If you're mostly answering "yes," it's probably a good candidate for AI.

Case Study: Nike's A.I.R. Project

Now let's watch these ideas play out in a real workflow.

☑️ Key terms for the AI era

• Hyper-personalization: designing an experience precisely tailored to one individual, using data about their behavior, preferences, situation, and context.
• Co-creation: building with users instead of handing them a finished product. Their input shapes what actually gets made.

Nike's A.I.R. project — short for Athlete Imagined Revolution — is a great example of generative AI working inside a real design process. Unveiled in Paris in April 2024, ahead of the Summer Olympics, it paired Nike's designers with 13 world-class athletes across four sports — track, football, basketball, and tennis — including Zheng Qinwen, Rai Benjamin, Dina Asher-Smith, and Diede de Groot. Together, each athlete and their design team imagined that athlete's dream Air Max shoe.

Here's how the workflow ran:

1. Start with the athlete. Designers interviewed each athlete about their playing style, emotions, training environment, and cultural inspirations — then distilled all of it into detailed text prompts.

2. Generate to explore. Those prompts went into image-generation models, which produced hundreds of concept images per athlete — visual interpretations of each athlete's vision.

3. Humans select and shape. Designers and engineers picked the useful pieces out of the AI output, combined them into a coherent concept for each athlete, and refined it through 3D modeling (CAD).

4. Prototype and get feedback. The concepts became 3D-printed prototypes — produced in hours or days rather than weeks — and went back to the athletes for feedback.

So the flow looks like this:

Athlete input → AI-assisted visualization → designer selection & CAD → 3D-printed prototype → athlete feedback

The most important detail is easy to miss: the team didn't take AI's output at face value. When the images for the "Air" theme kept drifting toward the same fluid, organic look, the designers rewrote their prompts around different sources of inspiration to push for more variety. AI widened the range of ideas; the humans steered which ideas actually mattered.

That's the useful way to picture AI's role here — less a replacement for the designer, and more a sharper, smarter pencil. It doesn't have the ideas for you, but it helps you sketch and explore them far faster. Nike itself frames the whole project as co-creation: the athlete's imagination, the designer's interpretation, and AI's ability to visualize, all combined.

The takeaway: generative AI didn't finish the design. What made it work was the combination — imagination, human judgment, and fast visualization — which together made bolder, quicker experimentation possible. AI expanded the speed and range of exploration. People still made the calls.

Watch: How Nike Built A.I.R.

The video below follows Nike developing these sneakers with AI. The prototypes debuted at an Olympics-themed event in Paris, and it's a good look at how AI can reshape the early planning and development of a product.

NIKE unveils A.I.R — its 13 3D-printed sneakers designed using AI, math, algorithms and more

What's Next

In the next lesson, we'll build your own integrated AI workflow — a hands-on guide to designing a process that works for you.

If this helped, stick around — follow the blog for the next one, and send it to a friend who's just starting out.

Thank you! 🙌