The 90% Problem: Why other's AI's designs may become your problem
The unfortunate reality of how many companies use AI
Photo by Jake Ingle on Unsplash
“We can get to about 90% done with AI, but the last 10% of design still needs traditional methods. Otherwise, the systems just don’t work as intended.”
That’s what a CEO told me recently, as part of my interviews with 22 design leaders. Does that mean design has no future due to AI? Of course not. But this does reflect a changing reality: we may be taking over and reviewing sketches from non-designers.
To understand why, we need first to consider what we lost.
Designers Lost Our Exclusive Superpower: Visualization
Many designers have long considered visualization our defining skill or superpower.
Most designers are visual thinkers who can analyze feature spreadsheets and transform them into interfaces, mockups, and prototypes. That skill alone used to get us great jobs and a seat at the strategy table.
But this is precisely what AI is replacing. A product manager (or anyone) can now take preliminary ideas, feed them to AI, and generate visuals that capture their thinking.
“A designer on my team had a Miro session with a PM — wireframes, sketches, the usual. Then the PM went to Stitch by Google and created designs that looked pretty good. To an untrained eye, it looked finished. It obviously worried the team.”-Global UX Director
Our superpower hasn’t been taken away: it’s more like anyone can buy something similar at the store.
Visual thinking, on its own, has become commoditized. This is why many designers are struggling. Fortunately, that’s not all we do.
What Doesn’t Change: The Need for Iteration
AI may generate visual representations quickly for anyone, but one thing hasn’t changed: it’s never 100% right the first time.
There are always nuances that are missed, UI mistakes, or aspects of the vision that are not fully captured.
As one CEO put it, “We can get to about 90% done with AI, but the last 10% still needs traditional methods. Otherwise, the systems just don’t work as intended.”
That “traditional method?” Iterating on designs to create polished designs for the public. Even the most AI-positive executives know that releasing something AI-generated directly to the public is a bad idea.
But going from “draft” to final product? They’re unsure what needs to change or how to implement the necessary changes.
This iterative refinement remains the core offering of designers, helping them move from a rough concept to a refined solution. This means that our new superpower is something more fundamental: understanding why.
Design’s “New” Superpower: Explaining Why
The most critical question designers must now master is: Why?
This becomes essential because people outside of design increasingly do the initial sketching step. Those things might look okay to the untrained eye, but they can only exist as drafts unless someone can answer why:
Why is this good?
Why will users choose this version?
Why doesn’t that solution work?
Why are we doing this from a business perspective?
This is where designers can shine, because we’ve been asking these questions for decades.
These are the questions we ask ourselves as we iterate on designs, break down the user’s context, and consider business constraints to come up with effective design solutions.
We just need to make the process more visible.
What doesn’t change: Design decisions need justification. Understanding the “why” behind choices has always been valuable. Now it’s essential.
The “Why” Audit: Practice This Today
Here’s something you can do in your daily work: Go through your designs and force yourself to articulate the “why” around anything a team member commented on with three questions:
Why is the previous approach not good?
Why is the current design better?
Why does this help the business?
For example, imagine you redesigned a homepage and received a question about the sidebar menu layout last week. Take those comments and turn them into three questions:
Why was the previous approach not good? The previous sidebar menu had confusing categories, such as “Inventory” and “Catalog”, and users were unsure where to find specific items.
Why is this design better? We conducted card sorting and created categories that make sense to our users.
Why does this help the business? When people get lost in the menu, they abandon their task after 1–2 tries. Better navigation should reduce abandonment rates across multiple critical tasks.
Document these responses in a comment or somewhere accessible, because this offers you several advantages.
It doesn’t just make your strategic thinking visible. It’s also a way to be prepared with your presentations to answer questions stakeholders might ask you.
It’s also a way to gain a deeper understanding of the big picture. Doing this allows designers to have a crucial advantage: you see how all the pieces connect.
As a CEO told me, “I want the person who’s designing the thing from the start to understand the full business context.”
What doesn’t change: Someone needs to connect tactical execution to strategic vision. Designers are uniquely positioned to bridge the gap between user needs, business goals, and technical constraints.
Your New Advantage
It stings to lose a superpower, but we’re not powerless.
The designers who thrive won’t be those who cling to visual thinking alone. To the untrained eye, AI can make visuals that seems ‘pretty good’, even if they’re flawed.
But that doesn’t mean design doesn’t have a future. In fact, designers will be more critical than ever. Why?
Because what doesn’t change is what matters: deep user understanding, critical thinking, strategic perspective, and the ability to justify every decision.
AI can sketch. But it can’t yet understand why that sketch solves the real problem, or why a different approach might work better given your specific constraints.
That’s your work now. Not making the first draft, but making it right.
Focus on what’s unchanging, and you’ll find your footing no matter how much the surface shifts around you.
Kai Wong is a Senior Product Designer and Data and Design newsletter author. He teaches a course, Design Storytelling with Data: How to Pitch the Value of Why Your Work Matters, which helps designers communicate their value and get buy-in for ideas.
Visualizations are just one tool we use to support ideation, so rather than being defensive when others use AI to generate them, we should ask "why" they’re bringing these ideas to the table, to find deeper strategic alignment together.
As designers, our value isn't in protecting artifacts, but in asking questions that transform starting points into outcomes that benefit all.
Great points. Your "why" audit is strongest when it is backed by receipts and a quick way to check it. AI drafts can also look "done" too early, so clear ownership of "done" and a simple definition of done really matters.