This week I sat in on a product conversation. A Product Manager was walking business stakeholders through a new platform capability, still in discovery, intent not fully defined. And they were doing it with AI-generated screens.
They weren't ready for development. They weren't even close to what our design system would produce. But they looked enough like the real thing that everyone in the room was suddenly speaking the same language. The conversation that usually takes three meetings collapsed into one. Stakeholders were reacting, aligning, deciding. The PM hadn't replaced design. They'd done something smarter: used AI to establish a common visual reference fast enough that the strategic conversation could actually start.
My first reaction was surprise. Normally a designer would be in that loop, helping produce those conceptual renderings. Then I caught myself. A solid PM has always done their own wireframing for this kind of early intent work. AI just gave that instinct a faster pencil. And keeping a designer out of that pre-discovery phase isn't a loss. Those are low-leverage cycles that were never the highest use of design's time anyway.
The alarming version of this story would be a PM using AI-generated screens for user research, or handing them directly to engineering to build. That's not what I saw. What I saw was AI accelerating the front end of the process so that by the time design enters, the foundation is already solid. That's not a threat. That's a better starting line.
That said, not everyone is reading it that way. Scroll LinkedIn on any given day and you'll find designers convinced the role is being automated out of existence, and PMs convinced they no longer need design at all. Both are wrong, but the fear is real and the stakes are high enough to take seriously. I've been doing this for over twenty years. I've watched this movie before. And every time someone declared design was over, the designers who leaned in came out the other side more valuable, not less.
This is that moment again. And if you're a designer, it's one of the best things that's happened to your career. You just have to choose it.
We've eaten harder transitions than this.
If you've been in this field long enough, you have scar tissue.
You learned Quark, then InDesign ate the world and you learned that. You built comps in Photoshop, emailed them as flat JPEGs, and called it a deliverable. You learned HTML and CSS because there was no other way to show someone what you actually meant. Some of us learned PHP and MySQL because we wanted our designs to actually work, not just sit in a deck. Flash died overnight and took entire portfolios with it. Sketch replaced Photoshop for UI. Then Figma replaced Sketch. Adobe XD existed briefly, as a cautionary tale.
Every single one of those transitions had someone standing at the edge saying the role was over. The designers who made it through weren't the ones who waited to see how it shook out. They were the ones who got their hands dirty first.
But let's be honest about something. Quark to InDesign was a better tool doing the same job. This is different in kind, not just degree. AI isn't a better Figma. It's something that expands to fill the shape of whatever you need it to be. A research assistant. A thought partner. A faster pencil. It amplifies intention. It wants to be useful in whatever direction you point it. The dystopian narrative misunderstands what AI actually is. It's a force multiplier. Extraordinarily powerful, but a multiplier. What gets multiplied depends entirely on who's holding it and what they know how to do with it.
What AI does and what it can't.
A PM with AI gets wireframes. That's real. But here's what they don't get.
They don't get the pattern recognition that tells you why that navigation model is going to fall apart at scale. They don't get the instinct that says this interaction is technically functional but cognitively brutal for the person doing it eight hours a day. They don't get the co-design relationship with the subject matter expert that took three workshops to build. They don't get the judgment call about when delight is appropriate and when it's just noise.
A designer with AI gets all of that, plus production velocity they never had before. That's a different equation entirely.
I wrote recently about whimsy. The short version: delight is the feeling, whimsy is the mechanism that produces it. You can't ship whimsy directly. You design the conditions for it. AI can produce usable. With a good prompt and some iteration, it can produce pretty good. What it cannot do is know that your users have been ignored by enterprise software for a decade, and that the moment they feel genuinely heard by a tool is the moment everything changes. It cannot sit in a co-design workshop and notice that people are leaning forward. It cannot feel the weight of a quote like: an experience we actually want to use, not one that is demanded of us.
That requires a human. Specifically, it requires a designer who knows how to listen before they make anything.
AI does not have users. You do.
Now the uncomfortable part.
If you're a junior designer right now, the fear makes sense. The entry-level market is contracting and that's real and it's hard. But here's what's different about this transition compared to every one that came before it. QuarkXPress cost hundreds of dollars. InDesign too. If you didn't go to a school with a well-funded design lab you were behind before you started. Even Figma is a subscription. The tool transitions that defined this industry were never truly open to everyone.
AI is. Free tiers exist across every major platform. If you need more compute you're talking less than four days of Starbucks a month. A junior designer who can't land their first job right now can still spend the next six months getting genuinely dangerous with these tools, building the fluency the market will reward on the other side of this. The on-ramp has never been cheaper. The ceiling has never been higher.
If you're mid-level or senior, this is the moment to lead. Your experience is exactly what makes AI powerful in your hands. But that only works if you pick it up. Skepticism without experimentation isn't a position. It's a delay.
The designers who come out of this dangerous.
They'll move faster than anyone expected. They'll prototype at the speed of thinking. They'll spend the time AI frees up on the judgment calls only they can make. They'll still run the workshops, build the relationships with the SMEs, and notice when a pattern is technically right but humanly wrong. They'll still design the conditions for whimsy in systems that have never had any.
Here's what AI will never replicate: the belief that every person using your product is worth designing for. Not the median user. Not the persona. Every one. Star Trek called it IDIC, infinite diversity in infinite combinations, and I've written about why I think it's the most honest design philosophy going. The point isn't the sci-fi packaging. The point is the commitment underneath it. AI optimizes for the center. Designers who are paying attention design for the edges, and that's where the real work lives.
The designers who come out of this stronger won't be the ones who had the least fear. They'll be the ones who decided the fear wasn't a reason to stop. The tools changed again. They always do.
Get dangerous.