For my colleagues, founders, and brand leaders out there. This is my personal take on AI as a strategic tool. Argue against it.
Let's be honest about where we are. The tools have gotten genuinely impressive. AI can generate hundreds of styling variations in minutes. It produces polished mockups from a single prompt. It simulates users, audits interfaces, and writes product copy that sounds plausible enough to ship.
That impresses people. It also distracts them from the real question.
The real threat is not AI. It's complacency.
The designers who should be worried are not the ones being "replaced by AI." They are the ones who never learned to think beyond the screen. The ones who can execute a beautiful mockup but can't explain why it solves a user's problem. The ones who resist new tools because the old ones still mostly work.
Same goes for founders and product teams. Treat AI as a magic content factory and you'll end up with a lot of content and no point of view.
What AI still can't do
For all the impressive demos, AI fails at the work that actually matters.
It can't sit in a room with a VP of Engineering, a product manager juggling competing priorities, and a CEO reacting to a competitor's release, then walk out with a coherent plan everyone believes in. It can't sense that a stakeholder's real objection isn't about button placement, it's about a power dynamic between two teams. It can't look at a technically flawless screen and say, "we're solving the wrong problem."
AI doesn't understand organizational context. It doesn't navigate ambiguity. It can't make the call that the best design decision is sometimes to remove a feature three departments spent six months building. That takes conviction, diplomacy, and pattern recognition built from having been wrong enough times to know what wrong feels like.
The product is never just the object. It's the object plus the interface plus the politics plus the technical constraints plus the user's real workflow plus the business model. AI can generate the surface. It cannot hold all of that at once.
The bottleneck is innovation
For the first time in modern history, the constraint on what gets built is not resources, distribution, or technical capacity. AI dissolves all three at once.
AI data sets, by their very nature, are backward-looking.
Creativity is, by its very nature, forward-looking.
The bottleneck is innovation.
The limiting factor for what gets funded, made, and remembered is whether the person at the keyboard brings something worth building. Real innovation is tied to lived experience, both intellectual and physical. It lives in the space between knowledge, feeling, function, and context.
The only prerequisite for having a point of view is having a life.
Not a recycled reference library.
That cuts both ways. It's liberating for anyone with a real perspective. It's brutal for anyone hoping AI will compensate for not having one.
AI is a remix, not a creator
Strip away the marketing, and AI is an advanced remix engine. It doesn't work from experience or intention. It pulls from massive datasets of work made by humans, recombines patterns and styles, and presents new variations. It feels original. It's largely the familiar reshuffled.
That brings ethics into the room whether we like it or not. If AI builds on the work of others, then using it well means using it with awareness. Transparency, attribution where it belongs, and acknowledgement that the human contributions underneath are the actual foundation.
This is why AI belongs in the toolbox, not on the throne. Its strength is speed, scale, and the ability to explore directions you might not reach alone. Direction, judgment, and meaning still have to come from people.
The classical process vs the AI process
Traditionally, we work in a line. Research, idea, prototype, test. A clean model, often slow and expensive to iterate inside.
AI doesn't just speed up that line. It bends it into a loop.
With AI, the process becomes circular. You jump back and forth between stages. You test ideas earlier. You generate new directions mid-flight. The distance between thought and reaction collapses.
That collapse is where the strategic value lives.
Four roles for AI in your process
A mental model worth hanging this article on. AI doesn't have one role in design work. It has four, and it switches roles when you switch your prompt.
Researcher. In the research phase, AI expands your understanding. It helps you find answers and, more usefully, formulate sharper questions. Trend signals, user profiles, problem framing.
Idea generator. In ideation, AI is broader and faster than any brainstorm you've run. Many variants, many directions, no filter. Quality doesn't live in the volume. It lives in your ability to choose, combine, and refine.
Partner. As a partner, AI translates your choices into something concrete. Text, concept descriptions, design specifications. You set direction, AI gives it shape. Now there's something to discuss and improve.
Critic. This is the role most people forget. Ask AI to attack your idea, find weaknesses, simulate a confused user, or argue the opposite. A traditional idea phase takes four to six weeks. With AI in this loop, one to three days. Not because the quality is lower, but because you can reject ten weak ideas in a day instead of one a week.
The key word is on command. AI changes roles when you change the prompt. You stay in charge.
The control problem
Now, the part most promotional content quietly skips.
Midjourney, DALL·E, and the latest image models look incredible in the reel. "Premium water bottle, studio lighting" gives you something polished in seconds. Then you sit down with the tool yourself.
You can't freeze the grip and regenerate the label. Each generation is a new lottery. There are no layers, no parametric control, no way to lock one decision and iterate on another.
You end up choosing between twenty variants you don't fully control. That's selecting, not designing.
The question for any AI tool in a real product process is not "how impressive is the output?" It's "how much control do I have?"
Output without control is decoration. Design needs both.
Keep your skills, amplify them
Most AI tools today optimise for fast visual output. They generate impressive images. The full industrial design process still doesn't happen inside one tool.
In reality you're still hands on. You analyse the market. You look for insights. You build moodboards. You sketch. You prototype in Rhino, SolidWorks, Vizcom, KeyShot, Figma. You make decisions, refine, present.
The goal isn't to let AI take over the process. The goal is to keep your design thinking, your hands-on skills, your craft, and use AI exactly where it makes the work faster, sharper, and more creative.
Keep your skills. Amplify them.
Three takeaways
Control over output.
Personal taste over copied references.
Human direction over AI decisions.
The honest question
The question was never "will AI replace designers?" The real question is the one I'd put to any designer, founder, or brand leader reading this.
Are you doing work that can't be replaced?
If you're asking yourself that, you're probably already on the right track. If you're not asking yourself that, that's worth noticing.
What's your take? Argue back.