Every generation of web tools claimed you didn’t need a designer. AI is just the latest empty promise.

Over the past few years, a lot of people have asked me whether AI is going to replace web designers. Rather than keep answering one conversation at a time, I’ve pulled together the questions I hear most and done my best to answer them honestly.

You’ve compared today’s generative AI craze to earlier eras of web design tools—like Microsoft FrontPage, Macromedia Dreamweaver, and early drag-and-drop site builders. What sparked that connection for you, and what parallels do you see between then and now?

One of the first websites I ever built FOR MONEY ($150 + ice cream cones for my girlfriend and myself) was a Baskin-Robbins site (corbysimpson.com/clients/baskinrobbins) in the late 90s. I had no clue what I was doing, but I got ahold of FrontPage and suddenly I could build a website, which at the time felt like magic.

The site itself was horrible. It had frames, image-based navigation, tiny text, weird colours, a repeating tiles logo background, bright colours, animated graphics, all kinds of things moving around the screen and, to top it all off, a stroke-inducing pulsing lens flare. It was peak 90s web design, and I absolutely loved making it.

That’s why I can’t hate on FrontPage. It lit a fire. I built a bunch of sites, unknowingly churned out garbage, and learned something new from almost every one of them. Eventually I started running into the limits of the software, which pushed me into the code, and then four years of post-secondary education studying design, typography, motion, animation, storytelling and script writing gave me a way to focus all that ambition.

That’s why AI feels familiar to me. We’ve had FrontPage, Dreamweaver, Photoshop templates, HTML templates, Flash, WordPress themes, page builders and everything else in between. It’s all the same shit, just a different pile. Every generation removes another barrier and creates another round of, “Why would I ever hire a designer now?”

When tools like FrontPage or Dreamweaver first emerged, business owners often thought, “I don’t need a designer anymore; I can just use this.” How did that play out historically, and where did those DIY sites usually fall short?

The tools were never really the problem. They made things accessible, and that’s exactly what happened for me.

The problem was assuming that because you could now operate the tool, you understood design. A Photoshop template could make your site look polished, an HTML template could give you working code, and a WordPress theme could hand you an entire site structure, but none of those things knew who your customer was, what mattered to them, what they needed to understand first or what the business actually needed the site to do.

The site might look finished, but that didn’t mean the thinking was finished. That distinction has never really gone away.

What is one specific nostalgic memory or classic pitfall from those early days that captures the gap between “having software that builds a page” and “designing a site that actually works”?

For me, the nostalgic memory is discovering 2Advanced.com (be sure to select ENTER 2ADVANCE STUDIOS to see the full magic).

They didn’t just build a website, they built an environment. The music, the motion, the pacing and the energy made it feel like an experience instead of a collection of pages. I’d already spent three years studying film and had all this ambition around motion, animation and storytelling, and suddenly I could see how those things could actually live on the web.

Flash was a huge part of that. It didn’t really do video yet, so one of the tricks was looping image sequences to make them look like video. The limitations forced you to be inventive, and that was a big part of the appeal.

That site gave me something to latch onto. It showed me that the technology wasn’t the point. The experience was.

AI web generators can spit out a full page layout in under 30 seconds. On the surface, that feels like magic to a business owner. Where does that initial high-tech illusion end and the actual reality set in?

The first result can be incredibly convincing, which is exactly why people get excited. You type in what the business does and seconds later you’ve got a hero image, headlines, benefit blocks, testimonials, calls to action and something that looks like a finished site.

That part is real. It’s impressive. The illusion starts to crack when you ask why any of those choices are there.

Is that actually the right message? Are those really the most important benefits? Is the information in the right order? Does the call-to-action match what the customer is ready to do, or is it just there because websites are supposed to have buttons?

AI is very good at producing something that looks plausible and complete. That’s not the same thing as producing something that’s right.

Where do you see AI actually serving as a useful tool in the web design workflow today, versus where it’s being overhyped as a complete replacement for human expertise?

I use AI all the time and I think it’s incredibly useful. It can speed up research, generate variations, organize content, prototype ideas, write code and remove a lot of the repetitive grunt work from the process.

For someone who already understands design, UX and development, it’s an amazing amplifier.

Where it gets overhyped is when people assume that because it can generate the output, it understands the problem. Those are very different things.

FrontPage didn’t make me a designer, but it gave me a way in. AI can do the same thing for somebody today, and for someone experienced, it can make them dramatically faster. That’s where I think the real value is.

When a non-designer uses AI to build a site, what are the subtle user experience (UX), accessibility, or strategy elements that almost always get missed or mishandled?

Usually it isn’t one spectacular mistake. It’s a whole collection of little things that just feel slightly off.

• The hierarchy is wrong
• The important information comes too late
• The call-to-action asks for too much too soon
• The navigation reflects how the company is organized instead of how the customer thinks
• The mobile version technically works but feels awkward
• The copy sounds polished while somehow saying nothing

Accessibility is another one because a page can look perfectly fine and still be painful to use for someone navigating with a keyboard or assistive technology.

And then there’s the strategy problem, which is the biggest one. A website isn’t supposed to just look like a website. It’s supposed to do a job.

Small business owners are often tempted by the “DIY loophole” to save money upfront. What are the hidden long-term costs or risks of relying on an AI-generated site to represent a professional brand?

The biggest risk is probably opportunity cost because you almost never get to see it.

You know exactly what you saved by not hiring someone, but you don’t know how many people came to the site, didn’t understand the offer, didn’t trust what they were seeing or simply couldn’t figure out why they should choose you over somebody else.

They don’t usually send you an email saying, “Just so you know, I was ready to spend eight grand with you, but your website lost me.”

They just leave.

And we’ve seen the branding side of this before too. In the early 2000s it was the power-handshake stock photo, people pointing and laughing at a whiteboard, or five people huddled around a laptop having the best meeting of their lives. Different logo, different company, same site.

AI risks doing the same thing on a much bigger scale.

True UX design is rooted in understanding human psychology, audience intent, and specific business goals. How does a human designer approach a project in ways an algorithm simply cannot replicate?

A huge part of the job is asking questions and then noticing when the answer doesn’t quite match the problem.

A client may come in convinced they need one thing, but after a discussion you realize the real issue is something completely different. You listen to how they describe their customers, you notice where people get confused, you challenge assumptions and you figure out what actually matters.

Good UX is also about understanding what’s going on in somebody’s head. What do they already know? What are they worried about? What do they need to see before they trust you enough to take the next step?

There usually isn’t one mathematically correct answer.

Sometimes something follows every design rule and still feels wrong, while something that breaks a rule works perfectly because it fits the audience and the brand. That judgment is the part people underestimate.

When the dust settles on this current wave of AI excitement, where do you think AI will fit into the long-term history of web design tools?

I think it’ll become completely normal, just like every other revolutionary tool eventually did.

Nobody gets excited about the fact that you can install a WordPress theme or drag a component onto a page anymore. AI will probably end up the same way, built into everything and used by everybody.

And that’s fine. I think a lot of repetitive production work should disappear. But I don’t think design disappears.

There’s already a term people use for a lot of AI output: “AI slop.” What’s interesting is that the work doesn’t even have to be bad. It can be technically excellent, beautifully laid out and grammatically perfect, but there’s still something missing that you can’t quite put your finger on.

It has no soul.

You can feel when nobody made a decision they really cared about. That’s where taste, judgment, personality and experience matter even more, because once everyone can generate something polished, polished isn’t special anymore.

If a small business owner is currently on the fence—trying to decide between running an AI builder themselves or investing in professional UX design—what primary question should they be asking themselves before making that call?

I’d ask: “How important is this website to my business?”

If you just need somewhere online with your phone number, hours and a description of what you do, then honestly, an AI builder might be completely fine. But if the site is responsible for generating leads, selling something, building trust, explaining a complicated offer or competing in a crowded market, then you’re not really just building a website.

You’re designing part of your business. And a huge part of good design is knowing what not to do. I can’t emphasize that enough. A single take-away here is knowing this.

There are more tools, effects, components, animations, integrations and technologies available now than anyone could reasonably use, and that makes it incredibly easy for somebody building their own site to fall into the same trap I did with FrontPage: adding something simply because they can.

A slider exists, so they add a slider (bad idea!). The builder supports animation, so everything starts animating. There’s a chatbot, a popup, a parallax effect, a video background, six different calls to action and some clever interaction they saw on another site, so naturally all of it starts creeping in.

Experience teaches you that just because something is available doesn’t mean it belongs there. I’ll refer back to the movie that led me into Film School.

“Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should.” — Ian Malcolm, Jurassic Park

Sometimes the most important design decision is the thing you decide not to add, because restraint is part of the craft too. It takes years of building things, watching people use them and seeing what works to develop the judgment to know when a feature is helping the experience and when it’s only there because somebody thought it looked cool.

And that’s where experience still matters.

Send us your questions!