“We can just use AI for this” — Are you really sure about it?

AI in 2026 Is the New WordPress of 2006

I recently read something that stopped me mid-scroll:

“AI in 2026 is what WordPress was in 2006.”

Think about it.

WordPress templates democratised web presence. Overnight, anyone could launch a website. Small businesses, solopreneurs, side projects, and suddenly everyone had an online home. That was genuinely transformative. But the websites that actually converted visitors into customers, that retained audiences, that ranked on Google and grew sustainably… Well, those had a designer and/or a strategist behind them. The template was never the hard part.

We are at the exact same inflection point right now.

“These AI tools are getting really good!”

“We’ve got AI tools that do this already.”

“Can’t you just run this through AI?”

“AI can make 10 versions faster”

And yes, the AI tools are getting really impressive. It can speed up the process, so we can start further ahead. But the value isn’t creating a prompt or pressing a button. Now we spend our time tweaking the right choices to ship properly.

AI tools let you build an app, generate a landing page, write marketing copy, spin up a prototype — all from just a few prompts. And having such power is so exciting. Everyone should try it. Everyone should be proud of what they create. I use AI every single day in my own work, and I see its potential clearly.

But there is a gap between creating something and making it work properly, accurately, for your business, in an ethical manner.

AI: hype or a real thing?

Here is what I have been noticing — and I bet my fellow designers can sort of relate to it too.

We are living through an information flood. Every day, there is a new ad from someone who just discovered AI, promising to teach you how to build the best product in the world, with no caveats. The sheer volume of fast, confident, surface-level content makes it easy to believe that knowing how to use a tool is the same as knowing what to do with it.

It is not.

It is almost saying that just because someone knows how to use Figma, they automatically become a UX/UI Designer — a multidisciplinary profession that demands knowledge not only of the tool and design. But it doesn’t create any barrier to anyone opening Figma and starting to create a nice interface, regardless of whether it would really work in real life.

Data analysts with PhDs, people who have been working with these kinds of language models for a decade, still make mistakes using AI regularly, even though they make fewer mistakes, and have an idea about how to solve some issues that come up in a safer way. AI is unpredictable by default; it can hallucinate, break, consume unexpected resources, raise IP concerns, and still has endless faults — all of which impact business outcomes. AI can do great deep research, collect data, generate all types of content and automate workflows, but it can also make mistakes. The AI tool is powerful; it might even learn frameworks, but it does not come with a human judgment, ethical or creative mind baked in, so it can be really biased.

I say this as someone who genuinely uses AI on a daily basis. I am not sceptical. I am not against it. But there is some hype about AI, loads of potential, and we need to be aware of both — what is a trap and what is truly a shortcut.

From WordPress to AI: What is missing?

Last year, a client came to me asking for a better design — which, in his mind, meant a nicer-looking interface. He ran a growth-marketing business, and when he set up websites for his own clients using premium WordPress templates (or subcontracted a professional), the results were solid. Paid traffic converted. The sites performed.

For his own site, though, he went the budget route and built it himself. It looked ok on the surface — I mean “ok” considering a broader perspective, not a designer one. But when I checked the analytics, the picture was completely different: zero organic traffic, inflated engagement, and some other inconsistencies. Google had flagged the site in Search Console for bad practices. What he understood as “design” was essentially UI — colours, layout, visual polish. What I saw was a product design problem: his ideal clients were not clearly defined, the user journey had no coherence, the value of his offer was poor, the copy had no strategic foundation, and the design practices and SEO were working against him.

He needed direction more than just a prettier template.

This year, a client came in proud of the app he had been building with AI-assisted development. And honestly? There was a lot to be proud of. He had real features running fine from a technical point of view. He cared deeply about privacy, ethics, and safety on the back end. He could see his competitive advantage against the bigger players in the market. Very promising!

But he had zero insight into his actual users. No user research, no validated personas, not even a basic PLG strategy to sustain shipping on what he was building. He had fallen into the feature paradox — building more and more capabilities without knowing which ones mattered to the people he was trying to serve.

He needed direction to prioritise for an MVP that would actually generate value, make the product sustainable, and move toward the goals he had been chasing broadly, with plenty of ambition, but without a strategy to get there.

Same pattern, both times: the tool was not the problem. The missing layer was direction. And this is a 100% human skill.

So why does this keep happening?

When you can produce a working prototype in an afternoon, it feels like you have solved the problem. When you can generate ten versions of a landing page in minutes, it feels like you have done the design work. The output is tangible, polished, immediate.

And that creates a blind spot: if it looks finished, it must be finished.

It is pretty much like some people who attend a workshop to learn a specific skill and then act as if they hold a full master’s degree. It doesn’t mean the workshop is bad, but it is a glimpse of what a master’s degree provides. Of course, professionals in the field can spot the depth gap, but the general public doesn’t.

And here comes the enormous difference between an output and an outcome.

An app with 150 features and no user research is an output. A product with 10 validated features that drive retention and revenue is an outcome. A beautiful landing page with zero organic traffic is an output. A strategically built website that ranks, converts, and supports a sustainable funnel is an outcome.

Designers with a business vision tend to know this distinction intimately. We have always worked in this sort of space — even before AI. We use UI Kits, pre-made design systems, and established patterns to ship faster and smarter. Now we also use AI tools to optimise our workflow and deliver stronger results. But this is not all the “design job”, just a glimpse of it.

There is nothing wrong with using tools to be more efficient at creating not only outputs but also real outcomes. The problem starts when someone believes that mastering the tool is the same as mastering the discipline.

Is critical thinking the real skill of this era?

I believe the most underestimated shift happening right now, in the middle of this infodemic, is not exactly the AI capability but about discernment. It is so easy to confirm different perspectives with proven studies that only critical thinking can spot the problem.

We are surrounded by more information than at any point in human history, and most of it is algorithmically optimised to confirm what we already believe. That same pattern shapes how people use AI: you get back what you prompt for, filtered through the biases you bring in. The output feels authoritative, even when it is incomplete or wrong.

In that environment, the ability to evaluate, question, and direct becomes the highest-leverage skill anyone can have — whether you are a designer, a PM, a founder, or just someone building their first product.

AI is not a replacement for strategic thinking. It can be an accelerant, but it needs direction, supervision, and (loads of) revision. We sometimes spend more time reviewing AI-generated work than we used to spend producing it from scratch. The work has changed shape, but it has not become easier. It can become faster in the long run, even though you also expect to be fully functional as an asset.

WordPress templates in 2006 gave everyone a website. The businesses that thrived were the ones that understood what a website was actually for.

AI tools in 2026 give everyone the power to create. The products that will thrive are the ones built with the clarity to know what to create, for whom, and why.

The tool was never the hard part. Direction always was. And for directing in a biased world, critical thinking is the baseline.