We're About to See the Biggest UX Shift Since the iPhone

I watched Steve Jobs unveil the iPhone in 2007, where the audience laughed when he said we'd use 'a pointing device we're all born with, our fingers.'

Three years later, buttons were dead.

We're at that moment again, except nobody's laughing this time because most people haven't even realised it. The web as we know it, with its navigation bars, search boxes, dropdown menus, the whole clicking-through-pages experience we've been through ten thousand times, is headed for the same fate as the Blackberry keyboard. What will replace it is interfaces that think.

Still Feels Like It's 1995

We're still using a 70-year-old document model to do basically everything online. We click through pages like we're flipping through a card catalog at a library.

Let's say you're looking for running shoes. You still need to manually search three different sites, compare options across dozens of product pages, check reviews on another platform, maybe hunt down a coupon code somewhere else. Then you do it all over again for the next thing you need.

We've gotten so used to this fragmentation that we don't even register how unnecessary it is. Despite all the talk about 'personalization', bounce rates stay high because websites can't really understand what you actually want.

With the iPhone, the entire screen became the interface, as opposed to just replacing buttons for touchscreens. This allowed for new behaviors we couldn't have imagined.

What's coming for the websites is the same kind of shift, except bigger.

When Machines Start Shopping

McKinsey published something interesting recently saying: 'The buyer will be a model acting in someone's best interest.' Read that again. Not a person clicking around. An AI agent that shops on your behalf, replacing the menus, navigation bars and changing the entire experience into a conversation.

The agent can learn what matters to you and adapt in real-time to reshape the entire experience around your specific needs.

Three fundamental changes are happening simultaneously:

Instead of searching through categories and filters, you just say what you want. Or don't say it, the system picks up context from your behavior and even the weather where you live.

Every interaction literally rebuilds the interface. What you see isn't what the next person sees. It's not even what you'll see next time, because the system learned something about you just now.

Your AI is able to negotiate with merchant AIs for price, delivery and even returns, either via voice or text, all handled while you focus on whether you actually want the thing.

What makes this bigger than mobile is that the iPhone changed how we accessed information, touchscreen versus keyboard, while the agentic internet changes who accesses it.

The Three-Year Window

I keep thinking about the companies that saw the threat coming and still died, like Kodak. Entire newspaper chains that understood the threat, but waited too long to fundamentally restructure how they operated, clinging to the old model while swearing they'd adapt eventually.

We're in that window right now. McKinsey says agentic commerce will hit $5 trillion by 2030. Seventy percent of e-commerce leaders already call AI implementation critical. Early adopters are already reporting increased likelihood of exceeding revenue goals by 48% because of it.

But it doesn't come without its challenges, with 52% of companies citing lack of AI talent as a barrier to adoption.That's exactly where mobile was in 2008.

Then the App Store arrived and suddenly you didn't need to be Apple or Google. We saw an emergence of an infrastructure that let anyone participate in the shift without hiring a team of engineers. The same thing is happening now with platforms that handle the AI complexity under the hood.

Companies have maybe three years to make this transition. Move late, and you're competing against businesses that rebuilt from the ground up for a world where traditional websites are invisible to the AI agents actually doing the shopping.

What We Stand To Gain

Some people point to how agentic commerce makes shopping impersonal, losing the human connection that makes shopping more of an experience. But I believe that algorithmic shopping, done right, might actually be more personal than anything we've had before. Beyond just showing users 'customers who bought this also bought that', real personalization means you'd have an interface that remembers you mentioned training for a marathon three months ago and factors that into every recommendation. That adapts its entire communication style to mirror yours, understanding your priorities without you having to spell them out every single time.

As an additional benefit, voice-first, multi-modal interfaces can eliminate the barriers disabled users face trying to navigate traditional websites, allowing them to speak, type, show an image, whatever makes sense in that moment, and the system just handles it.

Generation Alpha likely won't even know what a 'website' was in the sense we understand it now. The whole concept of clicking through pages to find what you want will seem as archaic as using a card catalog.

The people who built the mobile-first web defined what came next. The same opportunity exists right now for the agentic internet, and those who will be the first to spot it and act on it will be the real winners in the agentic commerce.

About the author: Uladzislau Yanchanka is a CEO of Agentplace, a no-code platform for agentic commerce. He's an AI scientist and AGI enthusiast, and has programmed the first AI agent a year before ChatGPT.