User interfaces have changed enormously over the last few decades, but the really interesting change isn’t just what they look like. It’s how much work they ask us to do. When we think about the future of user interfaces, it’s tempting to focus on what they might look like. But the bigger change may be in how much effort they require from us in the first place.
Early computers expected people to learn commands. Graphical interfaces replaced that with windows, menus and icons, while touchscreens removed another layer by letting us interact directly with what was on screen. Each step made the technology work a little more like us, rather than making us work like the technology.
AI could push that much further.
For most of the software we design and build today, the user still has to understand something about how the product is organised. They need to know where to click, which section to open, what to search for, which filters to use and what information the system expects from them.
The next big change in UI may be that users increasingly stop learning how software is organised, because the software starts learning what they are trying to do instead.
The next big change in UI may be that users stop learning the software and the software starts learning the user.
We may spend less time navigating software
Imagine a piece of business software containing hundreds of reports, functions and data points. Traditionally, a lot of interface design goes into helping people find their way around it, so we think carefully about information architecture, navigation, menus, search, filters, dashboards and shortcuts.
All of those things matter because the user has to get from where they are to the thing they need.
But if someone can simply ask, “Show me which projects are running over budget this month”, some of that journey disappears.
The answer might still appear as a chart or table on a screen and, in many cases, a visual answer will remain the best way to show it. What changes is the amount of effort needed to reach it.
That feels more significant than another round of redesigning buttons. The interface becomes less like a structure the user has to learn and more like something that interprets intent, then presents the best way to understand or act on the result.
That could change what visual interfaces look like too
Most software currently presents roughly the same interface regardless of what someone is trying to achieve. You open an application and its navigation, tools and controls are waiting for you.
There are good reasons for that. Consistency helps people learn how a product behaves and gives them confidence that things will be where they expect to find them.
But it also means we often design for every possible task at once.
If software becomes better at understanding context and intent, that may start to change. An application could bring forward the tools relevant to what you are doing now, simplify itself for someone who only uses it occasionally, or expose greater control to an expert. It might even create temporary views around a request rather than forcing every possible function into a permanent menu structure.
That creates an interesting tension for designers. Interfaces need consistency because people need to trust and understand them, but interfaces that never adapt force users to learn the structure of the software rather than allowing the software to understand them.
Finding the balance between those two things could become one of the more important UI challenges of the next few years.
Voice will change interfaces without replacing screens
Voice is part of the same shift because it can be a much more direct way of expressing intent. If you know what you want, saying it can be quicker than finding the right menu, opening the right screen and filling in the right fields.
That doesn’t mean voice is about to replace visual interfaces, though.
You probably don’t want your bank statement read aloud on a train, and you certainly wouldn’t want to choose a colour palette by listening to descriptions of colours. Comparing twenty search results one after another through audio sounds fairly grim too.
That’s why we’re increasingly unconvinced by predictions that screens are about to disappear. Humans are visual creatures and screens are extraordinarily good at helping us scan, compare, edit, browse and understand information quickly.
The more plausible future is voice working alongside visual UI. You ask for something, the system understands the request, and the screen gives you the best way to inspect, compare, edit or confirm the result.
For people designing software, that changes the job. We are no longer thinking only about a sequence of screens, but about how different forms of interaction work together inside the same product.
The future of user interfaces may mean leaving the screen
Smart glasses make that even more obvious.
Once digital information can appear within someone’s field of view, the interface is no longer contained neatly inside a rectangle. Directions might appear over the street in front of you, instructions could sit alongside the equipment you are repairing, and translation might appear next to the words you are trying to understand.
Information becomes attached to the thing it relates to rather than living inside an app.
That opens up huge possibilities, but it also creates some awkward questions.
If a device can see and hear what its wearer sees and hears, what does that mean for everybody else nearby? If voice becomes a common way of controlling technology, what happens to privacy in public places? And if software becomes increasingly proactive, how do people know what it is doing on their behalf?
Those aren’t separate ethical issues to deal with after the product has been designed. They are interface design problems too.
People need to know when a system is listening, recording, interpreting or acting. Good UX has always been partly about feedback, transparency and control, and those things become even more important when the interface itself becomes less obvious.
Buttons, menus and forms are not going anywhere
None of this means traditional UI is about to vanish.
Buttons remain useful when people need to make a clear choice. Forms make sense when precise, structured information needs to be entered. Menus are good for exploring what is available, while dashboards can still be the best way to understand a situation at a glance.
The mistake would be to assume every interaction needs all of them.
A form is less useful if the system already knows most of the answers. A menu is less useful when the user knows exactly what they want. A dashboard doesn’t necessarily need to be the place where every task begins.
The future of UI probably isn’t one revolutionary new interface that replaces everything we use today. It is more likely to be software getting much better at choosing the right interface for the job.
Sometimes that will be a screen, sometimes a conversation, and sometimes it may be a gesture, a camera or information appearing in our field of view. Often, it will be several of those things working together.
For decades, much of interface design has been about teaching people how to use software.
The next generation may be about designing software that learns how people want to use it.