AI Has Infinite Patience (And Why That’s a Warning Shot for UX Designers)

For twenty years we made the maze easier to walk. We never removed the maze. AI doesn’t need us to.

The Garmin Seconds Dilemma

I wear a Garmin Fenix 7, and I use it constantly. I’ve been tempted by an Apple Watch more than once, but the Garmin is just the better fit for me. The heart rate tracking, the running, the cycling, all of it. So I’ve stuck with it.

The thing is, the watch face is ugly. I don’t really mind the overall look of the watch itself, but the face? No designer would ever look at it and call it beautiful, and I’m aware of that every time I glance at my wrist.

Then there’s the seconds. Somehow, and I genuinely don’t know how, I managed to change my watch face so that it now shows seconds ticking away. It’s not the end of the world. But it bothers me, and I’d switch it off in a moment if I knew how. That’s the catch. I don’t know how, and the thought of digging back into Garmin’s nested, archaic menu system to find the setting feels like more effort than the annoyance is worth. So I leave it. I just live with the seconds.

That’s the story I keep telling people, because of where it’s heading. Eventually I won’t need to know how to fix it. I’ll just say what I want, something like “get the seconds off my watch face,” and the system will know exactly how to do it. The interface I currently have to fight my way through will quietly disappear.

That’s the direction we’re moving in, and it’s a lot bigger than a watch.

When the Backend Is the Whole Problem

Take that same frustration and scale it all the way up to the cloud. Anyone who has set up infrastructure on something like Amazon AWS knows the feeling. S3 buckets, permissions, configuration layers, the whole thing can feel like a labyrinth. Standing up cloud infrastructure shouldn’t feel like performing surgery in the dark, but a lot of the time it does. There are entire DevOps careers built around taming complexity that was never made simple in the first place.

For about twenty years, the industry has poured enormous resources into design for a very human reason. We get impatient, we lose our place, we second-guess ourselves when a system is confusing. Those aren’t flaws to be embarrassed about, they’re just real traits of how people interact with software. So we brought in designers to act as translators, building thoughtful interfaces, wizards, dashboards, and sensible defaults so that people could actually get through the underlying machinery without giving up.

Here’s the thing though. We got very good at building better interfaces, and we never really fixed the complexity underneath. We made the maze easier to walk. We didn’t remove the maze.

AI Has Infinite Patience

This is where everything shifts. An AI agent doesn’t get frustrated by a confusing button. It doesn’t feel any fatigue staring at an AWS dashboard, and it doesn’t get worn down by Garmin’s nested menus.

AI has infinite patience. If a task takes fifteen steps, four API calls, and an awkward workaround, the AI will work through all of it in a fraction of a second without once losing focus. It is the ultimate power user.

Because AI can move through difficult, poorly designed applications without any of the friction a person feels, the messy backend no longer needs a polished front end wrapped around it for human benefit. The AI becomes the wrapper.

We used to spend a fortune designing interfaces so that humans could talk to machines. Increasingly, what we actually need is a way for humans to state what they want, and let the AI deal with the machine.

UI Is Not UX

People mix these two terms up all the time, so I want to be clear about which one I mean.

When I say the interface goes away, I’m talking about the UI, the user interface. The screens, the buttons, the dropdowns, the menus you currently have to learn and navigate. That layer shrinks dramatically, and in a lot of cases it disappears entirely.

The UX, the user experience, moves in the opposite direction. It gets better. The entire reason for removing the interface is that getting what you want becomes faster, smoother, and far less effortful. Telling your watch “lose the seconds” and having it simply happen is a wonderful experience. It might be the best version of the experience there is. There’s just no traditional interface sitting in the middle of it.

So this isn’t the end of UX. What we’ll likely see is the beginning of the end of UX being delivered predominantly through screens. Screens aren’t going away. Your watch face will always need to show the time. But the configuration layers, the navigation, the menus you fight through just to get something done. That’s the part that shrinks. The experience not only survives, it improves.

I’m being specific about this for a reason. If we blur the two together we’ll end up with a room full of people arguing about the difference between UI and UX instead of talking about the actual idea.

The Shrinking UI

That distinction is exactly why this is a redefinition of a craft, not just another tech update.

If someone’s main interaction with a product moves to a chat box, a single record button, or an ambient gesture, then a huge portion of traditional UI design simply isn’t needed anymore. You don’t need a beautifully designed dropdown if no human ever lays eyes on it.

Traditional design has always been preoccupied with how a person gets something done. The flow, the clicks, the navigation. The future is preoccupied with what the person actually wants done. The intent.

We designers don’t need to panic, but we do need to pivot. The future isn’t about designing screens. It’s about intent architecture. Getting to what someone actually wants, handling it when the AI gets it wrong, presenting the outcome clearly. That still calls for real design judgment. It just isn’t a menu tree anymore.

The Ultimate Outcome

One day soon, I won’t be searching for a YouTube tutorial to fix my watch face. I’ll press a button, say “get these seconds off my screen,” and it’ll be done.

As Golden Krishna put it back in 2015, before any of this AI wave had really kicked off, “the best interface is no interface.” The products that win the next decade won’t be the ones with the prettiest dashboards. They’ll be the ones that ask the least of us to get to what we actually wanted.