I’m designing for a browsers and operating systems that have a well-established visual language and pretty solid interaction design patterns. The increasingly popular Flat Design aesthetic is making everything look the same and that’s okay.

I don’t need to reinvent the wheel when designing a door handle; two or three types of handles may be enough to cover all the possible use cases. Innovation just for innovation’s sake, like trying to create a completely disruptive navigation system for your website or app, might bring you some usability problems in the long term. The question becomes: what exactly is the user need you’re trying to solve by introducing a new interaction pattern? Robust and comprehensive interaction design pattern libraries are gradually letting us focus our time on what really matters for the user: getting things done in an easy and familiar way.

Apps are not necessarily your user’s final destination anymore; they’re just an engine that translates raw data into actionable information. Some users might still occasionally open that beautifully-designed weather app to check the forecast, but the most useful thing the app can do is to send users a notification 15 minutes before it rains — reminding them to bring their umbrella as they leave. Yes, a notification. Mobile OS features such as the ability to take action right from the notification center on iOS or Google Now on Tap on Android are going to make people need to see an app’s UI less and less over time.

Artificial intelligence is becoming more and more popular in 2016. Smart algorithms such as Facebook M are soon going to respond to your texts within the Facebook Messenger app UI, and third-party services and companies will be able to get things done for you without the need of a proprietary user interface. Needless to say, someone still needs to “design” the logic and script of those conversations, and to build the artificial intelligence behind the product.

The interface of the future might not always be made of pixels