The Application Model is Outdated
The fundamental way we interact with computers hasn’t really changed since the 1980s. Sure, we’ve added more layers of complexity, polish, and buttons you wanna lick – but are things really that different?
My very personal opinion, but one that I am attracted to, is that the fundamental way we interact with computers hasn’t really changed since the 1980s. Sure, we’ve added more layers of complexity, polish, and buttons you wanna lick – but are things really that different?
Let me describe a familiar workflow for editing text.
- Decide which app to use.
- Find the app.
- Launch the app.
- Write text.
- Save the result.
- Do something with the result.
Now take a guess: does this workflow describe the actions you’d perform on Classic Mac OS (say, 7.5.5), a modern M2-powered MacBook Pro, or a phone? The answer is that it could be either of them.
User-driven UI
The human-computer interaction model we, the humans, have taken on when designing desktop, and later mobile operating systems, is that of an interface agent. It was a great decision: the designers at the time made it easier for folks working in offices to adapt to the new virtual tools and quickly become familiar with them. Files, folders, desktops, trash cans (or garbage bins, if you're inclined). Buttons you push, knobs you twist, windows you open and close. Great.
But then something interesting happened. Hardware and software started to become much, much more powerful. You can now use your phone to shoot, edit, and publish a full-length movie. Workflows got more complicated, spanning many apps at once. The world got more complicated.
Fast-forward to 2023, and the average mobile phone user has dozens of apps on their device. All those apps are waiting to be opened using a click, a voice command, or a scripting instruction – something explicit. They’re the first thing you see on your computer, phone, and tablet.
Managing complexity
The added complexity needs to be managed. How? By adding more functionality, of course. Got more apps? Here are various launchers to choose from. Too many windows? Swipe between desktops or collect them into piles on the side of the screen. Got a task you do often? Spend time upfront to automate it and save time later.
Those are welcome and sometimes even innovative ways of dealing with the added complexity. But one thing stands out: all those solutions still defer to the user. You still have to decide which apps to use. You have to open them, and you have to close them. They’re still something the user needs to worry about, to explicitly decide to do, and then actually do.
The user-driven model keeps working just fine. We, the humans, are currently busy adding more layers to that interaction model to make it more delightful and to enable new experiences. But take a look at the workflow above again. Wouldn’t it be great to not think about which app to use, and to focus on the task? ChatGPT will tell you that it’s not impossible, even in 2023.
A hybrid?
Looking into the future, I’m curious whether we’re going to stick with the current model forever. Maybe we’ll eventually do the same thing we did several decades earlier: take the familiar elements, combine them with what is made possible by the technology at the time, and iterate on making interacting with computers a more delightful experience.
This post was originally published on May 13, 2023.