It played out like any other ordinary home project. We started with a screwdriver, a drill, and a tape measure. Simple enough. But the longer my father-in-law and I hung window curtains, the more chaotic things became.

In no time, we found ourselves towing multiple screwdrivers, innumerable screws of all kinds, tape measures, levels, several types of pliers, vice grips (for pulling stripped screws), different size drill bits, a stud finder, pencils, packaging debris, and more.

In each room, we found a table—a nightstand, for example—on which to lay our tools. By the time we were in the final room, there were so many objects on the table, I started noticing how long it took to locate each object, each and every time I went to the table.

Then it hit me: that’s why I dislike Microsoft’s ribbon interface. But first. . .

Home projects, lost in space

My project evolved more or less like this:

  1. We started simple with only the tools we knew we needed in advance.
  2. We added tools to meet the growing complexity of the project.
  3. As we added options to our tool table, we gained choice, but it took longer to find each tool.

Logical, but self-defeating.

The workflow of a home project is a teeny tiny manmade manifestation of the most stubborn laws in the cosmos. As objects travel through time and space, they transition from states of low entropy (high order, high potential) to states of high entropy (disorder, uselessness).

Our tools, screws, etc. began the day organized in containers, where they were easily found. By the end of the day, they were scattered in the spaces through which we had traveled.

It’s just how the universe works. We can’t change it.

But should we tow the physical world’s frustrations into virtual worlds? Into software?

When what you see isn’t what you need

Interface bloat in software isn’t so much a physics problem as it is a psychological problem.

The size and resolution of computer displays have increased tremendously in the last thirty years. As high definition approaches real vision, it’s becoming possible to replicate the real-world, and all its clutter, on a screen.

Screen specs are accommodating the “more is more” approach to software design. The human mind isn’t.

Nightstand or screen, there are limits to what the human mind can visually process when the choices are many. Cluttered palettes of buttons may seem like an obvious way to eliminate search costs. In practice, cluttered interfaces beget search costs.

At some point, adding additional choices begins to conceal all choices. It was painfully clear in my home project.

Even when software designers develop what they perceive as a logical layout of buttons (and nested buttons), new users of the software will always be confronted with the same reality: A high entropy workspace at t = 0.

You can blame the second law of thermodynamics for why we generally go from low to high entropy in nature. But why the hell design high entropy into virtual ecosystems—where we actually have some say in the matter?

Rethinking time and (virtual) space

The first, second, third, fourth, and fifth times I realized I needed to walk down two flights of stairs to retrieve additional tools during my curtain-hanging project, I had no choice but to do so.

I was physically located on the third floor, while the objects I needed were elsewhere. Sure, I could have dumped every conceivable object and tool onto the floor where I was working. But who does that? Who would want to start in a workspace with hundreds of things they don’t need?

When I’m working in front of a document, spreadsheet, or other virtual object, there’s no appreciable physical distance between my tools and me. It’s really quite a profound and unprecedented reality of knowledge work.

We are the most sedentary humans to date. Never before have people sat for so long making things that aren’t real—in a touchable/holdable sense—while using tools that can’t be touched or held either.

Our tools are as imaginary and virtual as the work product we create. Position and distance have become irrelevant. Time is now by far the most dominant dimension governing the product of a knowledge worker.

Why are we still spending so much time looking for buttons that don’t physically exist?

The future is flat

Google.com and iOS are primitive compared to what we'll see later this century, but they are important precursors. They're both very flat. Imagine looking at them from the side:

Google vs ios pe

Both Google’s home page and iOS are minimalist and elegant. They’re in a low entropy (high potential) state. As such, it’s the simplicity of these interfaces that make them so powerful.

An empty Google search field has more potential per pixel than any manmade creation to date. There are tens of billions of web pages in Google’s search index.1

Google has nearly eliminated search costs in an information economy of unprecedented size—far bigger than your hard drive or the toolbox in your basement. When we google, we don’t have to think through possible folder hierarchies, tags, or even click through categories to drill down. We don’t have to know where servers are physically located on the planet.

We just think, type, and get results—in a few hundredths of a second. (I hope that never stops being remarkable to me.)

iOS imposes only slightly more search cost. We need to think about which app will contain the information we seek. Once we have an app in mind, we can search by Spotlight, or swipe across screens that have at most two levels (due to the presence of folders).

Once in an app, there are very few steps left to connect to information. Buttons are few; entropy is low (and steady); potential is high. Perhaps the greatest testament to the usability of iOS apps is that so few require instructions—or “Help” in the conventional computer sense.

Google.com and iOS approach a natural union of the human mind and digital information. That's what we want. Isn't it?

Great software is and will be founded on search

As I illustrated with my Microsoft Excel ‘Hide Sheets’ odyssey, good feature search—like that built into the core of OS X—can transform even the most cluttered software into something more immediately usable.

“Search” can be thought of as the activity that occurs between wanting to find something and actually finding it. One of technology's most fundamental purposes is to enable our innate addiction to pursuing information.

And so it logically follows that the most advanced software interfaces this century will be simple and flat—putting as little distance as possible between our tools and our imagination.