User Friendly?

(The following was written after a few hours learning my way around an digital design CAD program.)

Man, fucking computers, am I right? Mankind is currently in this frustrating position where computers underpin more or less the entire modern world, and we'd all be royally screwed if they stopped doing their thing for us. Unfortunately, this forces us to interact with them, pretty much constantly. And computers are kind of dicks, when you think about it. They take everything you say totally literally, even if that might be obviously contrary to what you wanted. If you accidentally do something to screw yourself over, they have to be painstakingly built so that they won't just go ahead and do that. I never really liked AI-turns-evil-and-takes-over-the-world scenarios, for reasons I may go into one day. I don't think robots would be evil. I'm not willing to dismiss the possibility that they'd be douches.

All of this is bad enough. Just to rub it in, nearly every irritating trait of computers is actually key to their very concept. We built a machine to do everything we told it to—and, to our mounting horror, we're discovering that they do. Of course, maybe it wouldn't be too bad if a computer did exactly what you said, but they can't actually do that yet because they can't talk back and forth with you. No, the computer instead does exactly what you type, which is even harder to align with what you mean than standard language, unless you literally learn a “language” to program in.

Interacting with a computer requires what's known as a user interface. Wikipedia defines an interface (in chemistry) as “a surface forming a common boundary among two different phases, such as an insoluble solid and a liquid, two immiscible liquids or a liquid and an insoluble gas.” Two inmiscible liquids sometimes does seem like my relationship with a particularly dense computer program.

We keep inventing new ways to interact with computers. A basic logic circuit is all wires and transistors—you could build a basic sort of machine out of them by hand if you wanted. (...alright, I wouldn't call this a computer, admittedly. But still! Point stands.) In ye olde days of ENIAC and the like you'd interact with the machine pretty physically—programs were essentially hard-copies, using cards. When you get to a real computer, one that can actually perform more than one function, you switch to programs on cards. Then, you get yourself a 1Mb hard drive the size of a house, and before you know it you're at the keyboard and mouse.

A keyboard and mouse, it must be said, work pretty well as an interface duo. The keyboard allows you to communicate in natural language (for when humans want to read it) but also allows discrete, specific input for when you need to control a computer. Meanwhile, the mouse allows us to take all that spatial reasoning we developed for hunting down elephant-lions or whatever they had in prehistoric Africa and transplant it onto a modern, useful stage. (When you think about it, graphical user interfaces are really quite weird—they don't entirely correspond to anything else the way text input is roughly like writing.) It's probably why they've been so fantastically successful.

That said, in 2010 we have more advanced ways of interacting with our computers. Most obviously, we have the increasingly-ubiquitous touch screen. Touch screens work fairly well, it has to be said. It has to be said often by Steve Jobs, who likes to remind people how successful he is. In some ways, they're extensions of the mouse concept that take more advantage of human anatomy and the way we interact with real, physical things. The other two inputs that seem to be on the horizon are voice control—which I'm excited for purely as a Star Trek fan—and, increasingly, full-body interaction a la the Kinect, Playstation Move, etc. Honestly...I don't think we'll see any of these supplanting the current keyboard-and-mouse combo.

Let's be honest. Kinect-style input seems mostly for games. I can think of very few situations where I've ever thought “Man, I want to tell this TV what to do, but I want to do it by dancing.” As for voice interaction, this has a bit broader range, but the problem with sound is that it has a bit broader range. You can't say “computer, clear browser history?” and stay secret. “YOU MEAN OF THE PORN, DAVE?” may as well, er, erupt from the computer's speakers, you know? Voice therefore works best for controlling large devices. I can easily imagine a room where my voice controlled the TV, lights, fan, etc. But my microwave? My PC? Ehhh. Not feeling it.

Touch screens, touch screens. They're big right now. Like I said, they replicate some of the great qualities of the mouse. Their main advantage is in portable devices, of course, because a touch screen integrates display and interface into a single package—allowing you to have more screen and less buttons. On the other hand, they also suffer serious disadvantages. For example, the fact that you have to (as they say) touch the screen. Obviously the physical contact poses problems. Another issue is that the touch screen only approximates actual object manipulation. Without actual haptic feedback, fine control is often difficult. I can't help but sigh at my iPad because I get the feeling that, while quite cool, being purely touchscreen really limits its application for the kinds of things I need computers for—like blogging and other absolutely vital things.

Ultimately, new ways of interface are nice, but I don't think they'll end up supplanting the keyboard and mouse for the vast majority of computing uses and time. Niche applications abound, places where keyboards and mice would look/feel/be stupid to have, but the personal computer itself? I think we found our winning combo.