I'm not sure why press, media, futurists insist on pushing the idea that the future of user interaction is via replicating 3-dimensional physical interaction. There's a reason the keyboard, mouse and (small) touch screen are so successful as input devices. They each create an huge degree of meaning with a high level of precision, for relatively tiny physical movements.
While a 3-d interface that mimics reality might be more intuitive to a first-time user, it is vastly less efficient for an experienced user trying to get work done. A 3-d desktop is an accurate mental model but it doesn't need to exist on a screen if it exists in my head. I can move through the z axis of my desktop faster with alt-tab than I could ever move, shuffle, re-order a stack of documents with my hands in a 3-d environment.
* EDIT: tv and movies (eg. minority report) probably have a clear incentive to favor input devices where the audience can more easily see what's happening. 3-d works better there.
I agree and I think it is impossible to create an interface that is optimal for both inexperienced and power users. For users that are using a piece of software frequently, say few hours a day, intuitiveness of an interface should not matter, just efficiency.
I'm often surprised when I see professional programmers turning on some fancy 3D destop effects. Sure, these look nice, but they are only distracting and counter productive when you are using an environment several hours a day.
For power users, the word 'intuitive' doesn't even make sense. There's nothing 'intuitive' about hjkl, but when I tried switching from wmii to i3, the very first thing I did was switch their navigation keybindings back from jkl; to hjkl. Yes, the reasons for hjkl are historic and arbitrary, but I don't care - I don't want to relearn yet another set of muscle memories when I don't have to.
Even vim on the whole is completely 'unintuitive'. That's the whole reason vimtutor exists. But, the couple of hours I've spent (combined) over the last few years learning how to use it has paid off in full and with dividends in terms of my productivity - and I hopefully have many more decades of life to recoup that investment many more times over.
> I'm often surprised when I see professional programmers turning on some fancy 3D destop effects. Sure, these look nice, but they are only distracting and counter productive when you are using an environment several hours a day.
I couldn't find these more annoying. Web interfaces tend to be the worst (because they're highly uncustomizeable). I don't want a slick, 3 second animation where the tab wiggles and slides every time I want to change the page (I'm looking at you, AmEx). I know where I want to go, and I just want to get there immediately. Every second that I'm delayed by flashy animations in something that I need to use several times a day just makes me despise the product a little bit more each time.
Outside of very specific applications/domains (gaming, simulations, etc.), I don't want anything to replace my keyboard. As GP said, nothing (including the mouse) can beat the keyboard for allowing maximum control and precision with minimal movement.
Maybe a better word than 'intuitive' is 'obvious', or even 'blatant'. For a user inexperienced with the application and unfamiliar with interface conventions, the most effective interface is one whose most basic features scream, "Here I am!" A skilled, experienced user needs the opposite — everything should be available at the press of a key, and he or she knows which key it is; anything that calls attention to itself and is not the work at hand is a distraction, and needs to go hide.
> I'm not sure why press, media, futurists insist on pushing the idea that the future of user interaction is via replicating 3-dimensional physical interaction. There's a reason the keyboard, mouse and (small) touch screen are so successful as input devices. They each create an huge degree of meaning with a high level of precision, for relatively tiny physical movements.
While the advantages you mention of a keyboard/mouse/touchscreen are correct, I would not say that is the reason they are so successful. The reason they are so successful is because up until recently there was no competition to either of these. They are the default winners in the world of today, because they were the only ones racing yesterday.
When Doug Engelbart presented the Mother of All Demos back in 1968, he introduced both the mouse and the chorded keyboard, yet the chorded keyboard never caught on. And in 1968, the mouse was a new competitor to the light pen, which had been introduced in 1952. And just in case you don't think the light pen got a fair shake, remember that the IBM PC came with light pen support on both the CGA and EGA cards, and with support in the BIOS and in BASIC. There was no support for the mouse.
It's a natural progression in tech predictions. I remember the same thing in the late 90s with VRML (Virtual Reality Modeling Language, with a plugin available for Netscape). I never quite got how walking through a "virtual 3D store" was a good experience, but PC Magazine and others at the time seemed to think it was somehow the future of the web.
> People are used to gently flicking computer mice and grazing keyboards and tablet screens; do they really have the stamina to reach into their computers and flail their arms around?
That's not the key question. Traditional interfaces have some sort of physical contact. Moving from physical keys to touch screen removed most of the tactile feedback. But still many phones simulate it via haptic vibrations.
Imagine trying to grasp an object in the air purely out of visual feedback. I would imagine it to be extremely strenuous especially in 3D space. Unless there is a breakthrough sci-fi skin which can simulate this tactile feedback, we would be, quite literally, hand-waving in the air.
Mass Effect "lore" tries to justify graphical rule of cool (displays projected into thin air with a few shades of orange as their only colors etc.). I wouldn't take it seriously as something exploring the practicality of such things.
Maybe it would be possible to fake the tactile feedback by using some visual aid? There have been experiments where people "feel" things that are happening to a fake hand ([1] and [2]).
If you have an augmented reality setup maybe some visual effects applied to your hand would be translated into feelings.
I personally feel that the problem is not that of having a tactile feedback. While I don't think it can replace mouse/keyboard, I don't see too much of a problem of it replacing touchscreens.
The problem with most of the existing gesture based interfaces is that they are interacting with the interface in an indirect way, as in the gestures are used control something like a pointer on the screen, and users have no way to see how they are positioned relative to the interface they are interacting with. If they are used in ways similar to a touch screens, I think it would be pretty natural, without too much hand-waving in the air.
The best 3D interface isn't going to look like a computer interface at all.
It's going to be a small patch on your skull that lets you look at a link and think : click / go. Your eyes and focus will make your hands almost entirely pointless for using a computer.
Within 10 to 20 years it'll be possible to wear said small patch, look at a url bar, think google.com, then think: go or enter, and it'll all just work. I think it'll be available in 10, but larger consumer adoption within 20.
That same patch will enable true interfacing with your physical environment. Walk into your house and think kitchen lights on (plus a safe confirmation word or phrase to make sure you mean it), and they'll come on based on the associated phrase.
By the time companies like Microsoft get around to trying to perfect a ridiculous hand > 3D interface, this concept will already have destroyed it.
Minority Report interfaces also will never become mass consumer adopted because of this.
I appreciate your optimism, but I think perhaps you're underestimating the difficulty associated with decoding brain waves.
Right now, the bleeding edge is just coming to grips with decoding signals in the somatic nervous system (the one that's responsible for voluntary control of skeletal muscles).
I think we're still a long way off from being able to read the contents of thoughts from an EEG.
I don't think the point is that Minority Report type interfaces can't or won't be invented, but rather that they are more cool looking than they are practical for doing real work and thus won't generally be used outside of very niche areas.
I don't think you've looked into the research regarding MEG and EEG, especially regarding noise. It's possible you are right, but without a room temperature superconductor and/or implantation, it's going to be hard.
If we get a body temperature superconductor, a whole lot of other things are going to change and a SQUID patch on your brain is going to be one of the more boring advances from that technology.
First I want to get off my chest; are we still using the phrase "computer genius" in 2013? Something about it just rings 1980's to me. I dunno. It seems quaint and silly.
3D interfaces may eventually be the future, but I don't think this version is it. First of all, it's gonna eat up your desktop space like crazy, for what advantage really? Wow factor? That wears out quickly.
Which brings me to my next point; there has been technology to do a variety of different types of 3D desktops for a while now, but it hasn't been pursued because so far going 3D doesn't provide much in the way of practical advantages for most applications. If you're browsing the web and reading email, 3D is doing nothing for you except "Oh man isn't this cool?"
The one place where it makes sense is where you are dealing with actual 3D objects, like 3D printing and CAD/CAM software.
'First I want to get off my chest; are we still using the phrase "computer genius" in 2013? Something about it just rings 1980's to me. I dunno. It seems quaint and silly'
> First of all, it's gonna eat up your desktop space like crazy, for what advantage really?
That was my first impression as well. They want to make it enter the third dimension. Their solution for that is - hog more of that third dimension. No thanks, my desktop real estate is quite valuable.
I could see 3D going somewhere in the future, but only if it augments the desktop that is already there, not by trying to replace things that are still superior in their tangible form.
I think the disconnect is that it seems like they're trying to improve upon screens when they don't realize that they're competing. With things like this[0], this[1] and of course this[2]
The interface (using cameras to track hand / eye movement) just reminds me of the interface described in Snow Crash. It's possible that this full solution won't be the future, but maybe it will parts of it will push the 3-D desktop forward.
It shouldn't be taking up too much space. I can imagine its transparent screen is clipped in front of the monitor when not used. When used, you pull it toward you to create the volume for the 3-D activities.
This idea goes back further than you might think. The "Stereoscopic Workspace", from the Architecture Machine group at MIT in 1983, was a prototype of this idea, surprisingly close to this implementation 30 years later.
The effort put into backwards compatibility at Microsoft is always interesting... here is a presentation on research into supporting legacy 2D custom controls on arbitrary 3D surfaces:
So the writer's idea of progress runs: command line → WIMP → touch → 3D. I think at least the first three are in order of effectiveness — descending. I suspect that the fourth, at least by my understanding of this idea, will continue the trend.
Certainly there are areas where command line beats a GUI, and where keyboard/mouse beat touchscreen, but that doesn't mean they are less effective. Would computers be doing as much as they do for many, many people if they were still 100% command line?
You know, I bet they would. I think what's made so many people able to use a computer is not the new modes of interaction but something more powerful: time. As people have more exposure to computers, and as they become more aware of the interface conventions that develop over time, they get better at using them no matter what the interface style is. I wonder if people might be more useful to people if they had preserved a higher ratio of critical to non-essential visual data — which I think terminal applications do.
While a 3-d interface that mimics reality might be more intuitive to a first-time user, it is vastly less efficient for an experienced user trying to get work done. A 3-d desktop is an accurate mental model but it doesn't need to exist on a screen if it exists in my head. I can move through the z axis of my desktop faster with alt-tab than I could ever move, shuffle, re-order a stack of documents with my hands in a 3-d environment.
* EDIT: tv and movies (eg. minority report) probably have a clear incentive to favor input devices where the audience can more easily see what's happening. 3-d works better there.