3D movement tracking a la what Johnny Lee is doing with the Wii hardware is just insanely cool, and I have to think that it will find a useful niche in human-computer interactions somewhere--even if only in games.

Some of these ideas are fun and some are clunky, but I like playing with all of them because I think it shows you where things are headed long-term--once the kinks are worked out.  Witness the multi-touch interface called iGesture I bought like 6 years ago and the current raging success of the iPhone/iPod Touch.

Of course, SOFTWARE user interface innovations are much easier and cheaper to make as they rely solely on good design taste.  It's actually funny to me that Apple's eventual insane profitability and Web 2.0 have transformed the computer industry from a bunch of nerds who like keyboard shortcuts and command lines into ar-teests who discuss minimalism and color palettes as much as any interior designer.  There could easily be a Web Design Star TV Show.

"Ohhhh my, look at that menu, is that actually just TEXT?  We'll need to get some rounded buttons in here to soften things up, STAT. Oh, a grey background... no he DIDN'T.  Somebody spilled 1998 all OVER this place!"

But aesthetics aside, there have actually been some great user interface concepts to come out of the Apple/Web 2.0 era.

Here are some of my favorite user interface innovations I've seen lately.

Natural Language Processing

The Mozilla bookmarking and history concept video demonstrates looking for activity history using queries such as "yesterday, office, around 2 pm, eric sent"  The natural language search is a really nice touch whenever it's implemented, and one that could be useful in many more areas than it currently is used--life streams definitely seem like an ideal area here.  In fact, really good text indexing like Google does will usually suffice in this area.  It reminds me a lot of how (why) I use Gmail (I'm looking for an email sent to Jessica last year, regarding taxes?  Gmail can do that), and why I have come to loathe Outlook in comparison.  When we were designing our business intelligence application we quickly realized that our users loved how easy it was to run a report for yesterday, last month, etc, rather than having to spell out dates.  People like the ability to interact as they would speak to another human, but it assumes a certain level of intelligence on the side of the engine.

Natural language search also opens the door to using voice as an interface, which I'm really high on as well (see Jott). Voice is really a terrific low-effort way for humans and computers to interact, heinous uses of IVR systems have just numbed us to this fact.

Visual Item Recognition

The human mind really is amazing at image recognition, and that's rarely used.  Mozilla's new concept videos use thumbnails to display search results and history.  When I first looked at the concept video, I reflexively thought that visual search results couldn't possibly work better than text.  However, after further thought, I think that if you incorporated real-time zooming and panning, a la the iPod touch interface, it would actually be a great way to cover a much bigger swath of search results than textual results like Google's allow, relatively quickly.  Photosynth technology, which allows you to start from the 50,000 foot view of thousands of documents and zoom into specific high-resolution areas would be very helpful here as well.

Network Graphs

Being able to graph your network is just plain old fun.  I almost think it's hard to grok a network without seeing it.  If you use Twitter but haven't checked out www.tweetwheel.com yet you should go do it--it graphs your Twitter network and it's interesting to see how many people you follow also follow other people you're following.  Networks which grow exponentially, like social networks, are inherently hard to grasp--displaying them in a visual format like this makes it easier.

OpenLink Software has created a nifty Linked Data add-on for FireFox which shows you the data behind the page in a data graph, exposing the relationships between concepts in the document and other data.  It would be really cool to work an RDF relationship browser into an interface as polished as this Mozilla concept.  I'm envisioning that in the middle of browsing a page, I could zoom out to a network graph with that page at the center, and getting a visual thumbnail cloud of all the pages and concepts that the page links to.  That would be just groovy.

Mobile Interfaces

The iPhone/iPod Touch has done a great service to humanity by showing us that intuitive and easy to use interfaces are indeed possible on a small piece of screen real estate.  The touch interface and accelerometers have also shown us that there doesn't HAVE to be a barrier between the phsyical world and the digital, and in fact the more you blur that line the easier a device is to use.

As I mentioned in a previous post, I think that all user authentication will eventually be done via your mobile device.  Log into a Web site on your laptop and you will be prompted to authenticate via thumbprint on your mobile device.  You shouldn't need to have any kind of biometric input device duplicated on your desktop or laptop. To death with USB dongles, I say, let's converge the important stuff in our mobile devices and be done with it.

Because of the limited real estate on a mobile device, touch screens are almost a necessity.  Some people still favor a phsyical keyboard, but I've used a Blackberry and an iPod Touch, and I have BIG hands, and I think those people are just reluctant to change.  Apple's auto-complete based on QWERTY works really well.  Actually, though, probably the fastest way to do text input right now is via fold-out keyboards.  I had a Nokia 6800 with a fold-out keyboard once, and it was really cool.  However, I would place my bets on mobile devices eventually projecting a full-sized keyboard onto the surface they're sitting on and doing heavy-duty text input that way.

But for now, I think Apple has done a commendable job with mobile text input.  Touch screens also allow multi-touch interfaces, which brings me to...

Multi-Touch Interfaces

Multi-touch interfaces are still in their infancy, but already they have completely obliterated every other type of interface, especially in the mobile space.

One of the nice things about multi-touch interfaces is that they allow you to drag things off of the screen to access hidden functionality.  Aza Raskin uses this to great effect in the FireFox Mobile concept video to hide browser functionality offscreen while viewing a page, leaving more screen real estate available for content display.  I love the idea of not having to look at my options until I'm done with the task at hand.

Gestures

Intuitive, self-teachable gestures are definitely where touch interfaces are going.  The iPhone uses these and when you pick it up you don't even realize you're issuing a command to the computer.  You have to think about pressing Ctrl + to zoom in, but pinching your fingers on the screen is intuitive and natural.  The iGesture pad can learn new gestures, so once it's configured you can pretty much manipulate things on the screen Minority Report-style (albeit without the nice display).

It'll be interesting to see how this space evolves and which of these ideas go mainstream.  I'm taking apart Wii controllers while I wait :)

UPDATE:  Coincidentally, I just saw this really cool article along similar lines, and they have a few interesting interfaces that I hadn't seen before.