More Honours Considerations
So I haven't been keeping this blog updated. It's bound to happen with all blogs, eventually, unless there's a strong readership or a series of comments that make their way through here to push me into updated more. Or I could just keep it updated more to make me more interesting to future employers.
Anyway, I had been considering scalable interfaces, focusing specifically on touch-based interaction. Sure, I think touch-based interaction is The Future (2010), but I've also realized that Microsoft, while they get touch (which explains why Windows Phone 7 is so sleek), their fantastic Kinect product moves into the gesture group of their concept of a NUI, or Natural User Interface. However, to try and explain what a natural interface is, is incredibly difficult. Just ask these guys. The main problem, I believe, is context - what is natural to one person is incredibly unnatural to another.
The NUI Group are against the GUI because it requires a process to be explained and learned first, and then implemented or adapted into action later. They want people to essentially walk up to a device and understand how it works immediately. Whether this will be possible or not is another debate. For the record, I don't think it will, because a truly intuitive interface is based on previous familiarity - for example, in the paper I just linked, he explained how he saw a person first used a mouse for the first time ever, and got it wrong on several accounts. However, once that process was learned, it was (relative) smooth sailing. Even an interface like Microsoft's Kinect requires people to cognitively understand that their action has a reaction on a computer, even when they're not touching it. Also, for some people, walking up to their TV and pressing the buttons seems natural, yet it will entirely not work. In the future, the idea of the NUI will essentially become a mish-mash of many types of interaction that can all do roughly the same thing, but some are better than others. It is up to the individuals to decide upon what they believe is the most natural interface for them in that context.
I know that, in that last sentence, I placed italics on quite a few different portions there, but they do matter; each individual must look, understand, process, and then finally carry out their actions for every computational interaction (or any interaction, for that matter). Whether that's considered natural or not will be decided then. I personally think people will still complain about something when that time comes, though. People tend to.
Regardless, I'm moving further from my point. So I'll dive straight back into it:
I started thinking, if I were to do a survey of some sort, what questions would I ask? After some thinking, I was curious to ask people what the working desk would look like in 2010. Generally, we've been notoriously wrong when predicting future interfaces (according to some sci-fi movies, in the future we will all use FaceTime (or another variant of face-to-face chat), in every situation, including the car), and how they would look, and how they would feel. This is because I'm interested in text input, software keyboards etc. in a touch-screen world. If by 2025 everyone uses an Internet Tablet and does all their technical writing on that, how would they do it? Writing recognition? Wouldn't that be less efficient? Wouldn't that be going backwards?
It seems, as we inevitably make our machines easier to use for the general populous, their advanced features start to become more and more difficult to utilize. How do we strike a balance? Just have two devices - one for general stuff, the other for customization and advanced things (such as large amounts of text input)?
I think what we're looking at here is context. Context matters. Context is everything. A mobile user's context is different to a TV viewer which, again, is different to a desktop user, and so forth. This is not just in the form of input mechanisms (remote control, touchscreen, mouse + keyboard etc), but also their state of mind, their location in the world, their comfort, their social situation, their connectivity to the Internet, and so on. This all matters, and needs to be taken into account.
So this lead onto my idea of Scalable Interfaces. I think I should possible describe it as Contextual Interfaces.
Essentially, they start off small. They learn about the user and what they like. Google is already doing this with Web History and search suggestions, and also their Stars--of which I still have no idea how they work (are they just a contextual pseudo-bookmark?), I just hope they do. Very few others are really going deep for a contextual approach (unless you consider Facebook, Twitter, Gmail's Priority Inbox etc contextual as they show more relevant information readily etc). While these systems have the ability to learn a lot about me in a short space of time (the data that can be scraped off a single user's visit on Facebook would be enormous), they want consistency between users. Of course. It makes it easier for the designers to create a website when every user is practically the same.
For example, computers have several input devices that are severely under-utilized for various reasons. Sometimes, even if my phone (or tablet) was set to leave its screen on for 10 minutes, I think I would like it if it could snap open the front-facing camera for a quick second, have a peek out to see if it's being used, and turn itself off. Now, there certainly would be other contextual problems, such as if you're using its screen light as ambient light for other things. But if such a thing happens, and you don't want it to, you should be able to scold it (tell it to stay on etc), and it won't do that as readily as it normally would.
That's a simple complaint I have about my computer, tablet, and phone, and other devices; they don't look at me, or listen to me, at all, even if they have the capacity to.
This is where the creepy factor may go too far, and some people won't appreciate it.
I want my computer to be able to see where I am in the room. If I'm sitting on my chair, that's fine. But if I go to lie down on my bed, the text suddenly becomes too small, and unreadable. This is obvious; the further away we are, the harder it is to distinguish small objects. But what if said objects knew I was a distance away and changed themselves accordingly? Or showed more contextual information depending on your distance? If I want to read text from bed, I shouldn't have to have two users with separate text-DPI settings and change from one to another when I want to watch a movie without getting out from my bed (on cold mornings especially). I can currently use my Logitech MX Air Mouse (which is very nice, mind you) to turn my monitor on and try and find the music I like - but the text is small and unreadable, I've just got to hope that I can squint and read, or recognize the album covers. Why do I deal with this? My software should just know that I'm in my bed and change itself accordingly. The icon changes larger when I use the MX Air, but nothing else, because it has nothing else to base itself on.
Contextual uses of technology can go in other directions. My Microwave should have a tiny camera which sees if a person's there. Hell, even add some facial recognition. If you cooked the food, it will recognize you as The Cooker, and if you are in its field of view when it's done, and facing it, it will make a quieter sound than if it couldn't see you at all. If it could see you, but not recognize your features (i.e. you're looking away), it will make a medium-volume sound to alert you.
So there you go. My idea to tackle: Scalable, Contextual Interfaces.
