I think they call it "ambient computing"?

I just read iamdanw's retrospective on his pachube internet of things hackday project, DisplayCabinet. It's a beautiful bit of work - a simple projected circle, which when an object is placed within, shows information.

But here's the biggest limitation, as I see it. You have to have a little lump of wood representing the fridge to show the information from the fridge. You have to have a little lump representating _____ to show data about --- well, not quite ______. It's not literal, despite the presence of atoms - each object represents a sphere of information. Keys, for instance, represent the entire home. But what if you want to know what state the locks are in or if the alarm is set? The 'keys' mapping means something else already. You place the Dad maquette in the circle to see his tweets - what if you want to check your @s? What shape a lump of wood to represent that? Or "the girl I met last night"'s tweets? Sure, there's no reason this one interface should show everything, but expanding it's scope is such an enormous effort, I can't help but think you wouldn't.

There's an almighty world of data out there, increasingly, thank the Berners-Lee, queryable in a sensible semantic way. I do get that the point of the excercise is to humanize this data, to work it under the skin of our lives, but it seems such a great cost to lose the power of that vast sea entirely. If I can see my power consumption graphed, maybe I'll want to explore typical power consumption graphs for people of my demographic type. Maybe I'll want to share that data with more ease than taking a photo, or recreating the data elsewhere. In the example given, the data was explored in more detail by bringing in another token - it just seems overly, inherently, bounded. Unless you have more abstract navigational tokens, but this makes the metaphors stretch yet further. I might be wrong on this point. Maybe this sea of data is exactly what we're trying to exclude. But I still think I want it, now. For one thing, I'm used to having it around, and seeping, in it's full power, into everything I interact with.[1]

Where are these lumps going to be kept? The keys live in your pocket, or on the side, sure. But the place I would imagine putting a set of figurines representing your household appliances and family members is in a thimble case next to the table. And if you're going to do that, this whole system only gains over a touchscreen the joy of moving little wooden tokens[2]. It feels cumbersome, it feels less like wallpaper.

But! The explorable status system, that doesn't demand your attention. That feels like it has legs. Technology that we can use while not thinking we're staring into a screen like we do all day at work. That's nice. A picoprojector onto wood looks a treat, but they're expensive and require carpentry. So I think a natural place for this to emerge is the ebook cover [3]. As detailed by Tom Armitage, it's happy at rest. It doesn't call for you, but fits in. Somehow it doesn't make itself known as technology until the page changes (I think that might be the magic of it, that they change between object and technology so easily). People already have them and they leave them lying about. Always on top of things, always face up. If you want to explore what they're showing, they have buttons for now. Or wait a year or two and they'll all have touchscreens.

Or - you can go further, if you hate buttons and touch screens and boring technology. You can make the ebook super location aware. Why use a fridge token, when you can just go to the fridge? It can be a magic window on the metadata behind the actual object. Walk past a thermostat, and see it's set high. Look at your ebook, and see 24 hour and 7 day temperature graphs, and that the thermostat was changed last Tuesday. You take it to the kitchen, it tells you stuff about your food, then becomes a cookbook.

Of course, this only naturally maps to exploring the metadata attached to lumps of atoms in front of you. There's still metaphor involved, inevitably. But hopefully it should emerge naturally as a result of showing you the thing you're most likely looking for, if you're standing where you are. And of course, you can't view far off objects without resorting to symbols again. But at least these symbols can be scattered, a disparate set. Place your ebook next to a photo of your Dad, you see his Twitter account. The only real gain is that you're more likely to have a photo than a maquette, and it can carry in sitting on the side where it's already sitting. And it's easier to register new links.

Huh. I guess I just got halfway to reinventing the Chumby...

[1: for this reason, it would seem totally reasonable if this just displayed information. But as soon as you can use a thing to investigate data, it seems to call out for Google. My watch doesn't want it, but my Wii does.]

[2: a cruise on boardgamegeek will reveal this isn't an unappreciated joy. But it doesn't feel like it should be enough.]

[3: or we could cover yet another surface with ads, as Amazon is exploring. They'll be there, lurking on your bedside table, fucking display ads. It'd be a shame if that's what they end up being used for, but it does show there's money that's noticed their untapped potential.]

Sent from my phone.


15 April 2011