Even in the seemingly remote burg of Lawrence, KS, I am all over the place.
I could be at home, at the office, the coffeeshop, en route to Kansas City, at a show, restaurant, art gallery, or bar. I could be at the park with the kids, I could be at my folks’ house in Leavenworth. I could be at the antique store, the scooter shop, or getting my guitar repaired. Maybe I’m getting my haircut, or an oil change for my car.
So, what if I tweeted or sent a status update to any of my social networks stating something like the following:
“I can’t believe I just witnessed such a heinous misstep of personal grooming.”
Great (even brilliant) content, posted all over the ethers. But….so what? It’s a bon mot that doesn’t amount to much, and is an opinion shouted to a crowded room with almost guaranteed irrelevance. I may have thousands of people “hearing” it, but… so what? It’s gone, and it’s voice filled over by the thousands of other great one-liners filling your friends list.
Well, how do I change that? How do I create some relevance, some context, some social understanding of the “where” to my “what”? Would it convey more, to more people, for more reasons? Would a simple identifier of place give my shout-out more gravity?
I could do this: “I’m at the barber shop and I can’t believe I just witnessed such a heinous misstep of personal grooming.”
Ok, I had to type out “I’m at the barber shop and” just to make it contextual. Eh. Should be an easier way, right? I mean, there’s a shorthand paradigm for updates, for referring to other people, for mentioning that it’s a forward of someone else’s thought (re-tweet), there’s tagging for relevance (hashtags). Why isn’t there something like a location tag?
So, we were talking at Little Springs about this very idea. I think it was Barbara who brought up the observation that on my personal status updates, different “spheres” of people (friends, acquaintances, family, coworkers, etc.), could all read my content differently if they made any assumptions about the context, and more importantly, the relevance of that content based on their understanding of our relationship. This was very similar to the point made in a talk by Jared Benson at D4M 2008.
It’s a great point: Personally, I have multiple layers of audience that can construe my content with whatever filter they have that is most prominent in our relation. Even time of day or day of week cause us to make assumptions that aren’t always fair, and even presumptive in a manner that can often cause a lot of miscommunication issues. For example:
“Oh, I thought you were at work today when you mentioned in your status update that you wanted to kill someone. Is everything ok at the office?”
Well, no, I was actually at the store, during what would normally be a work hour.
Again, why update at all, why have a window into your world, if you don’t even have enough information for your audience to understand why you’re even saying it? I mean, really, are my status updates, tweets, text messages all just “A tale told by an idiot, full of sound and fury, signifying nothing”?
Nope. So let’s fix that. Like this:
^MyBarbershop – I can’t believe I just witnessed such a heinous misstep of personal grooming.
Ahhh, that’s better. The cognitive dissonance has been dissipated. I can see clearly now, the pain of trying to understand the context is gone.
Plus, it’s kinda funny.
So, we’ve outlined the details here about this whole deal. Basically, we need SOMETHING to set the stage for the context. Following the precedent of #subject and @identity, respectively, we’ve chosen to follow suit with ^location.
And, it works! @shoobe01, @barbaraballard and I have been using it for the past few days, figuring out the variables and rules, the refinements and usages. People from all my different “audience spheres” have let me know that using this tagging system actually creates more relevance and understanding of what i’m posting about. It seems pretty obvious to me, actually, but just like Marcel DuChamp, someone just needed to do it first.
So go try it. Read the article we’ve created about it in our design patterns wiki. See if you find that it makes a difference to how you post, and how your audience responds to it. And if you have more thoughts, references or suggestions, it’s a wiki. Add it in.
Stowe Boyd suggested this a couple of months ago and suggested a prefix of / – see http://www.stoweboyd.com/message/2009/05/a-modest-proposal-for-more-microstructure-location.html for details.
As I mentioned there, my friend has been tracking his whereabouts through twitter (as well as a lot of other ways) using his mapping service http://mapme.at for a year or so now. Once you’ve signed up then mapme.at watches your tweets for “(my location)” and derives your location from that (which lets you see where you’ve been, and also update things like FireEagle).
Thanks, Adrian! Its interesting to see the differences and similarities of our two versions of this idea, especially since our notion stems from social relevance versus Stowe’s suggestion of almost GPS-like specificity. In essence, they’re doing the same things, but coming at it from a slightly different “spirit of information”.
We’ll check out the mapme service as well, and add it to our wiki page for consideration of those types of locational technologies.
I also did search for a bit after the initial in-the-office brainstorm buzz wore off, and found neither this nor anything else. Which is a little too bad.
Note that there are some open issues for the caret, one-word system we are using about long addresses. Enclosing seems possible, but for length issues of (non-fragmented) sms and twitter messages.
I also have tried the caret on some other services, and it seems to be valid. Like, I tagged a few images on Flickr with it. A slash is often gonna be reserved, or will be converted to another character or for poor systems will actually break the data store. I’ve seen all three of these in the wild, in the past year.
Mapme is, possibly sadly, not that well used. No one I know is signed up for it, for example. Latitude is quite well known, but even then causes to much fear so only about half of those I have invited respond, and only half of THOSE actually share info.
Anyway, while I’d most like to see location info in twitter the opposite way (stick something like ^loc into the tweet, a service intercepts the macro and inserts a location based on your mobile sharing) at least better formatting should be able to help services like mapme. Having it scan every message for “things that look like addresses” seems to not work very well right now.
Interesting timing, Twitter just now announced that they’ll attach locations to your tweets, though it is opt-in. http://mashable.com/2009/08/20/twitter-local-twitter-to-attach-locations-to-your-tweets/
This was inevitable. If anything, this accentuates the benefits of your approach as GPS data is just that, hard data; and only gives a vague context rather than what you’ve got going – social and contextual relevance of your tweet’s content.
In…teresting. Wonder if they thought of the implications of privacy, not just storage and bad people (see EFF link at the bottom of the wiki), but the social-distance thing we’re trying to solve with the caret tag.
Not totally following how it’ll work yet so not much more I can say now.
(Oh, and I sorta hate lat/long. UTM forever!)
Watching blog entries from others, tweets, etc. I’m seeing a lot of worries about privacy that indicate not everyone is following this concept entirely. I’ve just added something to the Privacy and Security section of the wiki which I think covers it.
If that doesn’t make you feel warm and fuzzy, please bring up your worries and we’ll try to address them more.
Good thinking guys…
I’ve been experimenting with automatic context resolution, a complicated problem
(the introduction of context tags or similar take away from the limit space allowed by messaging such as SMS or Twitter)
But I like this idea, and its simplicity to quickly override context “settings” as necessary…
ceo
Indeed. I think that’s why I can get my head around this thing (I’ve been legitimately using caret tags all week, not just because we want to experiment and promote it). Before, I usually had similar context, but it had to be just part of the text. This is a shorthand, so is, if anything, shorter to type.
I do like, say, Latitude. But only in principle. Since almost no one else uses it (due to privacy fears mostly, and poor data plans otherwise) it’s not that valuable. In the ideal case, an automatic service might be able to do all this for you, but that future seems impossibly far off with those privacy concerns, and almost everyone’s unwillingness to play nicely together.
To tell more (if off channel) about what you’ve been thinking of. I for one love navigation, geo, location, etc.
After watching them in use this week, I’m noticing many serve a dual-purpose of location and hashtag.
Hashtags have been used for a variety of things, including locations. Most commonly, they are used as a “mention” of a subject others may be interested in and could discover independent of following you.
So an example, Steven tagged ^IndiaPalace yesterday. Not just a location, it’s easily a worthy subject others may have mentioned – whether or not they were/are there. e.g. “Been hankering for some #IndiaPalace all week.”
So how do we get the two tags aggregated together so the connection isn’t lost – ^#IndiaPalace maybe? Needs to be thought through; not sure if Twitter search results for a string of #IndiaPalace would return ^#IndiaPalace. Consequently, searches for ^#IndiaPalace would certainly not return #IndiaPalace.
Brillant + good story telling. Without context I believe that most tweets are senseless chatter. There may be something quotable in a tweet but little relevance and marginal value. This is especially true when following someone or some organization that one does not have a previous real world relationship with. Twitter is a grand experiment that connects synapses of the collective mind. Without context we have the danger of creating a whole new class of false knowledge.