9/02/2009

Ditch Mercator, go with the Interdome Projection

As with many new technologies, it seems that the fanfare about Augmented Reality has gotten caught up in the magical, special effect aspects of the user experience, and is ignoring the way the technology actually works.

Lucky for you, you have a dedicated semiotician and materialist at your disposal. Me! Allow me to fly into your view-finder and set a few things straight.

For a long time now, we have been finding ways to organize data inside of our physical space. It actually does this by completing the inverse; geography seeks, by way of data, to orient our physical bodies in a ideal dimension, which via the data, supposedly corresponds to the real world.

I often go on at length about the Cartesian conception of the world, and geography is the premier example. We have developed a pair of axes known as longitude and latitude, and various projections which allow us to portray the topology of the earth, and by microcosm, all physical space we currently inhabit, in a two dimensional Cartesian projection. (And by extension, outer space, because right ascension and declination are a sort of radial integration into 3-D projected space out from our datapoint of datapoints, good old geocentric earth.)

This worked great for a long time, because the data was only a reference, allowing us to navigate over the topology through our relation to this ideal Cartesian plane (which I'll refer to as Geography, from now on). If I can find my own current position in Geography, I can then move myself to a different point in Geography, which, if I've saved my data correctly, will be where I buried my treasure. Point A and Point B in ideal Geography only matter to me if I have been able to accurately label the actual, physical ground as such in my mind, and continue to do so, repeatable as many times as I wish to go from Point A to Point B.

We don't need a map to walk to the store, because we can visualize the ideal Geographic journey from A to B within our head, using cues from the environment rather than a 2-D projection. The point of Geography is not to help us walk down the street, but to extend a realm of potential travel, which we could navigate to, using nothing more than correctly collected ideal Geographic data. If you are visiting my house, I could draw you a map to the store, and if I do a good job, you will arrive where you intended, even though you had never been there before. I did not transfer my memories of "the way to the store" to you. Instead, we communicated via the ideal data of Geography.

But, as we all know, Geography is imperfect, because it is ideal. I would hardly give you a compass, sextant, and two sets of Cartesian coordinates, and expect you to find the store. Instead, I would give you information that relates more to the specific, actual environment, rather than the ideal plane. I would say, "walk south on 50th, then turn right on Division. The store is right past the bus stop." I use a combination of Geography (south), locally specific Geographic conventions (street names), personal locution (right/left), and actual physical indications (near a unique, recognizeable object). These are all elements of our physical and metaphysical ideal perception of space. But the problem is, they are not unified in a Geographic plane. They are only unified within us.

Geotagging, and hence, most of the data AR systems are using, are based on the single plane of Geography. Using various geometries and measurements of radio signals, our devices can reasonably place us on the Geographic Cartesian plane, and maybe even draw us a fairly accurate diagram of other local Geographic conventions. But although this may satisfy the device's programming, what does this mean to us?

Sometimes, a great deal. Watching my phone trace my current position on Google Maps, I can find my way to almost anywhere I want to go. Of course, I must still look up and compare the map with my current location to see if I've actually arrived. And the maps do break down. I've gotten caught at numerous dead-end streets that were drawn as if they went through. And in the Peloponnese, I watched my blue dot snake for hours across a blank expanse of land with no apparent roads. Luckily we were looking out of the windows too, or we would have hit a herd of goats.

Now, one could attempt to improve the data keeping, and the technology by which the device calculates its position. This will achieve better and better results for the Geography. But the real question is, will this change the way we interact with our physical space?

I don't think AR will change our sense of physical space, at least as long as it continues to function via a Cartesian Geography. We have already have the ability to accurately identify our ideal position. This is not a new skill. A digital map in the palm of one's hand is pretty amazing, but semiotically, it is the same technology seafarers have been using for centuries.

But, digital technology is changing our worldview. Our sense of time is radically evolving, as we enter an age when all information that can be said to exist is simultaneously available from anywhere, at any time. Our place in time, and our place in history, is defined by the information we can express about ourselves, in relation to any other period of time. As the ubiquity of the Internet advances, the timeline of history reduces to a singular point. Our sense of history becomes what we can access from the palm of our hands, our vast (but limited) expanse of digital memory, in addition to the memories we already have in our minds.

Access has changed things, and will continue to change them. But we could always instanteously find our cardinal direction. What else is digital Access changing about our physical space?

Here's something that is actually completely magical. A week ago, when I went camping on the Salmon River in the Kalamath National Forest in Northern California, my digital device completely quit functioning for three days. It was not a concrete building, or bad data, or a downed server. It was that there was no data access at all. From here I can call up a digital map of the area from the Internet in about three seconds. But from there, I could not access a thing.

The infrastructural grid may be ever-expanding, but availability of physical access to the network will always be an issue. Actual gaps in service might be fixed, but the proximity to a transmitter will always affect access, and speed will never cease being a problem. There are access nodes strewn across the physical space, and we still must be close to them to access them. These access nodes create their own data as well, and change the way we access the network. Much of the Geographic data we use to identify our location is actually pulled from the access nodes themselves. Our place in the ideal Geography is defined by the actual physical facts of our access.

In a sense, this "closeness", this nearness to access, is more important than the ideal Geography. We are not looking to draw a straight line over an ideal plane, we are looking for things in our immediate physicality. I want the closest coffee, the nearest 24-hour grocery store, or the easiest southbound freeway access. Our programs might be able to find us these things, but they do it by comparing all the data to the ideal plane of Geographic data, and then by running a complicated algorithm to find which is the closest to us.

What if physical things were not tagged with Geographic longitude and latitude, but their locution to particular access points? What if each cellular tower and wifi hotspot was its own axis mundi? Each point of interest was identified by its radius and directional from the access point, and spontaneously provided from this access point to the device as soon as it was accessed?

We already do this; we call them cities. We break down the land into countries, states, counties, cities, municipalities, metro areas. Depending on the use of the structure, we define it differently, and access it accordingly.

Imagine coffee municipalities--all the known coffee shops are broken down into their neighborhoods. So in the Inner Eastside coffee district, you know that the best spot is Bob's Beans. But if you crossover into the Westside River district, then the best choice is Jill's Java.

You see, we already do this all the time in our mind. We think of where we are, and imagine the best restaurant/bar/coffee in the area. We think in terms of access. If for some reason we want a particular place, our spatial dimensions shift. Rather than looking for the best ____ that is easy to get to, we think of the easiest way to get to that one particular ____. The real axis is always access, because either in our minds or in the physical world, it is what controls our experience. This is why we have favorite cities, favorite ways to get in and out of town, favorite neighborhoods, favorites subway lines, favorite pedestrian streets. Our favorites are not categorized as data in terms of longitude and latitude, or in a 5-star tier, or even in an algorithm of these combined. We orient ourselves in the world by what we can access at any particular time. This praxis is our true augmented reality.

In the same way as the semantic web attempts to mediate between the dimensional databases and our sense of lived expression in language, augmented reality attempts to replicate our conception of accessible space. However, the databases do not translate to our lived experience. They are databases, and are created in dimensions fit for cataloging and searching data sets. However, if we begin to move our navigations through the data of the network in terms of access, we might be able to find a way our consciousnesses can merge with the network, that works better than staring into a little visual portal, or even looking at down at a map.

The praxis is the key. The state of immediate consciousness in which our metaphysical conceptions of space and time intersect with our expressive awareness of specific phenomena in space and time. It is not a singular point, or a pair of eyes, but a looped and tied web.

The technology should mimic our consciousness, in way that we can use. If we developed "digital phenomena", access point tags that were present as we walked down the street, we could interact with these phenomena, and remember certain things. It doesn't need to be an RFID tag, or even a barcode. It could be a set of characteristics. A geometry of a store front compared to the height of the building and the depth of the sidewalk. What would you remember best? A logo? A face? Your device remembers it. Then it asks how you feel.

What is this place to you? Yes or no? Hot or cold? Hungry or thirsty? Do you have money? Do you need money? Who are you thinking of now? Who would you most like to talk to?

Then, our devices share on the network.

Are these meaningless bits of data? Only if they have no way to relate to others data. They don't need a univocal Geometry, or a rubric of rating. They need a way to interrelate. What did other people feel about this place? Do you feel that they felt the same way as you? How does this make you feel now?

And then, where do you go next? How does it all fit in to your day? Will you go to the same places again tomorrow?

Part of the beauty of the network is the availability of access not just to the network itself, but to the constituting substance of the network--millions upon millions of anonymous data points, that only become meaningful where shared, added to, and commented upon. Access is nothing unless you are accessing something. We need to begin sharing our praxis as well as our thoughts and times. We need to unite all our conscious thoughts and perceptions. We need to access each other, and not just a Geography.

Will it mean something immediately? Probably not. I'll open my device to all access points in range, and I'll get a firehose of data. Red, yes, hungry, bad service, scared, blue, too expensive, lonely, sex, newness, big, west, no, green. But maybe we can scrape something from this flow, something not just projected upon a map, but something that is immediate and expressive.

Once my device tells me, as an accumulation of all the access available:

"There's a good place, not too pricey, near the mailbox on 2nd avenue right past the store with the cheap beer with a real pretty cashier. I think maybe on the left-hand side of the street went walking downtown."

I probably wouldn't know it off hand. Maybe I would be able to find it. But finally, the device would be augmenting my reality, by networking our praxis.

No comments: