Augmented Reality: Open, Closed, Walled, or What?
Sunday, June 28, 2009 at 11:15PM Joel Ludwig recently blogged about augmented reality being open, a brief history of the web, and a number of observations and problems. He mentioned on twitter he would be interested in hearing what I thought, as well as a few other bright folks in the industry. Ok, Joe, here are some quick, late at night, after a long day, half brain-dead reply. I’ll probably want to edit this later : )
First, let me preface this all by saying that in general I prefer open systems that are extensible and expandable, which also facilitate the development, creation, design, and deployment of content, applications, and so forth by other people. If you intelligently empower the end-user, you accelerate market penetration and user adoption.
Ok, now the fun stuff:
When I talk about augmented reality, I am usually referring to it with a much wider scope and definition than simply 3D objects on a video feed. Rather, I mean something that is also in the realm of ubiquitious/pervasive computing, mobile internet devices, wearable displays, and all of the other fun stuff. I realize that some of this extends beyond the usual definitions of AR and into other spaces, but for the purposes of this discussion, I’m grouping it all together.
Some comments:
- Augmented Reality is not a destination.
- You go to a website, or you go to a virtual world, or you download content from somewhere else. AR is not somewhere you go to…it is everything around you, enhanced, augmented, intelligent, interactive, and dynamic.
- Augmented Realityis not global, it is local.
- AR content in Times Square is irrelevant to AR content at the Eiffel Tower.
- Augmented Reality is not 2D or 3D;
- AR content has other dimensions and axis…like time, context, and location. Simply taking a photograph or a 3D model and associating it with a GPS coordinate is not enough.
- Augmented Reality is not an extension of the web;
- AR is something completely different. Thinking about it in the same way we think about the internet or web pages as far as methods, business models, and interface is a fundamentally wrong approach.
- The consumer is not anonymous.
- On the internet, you can be anyone. AR, if implemented properly, is going to be accessed via a mobile device (in most cases), and each mobile device is going to have unique identifiers, and will be personal to the user (like your smart phone)
- The consumer is not a credit card number.
- Due to some of the benefits of mobile as the 7th mass media, consumers can no longer be considered as just a credit card number and a shipping address. AR, if done right, will leverage the power of WHO you are, as well as the other things like WHERE you are, WHAT you are doing, WHO is nearby, etc.
- The browser may be the wrong metaphor or model for AR.
Would you say the internet and the web are open, or restricted to walled/closed platforms? We only have a handful of browsers with any real market share (various versions notwithstanding), one gorilla search engine, three dominant operating systems, etc. The web is not as open as we think it is. We are at the mercy of ICANN for domains (how many millions of domains are wasted and useless because of cash parking?). Mobile phones are restricted to operator networks. E-commerce is ultimately controlled by credit card companies, gouging us on fees and interest. Apple is locked up tighter than (insert something funny here). Windows is bloated and expensive. Spam is so intrusive and overwhelming and has been for so long that we don’t even notice how bad it is anymore.
I think what I’m trying to say here is that everything changes with AR and we can’t assume the old methods and models that work for the internet, the web, or half a dozen other industries will work as well or even be passable for AR. I also don’t think that there will be one singular platform, one mobile device, one browser, etc.
What will likely happen is that we will experience a flurry of competing platforms, browsers, devices, etc. etc. and tons of formats. Eventually some type of protocol that governs how the data is all sorted out will win, and there will be tons of tools, apps, SDKs, and APIs to create content and other apps. All of the access devices (smart phones, sensors, hardware, etc.) will eventually become a commodity, like the PC is today. AR will simultaneously be open and closed at the same time…much like the internet is, or the PC industry is. And that is about where all of the similarities stop. One danger to watch out for is the virtual world model…where you need to download some custom application every time you want to experience new content.
AR is something new and it will be the centerpiece of a convergence of a multitude of other technologies. We need to keep things open while keeping them closed at the same time. Too much of one or the other will spell disaster.
The standards of the internet and the web today, including all of the communications protocols may not be the best solution for the ultimate mobile ubiquitous augmented reality. Square pegs do not always fit in round holes. Sure, some things like HTTP or KML will be useful early on as we experiment, iterate, and grow, but ultimately the inherent nature of the data, experience, and interaction we are talking about for AR will surpass these standards designed for a two dimensional old media link and page driven model.
We have only begun to imagine what is possible and how to get there. To be sure, we, as an industry, are going to make some mistakes along the way, and it will take a lot of baby steps until we can achieve the “big vision”, but we are all hungry for it and anxious to innovate and aspire for something great. Let’s keep the conversation going, and make sure that some 800lb gorilla doesn’t drop some backwards ass bloated user-unfriendly mega-expensive, buggy as hell, augmented reality solution on our heads before we notice it.
I need some sleep. I’m starting to see polygons floating in front of my eyes.
Reader Comments (8)
What I'm worried about is a combination of this: "One danger to watch out for is the virtual world model…where you need to download some custom application every time you want to experience new content."
And this: "Let’s keep the conversation going, and make sure that some 800lb gorilla doesn’t drop some backwards ass bloated user-unfriendly mega-expensive, buggy as hell, augmented reality solution on our heads before we notice it."
Today's brand new startup can be an 800lb gorilla five years from now, so it's never too early for people to at least think about how stuff will work together.
AR is local, but the data that forms it is probably stored on a server online somewhere. That's where it is most web-like IMHO. What is completely unweblike is everything else: how the data is presented to the user, how dynamic the data is, how tied to place the data is, and how many things the user is expected to do at one time. In those areas AR is much more like a virtual world/MMO than it is like the web.
I suggest HTTP and KML not because I think they are adequate for our needs (or even really all that appropriate.) They are just pretty close and easy on-ramps. The HTML of today is not very much like the HTML used in the first few web pages in 1990. Whatever protocols and formats we use for AR will follow a similar trajectory.
It's true that the web is pretty set in its ways at this point. That was much less the case when it was going through its hyper-growth phase 12 or so years ago, and it was not the case at all for the PC boom of the early 80s. That initial growth period is critical. I don't want vendor lock-in to do to AR what it did to the pre-iPhone mobile market.
Personaly, I agree the metaphor of websites is not a good match for what AR is or should be.
I think the idea of "channels" of data is more appropriate, which each channel being a layer to our world we can view how we like. (opacity, range etc could be variable).
But nevertheless, even with this metaphor, you still need a standard. We still need a method where multiple streams of data from multiple sources can co-exist in the same viewpoint at the same time.
Moreso, we need it so the user can use one application to view these multiple streams, and the experience will work on a wide range of devices regardless of specs and resolutions...so it all needs to be scalable, and rendered at the correct quality for the device.
To me, this sounds like we very much need an AR Browser.
Or, more specifically, and AR Browser standard.
Its not that the standards established now will be the standards forever, its more that by establishing standards now, AR can grow and flourish in the right direction.
That is, a direction where -everyone- can submit content, and that, because it only requires one application download, the number of user's can increase expotentialy, rather then being niche.
Once again, I honestly feel IRC, despite its age and primitiveness, is actually a great starting block for a open system we can build -right now- for non-persistant AR games and applications.
We all think of map examples when talking of AR, but a lot of other stuff *shouldnt* be persistent and perminatenly in public view-space. Chats, Games, and even just having a virtual screen of your own data in front of you.
For that sort of functionality, a standard IRC chat channel could be automatically opened and run by the app. Permissions could be set so that friends, colleges, or the public at large could see what you were doing. All the IRC channel does is exchange links to mesh data (with timestamp for last updated mesh, and a GPS location)....most links in this case being just IP addreess of the device hosting it.
This is the easiest way I see to build from current protocals something like the depictions of AR in "Dennou Coil" or "Halting State" (or rainbows end, I assume, but I havnt read that yet).
I could walk around the street with, say, my virtual pet running along beside me, anyone logged into the "#Public1" channel would see it.
I think a standard browser is a mistake. Rather, standardize the data types.
Imagine if IE was the only browser we had access to, or it was the standard. It would eventually get bloated trying to be all things to all people since it was the standard (there is a second thought here related to why the whole server/thin client model in the past has been stupid, but I digress).
However, with standard data formats and protocols, we give ourselves *choice* in browser, browser features, functionality, flavors, etc.
We will need standards, to be sure. It will take some time though, there is a lot of growth and maturation that has to happen first (one reason why I think AR is getting mass market slightly too fast). Besides, It is hard enough thinking about software standards, we haven't even started talking about hardware standards, much less fixing the problem with mobile operators and their broken models.
I agree with Robert on this; we need data standards, and communication standards, and then let people innovate on what they do with the data!
What we _don't_ need is one browser or, even worse, and bunch of special purpose apps (e.g., I hope Layar, Wikitude, Nearest Tube, etc., disappear as fast as they appear.) Imagine needing separate programs to access each web site on the internet....
Just to clarify, I did mean a browser-standard (ie, like html is), certainly not a standard browser.
As for hardware problems, I think we should just defer them as much as possible for the hardware people to deal with. That is, if a standard is made which defines ways to present mesh data when and where its needed in the users view, it could be interpreted by all sorts of different types of hardware, and (hopefully) competition would quickly form around which devices view it best. Idealy we want a situation where the software drives better and better hardware....like the games market.
IMHO what RDF was to HTML, Photoshop layers are to AR.
This discussion really does remind me of the origins of the world wide web.
It's important to realize that Microsoft "got" the Internet -- after all, it did give us WinSock. MSFT overestimated its market power to coerce gazillions of Microsoft Office users to create interlinked documents.
For every Marc Andreessen there were hundreds of companies - even Apple - who got the future, saw it was going to be dominated by lightweight documents - and STILL ended up in second place.
Which is to say, last place.
Augmented reality is going to be a tricky little debbil. And quite frankly, I think that any bolted-on AR standard is going to find itself in the same position as Microsoft's ill-fated attempt to force Office into the role of lingua franca of the Internet.
True AR adoption will probably end up starting off in an unexpected place, like a little "Busy" halo appearing above the head of anybody using their Bluetooth handset.
I can't disagree.
We will need standards, to be sure. It will take some time though, there is a lot of growth and maturation that has to happen first (one reason why I think AR is getting mass market slightly too fast). Besides, It is hard enough thinking about software standards, we haven't even started talking about hardware standards, much less fixing the problem with mobile operators and their broken models.