A Social Network for Google Earth?
(via A Social Network for Google Earth? · GoogleSystem Blog)
Arizona State University students (well, at least one) have received a web questionnaire that strongly hints at a new Google social network using avatars, modeling, and so on. It’s called "My World" and also hints at something large-scale, as in the oft-rumored Google Earth / Second Life hybrid.
Arizona State University has a very close relation with Google: it’s one of the first large universities in the US that uses Google Apps, the site search is powered by Google Search Appliance, the university uses Google Maps and the ASU campuses already have 3D models in… Google Earth. But there’s actually more than this: the university offered photos for the Google Mars project, Google employees serve as guest speakers or adjunct lecturers at ASU and Google has an office on the Tempe campus of Arizona State University.
As always, I don’t know Google’s plans — my former colleagues from Keyhole are not returning my emails lately (possibly because I blog stuff like this*), which has effectively put the kabosh on my "How Google Earth Works" series for now… But I do know that their CTO has publicly stated that Google Earth will remain true to the real world. In other words, it won’t turn into Second Life with new islands, fictional fantastic places, furries, and so on.
He didn’t rule out avatars at some level, and I do think they’re likely at some point — specifically, when GE offers the ability to walk down the street and enter virtual stores. Once you can do that, it makes sense to be able to do that with friends, or just meet on a street corner in virtual Paris and go sightseeing. Street View is a step in that direction. The same data could be used to make 3D models at street level. And there’s an effort paying people to digitize the insides of stores in major cities, at least in 2D. I’d give it two years or less for the technology issues to be worked out and for this to be possible.
However, I still don’t expect that to be a full-blown social network. The next generation of 3D social networks will have the concept of your "home" in the virtual world in the same way that MySpace offers you a home on the web (or at least a bedroom, messy as it may be). You need some kind of anchor point to tie all of your social links together and form a big social space (a graph without nodes is just a bunch of lines). And making a virtual home for yourself is most likely to be a highly creative (read: fictional) activity, not in GE’s main mission to represent the real world.
That’s not to say a "Social GE" is a bad idea, even without the element of role-play and fiction. Imagine the world’s biggest spatial matchmaking service, linked to cell phones and personal profiles (interests, etc..), and you can see at least one 3D social application that’s more or less grounded in the real world. That’s possible, even likely to come from Google or others. But that’s not really what this questionnaire is hinting at.
There were rumors of a number of distinct teams at Google, all working in or near this space. And the wording of the questionnaire, and the target audience, indicates something much more creatively and socially focused, like a 3D MySpace, that pretty much everyone (except maybe MySpace itself) is trying to build this year. This could easily be a ground up effort. Or there could be a new version of the old GE technology that’s rigged for a separate, purely fictional world — a new world, its users get to define. That would fit with the theme of Google Earth, Mars, Moon, and Sky. It’s just one more set of data to visualize (sort of, see below).
Imagine it more concretely: repurpose the GE sidebar for social networking functions, "friendmarks" instead of (or adding to) placemarks, "stuff to do" instead of (or adding to) layers to view. And then let people populate and build this new fictional world however they wish, using SketchUp as the main modeling tool.
I’ve talked before about the issues, mashing up Google Earth and Second Life conceptually, so I won’t repeat all of that here. To sum it up, Second Life would need literally billions of simulators (or a new approach) to cover the area of the real earth. Google Earth would need to support much more dynamic content than it currently does, and deal with the inevitable fictionalization of the world. There’s also the real danger of a million avatars crowding into a virtual Times Square, sending your PC into neverland (even NYC slows down my PC with 3D buildings alone). Plus, letting people build in SketchUp and import into a world is generally fine for adding 3D buildings. But when it comes to inhabiting your virtual living room with your custom avatar, you really want simple in-world editing tools that let you move your couch an inch to the left. They’re not quite there yet, nor does it really fit with the GE user interface as it is currently defined.
And while it would be very interesting to see the GE platform turn into a more generalized 2D/3D content editor and browser (that was something I strongly pushed for, way back when, and John was always keen to pull in full HTML rendering in one way or another), it’s not quite ready to replace your old browser just yet. Someday, maybe. And if any application has the potential to stealthily come in and become your browser of choice for Web 3D, that’s what I’d pick as a starting point.
For this ASU student beta, I’m leaning towards something web-based, 3D in the [current] browser, using Flash or Java with a significant portion in 2D. And the reason is simple: social interactions happen at the scale of people, not planets; living rooms and coffee shops, not continents. While it’s cool to imagine an earth-sized virtual world, capable of pulling in 10 billion people, it’s not yet necessary to make one that big. We’re still at the "city" scale of social virtual worlds — and if you’ve ever lived in New York, that may be plenty big enough to keep people busy solving technical and usability problems.
However, with 100 250 million downloads, I don’t underestimate the Google Earth team or ever rule them out.
________
* for the record, I don’t even ask Google employees about Google’s internal projects. All my speculation is based on my own experience in the field, public tidbits, and industry veterans I’ve talked to outside of Google. In fact, my friends know the best way to get me to not blog about their project is to tell me all about it. It effectively ties my hands until they announce.
We seem to be thinking in and around the same kind of concepts, which is encouraging, not least because that I now know someone else is thinking the same way as me, validating my own ideas.
There are all kinds of technical issues associated with opening the ‘real’ world up to a Social Network.
How many people are going to claim they live at 10 Downing Street, or 1600 Pennsylvania Avenue?
These and other issues would persist throughout. So maybe something purely fictitious like Second Life would make more sense.
Incidentally, the graphics engines are in place for the 3D stuff, which I’ve seen used for web 3D elsewhere, albeit experimentally…
Personally, I think two major things need to occur for GE. 1) Integration with existing standards, such as RSS, Atom [eg: communicating with existing sites, to present that data from the entire Web], and 2) radical optimization of their engine and rendering ability.
For instance, at the moment — I’ve been desperately trying to find a ‘best’ SuperOverlay method for imagery in GE — because for whatever reason, the way they parse tiles dramatically slows rendering. This is a huge concern when you’re talking about custom environments, let alone that it disrupts the vertical-market transition that I thought GE was more in tune with than the rest.
But the future — I think we’ll see something more along those lines. Avatars, I’m not so sure are ‘necessary’, since you’re inherently in a first-person view. GE is essentially a world browser. Integrating things like Chat, based on proximity to others roaming the area would be intriguing — which I’m sure is the reason for thinking of Avatar representations…
I don’t know, I’d have to give that a little more thought. I think GE could do similar ‘things’ without looking ‘similar’, if you know what I mean.
Agreed, optimization is a big issue. There’s a good reason we dealt with the base layer as we originally did, but it also imposed certain constraints, such as cutting data in a way that makes dynamism hard.
I don’t know why things get so slow right now — there could be a number of reasons. But I doubt any of them have to do with your 3D accelerator’s actual performance limits.
But however bad it is now with buildings and overlays, it would be 10 times worse with avatars. That’s why Unype is doomed to not scale without underlying engine changes, and why Google needs to implement avatars themselves.
[...] between Snow Crash and Google Earth. But if you read any of the other blog entries about the intriguing Arizona State University beta test without also coming here, you might have been lead astray on at [...]
[...] almost certainly not avatars in Virtual Earth, though that’s always possible (see my comments on GE avatars for why it’s still premature). On the other hand, I would have thought [...]
[...] RealityPrime ยป A Social Network for Google Earth? On the feasibility of a Google Earth powered virtual world. (tags: google googleearth virtualworlds MMOGs socialnetworks) [...]
I do not think it will work( Too much things are better around.