WSJ Opinion: The 3D Internet Changes Everything


Call me a cynic, but I have still not bought into the notion that 3D Avatar Chat (called the 3D internet by some) will, by itself, change how we live. Still, futurlologists (a typo I’m going to keep) like to point to investment in virtual worlds as a barometer of how big they’ll be. Big, huge, and 3D!

Well, I take investment in virtual worlds as a much more basic barometer of how much money people think can they can make in the next 2 years, assuming, of course, their particular project turns into the 3D equivalent of MySpace, Facebook, or even Orkut (at the lower end of success, at least in the US). I don’t know how many people invest $7-10m in a 3D startup hoping it’ll become the next BlabNote, for example, though many wind up that way — goofy and marginalized.

And while we’re talking about this $345 million number that’s bandied about as fact, keep in mind, those numbers are generaly rumored or self-reported by private companies. As it happened, last year’s $1B number included at least one company whose CEOs told me the numbers ascribed to them were not correct; not to mention the Club Penguin deal which is fueling a lot of this new gold rush — it included a $350M acquisition of a whole business, not exactly an investment, plus an other $350M as incentives for later. And that, right there, was 70% of the $1B number.

Anyway, though I really like the look of Lively, as I mentioned, I got bored. It remains to be seen how fast and deep a gaggle of 15 year olds will jump in — whether or not 3D will be the new "cool," in the same way that Harry Potter got kids reading big fat books with "words and stuff." There’s always potential, but throwing more and more spaghetti at the wall does not guarantee that any of it will stick. In fact, throwing it all at one small spot pretty well guarantees a fistful of fail.

Now, there’s much ado about open standards vs. walled gardens. And we’ve talked about this before. An enterprise solution wants to be walled for the most part. The open internet does not. A kid’s world wants to be walled and patrolled and indeed limited in interaction. But object definitions, inventories, reputations, networks, and so on would do much better if they were free and portable. However, just look at how hard it is to get the big "1D" social networks to give up their core data — your friends list — to outside apps, because that’s their bread and butter, not that they really own it, but they do covet it as if they do.

Let’s just agree that we do need open standards — high standards are even better — and that standards add a multiplier effect when they enable a million different ideas to come together into a bigger whole. But before HTTP changed the world, Gopher and FTP were also standards. One could have built the web using FTP instead, though it would have rightly sucked. So I’ll just emphasize this basic point — the right standards, and the appropriate level of inter-connectedness is what’s needed. And beyond that, we also need the Mozilla of 3D — the app that makes it all accessible — and that’s not quite there yet.

So I’ll put a stake in the ground and claim that the 3D Internet is still waiting for its killer app. It is not avatar chat. Avatar chat has been around since the 1980s at least.

The qualities it will have (let’s try to be positive) are more like:

  • Re-imagines the web as a series of connected spaces, some of them 2D, some 3D, most mixed. The Metaverse and SL got the notion of one big Euclidian space wrong IMO — proven, I’d argue, by the fact that SL had to add teleporting and remote islands to break the continuity anyway. There’s nothing wrong with portals. Many people, in fact, nagivate by landmarks and see the world that way. And portals better match the existing notion of websites and links, where you don’t need such a tight coupling of "adjacent" sites. If someone wants to build a giant 10,000m2 site, that’s fine. But their internal load balancing schemes, as with the web, should be hidden from view and they should connect up like everyone else.
  • Makes all content, text, 2D, 3D, etc.. accessible using the same basic protocols. However, those protocols need be done right. For example, we do not [usually] ship whole web pages as bitmaps — we ship it as text with markup or even script (e.g., flash) that is procedurally interpreted on your computer to render a final page. Same idea needs to happen for 3D content. No more polygons and vertices, much less hard-coded "put this object at XYZ" type stuff and more semantic relationships, more akin to "the art-deco lamp sits on a glass and marble table," where geometry is built on the fly.
  • User interaction goes way beyond chat, subsumes email and voice mail (your avatar represents you, even when you’re not on-line), and conveys more emotion than can be driven with a 2D mouse and F1-12. Facial recognition, speech, gesture, and so on are critical elements, but not the whole story.
  • Helps spaces come alive. Enough with the "build your 3D living room and host a soiree" crap. If the host of the party is not home, why would you go? On the other hand, if I make a theme-park with games I design for you to play, stories I write that you can jump into, now we’re talking about something you might find fun even if I’m not virtually there to entertain you all the time.
  • Enables accelerators like indexing and archiving to both find existing content and build on it, rather than always starting from scratch or re-using a small palette of specially crafted items. This goes back to the 2nd item above. When we move from polygons to a language of 3D shapes and relationships, we can begin to more easily search the set of existing and potential objects along intuitive lines, find similarities, and even morph objects into something new. For example, "Find me an art-deco lamp. Now make it more like this other one."

Try doing that with textual webpages and you’re lost in semantic hell. But it’s actually easier to do in 3D, where geometries and relationships are much more fundamentally quantifiable.

Do all that, and we have the beginnings of something really game changing. Do it right, and it becomes the basis of the next level of human-human communication, a new visual language that mixes literature, gesture (originally called dance), movies, and games and accelerates your everyday interactions to a degree we can only begin to imagine.

  1. #1 by James on July 17, 2008 - 12:44 pm

    I have nothing deep to say, simply ‘your blog rocks’.

  2. #2 by Yohan on July 26, 2008 - 6:53 am

    Hi,

    This post was really interesting and very close to what I have been preaching for years at my company – see my papers section : http://www.conceptsl.com/papers about 2D,3D, Text, … subtle mix and tell me what you think.

    - Yohan

  3. #3 by Joaquin Keller on August 13, 2008 - 11:35 am

    I agree with the skepticism about 3D being the future of the web. VWs are a way to socialize and communicate while the web is for accessing trillions of pages. But a virtual world divided in many disjoints rooms or 3D spaces will surely fail since most will be empty and hence useless for communication.
    See my post about that:

    http://joaquinkeller.blogspot.com/2008/07/why-we-are-alone-in-internet-and.html

  4. #4 by avi on August 13, 2008 - 12:23 pm

    Joaquin, I don’t intend to advocate just a series of small rooms, as with Lively. I think the spaces can and should be bigger (esp. in capacity) when they’re intended as social areas.

    I’m making two points: first, is that load balancing using multiple servers to foirn the appearance of “one big space” should be handled much like load balancing of websites — opaque to the world, not part of the web standards themselves, and not promoted as the way everyone should do it.

    Second point: portals are fine, and there is no need to turn a series of spaces which might be run by different individuals or companies into one big cartesian (and gridded) space just to let them connect up. Ad-hoc connections are fine, as are doorways to other realms. If not for this quality and topology, the first point would be moot.

    I don’t think these two points result in lots of living room-sized spaces, but rather spaces that are appropriately sized to fit the number of people intended, from one to a million or more, depending on what kinds of servers are put to the task.

    I hope that clarieifs my position a bit.

  5. #5 by Joaquin Keller on August 13, 2008 - 5:13 pm

    The conversation is getting interesting, I cannot resist to continue.
    Yes, you finally convinced me, but my feeling is that if there are spaces big enough to host an unlimited number of avatars, few -maybe only one space- will attract the vast majority of the conversations.
    The reason is that there is a threshold for chat spaces. A webpage with few visitors per month is valuable, not a chat space.
    So I agree with your scenario, but with a subtle difference there will one or few big spaces and a lot of small ones. With a lot more sharper distribution than web sites.
    And, sure, there is no need to stitch them together in one unique space.

    Question: Do you envision a solution for the transparent load balancing to form one big conversation space ?
    SL does not know how to do it, if so there will be no <100 avatar per island limit.
    You should read this short paper: http://twinverse.com/TwinverseTechnology.pdf
    Twinverse is virtual world that is browser-based -as defined in one of your posts. So for now it’s 2D only. A limitation of actual browsers.
    And it can handle an unlimited number of avatars using this kind of load balancing you suggest.

    It is a very huge space indeed, it’s build on top of google maps and avatars can move around the whole planet.

    I am happy to find someone to discuss seriously about virtual worlds without having the mind obscured by the second life kool-aid
    Thanks Avi.

  6. #6 by cyberbian on November 5, 2008 - 4:12 am

    The 3d web will most likely not come from some 3d web project, they have a sad history.

    There are several contenders now in the gaming world for a seed protocol.

    One is Spore, and the other is PS3 Home.

    Once one of these extensible universes grows successfully into a mature profit center in the sanctuary of the in-house garden. If the owners are willing and clever enough to open the franchise of the underlying architecture to wider distribution, the 3D web will be off and running.

    It will start with an engine which allows multiple games to tie their worlds together.

    No one does 3D Graphics and massive scaling like gamers. Success is not about throwing money at a concept, it is about providing value for money to an existing market and growing that market. That is why I think that Spore has the edge, even over a giant like Sony. Anyone trying to sell 3D web is going to be run over by a gaming engine. Get off the tracks!

  7. #7 by avi on November 5, 2008 - 5:21 am

    Game developers are smart, but not even FlightSim had built an exponentially scalable mirror world in the way Keyhole had. It takes a different kind of engineering to go to web scale. Believe me, I’ve done both.

    And game developers are plenty smart enough to know that a game engine, optimized for its original scale (single, MP, or massively-MP) will need to get completely rewritten to move to the next set of scales.

    Wil Wright wouldn’t put Spore out there as the proto 3D web, nor would Sony even claim that (they can’t voluntarily shake the proprietary handcuffs anyway, so no dice there).

    And the engine that ties multiple games together today already exists. It’s called the Operating System. It simply lacks features to make the transition attractive. I would argue that’s because the demand for moving your character from WoW to Club Penguin is actually low, unless you’re talking about actually clubbing penguins.

(will not be published)


  1. Pages tagged "mozilla"
  2. Virtual Paris 2007 » Mes liens du 24 juillet 2008 au 27 juillet 2008