RealityPrime

Conclusion

So does all this discussion prove Google is on the wrong track? Not necessarily, though I think they’re still a ways away. There’s a growing hunger for basic points and polygons that SketchUp can help meet. And if you really think about it, you can claim a functional equivalence between a database of every 3D chair created by Man and a program that creates any chair we seek — both are tools for exploring a universe of possible configurations, both rely on accurate "meta" information to find what we want. So to the extent Google helps you search the non-semantic web, they can certainly help you search the non-semantic set of 3D objects too. And they’ll succeed, to the extent there’s some value to add beyond simple keyword searching (think PageRank). But is it world-changing? Not until we change the fundamental properties of the virtual world.

Lastly, there’s this idea out there that with 3D models coming in, Google Earth may become the basis for the "Metaverse" some people seem to long for. Despite the fact that Google has people like John Hanke (who helped create Meridian 59) and a cadre of Snowcrash-loving ex-SGI graphics gurus, GE is still, so far, primarily a spatial engine, designed for 3D search and rendering and exploration of the real world.

Now, I’d love to see Google search become more like Google Earth–highly interactive and better mapped to our own cognitive spaces. And I’d love to see GE speed up the side-bar updates as we fly around too (as well as a few unnamed features I wanted but didn’t have time for way back when). But to the extent that these two apps better integrate over time, I imagine the result will still be mainly a useful tool within a 3D metaverse, a way to navigate and find things and information, but not the metaverse itself.

The metaverse itself requires an active simulation and a dynamic writable world (beyond layers or overlays), which GE doesn’t even try to do. SecondLife does this already, and to a significant degree. Short of having that new object language I desribed, what you’d really need is a mashup of both applications. And having seen the code for both (a long time ago, granted) I can say that’s no trivial task. I probably shouldn’t go any deeper than that due to old NDAs. But it’s true that even the best designed systems will only do what they were meant to do. They chose optimizations that suited them. And I’d argue that the best route to a world-scale Metaverse is still to design and build one from scratch, using the best ideas of both.

(and that, by the way, is not what I’m working on — but I still obviously need to work on my pith. Three pages is a record. Thanks for reading to the end.).

Pages: 1 2 3

Tags: Articles, Featured //