Can Startup Bubbli Turn Geotagged Photos Into Matrix-esque Augmented Reality?

Bubbli

Startup Bubbli just raised $2 million in funding from August Capital, which will help it in its mission to build "the Matrix (minus the sentient AI)." What's it all about?

Bubbli's home page is mystical, to say the least. It really does say its mission statement is to build the Matrix, but there's a string of text that gives us some big clues: "Take a look in your pocket," it suggests, "You probably have a smartphone. You're looking at something that governments used to pay millions of dollars to launch into space. You presumably launched your device into your back pocket for merely a few benjamins." So we know this is smartphone-centric. But then the text goes on:

While you were out shopping: Thousands of brilliant papers have been published in computer vision about understanding the world around you through a camera chained to a workstation in a basement. The algorithms need to be set free. Why liberate the algorithms? The better we understand reality through a camera lens, the better we can replicate it elsewhere. After all, our eyes are just light sensors, what does it matter that the light that goes into your eyes is reflected off of an object from the sun or comes from a digital display?

The company is trying to recruit three programmers, including a computer-vision expert, and will soft launch at the upcoming TED event.

Bubbli's site has resulted in some speculation online thanks to this mystery, and the fact that some of the site's support has come from John Doerr, who injected massive amounts of cash into Twitter—a venture that has done pretty well. Doerr noted he'd seen Bubbli, and thus had "seen the future."

But what can we infer from Bubbli's site? It seems the firm is planning something pretty impressive. Bubbli has realized that when all of us, by the million, snap photos and videos (by the billion) of our daily events, we're recording a rich digital story of the world—with accurate GPS locations, angular information and digital compass data (in some cases), so that it's possible to work out precisely where the images were created.

Attempts to usefully process these sorts of data have been carried out before—Google's Goggles system is a notable one that tries to identify real-world objects from a smartphone image based on previous searches and uploaded images, and Microsoft's Bing tries to out-do Google's Street View system with 3-D modeling of real-world locations based on crowd-sourced photos. These systems are flawed, but smart—yet they rely on massive computer power on remote servers.

Bubbli Bubbli would seem to be attempting something different—the "next generation" of augmented reality (it teases) that constructs a representation of the real world in a digital space, using billions of images from smartphones. It's fascinating because Street View would've been impossible just 10 years ago, and now we rely on it—but Google's cars can't go into buildings or parks, and they can't generate a real world image in real time.

Bubbli's hints at putting the computer vision algorithms into smartphones imply that it could generate near real-time representations of real-world places by updating previous images with ones that are acquired and uploaded to the Web. Imagine a Street View that covers inside museums, bars and so on—with almost real-time updates. Then imagine the potential for real-world AR gaming, from the safety of your desk, or the potential for detecting news events by seeing crowds forming, or the potential for abuse by law enforcement officials or terrorists, and you'll see instantly why Bubbli is fascinating.

To read more news like this, follow Fast Company on Twitter: Click here.

Add New Comment

0 Comments