When Sergey Brin and Larry Page built the search engine that would become Google, they started by making maps. Mapping the structure of hyperlinks across the Web enabled them to rank Web pages by the number of links that pointed to them. That turned out to be a pretty good measure of those pages' importance, and it formed the basis for what today is a $380 billion business.
For a company whose mission is to "organize the world's information," however, organizing the Web was only the beginning. Within a few years Google's efforts had branched into news, books, academic research, email, images, and videos. In 2005, it leaped into the terrestrial realm with Google Maps and Google Earth. Then came Street View, real-time traffic data, and turn-by-turn directions. In recent years it has begun building robots, self-driving cars, and smart home appliances, all of which sense, analyze, and transmit information about the physical environment.
Now Google is going for the bird's-eye view. Last year it announced Project Loon, a seemingly quixotic foray into high-altitude balloons. Two months ago it acquired Titan Aerospace, a startup that makes high-altitude solar-powered drones. And this week it paid $500 million for Skybox Imaging, a startup that builds tiny satellites that can shoot high-resolution photos and videos of the Earth below. The latest rumors have Google in talks with Richard Branson's space-tourism venture, Virgin Galactic.
Ask Google what all the balloons, drones, and satellites are for, and you'll get an answer straight from the mouth of a Miss America contestant. Why, they're for bringing Internet access to the world's poor, improving Google Maps, fighting deforestation, and aiding in humanitarian relief efforts!
I don't doubt that Google is sincere in those goals. Project Loon in particular seems to have hatched from a genuine desire to find new ways of providing Internet access to rural areas. (At this point, there are no downward-facing cameras on the balloons, unlike on the drones and satellites.) But it's hard to fathom that a company as strategically minded as Google would buy a drone company and a satellite company just to lend a helping hand to some NGOs. As for Google Maps, it already gets high-quality satellite imagery from a company called DigitalGlobe, and as the Atlantic's Robinson Meyer points out, the two firms just signed a new multi-year agreement.
It's important to understand that Skybox is not just another satellite-imaging company. To fully grasp its potential--and its natural affinity with Google's mission--you have to look more closely at what it was doing before Google bought it. Fortunately, we have a marvelously high-resolution snapshot in the form of an in-depth feature article that David Samuels wrote for Wired a year ago. The headline: "Inside a Startup's Plan to Turn a Swarm of DIY Satellites Into an All-Seeing Eye."
Commercial imaging satellites tend to be big, expensive, and seriously high-tech, which is why there are fewer than a dozen in orbit today. But Skybox has found a way to make them cheap and light, using ingenious image-processing software to cover the hardware's shortcomings. So far, it has only launched one, which can only shoot 90 seconds of video at a time. With enough resources, though, Skybox could launch dozens of them, generating near-real-time images of any spot on the globe throughout the day. That's like having a closed-circuit TV network that covers the whole world.
That doesn't mean Google is bent on rivaling the NSA as a surveillance agency, even if Skybox does sound a little like Skynet. Google has never shown much interest in government contracting, and spying on individuals runs counter to its business model. (Because Google's advertising network relies on people voluntarily giving it as much information as possible, chairman Eric Schmidt has joked that the company's policy is to "get right up to the creepy line and not cross it.") Besides, the federal government already regulates the satellite-imaging industry to make sure it isn't capturing classified information. It will surely be watching Google and Skybox closely.
In a phone interview this week, Samuels told me that Google's acquisition of Skybox confirms a hypothesis he developed in the course of his reporting last year. "I don't think this is about improving Google Earth," he said. "I think this is a data play."
"Once you've got real-time coverage of the entire globe from space, you start to get data sets that seem like science fiction-;data sets of the kind that were formerly available only to the NSA, and only in theory," he continued. Imagine Google putting its image-recognition capabilities to work to identify and track all of the world's oil tankers, or count the number of cars in the parking lots of all the Walmarts in the United States,and making that information searchable via the Internet. Conceivably, you could reach a point where Google can do things like estimate changes in a country's GDP on a daily basis.
More broadly, Google could use insights from its satellite data to improve its search results, hone its Google Now personal-assistant software, and enhance the navigation capabilities of its self-driving cars. As Alexis Madrigal has explained, those cars rely heavily on ultra-precise, digitized maps of their surroundings. And, of course, the more information Google's computers can glean about what's going on in the world at any given time, the more the company can refine its targeted advertising programs.
Google insists that's not what it has in mind, at least for the time being. But then, a lot of what the company does today was not what Page and Brin had in mind when they set out to map the Web.