by Adam Fisher: Fifty-five miles and three days down the Colorado River from the put-in at Lee’s Ferry, near the Utah-Arizona border, the two rafts in our little flotilla suddenly encountered a storm.
It sneaked up from behind, preceded by only a cool breeze. With the canyon walls squeezing the sky to a ribbon of blue, we didn’t see the thunderhead until it was nearly on top of us.
I was seated in the front of the lead raft. Pole position meant taking a dunk through the rapids, but it also put me next to Luc Vincent, the expedition’s leader. Vincent is the man responsible for all the imagery in Google’s online maps. He’s in charge of everything from choosing satellite pictures to deploying Google’s planes around the world to sending its camera-equipped cars down every road to even this, a float through the Grand Canyon. The raft trip was a mapping expedition that was also serving as a celebration: Google Maps had just introduced a major redesign, and the outing was a way of rewarding some of the team’s members.
Vincent wore a black T-shirt with the eagle-globe-and-anchor insignia of the United States Marine Corps on his chest and the slogan “Pain is weakness leaving the body” across his back. Though short in stature, he has the upper-body strength of an avid rock climber. He chose to get his Ph.D. in computer vision, he told me, because the lab happened to be close to Fontainebleau — the famous climbing spot in France. While completing his postdoc at the Harvard Robotics Lab, he led a successful expedition up Denali, the highest peak in North America.
A Frenchman who has lived half his 49 years in the United States, Vincent was never in the Marines. But he is a leader in a new great game: the Internet land grab, which can be reduced to three key battles over three key conceptual territories. What came first, conquered by Google’s superior search algorithms. Who was next, and Facebook was the victor. But where, arguably the biggest prize of all, has yet to be completely won.
Where-type questions — the kind that result in a little map popping up on the search-results page — account for some 20 percent of all Google queries done from the desktop. But ultimately more important by far is location-awareness, the sort of geographical information that our phones and other mobile devices already require in order to function. In the future, such location-awareness will be built into more than just phones. All of our stuff will know where it is — and that awareness will imbue the real world with some of the power of the virtual. Your house keys will tell you that they’re still on your desk at work. Your tools will remind you that they were lent to a friend. And your car will be able to drive itself on an errand to retrieve both your keys and your tools.
While no one can say exactly how we will get from the current moment to that Jetsonian future, one thing for sure can be said about location-awareness: maps are required. Tomorrow’s map, integrally connected to everything that moves (the keys, the tools, the car), will be so fundamental to their operation that the map will, in effect, be their operating system. A map is to location-awareness as Windows is to a P.C. And as the history of Microsoft makes clear, a company that controls the operating system controls just about everything. So the competition to make the best maps, the thinking goes, is more than a struggle over who dominates the trillion-dollar smartphone market; it’s a contest over the future itself.
Google was relatively late to this territory. Its map was only a few months old when it was featured at Tim O’Reilly’s inaugural Where 2.0 conference in 2005. O’Reilly is a publisher and a well-known visionary in Silicon Valley who is convinced that the Internet is evolving into a single vast, shared computer, one of whose most important individual functions, or subroutines, is location-awareness.
Google’s original map was rudimentary, essentially a digitized road atlas. Like the maps from Microsoft and Yahoo, it used licensed data, and areas outside the United States and Europe were represented as blue emptiness. Google’s innovation was the web interface: its map was dragable, zoomable, panable.
These new capabilities were among the first implementations of a technology that turned what had been a static medium — a web of pages — into a dynamic one. MapQuest and similar sites showed you maps; Google let you interact with them. Developers soon realized that they could take advantage of that dynamism to hack Google’s map, add their own data and create their very own location-based services.
A computer scientist named Paul Rademacher did just that when he invented a technique to facilitate apartment-hunting in San Francisco. Frustrated by the limited, bare-bones nature of Craigslist’s classified ads and inspired by Google’s interactive quality, Rademacher spent six weeks overlaying Google’s map with apartment listings from Craigslist. The result, HousingMaps.com, was one of the web’s first mash-ups.
Google never imagined that its service, which it called Maps, could be co-opted like that: its product was designed to be a Google brand extension, not a database that outside developers could use without permission. “We were faced with a choice,” Mano Marks, one of the engineers responsible for early versions of Google Maps, recalls in a conversation with Rademacher that Google has put on YouTube. “We could either sue him or hire him.” To Google’s credit, Rademacher was hired.
Rademacher’s mash-up showed Google that the map could be more than just something that people glance at to keep from getting lost. By opening up its map to everyone, Google could perhaps make itself into the one indispensable cog in the giant collaborative computer that was emerging. “HousingMaps was when people realized that making [map] data available to other programmers was incredibly powerful,” O’Reilly says. “Google never looked back.”
Rademacher helped Google develop and publish what’s known as an application programming interface for Google Maps. Think of an A.P.I. as a programmers-only side entrance into the Google mapmaking machine. No longer did they have to repeat Rademacher’s hack; instead, with access to the A.P.I., developers could combine Google’s free map with their own data and end up with a cool mash-up like HousingMaps — or build an entire company based on Google Maps. The real estate site Redfin, for example, is basically just that: pictures of and information about houses for sale layered over a map from Google. The same goes for AirBnB, but with room rentals. Uber and Lyft, the quasi-taxi services. RelayRides. TaskRabbit. NeighborGoods. They may not be household names (yet), but there’s an entire Google Maps-based ecosystem out there.
Behind Vincent and me, near the center of the raft and mounted about 10 feet above the surface of the river, was our expedition’s payload: a green orb, about the size of a soccer ball and dimpled with 15 lenses pointing in different directions. This custom-made panoramic camera is what has made Google’s Street View possible. Street View is the feature within Google Maps that allows you to pull up a panoramic photograph taken from a particular spot on a given street. For years now, cars with roof-mounted panoramic cameras have been driving the world’s roads while taking pictures every yard or so.
There is a version of the car-mounted Street View camera that is designed to be worn like a backpack — that’s the Trekker. For the raft-trip, the Trekker camera-orb was programmed to snap its 15 (virtual) shutters every few seconds. These pictures would be stored in the camera’s computers; tagged with precise coordinates of latitude, longitude and altitude; and then later digitally melded into one 360-degree image. Once the pictures collected on the raft trip are incorporated into Google’s world map, anyone with an Internet connection will be able to access an immersive virtual-reality view from anywhere along the bottom of the Grand Canyon.
As a light rain started to fall, I wondered aloud if on this trip we had already taken more photos from the bottom of the canyon than all the previous trips combined. Maybe around 20,000 people in a given year run the Colorado River through the Grand Canyon, and it’s the sort of trip that was all but unheard-of before the 1960s.
Vincent performed a quick calculation of our trip so far without a calculator or even pencil and paper: two cameras on two rafts, each taking 15 shots every two seconds, for three eight-hour days, versus 20,000 people a year taking snapshots for 50 years.
“Not yet,” he said. “By the end of the trip, I would think so — possibly.”
As we talked, lightning struck behind us, then to one side, then to the other. The orb and its associated computers and copper cables were all lashed to an aluminum mast, forming a conductive path reaching down to the several inches of water sloshing in the bottom of the raft. The consensus was that the Street View camera would make a decent lightning rod. Word was passed to the boss.
“That’s why we brought two,” Vincent said, shrugging.
One perk of being a Google engineer is being encouraged to devote 20 percent of your time to your own project. Back in 2004, Street View was Vincent’s. The idea was to photograph every inch of every street in San Francisco and put those pictures inside the map. It was a big job, and Vincent had a lot of people at Google pitching in to help. (Larry Page, one of the company’s founders, was a trailblazer; in 2001, he collected images by driving around town with a video camera mounted to the side of his car.) Eventually, Street View would become the next breakthrough for Google Maps after the introduction of its programming interface. But Google was not the first company to turn this idea into reality; Amazon was.
In 2005, A9.com, Amazon’s skunk works for search technology, unveiled an innovative feature called Block View. It was meant to be a newfangled Yellow Pages where you could find the phone number and address of a local business — as well as a photograph of its storefront. Block View was discontinued after only 20 months, but not before Microsoft introduced its own version, Streetside, that was essentially identical, except that Microsoft’s pictures of streets and storefronts were seen through a digitally created framing device. Though the photos were taken from car-roof-mounted cameras, they were presented online as if you were looking through a windshield. The result was dorky, but it was one solution to the vexing problem of coming up with a user interface. How do you move through a map made of photographs? Microsoft’s answer: In a virtual car.
Google ultimately developed a more elegant user interface. Instead of representing movement along a street as flipping through a filmstriplike series of photographs, as Block View and Streetside did, Google pursued the idea of a panoramic camera — what would become the green orb — and used it to take a panoramic photo every few feet. The effect of hopping from one photo to the next in Street View is one of walking through virtual space.
Microsoft’s Streetside debuted in 2006 with a photographic rendering of parts of Seattle and San Francisco. Google’s Street View arrived a year later, with five cities: San Francisco, New York, Las Vegas, Miami and Denver. Google eventually overwhelmed Microsoft with a more aggressive surveying program. Street View now covers 3,000 cities in 54 countries, and it has gone beyond streets and onto train tracks, hiking trails, even rivers. A section of the Amazon was the first river, appearing last year; the Thames made its debut in October; and the Colorado will be available by the end of the year. “We want to paint the world,” Vincent says. When I asked him what level of resolution we were talking about, he said, “About one pixel to the inch.”
By threading photograph after photograph along the lines that mark the byways and highways on the map, Vincent and his team are making, in effect, one large photograph of the globe. It’s a neat trick, perhaps even the next conceptual leap for cartography, but like most things Google spends a lot of money on, very likely to be more useful than it first appears. Like most people when they first encounter Street View, O’Reilly used it to check out the photo of his house. But then, he says, he later began to see the potential of the data collected by Google and to imagine more and more uses for it.
Street View turns out to be incredibly valuable for all sorts of things — but above all for mapmaking. By 2008, Google was ready to wean itself from the licensed data that underpinned the first generation of Google Maps by greatly expanding its database of geographical information instead, which was called Oyster. The team added terabytes worth of raw data tagged to locations, everything they could get their hands on. In the United States, some of the best information is free and comes from the federal government: U.S. Geological Survey and Forest Service reports, census records and the like. Google bought other map data outright, from both the United States and abroad. But in most of the developing world, there was simply no good map data to be had at any price. In places like India, Oyster made do with only poor-quality tracings of the streets taken from satellite photos.
Creating one big map from hundreds or even thousands of other maps means comparing each map with all the others to see how they line up. They never do. Including crucial details about address and turn-restriction information — necessary for generating driving instructions — has traditionally been a matter of sending cars out to drive the roads in question and waiting for the drivers to file their reports, a process called ground-truthing. Street View provided Google with a shortcut. Not only were the GPS tracks from the Street View cars great for reconciling map data, but the pictures taken by the panoramic camera also made it possible to go into Street View and look around for turn-restriction information. Google can ground-truth its data in virtual space. In Hyderabad, India, Google has a staff of more than 2,000 ground-truthers “driving” through cyberspace every day, cross-referencing map data with the Street View pictures.
In addition to the human operators, pattern-recognition bots search the archive for addresses: Google’s computer-vision programs look for house numbers, street signs, even the bespectacled face of Colonel Sanders — in which case the bot will flag the corresponding point on the map with a note that there’s probably a KFC franchise located there. “When we started, Street View was just some sci-fi idea,” Vincent says, “but now, it’s the backbone.”
The rainfall hitting the hot canyon walls produced a vaporous mist that put the entire canyon into soft focus. Vincent called back to the crew working the camera: “These panos, we must keep them. I think they will be quite artistic.”
During a routine Street View mission, pictures spoiled by rain are rejected. Street View drivers are instructed to drive only in the summer months, when the sun is high, in order to keep the light relatively consistent from region to region. If it rains, they have to pull over and wait out the storm. But a raft trip is a different story. And besides, Vincent was right: the scene before us was incredibly beautiful. Everyone was wide-eyed. “I’m trying to burn these images into my retinas, so I never forget this place,” I said.
“You never will,” Vincent said, “because Street View is there to help you to remember.”
It was a trippy moment, the realization that I was going to be able to look back at my own outsourced memory one day. It brought to mind the writer Jorge Luis Borges. In a short story entitled “On Exactitude in Science,” Borges tells of a long-ago empire where “the Art of Cartography attained such Perfection that the map of a single Province occupied the entirety of a City, and the map of the Empire, the entirety of a Province.” In Borges’s empire, the importance of the cartographic guild grew as the map grew, until finally the empire was completely covered by a map of itself.
After the downpour faded, I suggested to Vincent that there was something Borgesian about this project of his. This raft trip was effectively sucking the Grand Canyon into Google’s vast cartographic oyster — indeed, his green orbs were capturing the entire world. “What happens,” I asked, “when Street View grows to be as big as the territory it covers?”
Vincent answered with a question of his own: “How many photos would you need if you wanted one picture taken every 10 meters across the earth’s surface?”
“Um, a googol?” A wild guess.
“The answer is easy once you know how much land there is in the world.”
I didn’t, so I had no idea how many individual panoramic photos you would need to get the entire planet inside Street View.
“Well over a trillion,” he said, “and we are nowhere close.”
Vincent went on to point out that the two largest and most populous continents have barely been touched by Street View. “Africa and much of Asia are big holes right now.” And Street View clones are popping up in all the places where Google is not active. “There are three in China, two in Russia, one in Turkey, another in Korea and many others as well.” Vincent doesn’t worry much about competitors like Microsoft, but he takes the clones seriously. “They all have copied our user interface beautifully,” he said, “It’s a form of flattery.” He laughed, but it was clear that he regarded such copycat behavior as a form of theft. “We are behind in those places,” he added.
Vincent’s Street View cars have already mapped six million miles. Depending on your perspective, that’s either a quite a lot (equivalent to 12 trips to the moon and back) or not much at all (only one-tenth of the world’s estimated 60 million miles of road). Either way, Google’s huge investment in the camera-equipped cars — not to mention trikes, boats, snowmobiles and, yes, rafts — has yielded the most detailed street atlas on earth.
Early last year, Google’s United States market share for where-type queries topped 70 percent, and Google started to get serious about recouping the fortune it has been sinking into making its map, putting a tollbooth in front of its application programming interface. Henceforth, heavy users would be charged for the privilege. (The very biggest users — which Google wouldn’t identify — were already paying.) The use limit was carefully calibrated: it would start at 25,000 map-related requests a day for 90 consecutive days. More than 99 percent of the users of the A.P.I. — small, boutique sites like HousingMaps.com — would be under the limit and thus unaffected. Even so, that left approximately 3,500 sites, companies that actually have a real business dependent on Google’s maps, which would have to pay. The change prompted an exodus.
Foursquare, an urban-exploration app used by 6 percent of smartphone users worldwide, was one of the first big players to leave last winter. Additional high-profile defections followed in the spring: Wikipedia left on what could probably be described as ideological grounds; it simply doesn’t like the idea of proprietary data. Craigslist wanted more control. Apple defected in the summer. Its motive was strategic, even paranoid. The arrival of the tollbooth made it clear that Google saw Maps as a crucial part of an operating system for mobile devices. Could this lead to its having too much power over the iPhone itself?
Those four companies all turned to the same alternative: OpenStreetMap, a nonprofit based in Britain often described as the Wikipedia of mapping. Founded 10 years ago by Steve Coast, a cartography-obsessed computer-science student at University College London who liked to bicycle around town with a GPS taped to his handlebars and a laptop recording its data in his backpack, O.S.M. has since grown into a collaboration among some 300,000 map enthusiasts around the world. The resulting map is one that anyone can contribute to and use, free of charge. But it wasn’t until Google Maps started locking down its data that O.S.M. became what it is now — a potential challenger to Google’s cartographic hegemony.
On the last day of my ride-along, Vincent beached the rafts in order to take the two orbs up to the site of a prehistoric Indian ruin. He and a colleague, Daniel Filip, unstrapped the Trekkers from their masts and restrapped them onto their backs. Each pack weighed 40 pounds; the orbs, fixed at the end of a mechanical stalk, hovered at just above head height. Together the two men started zigzagging up the North Rim, a pair of eyeballs going to see what was at the top of the trail.
Filip is the most senior engineer on the Street View team. He was the one who came up with the idea of using the 360-degree panoramic camera in Street View, and he built the software engine that allows you to navigate from one panorama to the other. At one point Filip managed to become separated from the group. He didn’t see anything funny about someone on a mapmaking expedition taking a wrong turn. “The trail is just not very well marked, is all,” he said.
The vista down the Colorado River from the Indian ruin is the same view that appears on the back of Arizona’s 2010 commemorative quarter, and after Filip arrived, Vincent handed him his phone and, his orb still overhead, posed for a portrait. Crouching down for a better camera angle, Filip suddenly lost his balance. The orb puts the wearer’s center of gravity high on the body. For a long moment, Filip teetered. The trail was a mere shelf in the steeply sloping cliff-face, scarcely two feet wide, 700 feet above the canyon’s bottom. His foot slipped, sending a shower of gravel over the side. His arms pinwheeled. It was the closest of close calls.
For a long while afterward, Filip told me, he couldn’t stop thinking about his son and the long-overdue appointment with his estate lawyer. There are still dangers associated with mapping the world.
Today, Google’s map includes the streets of every nation on earth, and Street View has so far collected imagery in a quarter of those countries. The total number of regular users: A billion people, or about half of the Internet-connected population worldwide. Google Maps underlies a million different websites, making its map A.P.I. among the most-used such interfaces on the Internet. At this point Google Maps is essentially what Tim O’Reilly predicted the map would become: part of the information infrastructure, a resource more complete and in many respects more accurate than what governments have. It’s better than MapQuest’s map, better than Microsoft’s, better than Apple’s.
“You don’t see anybody competing with Google on the level or quantity of their mapping today,” says Coast, who now works as a geographic-information professional. But, he adds, “that’s because it’s not entirely rational to build a map like Google has.” Google does not say how much it spends on its satellite imagery, its planes, its camera-equipped cars, but clearly it’s an enormous sum. O.S.M., by contrast, runs on less than $100,000 a year. Google’s spending is “unsustainable,” Coast argues, “because in the long run, this stuff is all going to be free.”
The O.S.M. map data is free now — but using it comes with a catch. Any improvement, or any change at all, that a developer makes to O.S.M.’s map must be sent back to O.S.M. It’s a clever tactic, forcing competitors of Google Maps to choose between fighting Google alone or joining a coalition that, if it prevails, will ensure that no private company will ever be able to establish a mapping monopoly.
So far Coast’s coalition is doing pretty well. In some places, he says, O.S.M. has grown to be even more information-dense than Google Maps — in North Korea, for example, but also parts of Europe. One limitation, though, is the questionable utility of some of the details. The cities that O.S.M. has mapped are sometimes charted down to every footpath, bench and tree, yet they can still lack accurate particulars about addresses and traffic rules. It turns out that for the unpaid map nerds who make up the bulk of O.S.M.’s volunteer staff, Coast says, “entering turn restrictions is just not as fun as entering trails.”
For-profit companies have started contributing data and in some cases even money to the O.S.M. cause. Microsoft was an early supporter, opening up its A.P.I. and giving access to aerial imagery that Coast values at “approximately priceless.” One of the smaller in-car GPS companies, Telenav, where Coast is currently employed, has lately provided turn-restriction data and hired professional mappers to work with O.S.M.’s cadre of amateurs. Foursquare, whose map uses data from O.S.M., has a map-correction app that potentially adds its 40 million users to the O.S.M. coalition.
Coast is confident that, given time, Google’s map will be surpassed by the O.S.M. map: “You don’t see any proprietary competitors to Wikipedia, right?”
O’Reilly is more skeptical. “An open-hardware play broke the IBM monopoly, an open-software play broke the Microsoft monopoly, and eventually an open-data play will prevail,” O’Reilly admits, but he points out that those earlier cases were not instances of direct competition between rival companies. “It wasn’t a plug-compatible mainframe clone that dethroned IBM; it wasn’t a free operating system like Linux that dethroned Windows.” Rather, he says, “it was this toy, the personal computer, it was the global operating system that we call the Internet.”
Google, for its part, is committed to its strategy of having the best map, whatever the cost. Brian McClendon, a vice president who oversees all of Google’s Geo products, disputes even the idea that the free-spending map division is a money loser. Because 20 percent of Google searches produce where-type results, he argues that his team should be credited with a commensurate portion of search revenues. Revenue from ads on local where-type searches, McClendon says, are “already valuable enough to justify the investment — plus, plus.”
In June, Google bought the popular social-mapping app Waze for close to a billion dollars. The product can be thought of as a Twitter for traffic jams, and the acquisition was widely interpreted as a defensive move — a way of keeping valuable map data out of competitors’ hands. Then in the summer, Google released a new Maps interface, code-named Tactile. The redesign, which Google officially refers to as “the new Google Maps,” is currently accessible in preview mode (and is expected to replace and take the name of Google Maps sometime in the next couple of months). Zoom in on more than a hundred cities around the world and see not simply a photograph of the rooftops, but also the buildings themselves rendered in 3-D and viewable from any angle. Zoom even lower, switch to Street View and you can enter public buildings. Pull back to the stratosphere, and clouds can be seen encircling the earth, rendered from real-time weather data. Pull back even further, and there is the big blue marble at the edge of the Milky Way, our planet rolling like a trackball under your fingertips.
The new interface is as significant as any change to Google’s mapping products since Maps debuted nine years ago and one that makes Apple’s rejection of Google Maps seem like an understandable business decision. Tactile is beautiful and graceful and poised to dominate its world — Apple-like, inother words.
In most tellings, Apple was the big loser in its 2012 clash with Google over maps. The public outcry over the many shortcomings of Apple’s Maps — mismarked hospital emergency rooms, whole towns gone missing, twisted and disfigured aerial imagery — prompted a public apology from Apple’s new chief executive, Tim Cook. The Apple executive responsible for mobile software, Scott Forstall, was dismissed. Possibly the most lasting damage was the blemish the episode left on Apple’s reputation: Where was the company’s reliably elegant design?
At the same time, Google seemed to be blindsided by Apple’s move. Google Maps had been the default map on the iPhone — part of the operating system, not simply an app — but when Apple issued iOS6, its upgrade to the iPhone’s operating system, Google’s map was suddenly replaced with Apple’s homegrown version. Overnight, Google Maps lost 300 million iOS users — approximately 20 percent of the global smartphone market — to Apple, not to mention the data that those iPhone users had been generating for Google. Such data is precious. It can be used to refine the map. It could also be mined for hidden correlations and moneymaking opportunities. It’s possible to imagine an analysis of where, when and how long people shop at some stores compared with others or getting an answer to the question: How many potential customers who are headed to one particular store end up in the competitor’s store across the street? What’s more, the data from iPhones is particularly valuable, because it comes from people who are known to pay a premium for technology and convenience.
What really made the experience sting, though, was that Google had no contingency plan. After Apple’s surprise switch, iPhone customers were clamoring to dump its product and return to Google Maps, but Google had no external Maps app ready for the iPhone. It took the company three months to make one. Google should have been prepared for this possibility: it had been no secret that Apple was up to something. The first iPhone debuted in 2007 with Google Maps built in, but since then, Apple has been buying up promising little mapping-technology companies. Industry data should have prompted suspicions: “We keep a database of all online job postings,” O’Reilly says, “and I remember seeing a huge spike in Apple hiring developers with mapping expertise.”
The blows suffered by Google and Apple were seen as opportunities by the other two players still left in the game: Microsoft and OpenStreetMap. Microsoft knows better than most that a monopolistic position in the technology sector is not unassailable. It has itself toppled giants like IBM and seen its own operating system’s dominance unwound by the Internet.
No one knows what the next new thing will be, but it’s very likely that there will be one, some technological innovation or legal event that shakes up the Internet again. Microsoft is hedging its bets, in case privacy concerns lead to changes in consumer behavior or regulations that upend the communications-technology industry: it asks users to opt in before it collects GPS traces from mobile phones in order to incorporate that data into its maps. Its many businesses — Windows, Office, Xbox, video games, consulting services, mobile phones and advertising — offer potential hedges against unpredictability as well. Google, on the other hand, depends on a single extremely profitable business — selling advertising — to subsidize the rest of its enterprises. Microsoft is betting that its diversified, conservative approach will enable the company to endure and prosper should Google be brought low.
OpenStreetMap, by contrast, is rushing headlong into Google’s territory. Steve Coast recently showed me the latest innovation: iPhone attachments that look a bit like kazoos or doll-size French horns, made of plastic. “They’re snap-on panoramic lenses,” he said. Coast intends to release an app soon that will enable anyone’s cellphone to function as an open-source version of the Google orb. The resolution of the panoramas it will produce will be nowhere near orb-quality, Coasts conceded, but he claimed that the metric that really matters is the price-quality ratio. “For $60 anyone can have their own Street View vehicle!” He did add, sotto voce, that “the real barrier to entry is that you have to be willing to duct-tape your phone to the top of your car.”
Coast has a related plan for adding more and better aerial imagery to OpenStreetMap: it turns out to be relatively simple for a computer program to transform snapshots taken from a small plane into what look like extremely high-resolution satellite photos. And sometime this month, Planet Labs, a new space-imaging start-up, plans to launch the world’s largest privately owned network of earth-imaging satellites and make all the pictures they take publicly and freely available.
Borges’s story ends with the map of the empire becoming so big that it achieves a scale of one to one, at which point it — along with cartography itself — fades into irrelevance. “In the deserts of the West, still today,” Borges writes in his last line, “there are tattered ruins of that map, inhabited by animals and beggars.”
We’re fast approaching an endgame in which the capacity to read a map could become a lost art. The online-map era started with a flowering: Rademacher’s HousingMaps.com. Foursquare and others took the concept to its logical conclusion. It’s no exaggeration to describe the smartphone as the equivalent of a cursor moving through a one-to-one-scale map of the world. Today, turn-by-turn navigation is the quintessential map app. Already some maps exist as voices that tell you where to go: Turn left, turn right. When cars drive themselves, the map will have been fully absorbed into the machine.
Right now Google has about 25 experimental self-driving cars on public roads in California and Nevada. So far they have driven more than 600,000 miles without being involved in a serious accident. The self-driving algorithms do not work because there has been some breakthrough in artificial intelligence; they run on maps. Every road that Google’s robo-cars drive on was first surveyed by a human-driven pilot car outfitted with sensors accurate enough to measure the thickness of the painted lines in the middle of the road. Every detail of the road has been mapped beforehand. According to Peter Norvig, Google’s head of research, it’s a hard problem for computer vision and artificial intelligence to pick a traffic light out of a scene and determine if it is red, yellow or green. But it is trivially easy to recognize the color of a traffic light that you already know is there.
In effect, the robot car is not driving through the real world so much as it is moving through, in Borges’s words, “a map of the Empire, whose size was that of the Empire, and which coincided point for point with it.” When the real world is transformed into a data set, it starts to take on some of the aspects of the virtual.
Sergey Brin, Google’s co-founder, has promised to release self-driving technology within four years, and Google’s maps will then be a standard feature in its robot cars. Not to be outdone, Elon Musk has promised that Tesla Motors will deliver a self-driving car in three years. It’s too early to know whether Tesla will use O.S.M.’s maps — but the indications are that it will not use Google’s.
The map, at that point, will just be data: a way for our phones, cars and who knows what else to navigate in the real world. Whose data will that be: Google’s? Ours? Our car company’s? It’s too soon to tell. But one thing seems certain, O’Reilly says. In the end, “the guy who has the most data, wins.”
Adam Fisher lives in San Francisco and has written for Wired, Popular Science, Outside and other publications.
Editor: Dean Robinson