Two weeks ago, when I asked Blaise Agüera y Arcas, one of Microsoft’s most prominent engineers, for his thoughts on the company’s fierce rivalry with Google and Apple in the digital maps space, he was careful not to disparage his competitors too much—even when discussing Apple’s much-publicized flop when it entered the maps market. "I don’t have schadenfreude about these things," he said. "I believe in competition—it’s a race to some degree. It’s not a large set of companies that have the scale to have a serious go at this, but it’s a set of more than one."
Then Agüera y Arcas added, "We’re all going for these moon shots."
Call it a forewarning. Yesterday, The New York Times reported that Agüera y Arcas had departed Microsoft to join Google, a company most known for its "moon shot" approach to innovation. The high-profile hire by Microsoft’s arch rival is likely a bitter pill for the company to swallow—not least because Agüera y Arcas so recently talked up how Redmond was better positioned than its competitors in the technology space. ("Not to diss Google but we actually had that backpack camera before they did," he told me earlier this month.) But Agüera y Arcas doesn’t just provide Google with a PR victory—he’ll also bring years of experience to the search giant’s products, especially in mobile.
Ever since Agüera y Arcas delivered a flashy presentation at TED a number of years ago, he’s become something of a celebrity in the tech world. In his demo, Agüera y Arcas showed how his team could reconstruct 3-D geometry from a set of flat photographs—a seemingly magical feat. He went on to make significant contributions to Microsoft’s beleaguered Bing Maps service and also helped launch Photosynth, a popular smartphone app that enables users to easily stitch together interactive panoramas with photographs.
At Google, according to The New York Times, Agüera y Arcas will develop machine-learning technology. But perhaps more significantly, he’ll bring a style of right-brain thinking that could benefit Google. At Microsoft, Agüera y Arcas was responsible for some of the more outside-the-box experiments, such as strapping image-capturing technology to scooters, bicycles, quadcopters, and, yes, even backpacks. What he was most set on, though, was advancing mapping technology beyond the arms race of satellite and street-view imagery—a growing issue for Google’s engineers, whose map products have lost some of their novelty since Google Earth first launched.
"When you take the map as an app and you apply some sugar on top via more pixels or more imagery—that in itself doesn’t make the map many times more useful than it was," Agüera y Arcas said.
It’s perhaps the reason, Agüera y Arcas continued, that Apple received so much criticism for its maligned map product. "A lot of the negative reaction was, in some sense, because when people actually zoomed in and looked at this, it was, 'Well, it’s pretty, but in the end, it’s a map product, and has it gotten a lot better or no?' The answer was no—it hadn’t gotten a lot better as a map product," he explained. "We’re not thinking about just this sort of global 3-D model—where we image in as many ways as possible with our planes and cars and so on."
Indeed, what Agüera y Arcas seemed most excited about was the impact the smartphone could have on digital maps. He envisions a world where the "huge majority of [map] data" would be crowdsourced, gathered by the devices consumers carry everywhere in their pockets—and "not stuff that we’ve imaged ourselves." While Agüera y Arcas was highlighting the potential of the camera technology developed by Nokia, which Microsoft is in the process of acquiring, the scale of Google’s Android platform gives him a much greater and more promising trove of data to work with. It has the potential to make Google's maps more personalized and social.
"If you start thinking about the relationship between that graph, and what people have in their own images or in their own email—there’s this sense that your map can be really very different from the world map," he said. "Think about it as a mash-up between what is known about the world, and what is known about your own personal universe."
It’s too early to know what that product would look like in the future. But Agüera y Arcas stresses it’s no longer "just a matter of pixels."