How the Google X Moonshot Idea Factory Works

It’s not often that a company actively seeks out new problems. But that’s exactly the purpose of Google X, the secretive lab that’s hatched Google Glass, package-delivering drones, and high-altitude balloons that beam Internet service to people not yet online.

Astro Teller, the captain of moonshots at Google X, says the company created the research lab four years ago to tackle hard societal problems that aren’t directly related to Google’s core businesses. The goal is that these “mini Manhattan projects” that improve people’s lives will eventually lead to profits for Google, said Teller, who spoke at the EmTech conference in Cambridge, MA today. (Teller happens to be the grandson of Edward Teller, a physicist who worked on the original Manhattan project, according to his Wikipedia page.)

“It’s not philanthropy at all,” Teller said. Instead, Google X comes from Google DNA as a product company. “Find something that would make the world a radically better place, that will create a lot of incremental value for users. Go and build that. If we do that, then the money will come back to us,” he said.

The most obvious case of this thinking to come out of Google X is Project Loon, an effort to build a network of high-altitude balloons to provide Internet connectivity—without the need for ground base stations—to the billions of people in the southern hemisphere who aren’t connected to the Internet. Google is currently testing several of these balloons in the air now and in the “next year or so” expects to start testing the service, Teller said.

The other publicly disclosed Google X projects are the self-driving car, a flying wing that can capture wind power, and a contact lens that can be used to monitor patients’ blood sugar levels, which has been licensed by Novartis.

Often, these projects work on marrying the digital and the physical world, said Teller. The point of Google Glass, for instance, is to be smart eyewear, not a computer that fits on your head that people need to learn how to operate and manage.

Teller

Teller

“We believe the real calling of wearables, in general, is to get out your way,” he said. “It’s a big ask to ask people to wear technology of any sort on any part of your body and particularly your head. In exchange for that, what wearables need to do is to bring you value, is to improve your life.

Google chooses projects for incubation at Google X if they address a well defined problem, there’s the potential for a radical solution, and it’s based on science or technology that “barely looks possible,” Teller said. The X is shorthand for the goal for making things ten times better than current solutions.

The philosophy is to pose such a big technical challenge that people need to rethink the problem from the ground up. Asking engineers to improve fuel efficiency of a car by 10 percent will lead to incremental improvements on existing designs; asking for a car that goes 400 miles on a gallon of gas forces them to “break a bunch of basic assumptions,” he said.

The ideas are generated internally and borrowed from elsewhere. The sensor contact lens work originated at the University of Washington and was further developed at Google X. The lab has also passed its work onto Google product divisions and spun out companies funded by venture capitalists. Teller confirmed that the Google X actively considered building space elevators but, after a few days of vetting, passed on the idea.

As interesting as Google X’s work is, one has to ask: can anybody but Google actually do this sort of work? Teller says he wants others to do Google X-like work. But not many companies have the financial resources and talent to pour into such speculative efforts. Also, the trend among many venture capitalists is to place many bets on “capital-light” startups pursuing more incremental improvements on existing products.

Still, Google X’s experience shows that intriguing things happen when technologists and entrepreneurs rethink a problem and pursue breakthrough technologies. “Often you find solutions that are easier, or at any rate no harder, than doing something that’s ten percent better than super smart people before you have already done,” he said. “And the payoff tends to be ten time better when you do things that are ten times better.”

Trending on Xconomy