The Case for Secrecy in Tech

Sharing. Openness. Transparency. These are things we value—not just in the startup spaces of Silicon Valley, but in society at large. Whether you're talking about information or weather, what could be better than sunlight?

Darkness, for one thing.

In a talk at the Washington Ideas Forum today, Astro Teller, head of Google[x], made the case for opacity. (Well, for contextual opacity.) When your name is Astro Teller, and you head up Google[x], and your official work title is "Captain of Moonshots," you tend to split your time pretty evenly between the present and the future. And the latter of these places, courtesy of Google's generally secretive skunkworks, may or may not include: self-driving cars, balloon-powered Internet, and augmented reality-bearing glasses. The primary mission of Google[x], Teller told Hari Sreenivasan, is "to go out in the world and find new problems for Google to have."

Here's where the secrecy comes in. "It is the essence of innovation to fail most of the time," Teller pointed out. Google[x] is big on crazy ideas; one of its main challenges is to figure out which tiny proportion of those crazy ideas can become—should become—reality. (Or, as Teller put it: "I'm trying to work with the people at Google[x] to kill projects as fast as we can.")

Letting a thousand flowers bloom, the thinking goes, can lead to some pretty awesome gardens. To get things right, you first need the freedom to get things very, very wrong.

Publicity, however, changes the dynamics of failure. It brings accountability to experimentation. "Once you're in the public eye with any project," Teller said, "people start to rely on it." They become invested. Their egos become invested. And while Teller and his team see ending a project as "a success moment," he said, the result of the world .... doesn't. A non-released project, most of us assume, is a failed project. And no one, even and especially a Googler, wants to be seen "as a quote-unquote 'failure' in the public eye."

The result? "Public expectation," Teller said, "stifles creativity."

Ergo: secrecy. At least for a while. At least at first. And then? Once you've decided on which products and experiments to bring to market, you invest in transparency and publicity and, in general, public buy-in. (See, Teller pointed out, Google Glass's "Explorer" program, the waiting list for which just went live.) You bring users into the process. You take your initial introversion and flip it, becoming extremely extroverted.

Before that, though, you take your time—to think, to consider, to build. To decide which crazy ideas should actually become crazy projects. Until they do, Teller said, "I think it's important for us to have the quiet freedom to kill them."