Why Pentagon And Silicon Valley Find Working Together Hard

Many years into the Pentagon’s outreach campaign to befriend Silicon Valley and the technology community, headlines this month would lead you to believe things are not going well 18 June 2018, 12:58
Alphabet’s Google reportedly will not renew a partnership with the Defense Department for Project Maven, an artificial intelligence (AI) effort to help unmanned combat aircraft find and fix targets. What is more, Google CEO Sundar Pichai on June 7 published the tech giant’s new AI-related principles.
The tech community quickly recognized the worldview espoused by the principles. Among the areas for which Google would not deploy its AI: “Weapons or other technologies whose principal purpose or implementation is to cause or directly facilitate injury to people.” Other forbidden uses include “technologies whose purpose contravenes widely accepted principles of international law and human rights.”

So much for the vaunted defense tech outreach, right? Guess again. As the Google news was breaking, several other defense and technology experts and authorities in San Francisco and Washington were busy hailing the Pentagon’s outreach and declaring success in establishing bridges of communication.
Take Chris Moran, vice president and general manager of Lockheed Martin Ventures, who sat down with Washington reporters recently to discuss his two years on the job. “I was pleasantly surprised to see how most of the [tech] companies embrace what we’re trying to do,” he says. “I haven’t been . . . refused meetings in any way. I think the relationship in general is pretty good.

“Engineers love problems, and we have really great problems, really difficult problems to work on, and they feel challenged by them,” he adds.

Meanwhile, several other participants of a recent Wharton Aerospace West Coast conference echoed the sentiment. They confirm that bicoastal conversations are happening (as well as all-important funding), albeit not as much as compared with major defense acquisition programs (MDAP). Still, they stress that no less than former Alphabet Executive Chairman and Google CEO Eric Schmidt now helms the Defense Innovation Board. The Pentagon is to be commended for rebuilding bridges of communication and establishing “beachheads” or outposts in the tech world, namely in Silicon Valley, Boston and Austin, Texas, they say.

“There’s definitely a new tone; that tone is, ‘let’s just go do it,’” says a Silicon Valley investor about the Pentagon’s drive to access commercial technology.

And there lies the rub. The problem—the real problem—is when the two sides try to do “it.” They remain a world apart when it comes to the speed and the processes they follow in doing business, and this more than anything else is hindering a working relationship. Washington takes decades to develop and field MDAPs, while the product cycle in tech can be months. Worse yet, the philosophy in Washington is “never fail”; Silicone Valley’s ethos is “fail fast.”

One glaring example of where the two worlds collide is over concurrency in development—the process of developing and producing something at the same time. One point of contention arises whenever government types try explaining to techies why concurrency has such a bad reputation in Washington. Just look at the Lockheed Martin F-35, once identified by a defense acquisition czar as the epitome of “acquisition malpractice” for being rushed into low-rate initial production while still being years away from meeting requirements and getting declared operational. But in the tech world, the Joint Strike Fighter sounds like the only legacy MDAP that did development right.

Unfortunately, that is just one of the cultural differences between the two sides. Idea generation is another. Some conference goers lauded how the tech world starts by identifying the big, hard, real, world-changing solution and then works toward it as quickly as possible to lock in marketshare. Think Apple iTunes and iPhone.

Whereas in government, the end-goal is an overly scripted “requirement” that takes years to generate—and may no longer be as applicable to the battlefield—but then is handed to like-minded people whose goal is to create a perfect prototype they can use to lock in a long-term production contract.

Indeed, many insiders lament how the vast majority of legacy industry’s responses to broad agency announcements (BAA) or requests for information (RFI) coming out of Washington mirror the same inside-the-box thinking of the people who issued the BAA or RFI in the first place. That is because legacy industry is heavily populated with ex-military and ex-government people whose frame of mind is the same.

Clearly, the two worlds have much to overcome before they are working seamlessly. But at least they know, and as G.I. Joe used to say, knowing is half the battle.