We’ve been strategists our whole adult lives — as senior officials in the U.S. government, in the corporate world and in academia. But it didn’t take those lifetimes of experience to know the fundamental rule of strategy: Seek to base your strategy on as unique and sustainable a competitive advantage as you might have.

For the United States, and its competitiveness in the global arena, that best source of an enduring competitive advantage is the innovation ecosystem of our private sector. Silicon Valley and other American pockets of intense technological advances illustrate the genius of U.S. innovation — especially in the past half-century’s maturation of the ecosystem of capital, ideas, experimentation, mass markets and rapid innovation. In strategic competition today, where competitive advantage comes from how you use data rather than how fast your plane might fly, the U.S. private sector is particularly strong.

This was among many reasons why it was so painful for us to watch the short-sighted and self-referential decision by Google to not continue its artificial intelligence work for the Department of Defense under Project Maven. In addition, according to scholars PW Singer and August Cole, Google bought several companies that were undertaking innovations in military robotics and then steered them away from continuing that work for our government. Our disappointment over Google’s Project Maven politics turned to irritation when it was revealed this week that Google has been developing a censor-friendly search engine as part of its effort to re-enter the China market.

We could bemoan Google’s narrow vision or wrestle with the cultural questions here forever, but we are interested in pragmatic solutions for the rest of the innovation ecosystem.

The Pentagon, and the U.S. government more broadly, has to invest in advanced technologies such as AI, microelectronics, hypersonics and quantum computing because we are entering a time of renewed strategic competition. In particular, China is investing hundreds of billions — through companies, state-owned enterprises and government-backed investment funds — in the development of advanced technologies to strengthen its economic and strategic position in the world.

The U.S. government has clearly recognized that the status quo of innovation is not going to address today’s security challenges. Over the past few years it has launched a number of efforts to up its game. In addition to Project Raven and a reorganization, the DoD has established outposts in Silicon Valley, Austin, and Boston to directly engage with startups and tech companies; and has conducted numerous hack-athons to help bring commercial approaches to bear in addressing national security challenges. And the Army has launched its Futures Command, with its headquarters far outside the confines of a military base.

These are important first steps, but they are frankly tiny experiments when compared with the size of the Pentagon’s budget and, more importantly, the scope of the security challenge. The DoD needs to be much bolder.

Consider the bold moves by innovation leader 3M Corporation over the past few years. To spur an innovation culture, 3M does not tinker with side innovation experiments because that is always a temptation for managers clinging to current successful products, technologies and processes in order to give lip service to innovation while in actuality living off the present. To fight this natural phenomenon — one that seizes the Pentagon acquisition structure, 3M employs the Thirty Percent Rule: 30 percent of each division’s revenues must be derived from new products put into the market in the past few years.

Division and managerial performance (and bonuses) are tied to this goal. Innovation managers at the company have wide authority over the allocation of seed capital and personnel assignment. Importantly, failure honestly come by is not punished.

The Pentagon needs to adopt these sorts of commercial approaches to innovation — and not at the fringes. The DoD should commit whole hog to large programs of record done with cutting-edge majority commercial companies to the tune of 10-25 percent of certain parts of the portfolio. Shipbuilding and radars would obviously not be appropriate for these kind of bets, but heavily commercial areas like microelectronics and enterprise IT architecture certainly would. Many of these investments will fail early and often, but some will succeed tremendously and the ultimate return of investment will be better both for the private sector and the DoD.

To conduct this level of change — call it scaled disruption — will require a fundamental shift in the way the DoD does business and a recasting of the political disconnect between some in Silicon Valley and the national security establishment. The best way to do that is to do interesting and meaningful work together on commercial terms — and not the slow, overly structured and bureaucratic normal approach of government. Many companies will be pleased to participate in this effort. Some won’t, but, hey, that’s the free market system at work.

We need to recast the system to harness the genius of America. We cannot centrally plan our way to global leadership in robotics or artificial intelligence — our strength has always been in incentivizing private enterprise to support national goals. As detailed in important works like “Freedom’s Forge,” our mobilization during World War II was successful because government fundamentally changed its acquisition policies and practices and unleashed the private sector’s energy and spirit to address the seminal challenges facing the nation. Today’s challenges are very different, but they are no less important. Let’s focus on getting the partnership right.

Jerry McGinn, a former U.S. Defense Department official, is the executive director of the government contracting center in the School of Business at George Mason University. John Hillen, a former U.S. State Department assistant secretary, chairs the center’s advisory board and teaches strategy at the school. The opinions represented here are their own.