Defense Department leaders agree the U.S. military must reinvigorate its technological edge. They just can’t agree on which technologies matter. Nor do they appear to be laying out arguments that would help the rest of the Pentagon, lawmakers, and industry understand which technologies will matter most in tomorrow’s wars, and therefore which should receive top priority in terms of effort and funding. In short, DOD needs a technology strategy.
Former Deputy Secretary of Defense Robert Work made artificial intelligence and autonomy the centerpiece of his “Third Offset Strategy.” Current Defense Undersecretary for Research and Engineering Mike Griffin has called hypersonics his top priority, with directed energy a close second. That would seem to be at odds with his boss, Defense Secretary Mark Esper, who has declared AI his “number one.” Yet the past two years have seen Griffin issue formal guidance about 10 to 13 “priority” technologies, with AI, hypersonics, and nuclear modernization occupying the top spot, depending on the list. But they can’t all be number one, and conflicting guidance from senior DoD leaders isn’t helpful for establishing real priorities in defense spending.
Even with a $106.6 billion RDT&E request for 2021 — the Defense Department’s largest ever — actual priorities are needed. The Pentagon cannot spend heavily on all potentially useful technologies. If Esper and Griffin want to ensure that the Department pursues their top priorities, they need institutional buy-in. And in a long-term technology competition against China, the U.S. military needs stability in its budgetary priorities.
Instead, it has something of a patchwork. In terms of spending, the DoD’s 2021 budget request reflects the undersecretary’s view that hypersonics are his “highest technical priority.” The budget request includes $3.2 billion for hypersonic development, a 23 percent increase from last year’s $2.6 billion request. Hypersonic missiles are valuable because they can sustain speeds in excess of Mach 5 and maneuver in hard-to-predict trajectories, allowing them to defeat current missile defenses. But the result is a valuable tactical capability, not one that revolutionizes warfare. Even once the technology is developed, hypersonic missiles are unlikely to be affordable enough to be anything other than an exquisite silver bullet for the highest priority targets. Hypersonics undoubtedly have value but are unlikely to change most military operations even against near-peer competitors.
The Pentagon’s current focus on hypersonics is largely a reaction to recent Russian and Chinese advances. The U.S. should compete in the field, but matching adversary investments does not explain its strategic value. Nor does it necessarily deter them, defend against them, or build pathways to fight and win.
Every defense leader is entitled to his or her priorities. However, Griffin has not articulated a clear vision for why hypersonics should be DoD’s No. 1. Instead, he has said, “I’m sorry for everybody out there who champions some other high priority, some technical thing; it’s not that I disagree with those. But there has to be a first [priority], and hypersonics is my first.”
If Griffin is right and hypersonics – along with directed energy – should be the DoD’s top priorities, then sharing his rationale would help secure institutional buy-in for the kind of long-term sustained investment needed to achieve a robust advantage. Regardless of which technologies are number one, DoD needs a transparent process to bring the institution along. DoD needs a technology strategy, not a list of buzz words. That strategy must first come to grips with the reality that DoD occupies a very different technology landscape than it once did.
U.S. military superiority today is built on a first-mover advantage in the information revolution. Defense investments in the 1960s and 70s laid the seeds for the microprocessor revolution, origins of the internet, Global Positioning System, precision-guided weapons, stealth, and other digital technologies that have transformed commercial industries and warfare. Yet the awkward reality is that the U.S. defense community no longer has the power it once did to shape the arc of military technology. In 1960, the U.S. Defense Department alone accounted for 36 percent of total global R&D spending, both military and non-military. By 2016, the combination of reduced federal spending on R&D and increased spending by other nations had driven DoD’s slice down to 3.7 percent of global R&D. DoD simply does not get to determine anymore which technologies will shape the future of warfare. DoD can make bets in key technologies, but these bets are tiny compared to the magnitude of technological forces that are exogenous to DoD. The challenge defense leaders face is not picking winners, but accurately seeing global technology trends and adapting.
The dominant technology trend today is the information revolution, which is sweeping in its scope and scale. Global spending on information technology is more than double global military spending. These investments have led to exponential growth in data, networking, bandwidth, storage, and computer processing power. Even as semiconductors approach the atomic limit and the rate of growth of raw chip hardware slows, the machine learning revolution has pushed computing power to stratospheric levels. Since 2012, the amount of computing power used in cutting-edge machine learning research has increased by 300,000-fold, a doubling every 3.4 months, compared to a 24-month doubling prior to 2012. Few technology areas are accelerating as rapidly as machine learning. Missiles aren’t 300,000 times faster than they were seven years ago; armor isn’t 300,000 times lighter.
DoD will have to adapt to the reality that these broader technology trends are shifting the art of the possible. Digitally enabled military capabilities are the ones most likely to see dramatic changes in the coming decades: cyber, electronic warfare, autonomy, robotics, AI, networking, sensors, command-and-control, data fusion, and biotechnology. DoD will need to invest in military-specific technologies like hypersonics and directed energy when they have clear operational advantages precisely because these are not being advanced by the commercial sector. But when envisioning how technology is likely to change the future of warfare, it is the information-based elements of the kill chain: sensors, data processing, networking, command-and-control, targeting, and decision-making that are likely to be the most consequential for achieving military advantage.
Air Force Chief of Staff Gen. Dave Goldfein recently outlined a vision for how these information-based technologies would work in practice. Last year, U.S. forces demonstrated a near-fully automated sensor-to-shooter kill chain. A satellite detected an enemy ship, then automatically cued an airborne surveillance asset to get higher quality target ID, which then passed data to a command-and-control asset that automatically selected a naval destroyer to execute a strike. A human remained in the loop to decide whether or not to authorize the strike, but the rest of the information processing chain was automated. This is precisely the kind of innovation DoD needs, working to spin-in and adapt information-based technologies to achieve military advantage.
Warfare is changing, and whether we like it or not DoD has only a limited ability to influence which technologies will be most consequential in future wars. What DoD can affect is its ability to adapt to a rapidly shifting technology landscape. To do so, the Department needs a technology strategy, not buzz words and conflicting guidance from senior leaders. To succeed in a long-term competition with China, the Department needs a transparent process to set spending priorities that can be sustained over time, outlasting individual leaders. And while there are many important technologies, the Department’s focus should be harnessing the exponentially growing digital technologies coming out of the commercial sector to change military operations faster than potential adversaries.