On August 28, Deputy Secretary of Defense Kathleen Hicks announced what she called the “Replicator” program, an all-out effort to build autonomous vehicles by adding artificial intelligence, autonomous driving, and relatively cheap weapons and equipment to modernize the U.S. arsenal. She describes the machines as “expendable,” meaning they can be subject to attrition without affecting the mission. Imagine hundreds or even thousands of drones communicating with each other as they gather intelligence on enemy movements, and you begin to understand the Under Secretary of State’s vision for Replicator. Even if a significant number of drones are shot down, the information they collect will be recorded and sent back to operators on the ground.
In one sense, Hicks’ announcement during a speech titled “The Urgency of Innovation” at a National Defense Industrial Association conference does not represent a completely new approach. For example, five years ago, the National Defense Strategy already called for significant investments in artificial intelligence, stating that “we cannot expect to successfully respond to tomorrow’s conflicts with yesterday’s weapons or equipment.” Since then, the Department of Defense has made significant investments in artificial intelligence. Billions of dollars have been spent; last year alone, it allocated nearly $900 million to support nearly seven hundred AI projects. Still, as Hicks points out, many such technologies end up stuck in the so-called valley of death—never adopted, even if they are successful in the lab or in the field. Her audience included numerous military contractors, whom she called on to “out-innovate our competitors.”
But in another way, the Replicator program is radically different from what the Department of Defense has always done. It aims to accelerate the invention of military technologies that will transform the way America wars and deters. Hicks claimed that Replicator will “deploy consumable autonomous systems at thousands of scale across multiple domains over the next eighteen to twenty-four months.” As she envisions them, these systems will be “thrown into space, many at a time.” Swarms of small, solar-propelled ships equipped with sensors to trawl the ocean and deliver real-time intelligence; and “swarms” of aerial drones, some conducting surveillance and others carrying weapons. Rather than focusing DoD resources on extremely expensive and complex equipment that must operate for decades to justify the cost, the goal of Replicator is to field equipment with a much shorter shelf life that allows for continuous innovation in technology.
“Throughout my career, military strategy has been to build these elaborate and expensive systems that are very effective,” said Chris Gentile, a retired Air Force lieutenant colonel who is now a vice president at EpiSci, a Defense contractor developing autonomous driving technology. The system told me. “I flew stealth fighters in the Air Force. I flew the F-22. It was a great airplane, but we only bought one hundred and eighty-seven.” (Just last week, era reports on the Navy’s continued investment in large, unwieldy warships that are often obsolete before they are even built. ) “As we look to the future, we just don’t think this economic model is going to work — so we’re going to increase the quality, just the quantity, of what we’re able to bring into theaters.”
Thomas Hamilton, physicist RAND Corporation The Corporation, a think tank that often conducts research for the U.S. Armed Forces, calls this approach iPhoneomics. “The idea is that while it costs billions of dollars to develop software, it costs very little to make each iPhone. From a military perspective, if I send out a lot of things that are cheap to make but expensive to software, there will be Huge advantage – so if the enemy shoots them down, they won’t destroy any of my expensive software.”
Over the years, Hamilton and colleagues RAND Corporation, has always advocated the collaborative use of cheap aircraft. He remembers watching the cost of microprocessors drop and capacity increase until the idea of swarms of drones, which the military had been discussing for more than a decade, seemed feasible. “You could have ten, twenty or fifty drones flying over the same transport, taking pictures with cameras. And, when they determine it’s a viable target, they’ll send the information back to Pearl Harbor or Operators in Colorado or somewhere else,” Hamilton told me. The operator would then order an attack. “You can call it autonomous because not every plane is piloted by a human. But eventually someone pulls the trigger.” (This follows the Department of Defense’s policy on autonomous systems, which is that there is always someone “skin in the game.”)
The war in Ukraine inadvertently proved that many small drones can overwhelm an advanced combat force. As Russia rolled out missile systems designed to shoot down expensive Western aircraft and munitions, Ukraine launched cheaper drones controlled by human operators to take them out. Drones give the Ukrainians a tactical advantage while also allowing them to do more with less. “You’re launching a $250,000 missile at a $40,000 drone,” Gentile said. “The goal here is to win the cost war – make sure that at the resource level and at the funding level, you show a positive attitude towards the deal.” Individual Ukrainian enthusiasts also fly small quadcopters (some of which have a wingspan of no more than 12 centimeters). Crash and destroy Russian weapons. Ukraine is losing an estimated 10,000 drones a month but is still able to continue the war — a real-time demonstration of expendable technology.
Hicks made several references to the People’s Republic of China (PRC) in his statement, and as Hamilton told me, there is no question that China is “a threat to us”—the adversary whose aggression the United States most fears. “Replicator is designed to help us overcome the People’s Republic of China’s greatest advantage, which is mass. More ships. More missiles. More people,” Hicks said.A week later, by defense news” she asked, before responding to a question posed by the timing of the Replicator program: Is there new intelligence suggesting an imminent attack on Taiwan or U.S. allies in the Indo-Pacific? While she said there were not, Replicator’s 2025 goals are consistent with some Military analysts believe the timing of a possible Chinese attack on Taiwan corresponds to that. “We have to make sure that the leadership of the People’s Republic of China wakes up every day, considers the risk of aggression, and concludes, ‘Today is not the time,'” Hicks said. Not just today, but every day, now and for the foreseeable future. “
Perhaps the biggest obstacle to the success of replicator programs in the Pacific is the mismatch between geography and technology. Currently, Ukraine’s small drones have limited range and power, and most are either made in China or use Chinese components.In contrast, U.S.-made surveillance drones such as Northrop Grumman’s Global Hawk are being NATO The troops in the Ukraine conflict cost millions of dollars and are about the size of manned aircraft, in part because they can stay in the air for more than a day. Even a mid-sized drone can cost over a million dollars. Neither is likely to be considered “expendable”. So until U.S. companies or the government itself can produce relatively cheap drones that can fly long distances, China will have the upper hand in the air. As Staci Pettijohn, director of the defense program at the Center for a New American Security, a nonpartisan think tank, told me, “It’s now easier for Chinese ships to spit out some small drones, or some drones.” Their ground forces have them,” she said. “You would expect to see them filling the battle space. She said the drones could send intelligence back to China, and the Chinese military has long-range missiles that can reach Taiwan.
Aerial swarms of drones are just one part of this new initiative. The U.S. Navy is developing a fleet of small unmanned vessels, some no bigger than children’s toys, as well as submersibles that can locate and neutralize underwater mines. It is also seeking a deal with the Air Force so each service can control the other’s combat drones. The Army has long used autonomous ground vehicles and is now testing a four-legged “robotic dog” armed with a six-and-a-half-millimeter rifle that can be fired by soldiers from the sidelines. The Air Force aims to build a thousand “robotic wingmen” to assist manned aircraft. Hicks also envisions launching thousands of “smart satellites” that would use artificial intelligence to navigate and track adversaries. Whether these technologies remain limited to the battlefield or are repurposed in new, unrestricted ways—for example, by increasingly militarized police forces for domestic surveillance—remains to be seen.
Eric Pahon, a spokesman for Hicks, told me that realizing Hicks’ vision for Replicator is a moonshot that will require new ways of thinking at the Department of Defense. “The Department of Defense has been doing things the same way since World War II. Sometimes when we have conflict we innovate. But what we are focused on here is being able to change the culture to keep up with this strategic competition pace while also keeping up with modern technology. We can’t be in a ten to twenty year development cycle. That’s not going to work.” Pahon said that previously the process of submitting the form could take thirty days, now the department This may be shortened to five days. During Hicks’ presentation, she said they have reduced the time it takes to deliver new technology.
Replicator could still end up in the valley of death, especially given the pressures of its accelerated timeline. “When it comes to innovation, you’ll never see the minister or I put out a ‘mission accomplished’ banner,” Hicks told the trade group. But she has started the clock and is counting down. ❖