But the most important experiment was Carter and Rosenbach’s creation of DIUx. Ultimately it placed about fifty officers and civilians in the heart of Silicon Valley with the explicit mission of finding existing cutting-edge technologies that could be rapidly given to war fighters, Special Forces operators, marines, or army personnel.
Shah and Christopher Kirchhoff, a veteran of the Obama White House and the office of the secretary of defense, were brought in as lead “partners”—the word indicating that the Pentagon was intent on blending in with the Silicon Valley natives. They were a good mix, but Shah knew from his own experience how hard it would be to get companies in Silicon Valley to participate in any effort sponsored by the Pentagon. First, there was the post-Snowden hangover. But there was also a very practical fear of getting wrapped up in the deadening, regulation-heavy Pentagon bureaucracy. “When I was getting a business start-up going, my investors wouldn’t let me even talk to the military,” he said. “They complained it took years to close a deal, and even longer to get paid. You’d go bankrupt just waiting for them to make a decision.”
As with any new venture, Shah needed cool office space. But while the Pentagon was ready to be a player in Silicon Valley, it wasn’t ready to pay the sky-high rents. So DIUx commandeered space in an old military building in Mountain View, just on the edge of the Google campus. Autonomous Google cars buzzed past the headquarters constantly.
But what DIUx really needed were some quick wins—off-the-shelf products that could save lives and demonstrate to the Pentagon that its old, slow acquisition methods left troops vulnerable and made war fighting harder than it needed to be. “The key is using technologies that are already available,” Shah told me, “and making the modifications we need for a specific military purpose.”
Soon his offices were filled with possibilities. Among the most promising first finds was a quad-copter—a tiny, autonomous helicopter that had been developed to help construction companies. The devices fly indoors and up stairwells to confirm, with laser measurements, that every wall has been built to precise specs. Upon seeing the copter, Shah and his team suspected it could prove immediately useful in Iraq and Afghanistan. When Special Forces were getting ready to clear out a house full of suspected militants, they needed to know where everyone was—and whether the place had been booby-trapped with explosives. Though it took longer than it should have, Shah got prototypes into the hands of troops on the ground.
But the most urgent project—and the one that ran afoul of the Pentagon and its biggest and most powerful contractors—involved launching a new kind of tiny satellite over North Korea to keep track of the missiles Kim Jong-un was placing on hard-to-track mobile launchers. The giant spy satellites the United States has relied on for decades were a prime example of how the Pentagon was dependent on technology that was outdated the day it was launched: they cost billions, took years to design and build, and were kept in orbit for years, until they fell out of the sky.
Shah and Kirchhoff envisioned something completely different: tiny, backpack-sized, inexpensive civilian satellites that had been developed to count cars in Target parking lots and monitor the growth of crops. They were launched in clusters, and would stay in orbit just a year or two. But they were also so cheap that when they fell out of the sky, the Pentagon could simply launch the newer, higher-resolution replacements. This would likely provide the kind of coverage needed to execute a new military contingency plan called “Kill Chain,” in which satellites would identify North Korean launches, or nuclear facilities, and provide the targeting data to destroy them preemptively if a conflict seemed imminent.
The urgency arose from the fact that the American coverage of North Korea from space was (and remains) terrible—the United States had eyes on the country less than 30 percent of the time. (The exact figure is classified.) William Perry, the former defense secretary, told me that if the North Koreans rolled out one of their new missiles, “there’s a good chance we’d never see it.”
“Kim Jong-un is racing—literally racing—to deploy a missile capability,” Robert Cardillo, the director of the National Geospatial-Intelligence Agency, told me in mid-2017. “His acceleration has caused us to accelerate.”
But the Pentagon was still not moving fast enough. The North was adding missiles and missile launchers, which could be hidden in caves and rolled out minutes before launch, faster than the defense bureaucracy could stomach the idea of letting small firms like Capella Space Corporation, a start-up named after a bright star, threaten the traditional contractors. Capella was one of a number of small satellite firms that had slashed the cost of space-based radars that can see through clouds, rain, snow, camouflage, and foliage, and pick out changes in the ground elevation that can point to hidden tunnels. Just ten years ago, building a constellation of those satellites was estimated to cost more than $94 billion. Now, Shah believed, it could be done for tens or hundreds of millions. It seemed like a no-brainer, especially with the warnings that the North was investing in solid-fuel missiles that could be rolled out of caves quickly, and launched in hours or minutes. Early detection was key.
Yet as Shah traveled back and forth between Silicon Valley and Washington, trying to get the satellites launched on an accelerated scale, he ran into one impediment after another. Members of Congress didn’t like the way DIUx was using a loophole in acquisition rules to short-circuit the usual, tortuous process of getting R&D projects approved. The big satellite makers, while faking enthusiasm, felt their multibillion-dollar contracts might be threatened by start-ups they had barely heard of. They were a powerful lobby.
“We’re trying to do something very different,” Shah told me with considerable understatement. “And that always ruffles people.”
The bureaucratic reluctance in Washington to move at entrepreneurial speeds was bad enough. But as Shah and Kirchhoff moved around the Valley, exploring possible new technologies, they kept running into a competitor with the same vision, less reluctance, and a much bigger budget: it turned out the Chinese had an informal kind of DIUx of their own under way—on American soil.
“I don’t think anyone realized the extent of it,” Kirchhoff told me one evening. “In the old days the Chinese bought up companies,” until the United States began to reject many of those deals on national-security grounds. So the Chinese were taking another path to the same destination. They opened up their own venture-capital funds in the Valley, as well as investment funds that could take a minority position in new companies—not big enough to trigger a review, but enough to get an early understanding of the technology.
But no one could quantify China’s involvement—there were no concrete numbers that gave a holistic sense of the Chinese investment strategy. So Shah and Kirchhoff found a savvy executive who understood Silicon Valley—Michael Brown, who had run Symantec, the company that identified so many of the details of Stuxnet—and asked him to write an unclassified report on it. They hoped this would wake Washington up to the new ways that the Chinese were scooping up Silicon Valley’s most critical technologies for both their state-owned companies and their own military. Brown quickly joined forces with Pavneet Singh, a former National Security Council and National Economic Council staffer who had prepared Obama for his meetings with President Xi.
“The first thing I learned is that the Chinese have done a beautiful job mirroring Silicon Valley,” Brown told me. Baidu is the answer to Google. Alibaba is the Chinese Amazon. Tencent is known for apps used by two-thirds of all Chinese for messaging and money transfer, apps like WeChat and QQ. By one venture capitalist’s measure, the Chinese spend 1.7 billion hours a day on Tencent apps.
But the second thing Brown learned was that the Chinese were essentially doing what DIUx was doing—investing in “early-stage” companies. They were just doing it on a far larger scale than the Pentagon had in mind.
What made the strategy so brilliant was that the Chinese were
flying in under the radar. When they bought an entire company, it triggered an official review in Washington, from a little-known, little-understood group called the Committee on Foreign Investment in the United States. It could recommend that the president block any sale on national-security grounds. And both Obama and Trump did so, several times.
But there were no rules about taking a minority stake in a company, which would give the investors an early, privileged look at the technology. This alarmed Brown, who knew that the Chinese investors were getting a detailed look even at companies and technologies that they turned down for investment. Nobody had ever imagined this possibility when the foreign-investment rules were written decades ago: venture capital barely existed, and it was unimaginable that the Chinese would be players in the game.
“People keep saying to me, ‘You are overreacting—remember Japan,’ ” said Brown, referring to the time in the late 1980s and early ’90s when some feared that the United States would become a “techno-colony” of the Japanese. But Brown became convinced this was different. “Japan was a staunch ally. It was never a military rival in that time. It never had a chance of challenging the US economically. And we had shared values.” In China’s case, none of that applies. (It is a sign of the change in economic fortunes that Japan, the country the United States worried most about in the ’80s and ’90s, provided only $13 billion of the venture capital in the DIUx charts—only a little more than half of what the Chinese were spending.)
The DIUx report’s findings, which began circulating confidentially around Washington in the spring of 2017, were astounding. They demonstrated that even while the Chinese were paring back on stealing the fruits of American industry—Obama’s agreement with Xi had begun to have some effect—they had found many perfectly legal ways to invest in it. A government that still gave lip service to communism had figured out venture capitalism—and concluded it was the shortest path to get the technologies the country needed.
The numbers that Brown and Singh gathered, all from public sources, told the story. China participated in more than 10 percent of all venture deals in 2015, the report found, focusing on early-stage innovations critical to both commercial and military uses: artificial intelligence, robotics, autonomous vehicles, virtual reality, financial technology, and gene-editing. When they broke down who was investing in US-based venture-backed companies between 2015 and 2017, American investors ranked first, with $59 billion in investment. Europe was second, with $36 billion. And China was right behind, with $24 billion.
Some of the biggest direct investments came from Baidu and Tencent, but there were also a surprising number from venture-capital firms with Western-sounding names—West Summit Capital and Westlake Ventures—that were wholly Chinese-owned. “They are private actors,” Brown told me, “but always acting with the approval of the Chinese government.”
Xi Jinping made it clear he was completely behind the strategy by sending subtle messages. In the fall of 2017, at the 19th Party Congress, Xi talked about how he was targeting these strategic areas. In his New Year’s speech in 2018, Xi had books on artificial intelligence placed strategically behind him, so they would be caught on camera. Chinese investors in the Valley didn’t need to see the speech to get the message: they did eighty-one deals in American artificial-intelligence companies between 2010 and 2017, worth $1.3 billion. More than a third of that—over $500 million—was spent in 2017 alone.
Brown and Singh’s DIUx report was soon in the hands of Gen. Paul Selva, who held the vice chairman post at the Joint Chiefs of Staff, the job once occupied by James Cartwright. General Selva had encouraged the study and used it to sound the alarm inside the Pentagon. But the report arrived in the early days of the Trump presidency, and rather than serve as a call for the United States to think in Chinese terms about how best to invest in research and development—and how to integrate those investments with defense projects—the report became another excuse for Trump’s calls for protectionism. The Trump economic team somehow convinced itself that it could cut the Chinese off entirely from investing in sensitive new technologies in the United States—even while China held $1.2 trillion in US debt.
By early 2018, Trump was looking for new ways to block Chinese investment, and even non-Chinese investment that might help Beijing. He stopped a Singapore company, Broadcom, from buying Qualcomm, a maker of vital but specialty chips that are used by US Special Forces, among other military units. The stated fear was that Broadcom, while not itself Chinese, would not invest heavily in research—and Huawei and other Chinese firms would benefit.
Anthony Balloon, an international lawyer concentrating on China, told my Times colleagues that this was a turning point in a growing technology war: “There is now a recognition in government that foreign investors, particularly from China, are getting more and more sophisticated on how they get access to technology in the US.” The message was clear: the United States would look anew even at minority-stake investments and other forms of Chinese capital.
And American officials became explicit about banning Chinese technology that they believed could give Beijing a back door into US networks. In March 2018, when Paul Nakasone was finally nominated to run the NSA and US Cyber Command, he told Congress with a smile that he would never even use a Huawei phone; around the same time, Best Buy stopped selling them. What Nakasone didn’t say was that even though Huawei had emerged as the world’s largest builder of networking equipment—and was wiring most of Asia and some of Europe—the NSA had quietly banned AT&T and Verizon from even allowing Huawei to bid on building parts of America’s 5G network. The companies, and even some in the intelligence agencies, argued that this was shortsighted: if Huawei bid, they pointed out, it would have to allow the United States to pick apart the details of how its networks are built. American officials shrugged and said no.
It was a remarkable progression. What had begun early in the Obama administration as a fear that China was using cyber techniques to steal American technology for the benefit of its state-owned companies had turned into a much larger technological cold war. The Chinese weren’t stealing as much. Instead, they were buying into America, perfectly legally. And the United States was struggling to figure out how to stop them without rejecting the principles of a free, global market.
* * *
—
The real warning contained in the DIUx document was not about what the Chinese were doing in Silicon Valley, but about what they were working on at home.
The Obama-era battles over the theft of intellectual property captured the headlines, and in the spring of 2018 President Trump revived the issue as part of his justification for the threat of massive tariffs against Chinese goods. While the pace of theft had lessened, without question many of China’s large and small firms were still looking to acquire research, development, and product blueprints by any means—legal or illegal. But the equally big concern was China’s focus on becoming the world’s leader in artificial intelligence by 2030, and its huge investments in the technology it needs to accomplish that task. A single, massive campus in Hefei, in Anhui Province, captures the scope of that effort: there, China is constructing a $10 billion research hub called the National Laboratory for Quantum Information Sciences, the centerpiece of its quantum computing effort.
Quantum computing—the ability to take calculations to warp speed by using photons instead of manipulating ones and zeros the old-fashioned way—holds the key to cracking any encryption by brute force. If successful, it could result in developing secure communications links, and navigation systems that do not rely on global positioning satellites, which can be jammed by an adversary or used to locate a hidden submarine. The Chinese have already tested quantum satellites.
“The question is, how should the United States respond to this challenge?” Robert Work, the former deputy secretary of state who pushed the Pentagon into competing in this arena, told my colleague Cade Metz. “
This is a Sputnik moment.”
Work’s analogy captured a critical truth: any breakthroughs produced by the concentrated Chinese effort will flow directly to the country’s military might. An equivalent to the Silicon Valley/Washington divide, which bubbled along before Edward Snowden’s disclosures and re-erupted in the battles over the government’s effort to get a back door into encrypted systems, does not exist in China.
Yet in the United States, the divide is widening. The Cold War model, in which breakthroughs in American military technology and the space program flowed to the commercial sector, is gone forever. The reverse model—using the skills of Silicon Valley to create the next-generation weapons—has run headlong into political and cultural opposition.
“Even if the US does have the best AI companies, it is not clear they are going to be involved in national security in a substantive way,” said Gregory Allen of the Center for a New American Security. The effects are already visible: the military edge the United States has grown accustomed to since World War II is eroding.
The Perfect Weapon Page 30