Vice Admiral Kevin M. Donegan speaks with Intelligence Specialist First Class Richard Hull on board the guided-missile cruiser USS Anzio (CG-68).
When then–Chief of Naval Operations Admiral Gary Roughead issued the memorandum formally establishing the Information Dominance Corps (now Information Warfare Community) in November 2009, he quoted Spanish explorer Hernán Cortés at Veracruz: “We’ve burned the boats . . . there’s no going back!” This is often seen as the act that led to Cortés’ historic defeat of the Aztec Empire in 1521. His men, as Cortés later put it, “then had nothing to rely on, apart from their own hands, and the assurance that they would conquer and win the land, or die in the attempt.”1
Today, the Cortés anecdote is a metaphor for bold, decisive action necessary to take organizations through fundamental change to new heights. It is problematic, however, because it confuses the motivation for the plan with its actual execution. Burning the boats did not advance Cortés’s men one step toward Tenochtitlán, the Aztec capital. The Spanish still had a long way to go—there were brigantines to build, alliances to form, and numerous battles to be won in the two-year campaign. After the boats were burned, Cortés’s men moved off the beach and got to work.
Nearly ten years into its time in the Information Warfare Community (IWC), naval intelligence has not “left the beach” with a sense of urgency to acquire and field cutting-edge systems that will vault the community into the era of big data and human-machine pairing. Instead, it largely has remained complacent while watching dramatic change occur in the information domain. The past decade has witnessed the emergence of mass digitization, artificial intelligence, robotics, and rapid technological change: the big data era. Yet naval intelligence persists in using the same tools, people, and tradecraft as in 2009. In a global security environment where “margins of victory are razor thin,” this must rapidly be addressed.2
The 2018 National Defense Strategy (NDS) challenges naval intelligence to play an integral role in the “lethal, resilient, and rapidly adapting Joint Force.” To do this well, naval intelligence must make priority investments in its people and big data–enabled technologies. And, perhaps most important, it must experiment with human–machine pairing, the most promising development for analytical tradecraft. The community has a proud history of out-thinking, outmaneuvering, out-partnering, and out-innovating adversaries. But naval intelligence cannot rest on its laurels in a renewed drive for prominence.
Maintain a Technological Advantage
Spanish conquistador Hernán Cortés burning his ships before the conquest of Mexico.
While Cortés is best known for burning his ships, he actually just scuttled them. He should be known for repurposing the materials from his ships and using them in innovative ways. Cortés salvaged parts and turned them into brigantines of his own design. These could be either rowed or sailed, and they were fitted with heavy cannons and could carry as many as 75 people.3 Over the course of the two-year Aztec campaign, Cortés’s brigantines conducted reconnaissance missions, evacuations, amphibious assaults, and mobile firepower strikes. These innovative multipurpose vessels were the “key to the war.”4
Today, instead of Cortés’s ships, intelligence professionals have information systems. Using these systems, naval intelligence receives and analyzes data and then creates and distributes intelligence products. Information systems are its “key” to victory. And yet, by comparison with cutting-edge collaboration software in the civilian sector, naval intelligence lags behind.
Naval intelligence seems to be proving Microsoft co-founder Bill Gates’ contention that there is “always an overestimation of change in the next two years and an underestimation of it in the next ten.”5 When it became part of the new IWC, naval intelligence rightly emphasized the growing importance of information in warfare, but underestimated the dramatic change occurring in the information domain itself. Today, exceptionally large sets of structured and unstructured data can be analyzed to reveal patterns, trends, and associations pertaining to human behavior. This is the very type of innovation that naval intelligence should have pursued aggressively. It has yet to adapt to this dramatic change.
Naval intelligence’s current information systems do not handle big data analytics very well. Rather than speaking the language of advanced analytics—nonrelational data stores, predictive modeling, social web, data mining, and forecasting—the community’s systems are stuck in the past of structured query language, stovepiped relational databases, and site-by-site trawling through individual command webpages. As the NDS warns, “we cannot expect success fighting tomorrow’s conflicts with yesterday’s weapons or equipment.”
Much like Cortés’s multimission brigantines, big data–compatible information systems offer many advantages. The technology can yield “superior information and unparalleled insights” into human behavior and improve decision-making.6 Intelligence products often are created and distributed with little objective awareness of their usefulness. With the appropriate advanced analytic software, naval intelligence could discover how many times and by whom a product was downloaded, read, forwarded, cited, and incorporated into other products. Just as companies such as Amazon and Netflix have done, naval intelligence could use this consumer-driven information to continually improve products and services while gaining efficiencies and eliminating redundancies.
In 2011, the consultancy firm McKinsey called big data “the next frontier for innovation, competition, and productivity.”7 Commercial companies that embraced these analytics early are now at the forefront reaping profit and market control. If naval intelligence does not make this a priority investment, it risks allowing the nation’s adversaries to gain a dominant position.
The New Analyst
One of the first things Cortés knew he needed to do after moving off the beach was raise a force with numerical parity to Montezuma’s. More important than superior technology, Cortés’s “people” strategy was at the heart of the successful invasion and challenge to Aztec dominance. First he brought in the Totonacs, a tributary tribe chafing under Montezuma’s rule.8 Next were the Tlaxcalans, a large tribe that led a military federation of other cities displeased with the Aztecs.9 Ultimately Cortés went from an army of hundreds to one of hundreds of thousands. Without a large and motivated force of skilled warriors, he could not have waged one of the most impressive military campaigns in history.
During the past decade, some in naval intelligence seem to have concluded that what is needed most is more information, not more and better trained people. With new collection platforms, sensors, and associated data links, the amount of information entering work centers is rising exponentially. Increasing the intake, however, does not ensure a qualitative improvement in analysis because, as statistician Nate Silver points out in his 2012 book The Signal and the Noise, “There isn’t any more truth in the world.”
Yet more data is not just useless “noise”—far from it. More information means more opportunities to uncover the “signal,” or truth. What matters in this era is having the tools to sift through the immense increase in information and continue to analyze it as well as naval intelligence always has done. Human analysts matter now more than ever. Despite all the advances in technology, there is no program, model, or algorithm that can accurately assess the future. Well-trained analysts comfortable in the language of big data know when systems or models have flaws.10 It is people who will make this era work for naval intelligence. But people with the right training and skills remain scarce.11
Far too much effort has been devoted to administratively adapting naval intelligence to the bureaucratic structure of the IWC, and too little has been invested in recruiting, training, and cultivating the types of analysts needed. Just as the nature of the information domain has changed, so too must the composite skill set of the naval intelligence professional. Naval intelligence needs more critical thinkers who can separate signal from noise and transform information into knowledge. Analysts need to be more than information warriors—they need to be knowledge warriors.
Time for a Tradecraft Breakthrough
Cortés’s well-trained soldiers were armed with the latest in military technology—steel weapons, horses, and ships. They applied a warrior tradecraft with lethal proficiency. In early battles against the Otomis and Tlaxcalans, Cortés’s army of “well-schooled divisions and disciplined ranks” used “crossbow volleys,” “well aimed artillery,” and “cavalry charges” to defeat the large armies.12 At the Battle of Otumba, his men, numbering only 1,000, were caught out in the open by a larger Aztec force. Despite the mismatch, victory came as a direct result of “swift and skilled warhorses and the strict defensive discipline of the Spaniards.”13 Advanced technology with highly skilled people were decisive.
Today, wise investments in technology should be informed by naval intelligence’s expertise in the tradecraft of analysis. The community’s history is steeped with examples of well-honed analysis leading to success. From victory at Midway in 1942 to the undersea contests of the Cold War, superior analysis has proved decisive for the Navy. To support a more lethal and agile joint force, naval intelligence will need to continue what it always has done—provide timely and accurate operational intelligence to commanders. In today’s dynamic information domain, analytical tradecraft must improve with the help of some promising technologies.
Analytical Tradecraft Today
From increasing use of Bayesian inference to applying the recommendations of Professor Philip Tetlock’s Good Judgment Project, where teams of amateurs beat intelligence professionals at prediction, several potential improvements to analytical tradecraft are worth examining in greater detail.14 Perhaps the most promising is human–machine pairing—combining the most advanced technologies with the best analysts.
Within the artificial-intelligence and robotics community, researchers consistently find that machines are good at the things humans are weak at, and vice-versa.15 Known as Moravec’s Paradox, this phenomenon manifests most clearly in games. Computers long ago beat human chess champions. Even in games such as Go that until recently were thought to be so complex a computer could never calculate the optimal moves, computers now convincingly beat the best players in the world. At first glance this may seem disheartening to those who have argued that people remain vital in the field of analysis.
Chess champion Garry Kasparov famously lost to the computer Deep Blue in 1997. Instead of fading into history, however, Kasparov committed himself to exploring what exactly gave computers the edge over humans at chess. He eventually devised an experiment to examine what happens if humans and machines play chess as partners instead of as competitors. The experiment, known as Advanced Chess, led to a significant discovery. While a computer regularly beat a human chess champion, a human paired with a computer always beat a computer playing alone. Kasparov determined that “human strategic guidance combined with the tactical acuity of a computer was overwhelming.”16 In other words, human–machine pairing was decisive.
Perhaps more relevant to intelligence analysis, two amateur American chess players paired with a superior computer and beat a grandmaster paired with an inferior computer. Kasparov concluded, “it was a triumph of process . . . weak human + machine + better process was superior to a strong human + machine + inferior process.”17 The process is the most important aspect when it comes to leveraging the power of technology. When a person is paired with technology and uses a superior process (tradecraft), the benefits for intelligence analysis are clear. These results and the relatively low cost to enable human–machine pairing should compel naval intelligence to immediately experiment with this emerging field.
As the Advanced Chess experiment demonstrates, artificial intelligence does not spell the end of the human analyst. With Moravec’s Paradox in mind, human creativity and curiosity are of paramount importance in the big-data environment. Computers are great at finding answers. They are terrible at asking good questions. Paired with machines built to find answers, the best analysts will know which questions will produce increased knowledge.
Get Off the Beach
Over the course of the two-year Aztec campaign, Cortés gained control over a territory that became the Spanish Empire’s largest addition ever secured by an individual.18 He succeeded through bold, innovative, and skillful actions over a very long campaign.
Despite the IWC’s founding memo stating it would “supplant an older order with new structures and more efficient processes to deliver transformational, game-changing warfighting capabilities,” naval intelligence wrongly assumed in 2009 that the information domain would not evolve dramatically. Naval intelligence should, as complexity theorist Alvin Saperstein suggests, “always be contemplating the future . . . always include attempts to change the field of endeavor itself.”19 This includes urgently embracing big-data technology and human–machine pairing to improve analytical tradecraft, innovative advances that will transform intelligence analysts into knowledge warriors.
As Niccoló Machiavelli wrote in his 1513 work The Prince, “there is nothing more difficult to take in hand, more perilous to conduct, or more uncertain in its success, than to take the lead in the introduction of a new order of things.” Difficulty does not excuse complacency. What were just potential threats a decade ago are today quite real. While naval intelligence may not have the technological advantages it once enjoyed, it still retains a superior workforce and culture. It is time to move off the beach and get to work!
1. Hugh Thomas, Conquest (New York: Simon & Schuster, 1993), 223.
2. ADM John M. Richardson, USN, “A Design for Maintaining Maritime Superiority” (Washington, DC: 2016), 8.
3. Buddy Levy, Conquistador (New York: Bantam Books, 2008), 134.
4. Levy, Conquistador, 277–79.
5. Bill Gates, Nathan Myhrvold, and Peter Rinearson, The Road Ahead (New York: Viking, 1995).
6. Phil Simon, Too Big to Ignore (Hoboken, NJ: Wiley, 2013), 23.
7. James Manyika, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles Roxburgh, and Angela Hung Byers, “Big Data: The Next Frontier for Innovation, Competition, and Productivity” (McKinsey Global Institute, May 2011), 1.
8. Thomas, Conquest, 207.
9. Thomas, Conquest, 238–39.
10. Nate Silver, The Signal and the Noise (New York: Penguin, 2012), 123–25.
11. Thomas H. Davenport and Jeanne G. Harris, Competing on Analytics (Boston: Harvard Business School Press, 2007), 131.
12. Levy, Conquistador, 76.
13. Levy, Conquistador, 198.
14. Must-reads on Bayesian thinking and the Good Judgment Project are Silver’s The Signal and the Noise and Philip Tetlock’s Expert Political Judgment (Princeton, NJ: Princeton University Press, 2017) and Superforecasting (New York: Crown, 2015).
15. Garry Kasparov, Deep Thinking (London: John Murray, 2017), 244.
16. Kasparov, Deep Thinking, 246.
17. Kasparov, Deep Thinking, 246.
18. Levy, Conquistador, 327.
19. Alvin M. Saperstein, “Complexity, Chaos, and National Security Policy,” in Complexity, Global Politics, and National Security, David S. Alberts and Thomas J. Czerwinski, eds. (Honolulu: University Press of the Pacific, 1997), 124.