What Would You Say You Do Here? Redefining the Role of Intelligence in the Information Age
Mike Judge’s 1999 satire Office Space parodies turn-of-the-century white-collar office work at a fictional software company called Initech. In an iconic scene, corporate management consultants referred to as “the Bobs” interview Initech personnel in a search for “efficiencies.” When they sit down with Tom, a middle-aged employee who works in customer management, the meeting quickly devolves into absurdity.
Tom’s job is bringing the requirements of the company’s customers to the software engineers. One of the Bobs asks, “Why can’t the customers just take their requirements directly to the software people?” Because, Tom replies, engineers aren’t good at dealing with customers — but he is. The Bobs assume this means Tom himself brings the customers’ requirements to the engineers, but, in fact, Tom’s secretary does that or, more often, the fax machine. Finally, one of the Bobs deadpans: “What would you say you do here?” To which Tom exasperatedly responds, “Well look, I already told you! I deal with the [expletive] customers, so the engineers don’t have to! I have people skills! I am good at dealing with people! Can’t you understand that?”
Tom’s difficulty justifying his job illustrates one of the more disruptive trends unleashed by the information revolution, as well as an emerging threat to the traditional model of intelligence analysis: the disintermediation of knowledge. Disintermediation means removing the middleman. Intelligence analysts — the members of the intelligence community who “deal with the customers”— risk ending up like Tom if they don’t successfully redefine their role in a changing world.
An Increasingly Obsolete Paradigm
In the slower-paced world of the 20th century, a customer could not access information unless they went through an intermediary like Tom. When information was scarce and knowledge took a long time to accrue, this model made economic sense. If someone wanted to find the answer to a difficult question they used to have to spend hours in a library. It was reasonable, then, to hire an expert to just answer the question for you. This is, in short, how professions came about in the first place.
Consider stockbrokers. Not so long ago, most people had neither the time nor inclination to master the complicated world of financial markets. Those who did were able to lower transaction costs for their customers while charging them a fee. In other words, brokers had a value proposition. But when electronic stock trading applications like E-Trade came along, almost completely removing the need for intermediaries, the halcyon days of plain old “stockbrokers” were numbered.
Similarly, intelligence analysts are the “hub” of the intelligence cycle, serving as an intermediary between intelligence collectors (the sensors and agents gathering information) and intelligence consumers (those who make decisions based on that information). In an almost industrial process, thousands of sharp, dedicated analysts work diligently every day to refine and repackage volumes of “raw” intelligence into a digestible, “finished” form for American policymakers and military commanders, ostensibly to provide them with decision advantage.
Indeed, at the outset this was seen as a feature, not a bug: Gen. Hoyt Vandenburg, the second director of central intelligence, explicitly stated that the then-new CIA “would, like a central assembly line, put [raw collection] and its own work together and present an overall picture in a balanced national intelligence estimate including all pertinent data.” Over time, the assembly line model became routine, its products commodified to the point where they could be dismissed as “CNN plus secrets.”
Although some experts think analysts will remain the “central element in the policy-intelligence relationship,” this proposition is doubtful if the current model is retained. As artificial intelligence (AI) expert Kai-Fu Lee puts it: “Much of today’s white-collar workforce is paid to take in and process information, and then make recommendations based on that information — which is precisely what AI algorithms do best.”
The Promise and Peril of Information Technology
During the Cold War, it was easy to justify the analyst’s role. They had exclusive access to exquisite information provided by expensive technical collection platforms that could peek behind the Iron Curtain. Armed with this unique capability to collect secrets, intelligence analysts were the lens through which this data was viewed. They built their reputation as reliably objective truth-tellers, even in light of several widely publicized failures. The “Cold War Consensus” lent their judgements respectability, if not always uniform acceptance. In effect, the intelligence community held a monopoly.
But disintermediation has been coming for the intelligence analyst for many years. Back in 1989, former CIA Director Robert Gates noted that the creation of the White House Situation Room decades earlier had initiated a great change in how intelligence was provided to the president, in a way that “had yet to be fully appreciated.” Fast-forward 30 years and the cutting-edge information technology that allowed President John F. Kennedy to receive raw reporting from collectors is available to pretty much everyone.
In a data-driven age, the future of this monopoly is seriously in question because transaction costs have been virtually eliminated. People can simply type questions on their keyboard or, increasingly, ask Alexa. As attention spans shrink and meetings expand, analysts spend less time interfacing with their customers, resulting in stacks of unread intelligence reports. Unfortunately, speed trumps quality in the information age.
Accordingly, it is fashionable to be sanguine about the dawn of practical machine learning and how it could upend the profession of intelligence analysis. For instance, many put their faith in search algorithms that can scan billions of words and find correlations instantaneously. But this speed leads many to overestimate the benefits of AI, perhaps because of a widespread confusion about how it works, combined with humans’ inherent difficulty understanding probability. Algorithms are useful tools — but they are greedy, shallow, and brittle. And, as Pentagon CIO Dana Deasy says, without accurate data, “AI is irrelevant.” Unfortunately, the vast majority of the world’s data is anything but.
What They Do Here: The Enduring Importance of Intelligence Analysis
Intelligence analysts who primarily perform the time-intensive and error-prone task of what one professor at the National Intelligence University calls “data sifting,” — think editing spreadsheets and writing situation reports — are indeed likely to be replaced by AI in the next few years. But these analysts are working at the bottom of the knowledge hierarchy, a conceptual model that connects data, information, knowledge, and wisdom. The true utility of analytic thinking is much closer to the top, in the uniquely human realm of cognition.
What is the value proposition of intelligence analysis in an age when information flows freely, answers are only a millisecond away, and algorithms make recommendations? Intelligence does what the fastest information technology and smartest artificial intelligence can’t: make sense of it all.
Sense-making is the process through which organizations understand the world. Organizational understanding emerges from the collaboration of all parties involved — collectors, analysts, algorithms, and the consumers themselves. Forward-thinking intelligence officers have urged the intelligence community to adopt a sense-making model for years, but as we move further into the 21st century, the risk of not making this transition becomes existential.
Soon, savvy political and military leaders will expect to have access to everything, anywhere, at any time, and to be able to call upon any fact or figure almost instantaneously. The language of contemporary knowledge professions reflects this shift. Professionals in the private sector are replacing the dead metaphors of the past — verbs like produce or deliver — with present participles like servicing, sharing, filtering, cognifying, and brokering, all of which imply continuous action rather than a linear process. Intelligence analysts should no longer think of themselves as building a product — AI can do that. Instead, they should recognize that they provide a service.
National Geospatial-Intelligence Agency Director Robert Cardillo says analysts today should be “coherence control officers” in an incoherent world. He told me in an interview earlier this year that the intelligence community “has to be different than Google.” While Google can retrieve answers to questions in milliseconds, it doesn’t provide context, background, or understanding. In a similar vein, Gregory Treverton concludes that intelligence is ultimately storytelling. Intelligence failures happen when the story doesn’t match reality or, if you like, when a new chapter begins but we’re still reading the last paragraph. The analyst’s challenge, then, is to tell the story of reality in a coherent manner that decision-makers understand and internalize.
To stay relevant in the information age, the intelligence community should abandon the product-delivery metaphors of the past and make intelligence a sense-making experience, not a thing that can be packaged and delivered. Intelligence is not the white paper or the slide deck or the overhead satellite image. Intelligence is the experience of understanding something more completely than before. What the analyst does here is facilitate understanding.
Zachery Tyson Brown is a professional intelligence analyst and U.S. Army veteran who graduated this year from the National Intelligence University’s Masters of Science of Strategic Intelligence program, where his thesis Adaptive Intelligence for an Age of Uncertainty was awarded the LTC Michael D. Kuszewski Award for Outstanding Thesis on Operations-Intelligence Partnership. He can be reached on Twitter and LinkedIn.
His views are his own and do not reflect the opinion of the U.S., the Department of Defense, the Defense Intelligence Agency, or the National Intelligence University.
Image: 20th Century Fox
