Messing with the Enemy

by Clint Watts

  The Philippines, where 97 percent of Filipinos have Facebook, may represent the most devastating example yet. Social media armies conducting “patriotic trolling” pushed President Rodrigo Duterte into office in 2016, and they have now been fully co-opted by his regime to attack his enemies and the free press. According to Lauren Etter of Bloomberg Businessweek, Facebook’s stance on the Philippines problem “is simply that the country has come online fast and hasn’t yet learned the emergent rules of the Internet.”6 For social engineers working with authoritarians, they go where opportunity arises, and there are dozens of countries like the Philippines that are ripe for dominating. Facebook now seeks to rapidly make policy adjustments to counter manipulative forces largely beyond its grasp.

  Authoritarian regimes, political campaigns, large corporations—those with time, resources, access to social media activity, purchase information, and emerging technology capabilities—will infiltrate preference bubbles and deliver tailored messages, using tailored messengers, to move audiences to their preferred goals, agendas, history, and identity.

  Social inception is the future. A hidden elite core will social-engineer an unwitting crowd into choosing the policies, politics, and preferences of the hidden elite, all without the crowd’s ever realizing it. Future social media influence will be dominated by those who can aggregate and harness the collective preferences of audiences, deceiving the crowd into willingly selecting the core’s agenda. If social media users believe that the core’s agenda is their own preference, they’ll wholeheartedly support an idea, a politician, or an organization. Social media nations and their members will cast doubt on experts who oppose the hidden core, insulate themselves from challengers through clickbait populism, and unwittingly support policies detrimental to their own well-being. In some ways this has already happened, with companies like Cambridge Analytica and the propaganda machine of Steve Bannon coming together during the election of 2016 to convince poor, working-class southern and midwestern whites to vote for a New York City real estate developer and reality TV star named Donald Trump.

  The “hidden core” conducting social inception will win over key influencers by mapping their every purchase, chat, post, and picture, creating a targeting profile to nudge unwitting “useful idiots”—those enticed by money and ego—to advance scripted narratives. Media personalities, celebrities, and those aspiring to fame will be easily drawn to these social media inducements. The hidden core will enlist, empower, and advance fellow travelers—those enticed by revenge and ideology—as de facto agents for the cause. Political candidates, CEOs, or even clergy may wittingly join in with information power brokers. These agents will create the strong ties in the physical world—real human relationships—to unite the weak ties of the virtual world, creating a symbiotic preference bubble in pursuit of the hidden core’s objectives.

  Some of the social inception methodology I’m describing already occurs. It’s what I wanted to do with Omar Hammami: nudge him to arrive at his own decision to denounce jihadism. My Twitter jousts with Hammami required reading spreadsheets of data, simple Microsoft Excel summations, and trial and error. Figuring out what worked and didn’t work when messing with terrorists took many years. I had little time and only a few engagements to test out different conversational tactics. Today, I can still perform most of these social media actions from my house with a laptop and Microsoft Office. But I’d still be quite limited in my reach, and it would take a good deal of time to scale beyond single social media targets.

  The social inception process will be relatively easy for well-resourced and motivated large-scale organizations around the world. They will be the ones who achieve data dominance, wield artificial intelligence, and rule social media’s future. Data dominance occurs when an actor, either through corporate consolidation or third-party aggregation, can bring together sufficient social media preferences to clearly demarcate preference bubbles and individual motivations. Those sitting atop social media companies, mainstream media outlets, online retailers, and streaming services will have unimaginable power of persuasion as they gain more and more information on each person’s daily life. Intelligence officers conducting espionage or targeting call this process a “pattern of life” assessment, but they’ve never had the capability now open to social media companies. The Internet of Things (IOT) records nearly every facet of one’s life. Health applications tracking height, weight, heart rate, and steps, combined with Google searches and Amazon purchasing patterns, can provide social engineers with the ability to deliver someone a message via social media at the exact time and place where they are most vulnerable to it psychologically. A social media user won’t know why they went for the content, but the big-data overlords will, which leads to the next emerging capability making this all a bit scarier: machine learning.

  Artificial intelligence enthusiasts have always overestimated the point at which computers will overtake humans, become autonomous, and dominate our world. AI hasn’t quite reached Terminator levels, but some of its tools have come into widespread adoption, and for social engineers, machine-learning advancements can be nuclear weapons for information warfare. Machine learning is the ability of systems to learn and improve from experience without being programmed by a computer operator. The information caches garnered through data dominance can be fed into machine-learning platforms, allowing the computer to learn more and more quickly—faster than any human propagandist ever could.

  Machine learning applications will be able to pore over old data and current information feeds to design the perfect message for an entire preference bubble and the precise variant for each individual in the bubble, as well as when, how, and where to deliver it. Advertisers, political campaigns, and Russian disinformation peddlers would be handicapping themselves if they didn’t use this approach to push their products and ideas.

  Disinformation and misinformation have been easy to create and proliferate as the barriers to entry for these technological tools have lowered. False print stories and graphic memes have had a devastating effect. The latest dangerous technological dimension coming online now is falsified audio and video. Researchers at Stanford created software that allows programmers to spoof the speech and appearance of any person if given enough video and audio data. Presidents, elected officials, media personalities, and celebrities can be made to say anything a social engineer chooses. Data dominance will enable machine learning to create highly convincing hoaxes, propelling video smears and audio pronouncements that will move preference bubbles to alternative realities and false beliefs, playing to the benefit of those who design them.

  I testified before Congress three times in 2017 and once more in 2018 regarding social media influence, and I immediately recognized from the experience that there are a few simple regulations that could impede these masters of manipulation. By law, political advertisements must announce themselves as such in print, radio, and television; these same rules should immediately be extended to social media, where political campaigns and super PACs increasingly focus their influence. But these regulations have gone nowhere, despite Congress’s grilling of social media companies in the wake of Russian meddling investigations. The rollback of net neutrality might also aid those with data dominance, as internet service providers could be coerced or influenced to shape content flows, targeting preference bubbles, denying users access, or restricting the content flows of those who challenge political propagandists. Emerging technology innovation and the regulatory environment, in sum, play to those able to dominate social media rather than users of social media.

  Despite the ominous signs on the horizon, we must remember that people still have a choice. The backbone of preference bubbles, social media nations, and global commerce is ultimately trust. Social media influence could become more pronounced and yet less effective if users don’t know what to believe. The educated and the mainstream may lose faith in social media, or maybe they already have, and in turn they could decide that the emotional crowd
s and social media mobs antagonizing each other in debates manufactured by influence peddlers aren’t worth their time. Those not addicted to the constant stimulation of social media may simply leave the platforms altogether.

  That said, there’s also a more frightening possibility. The erosion of trust in social media may lead the less educated to rely more heavily on their preference bubbles, and the more educated to become apathetic and disengaged from democracy. If that happens, democracies will lose ground to authoritarians, science will retreat under attack from zealots, and countrymen will turn on one another, to propel their personal preferences over the greater good.

  10

  Surviving in a Social Media World

  “We’re in the business of filling souls,” Howard Schultz told Scott Pelley during a 2006 interview on 60 Minutes. Schultz, as head of Starbucks, saw a broader vision for the coffee company beyond specialized lattes. “We’re in the business of human connection and humanity, creating communities in a third place between home and work.”1 The idea harked back to community gathering places of the 1950s, where people of all walks of life came together in the physical world and relaxed, talked, and even read books printed on paper. Starbucks innovatively offered specially selected music, which they repackaged, sold, and sometimes gave to customers as a gift while they waited in line.

  Strangely, though, Starbucks’s quest for community served as a harbinger of today’s preference bubbles. Baristas created coffee cocktails to the exact specifications of each customer. In 2010, a spokeswoman for the company, Lisa Passe, explained to a Wall Street Journal blogger, “If you take all of our core beverages, multiply them by the modifiers and the customization options, you get more than 87,000 combinations.”2 Nearly everyone could have their own unique flavor tailored specifically to their liking. Tom Hanks’s character Joe Fox, two decades ago, cleverly noted the challenge of so many choices in You’ve Got Mail:

  “The whole purpose of places like Starbucks is for people with no decision-making ability whatsoever to make six decisions just to buy one cup of coffee. Short, tall, light, dark, caf, decaf, low-fat, non-fat, etc. So people who don’t know what the hell they’re doing or who on earth they are can, for only $2.95, get not just a cup of coffee but an absolutely defining sense of self.”

  As usual, Nora Ephron got it right. Preferences aren’t about better taste, but identity. The movie also provided one of the first accountings of preference divergence and convergence between the virtual and physical worlds. The characters played by Meg Ryan and Tom Hanks meet and fall in love online, but they despise each other as corporate enemies in real life. The movie resolves with the two reconciling these differences—online preferences bridge real-world differences, ending in happily ever after.

  This ending would be unlikely on modern online dating sites. Apps from Match to Tinder allow users to pick their preferences in the virtual world in order to streamline their matches in the physical world. Tom Hanks, the corporatist, and Meg Ryan, the small-business owner, despite living close by, would likely never see each other on today’s social media dating apps. Their dating preference bubbles would ensure that the other never penetrated their social media fortress of personal preference. Even if they did appear in each other’s feeds, the other’s college, profession, or some additional Google research would likely lead one of them to “swipe left” on the other’s profile. Ephron’s brilliant script should instructively remind all of us, whether it’s 1998 or 2018, that we don’t really know what we want.

  Fast-forward twenty years. Starbucks in the social media era has become the place where people come together to communicate with other people who aren’t there. Lines to the counter fill with Americans donning headphones. It doesn’t matter what music Starbucks plays; few even hear it. Heads are down, peering at smartphones; rarely do eyes meet. Customers might stand in the same line with the same people hundreds of times each year and never utter a word or even remember each other’s faces.

  The proliferation of smartphones and the rise of social media combine to allow any of us, at any time, to be in a world of our own preference, where we control the environment to block out anything we don’t want or don’t like. This only appears to be getting worse as more Americans work from home, conducting more conference calls than in-person meetings, talking more to Alexa, Amazon’s in-home AI, than to real people and ordering food and groceries with smartphone apps rather than suffer the misery of shopping or dining out, surrounded by real humans. Self-imprisoned in our homes, governed by our preference bubbles, and yet seemingly having everything we ever wanted, Americans appear, at least on social media, to be absolutely miserable, downtrodden, victims of their excess and unfortunate to be alive. How do democracies, corporations, and citizens survive in this new and growing social media world?

  * * *

  In 1835, Alexis de Tocqueville, the French diplomat, political scientist, and historian, published his seminal book Democracy in America. The work detailed Tocqueville’s travels through the young United States in 1831, and his observations explained to the rest of the world why the experiment in American democracy might just be the best form of governance for balancing liberty and equality:

  “Americans of all ages, all stations in life, and all types of disposition . . . are forever forming associations. There are not only commercial and industrial associations in which all take part, but others of a thousand different types—religious, moral, serious, futile, very general and very limited, immensely large and very minute.”3

  Tocqueville’s study revealed much about the United States’ charm, but most notably the power of cross-cutting associations between political and civil society that prevented government dominance over citizens. His observations about these horizontal linkages among citizens, rather than the highly vertical linkages of patronage found in autocracies and theocracies with high levels of inequality, demonstrated the advantages of America’s melting pot.

  Tocqueville explained what would later be described as social capital—how Americans come together in pursuit of their interests. Associations of marriage, religions, businesses, unions, military service, and schooling overlap to help protect each citizen’s freedom and preserve opportunity for all, regardless of race, religion, socioeconomic status, or ethnicity. Robert Putnam, the renowned Harvard University professor who pioneered research into social capital, authored the famous and contentious book Bowling Alone: The Collapse and Revival of American Community. Putnam maintained that social capital was the glue for a thriving American democracy. By the late 1990s, though, membership in civic, social, and fraternal organizations was winnowing each year in America, and Putnam argued that the United States was headed in a dangerous direction. He pointed out that Americans were still bowling, but, rather than in organizations emblematic of civil society, they preferred to do it alone.

  Putnam further defined the social capital concept pioneered by Tocqueville, dividing it into two types: bonding and bridging. Bonding capital arises when we Americans associate with people similar to ourselves. Bridging capital comes when we make friendships and associations with people unlike ourselves. Putnam argued that these two types of capital, when combined together, power American democracy, and that the latter’s decline had already begun signaling an ominous future for the United States. Some challengers of Putnam’s research, whose book was published in 2000, argued that he missed the new opportunities of connection and association offered by the internet. But these internet optimists of political science, similar to “long tail” enthusiasts, misunderstood how connections, when hyperpowered on social media, would strengthen Putnam’s warnings of the degradation of social capital. Bonding capital has risen with the internet, but social media polarization has jarringly eroded bridging capital—the result being our current preference bubbles.

  Post-publication, Putnam not only defended his thesis but worked to identify solutions for increasing American social capital, particularly of the bridging type. His 2001 Social
Capital Community Benchmark Survey sought to discover approaches for increasing social capital but instead revealed more troubling indicators for American society. The study noted:

  “Our survey results also make clear the serious challenges of building social capital in a large, ethnically diverse community. The more diverse a community in our study, the less likely its residents are: to trust other people . . . to connect with other people, even informally . . . to participate in politics . . . to connect across class lines.”4

  These findings come more than fifteen years before the social media nationalism arising today. It turns out that Tocqueville also saw this vulnerability in American society, as James Wood noted in a 2010 piece in the New Yorker:

  In the book’s second volume, he warns that modern democracy may be adept at inventing new forms of tyranny, because radical equality could lead to the materialism of an expanding bourgeoisie and to the selfishness of individualism (whereby we turn away from collective political activity toward the cultivation of our own gardens). In such conditions, we might become so enamored with “a relaxed love of present enjoyments” that we lose interest in the future and the future of our descendants, or in higher things, and meekly allow ourselves to be led in ignorance by a despotic force all the more powerful because it does not resemble one.5

  Wood notes that Americans might not see the evil arising from a preference for the status quo, for confirmation of their beliefs, and for the implicit comfort of being among the majority. He continues with a warning directly from Tocqueville: “It does not break wills, but it softens them, bends them, and directs them; it rarely forces one to act, but it constantly opposes itself to one’s acting; it does not destroy, it prevents things from being born.”