What's a social engineer to do?
by John McCreery on
Social engineering is the idea that those who understand how society works will be able to control it. This idea was very popular in the years leading up to and after World War II. Governments, corporations and foundations invested heavily, providing the funding for the postwar growth of social science disciplines, including anthropology. They thought that they would acquire the knowledge they viewed as necessary to make social engineering a reality. Did they get what they were paying for?
In a recent discussion on the OAC, Mark Stahlman raised an awkward question: "Economics doesn't work anymore. Anthropology doesn't work anymore. What's a social ENGINEER to do? Move to China?" (He seems to think that in China, people in power with money to throw around haven't yet gotten the message).
I replied that we already know what social engineers are doing: big data and microtargeting.
A case in point: The story that was all over the Web a few weeks or months ago about the father who stormed into his local Target department store to complain about his sixteen-year old daughter being sent advertising about expectant mother and baby gear. Then he went home and talked to his daughter. Guess what.
The story broke because the father went back to the store to ask how Target knew his daughter was pregnant. The person who talked to him pointed to the CCTV cameras monitoring the aisles and the data on recent purchases from the POS and credit card systems. Target's social engineers are constantly analyzing this data in search of patterns that will help Target anticipate what customers are likely to be looking for, patterns that are then used to personalise advertising. The process just takes a bit further what we see every time we buy a book from Amazon.com, and when we search for one book, get others that we might be interested in suggested to us.
Politics is moving this way, too. In the USA, the Republicans got a leg up on the Democrats in the 1980s by being the first to take advantage of targeted direct mail with the targeting based on surveys and focus groups. For the Democrats, the Internet was a god-send, a chance to turn the tables that was first embraced by the Howard Dean campaign. The Dean campaign failed because, as my friend Jerry Bowles put it, that campaign managed to demonstrate that the wired population was still only about 13 percent of the electorate back in 2004.
The Obama campaigns took targeted mail to the next step, combining massive data crunching with community organising in a way that caught the Republicans' Mad Men era advisors totally flat-footed, with the McCain and Romney campaigns running TV era mass market and direct mail campaigns in the era of the Web and the long tail. Of course, the demographics helped, too, with Net-savvy younger voters increasingly leaning left instead of right.
To me the most plausible scenario for the way the wired world is going is Alberto Lazlo-Barabasi's Bursts: The Hidden Patterns Behind Everything We Do. It's a book with huge, and what should be terrifying, implications for anyone involved in traditional social science, quants and quals alike. Barabasi argues that the conventional assumption that mass behavior is predictable using statistical models while individual behaviour isn't (the theme of Isaac Asimov's Foundation series) is demonstrably ass-backwards. Mass behavior is like the weather, so infected with chaos and complexity that predictions beyond the next few days or weeks become increasingly likely to fail.
In contrast, the integration of systems like those that Target uses, ubiquitous CCTV cameras, transaction tracking, phone and Internet usage monitoring now make it possible to predict individual behavior in microscopic detail. Why? Because most individuals lead utterly predictable lives as they follow their daily routines.
Indeed, the behavior most likely to get you tagged as a possible terrorist is being more unpredictable than usual. Barabasi's book includes a marvelous story about an artist who had to spend years getting himself off Homeland Security watch lists because his installations and performances took him on short, one-shot trips to all sorts of places around the world.
If you check out recent academic discussions, you will see that the question of how ethnography and ethnographers will survive in a big-data world is the No. 1 most pressing issue now facing practicing anthropologists (the ones who try to do something instead of just talking about it). The most promising line appears to be that intimate ethnographic immersion makes anthropologists better able to come up with the stories that big data crunchers need to understand the patterns they see. Could be so. Time will tell.
Anyway—I add this for the sake of our younger readers—demand is rising for people who combine the intimate knowledge and empathetic awareness that anthropologists are supposed to bring to the table with the technical chops to understand the implications of the patterns the big data crunchers detect. One without the other won't get you a stable career. Let alone make you rich.
Demand is rising for people who combine the intimate knowledge and empathetic awareness that anthropology brings to the table with the technical chops to understand the implications of the patterns the big data crunchers detect.