The Guardian has a long article about how Cambridge Analytica acquired data on Facebook users and then used it to help get Trump elected. I figured as much. If you took a personality test on Facebook, which millions of users did, 40% of you allowed Facebook to use that information in order to target ads to you. In 2016, Steve Bannon teamed up with Cambridge Analytica to change your mind about Trump. Here’s what Cambridge’s Christopher Wylie says Brannon’s philosophy and goal:
“[Bannon] got it immediately. He believes in the whole Andrew Breitbart doctrine that politics is downstream from culture, so to change politics you need to change culture. And fashion trends are a useful proxy for that. Trump is like a pair of Uggs, or Crocs, basically. So how do you get from people thinking ‘Ugh. Totally ugly’ to the moment when everyone is wearing them? That was the inflection point he was looking for.”
How do they do it? How did they target Facebook ads, and probably the whole Trump message directly to you, their targets? It’s all psychology and statistical analysis:
“at Cambridge University’s Psychometrics Centre, two psychologists, Michal Kosinski and David Stillwell, were experimenting with a way of studying personality – by quantifying it.
Starting in 2007, Stillwell, while a student, had devised various apps for Facebook, one of which, a personality quiz called myPersonality, had gone viral. Users were scored on “big five” personality traits – Openness, Conscientiousness, Extroversion, Agreeableness and Neuroticism – and in exchange, 40% of them consented to give him access to their Facebook profiles. Suddenly, there was a way of measuring personality traits across the population and correlating scores against Facebook “likes” across millions of people.
The research was original, groundbreaking and had obvious possibilities. “They had a lot of approaches from the security services,” a member of the centre told me. “There was one called You Are What You Like and it was demonstrated to the intelligence services. And it showed these odd patterns; that, for example, people who liked ‘I hate Israel’ on Facebook also tended to like Nike shoes and KitKats.
So, if you are a Nike shoe buyer who likes KitKats, you probably didn’t get a pro-Israel message. On the other hand, if you were a mom who likes designer fashion and buys a status car, you probably got a lot of warm and fuzzy stories about Ivanka.
It was all very well designed to target you. I think the difference in what Cambridge did compared to what other political data and marketing companies did was that they were able to target people very specifically as well as all their Facebook friends. Most marketing companies go after just the neighborhoods or regions based on income etc. But who knows who will be scrutinized next? Maybe the info that you give Giant Eagle every time you buy a certain kind of apple or avoid the steaks can be used as well. That information is very important to people who want to influence you.
Well, at least Cambridge Analytica knew that Trump was an Ugg and they were going to have to push people to vote for him anyway. The question is, why were they trying to force this uniquely unfit man down our throats in order to ruin the country? Was it in service of Brannon’s chaos theory or was it in order to help the Russians? Or are we seeing a convergence of two efforts to get Trump elected that just happened to be reinforcing and synergistic?
Put that into your PLS model and spit out some principle components.
Filed under: General | 45 Comments »