Financial Crisis and Big Data
Article written by Victoria Maldonado
Weapons of Math Destruction, as Cathy O’Neil calls them in her book, are the destructive models that use mathematics to reinforce bias. In chapter two, she talks about the time where she first realized that she was disillusioned with WMD’s. It was during the time of the subprime mortgage crisis in the US in 2007 and she said that she was disappointed in the role mathematicians played in that big problem.
The whole issue, in summary, was that people used models to predict whether or not a mortgage was safe. These models started being corrupt since banks, investors, and ranking companies would only benefit if the rating for a mortgage was good. Otherwise, they would either invest in another mortgage, or go to a different company and try to bribe it to get a better ranking. Soon enough, the model that was supposed to calculate the real rating of an investment, was filled with unsafe ones, so it would make those unsafe qualification the new standard. Therefore, even mortgages that were really bad would still have a good score and people believed it. Since no one was able to pay the mortgages and investors would not make more money out of it, millions of people were defaulting on their payments, eventually ending up in a crisis.
I think that the biggest takeaway from the chapter is that people should not be blindly trusting algorithms without knowing how they work. There will always be people who would want to change the model to benefit them, and that should not be allowed. The bad thing about algorithmic models is that they hide behind complicated math so most people cannot understand them. However, it is important to reveal what those models have and open them to alterations to constantly update them and improve them.