EECS 649 Blog 10: Collateral Damage: Landing Credit

Author: Megan Rajagopal

In Weapons of Math Destruction, O’Neil talks about big data is used and how it can cause collateral damage. O’Neil mentions that big data impacts people’s ability to get loans, jobs, credit, etc. She also describes how different companies will use big data to determine the potential risk of future clients. A very well-known example of big data being used is through the algorithm known as FICO. The FICO algorithm only looks at a borrower’s finances, particularly their debt load and bill-paying record. This was used to minimize any biases a banker may have towards a borrower. However, this isn’t the only scoring algorithm out there. E-scores are a weapon of math destruction because they are unaccountable, unregulated, and usually unfair. For example, there is a large chance that an e-scoring system will give a borrower from a lower-class poorer neighborhood a lower score than a borrow from a upper-class wealthy neighborhood. This would cause people from less wealthy areas to have less credit available and higher interest rates. Another example of e-scores proving to be a weapon of math destruction is when a Virginia company would place callers in a hierarchy based on the available data on the caller. The callers who were deemed more profitable were quickly moved to a human operator. Because of the large-scale impact this algorithm has on society, it is considered a weapon of math destruction and it also creates a negative feedback loop. There definitely needs to be a change in this system because the process is very automated and heavily based on the algorithms results with little human verification. There is also little humans who are being negatively impacted by these algorithms can do to change it. As a result a change that should be made to e-scoring algorithms is looking at the appropriate data for the task. For example, financial companies should only look at debt, financial payment history, and ignore where a person lives/there economic status. This chapter was very interesting and showed how weapons of math destruction are used in our society. It also shows how with algorithms an automation becoming more popular, there needs to be ways to verify the information is fair and correct.