Author: Cathy O’Neil
Genre: non-fiction
Publication date: 6 September 2016
Publisher: Crown Books
Number of pages: 272 pages
Review
Have you ever watched or read The Big Short, a based on true story movie or book about the financial-industry in 2008 where the collapse of the real estate market happened? On that movie, Brad Pitt who acts as Ben Rickert, a retired ex-trader said, “If we’re right, people lose homes. People lose jobs. People lose retirement savings, people lose pensions. You know what I hate about f*cking banking? It reduces people to numbers — ever 1% unemployment goes up, 40,000 people die, did you know that?” Tiny numbers are significantly affects people in the real world.
Cathy O’Neil, a mathematician academia turned data scientist, began working for a hedge fund not long before the 2008 crash. This is the moment when she spotted the disaster raised by mathematical models called Weapons of Math Destruction (WMD).
We are living in a big data era. Algorithm encodes quantification and decision of critical life moments. Many of these algorithms are built to model large numbers of people, and do injustice to the poor and benefit the rich.
In this book, Cathy O’Neil reveals her disappointments in the part that mathematicians had played by writing people experience as the victim of unfair algorithm. Student who couldn’t land a job with minimum wage in a grocery store because of his answer in a personality test. Teacher with outstanding performance is fired because of low points on a teacher evaluation. Poor people who lives in high-crime neighbourhood are exposed with payday lenders advertisement that potentially lead them to drive their credit rating plummets further more. This model creates a death loop. Living in WMD is keeping the poor poor and the rich rich. WMD widens the wealth disparities.
Algorithms can also be gamed where a university offers an adjunct faculty position to a well-known mathematicians just so the university can claim the publications as its own in order to be raised in a university ranking. Some university administrators also desperately playing the system and input to boost their university position.
This book is highly recommended for us to realise and understand the flaw of the big data. However, it does not mean that we should just automatically leave big data-driven world. But we should noted that algorithms encode the past and does not calculate the future well and wisely. Any type of model should be based on transparency and they must be welcome to audit of assessments. At last but not least, data with moral foundation should be the principle of building models.