Weapons of Math Destruction by Cathy O’Neil, Crown Random House, NY, 2016

Harvard mathematician Cathy O’Neil authored Weapons of Math Destruction after a stint on Wall Street as a quant at D. E. Shaw, the hedge fund associated with former Treasury Secretary Larry Summers.

After observing the cultural priorities and the short-term values most financial decisions, O’Neil became disillusioned, searching for higher goals for the mathematical modelling she loved.  By 2009, she realized that even the lessons of the financial collapse brought no new direction and instilled no new values in the world of finance.  O’Neil left D. E. Shaw to join Risk Metrics Group, hoping to work on fixing the kind of financial engineering she called “weapons of math destruction” (WMD).  She worked on risk models, predicting the likelihood which tranches of securities or commodities would blow up next week, next year or five years hence.  She liaised between Risk Metrics and the longest most discerning connoisseurs of risk: the quantitative hedge funds, who considered themselves the smartest of the smart.  O’Neil soon found that analysts warning about risk were viewed as party poopers or a threat to the bottom line.  The refusal to acknowledge risk ran deep and a risk report might tumble a trader’s Sharpe ratio.

So, O’Neil recast her expertise as a Data Scientist and moved to Intent Media, a Big Data startup to design an algorithm to distinguish among consumers the buyers from mere window shoppers.  O’Neil found parallels between finance and Big Data.  Both gobble up talent from elite universities, MIT, Princeton or Stanford, with students expecting riches and success.  Money vindicated all doubts, even in these industries that seemed a combination of gaming the system and dumb luck.  O’Neil quit again, joined Occupy Wall Street and wrote this well-documented book.

In each revealing chapter, O’Neil shows how algorithms are increasingly running our lives, deciding which colleges or jobs we access, how prisons and criminal justice operate and how our personal data is sold to data brokers, advertisers, marketing, insurance companies and banks.  Her chapter 3 shows how college became an arms race due to unfair algorithms and standards ranking various aspects, but omitting data on costs and results for students.  Chapter 7 examines algorithms leading to more exploitive work schedules.  Chapter 8 looks at how Big Data shapes credit markets and reveals faults buried deep in algorithms like FICO and how peer-to-peer lending is being taken over and gamed by big legacy banks using arbitrary, unaccountable “e-scoring”.  Most scary is Chapter 10 showing how social media giants Facebook (FB), Google (GOOG), Apple (APL), Microsoft (MSFT), Amazon (AMZN) and cellphone providers AT&T (ATT) and Verizon (VZ) are targeting political campaigns and helping shape the outcome of US elections with micro-targeting groups of voters.

O’Neil’s conclusions are similar to my own in “Artificial Intelligence + Algorithms = Assumptions”, as well as in many other useful reports in New Scientist on misuse of mathematics and computers.  These follow an earlier critique, “Calling Wall Street to Account”, by internet pioneer Alan F. Kay, founder of AutEx, now part of Thomson-Reuters (TRI).

As the FINTECH 100 continue disrupting legacy financial incumbents, it is useful to dig deeper into the many mis-uses of Big Data and how the crucial issues of privacy v. security are evolving.  What global standards and new regulations will guide Big Data in our global markets?  This book is revelatory as well as a good read for those in finance