A review by george_odera
Weapons of Math Destruction by Cathy O'Neil

5.0

It is said that "the road to hell is filled with good intentions", and Cathy O'Neill's Weapons of Math Destruction (WMDs) is a befitting narration of how Big Data fits within this aphorism. In WMD, Cathy argues that math-powered applications powering the data economy are based on choices made by fallible human beings, choices made with the best intentions. Nevertheless, many of these models encode human prejudice, misunderstanding, and bias into the software systems. WMDs harbour 3 elements, namely Opacity, Scale, and Damage, supplemented by a feedback mechanism that may be at times pernicious. The tragedy of WMDs is that are they are deaf not only to charm, threats, and cajoling, but also to logic.
Cathy describes a select number of WMDs in Recidivism models, College rankings, Online advertising, Predictive policing, Employee recruitment, Credit reporting, Insurance policies, and Politics, as well as their destructive effects manifested by astronomical student loans, criminal (in)justice, racial inequality, poverty cycles, and fissures in democracy. These are not mere anecdotes, as Cathy is able to synthesise the destructive effects of WMDs by showing how they breed perverse incentives, yield unintended consequences, and generate pernicious feedback.
Cathy makes a strong case that WMDs present a trade-off between fairness and efficacy, and that mathematical models should be our tools, not our masters. I can't recommend this book enough, as it will prove to be one of the most prescient books of the 21st Century