Cathy O'Neil, Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy (New York: Crown, 2016), 259pp.
I learned about Cathy O'Neil and her best-selling book (long-listed for a National Book Award) in the excellent movie Coded Bias (2020), which movie considers the inherent biases of race, class, and gender in the computer models and algorithms that now impact pretty much every area of our lives. As both this book and the movie demonstrate, today we are all "graded" for a broad array of risk-reward factors by computer algorithms in hiring and firing, performance evaluation, bank mortgages, predatory lending, voter targeting, housing, insurance rates, building security, college applications, crime prediction and the risks of recidivism, and on and on.
These algorithms are not just biased. They can also be abusive, inaccurate, and unregulated. Most of all, they are opaque, and there is typically no court of appeal to protest how they characterize you. These powerfully efficient algorithms tend to be inherently unfair, says O'Neil. In particular, they punish the poor and corrode our democracy.
Consider this one example from her book (pp. 164–166), to get a feel for the scale and scope of Big Data, and how it is used. In 2015 Consumer Reports studied the disparities in insurance rates by analyzing over two billion (!) price quotes for every one of the 33,419 zip codes in the country. What they found was "wildly unfair." Another consumer watchdog group documented how Allstate had one hundred thousand micro segments or pricing tiers based upon its "price optimization algorithm." Some Allstate customers received discounts as high as 90% off the average rate, while others faced increases of up to 800%.
It's hard to see how or where this will stop. In her last chapter, O'Neil makes some suggestions. At least we can be aware of and educated about what's happening to us. And thank God for consumer watch dog groups, and for scholars like O'Neil for excellent books like this.
Dan Clendenin: email@example.com