The Wall Street Journal published an article by Jo Craven Mcginty back in October titled “Algorithms Aren’t Biased, But the People Who Write Them May Be“. The author discussed a book called “Weapons of Math Destruction” published by Dr.Cathy O’Neil and how algorithms are used in many different industries.

An excerpt from the piece:

The algorithms are what Dr. O’Neil calls “weapons of math destruction.” WMDs, she says, share three characteristics: They are biased. They are opaque. And they are scalable.

In one striking example, she describes a Consumer Reports analysis of more than 2 billion car insurance price quotes issued over two years that revealed how well someone drove had little to do with how much they paid for insurance.

In Florida, for example, drivers with a clean driving record but poor credit paid $1,552 on average more than drivers with excellent credit and a drunken driving conviction.

To decide how much to charge, insurers scored drivers using elements found in their credit reports to rate how likely they were to file a claim or shop around for cheaper insurance. Spending characteristics carried more weight than driving records. Insurers didn’t have to reveal scores to the customers whom they rated or even divulge the practice. And in the last decade, use of the algorithms has become pervasive.

Biased? Check. Opaque? Check. Scalable? Check.

Consumer Reports noted that only three states—California, Hawaii and Massachusetts—prohibit insurers from setting rates based on credit scores. In those states, premiums are established largely on driving records and the number of miles driven.

Source: Algorithms Aren’t Biased, But the People Who Write Them May Be, WSJ

Algorithms play a critical role in most US industries. However most of us are woefully unaware of how they affect our lives. Even highly educated Americans are stumped by the power of algorithms, There are many reasons for this situation a few of which we will discuss below:

• Algorithms are developed by humans. So there can be human errors or they can be massaged to produce the desired outcome.
• An algorithm is  a”Black Box” for most people other than the companies/people that created them. This is because most of these are proprietary and are closely guarded by the companies that own them. So the inner workings of an algorithm are a mystery to everyone.
• Sometimes the writers of these mathematical tools depend on external data sources to crunch the numbers. If those sources have erroneous data obviously the algorithm produces inaccurate outputs.

The entire journal article is worth a read. Readers may also want to read Dr.Cathy O’Neil’s book for additional insight into the world of algorithms.