There is a new research report from Pew, Code-Dependent: Pros and Cons of the Algorithm Age, and the results are surprising. There has been a lot of pushback against the algorithms that are controlling more and more of our lives, and this reports provides a great summary of what is going on.
One of the best books I have read about this is Cathy O’Neil’s Weapons of math destruction: how big data increases inequality and threatens democracy. One of her main purposes is to illustrate the serious consequences that algorithms have on people’s lives: from losing jobs to greater prison time, and often these decisions are made without any understanding of the algorithm and with without any possibility of finding out how the algorithm comes to its decisions. She tries to demystify the algorithms for the general public and wants greater transparency.
She says that algorithms reflect the biases of the people who build them. Since most people who build these algorithms tend to be the rich and powerful, it is their biases that are reflected in the algorithms. These biases may be honest, that is, they were not incorporated into the algorithm intentionally, but the biases are there nevertheless. For instance, an algorithm may use zip code as a data point and do so in all innocence, but the zip code can reflect on the racial makeup of the neighborhood, the local economy (how much money someone makes, the number of arrests and so on. As a result, zip code is not an unbiased data point.
As libraries go into the world of big data and linked data, librarians should be aware of how that data can be used. This Pew report is most enlightening.
There is also a talk by Cathy O’Neil that she gave at Google.