Review of Weapons of Math Destruction
The following is a review of the Artificial Intelligence book Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy by Cathy O’Neil.
Weapons of Math Destruction, by Cathy O’Neil is an artificial intelligence book that aims to show the dark side of the algorithms and machine learning systems that have become ubiquitous in our day to day life.
These software systems, which are meant to be tools to increase our overall quality of life, turn out to be flawed with biases and can lead to massive feedback loops that do nothing but further increase the misery of those affected by them. In theory, these machine based decision systems should lead to a greater fairness, having everybody be judged according to the same rules, however O’Neil reveals that these systems are used in an unregulated, uncontested and opaque manner, which leads to more discrimination rather than more fairness.
Welcome to the dark side of Big Data. Welcome to the era of the Weapons of Math Destruction.
Through the book many examples of this algorithmic discrimination are illustrated, highlighting the flaws of many artificial intelligence systems and the impact they have on the people that are being analysed by them, or simply the responses they give: accepting or rejecting loans, filtering candidates for job interviews, allowing students to be accepted into universities, and a whole lot of other examples that have a profound impact in many peoples life.
There’s no harm if a recommendation algorithm from Netflix advices you to watch a movie you don’t like, however if you need to insure your car, have a perfect driving history, but the Artificial Intelligence model that is in charge or accepting or rejecting your application says no, you can be in serious trouble without even knowing why you are being rejected.
Weapons of Math Destruction (how the author refers to this kind of harmful Artificial Intelligence models) are opaque (the person at the other end of the model rarely knows why the model gives the response it does), are built using proxies (data that is not an exact source of information for the task at hand, for example using your eye color to influence in the decision of whether you should get a loan or not), and are rarely questioned or debugged.
Who is this book for?
Altough the book can be understood and enjoyed by anyone without the need to have a technical or mathematical background, we think it is best for Data Scientists that want to understand how the world works in the era of Big Data and what a good data scientist needs to consider in order to create fair and useful models.
It is one of the best artificial intelligence books about fairness, and one of the very few that highlights that despite of the amazing benefits that AI can bring to us, it can also be deeply dangerous in an earthly manner. There are books like SuperIntelligence by Nick Bostrom or Surviving AI by Calum Chace that speak about the divine dangers of AI: what will happen when Artificial Intelligence Surpasses general level human intelligence.
Weapons of Math Destruction is way more down to earth in that manner, speaking about the present harms AI can do to us, without going into technological or ethical debates.
About the book
Author: Cathy O’Neil is a data scientist and author of the blog mathbabe.org. She earned a Ph.D. in mathematics from Harvard and taught at Barnard College before moving to the private sector, where she worked for the hedge fund D. E. Shaw. She then worked as a data scientist at various start-ups, building models that predict people’s purchases and clicks. O’Neil started the Lede Program in Data Journalism at Columbia and is the author of Doing Data Science. She is currently a columnist for Bloomberg View.
Pages: 259 pages
You can find out more about the way of thought of author Cathy O’Neil in the following Ted Talk: much of what she speaks about there is covered more deeply in the book.
Summary of Weapons of Math Destruction: An Artificial Intelligence Book on Inequality
Weapons of Math Destruction is a top Artificial Intelligence book for those interested in seeing the different uses of Big Data based Machine Learning models that despite creating a lot of hype can sometimes be highly detrimental for certain population groups.
It is an easy to read book with many real world examples across a wide range of industries of how AI should not be used, and the love and passion of the author for the topic can be felt in every single page. It can get a little repetitive at times, but it is a short read that is definetly worth it for any Data Scientist to see what impact the models he creates can have.
It is also a reminder for the need to educate the wider community and general public on how data is used, what can be done with it, and what needs to be taken into account for successful and fair applications to be built.
Buy the book from Amazon here:
- O'Neil, Cathy (Author)
- English (Publication Language)
- 288 Pages - 09/05/2017 (Publication Date) - Crown (Publisher)
Another very similar book to this one, is An Artificial Revolution: On Power, Politics and AI (Mood Indigo). If you liked weapons of math destruction, then you should definitely check it out. Also, for more Artificial Intelligence books like these ones, check out our section full of them!
That is all, thank you for reading How to Learn Machine learning, and have a wonderful day!