WEAPONS OF MATH DESTRUCTION Hardcover – 6 September 2016

Awesome book, Cathy brings her insights so well.. Where even thought data could be fact, but it can be interpreted incorrectly. Because, data supposed to closely represent the phenomena, and often phenomena is not measured but data which potentially assumed to represent the phenomena is measured. When that underlying assumption is not validated and if it fails, the whole result fails. She brought multiple examples to prove that.She also brings how insights can negatively affect the desirable potential positive outcomes, especially when incorrect insights drawn from misleading algorithms in the name of math!

Weapons of Math Destruction is a book which is smartly titled in a way that it would pique anyone’s attention.Cathy O’Neil ( PhD., Harvard) is an accomplished academic-mathematician , and it is a bit strange that a mathematician would so eloquently highlight the frightening reality of algorithms which could run amok and trample people of their choices, distort truths when they are expected to remove biases and reinforce the vicious feedback loop in an endless spool.The examples problems are wide ranging from ranking of colleges and their impact of universities , faculties and most importantly students, to credit scores and the way they are defined and regulate, the flawed banking system governed by statistical fallacies.The fact that Big Data processes codify the past and they do not invent future, but could masquerade inventing future. Defining the future would need moral imagination and that is something humans can provide. The fairness and morality of the Data model is often sacrificed in the alter of the God of Mamon , profits , efficiency rule the roost.For a mathematician to courageously describe the ills of misusing mathematical principles is something which O’Neil has demonstrated well with real life examples and the disastrous effects.The equivalent of Hippocratic Oath for Data Scientist is something which is a novel idea presented. In the same breath , she has mentioned that Data , Computers and much less mathematics would play a critical role in the mainstream of businesses. However the key is to make them work a neutral yet an inevitable force is our responsibility.Here, however there is a miss. Problems have been explained and the consequences of the proxy driven predictive models could be disasters in the making. However, how do we empower the disadvantaged, the techno-phobic population which is more often than not at the receiving end of the algorithmic tsunami has not been covered in any reasonable depth. Making the guys who define and run the data models do the algorithmic audit would not happen, so who would be the ethical champions to bring order to the chaos which could be created had to be better articulated.Overall the book gives a very good sense of what we are into , where we are going in this inescapable world of algorithms , but what needs to be done by the people affected could have rounded off the book nicely.

A must read book to understand and learn about the AI models

Well written in a simple yet powerful language

Overall this book gives good insight into how a well intended modelling still causes problems due to lack of fairness in the objectives. Length is a more abd after couple of chapters rest of it looks very similar. Blog Post could have made the same point as convincingly

As a software developer really opened my eyes towards the not showing ignorance towards good ethics in the kind of solution you develop

A great book which should be read by everyone. The author keeps the language simple so even people who don't have maths or technical background can also understand it. The author also mentions ways by which to avoid being an WMD victim.

Provides a new perspective. Well supported by examples.