Although it could benefit from some editing (it can be repetitive), this short book helps clarify the benefits and drawbacks of "Big Data" and has some solid (if early) proposals about how to handle it.
The lines of argument that the authors draw are clear and very helpful in thinking of this new way at looking at the world. For example, Big Data shows us the "what" not the "why"; the growing influence of correlation with respect to causation; the different types of Big Data such as mechanical analytics and personal information; and the dangers in using Big Data for punitive measures. The distinction between certain types of analytics data (ie about auto parts or the weather) and personal information is particularly helpful in thinking about the damaging effects of Big Data. This book does a good job in drawing that distinction, but it does not tackle the large grey area in between. An example in the book has car companies recording GPS data when a car's ABS is engaged, ostensibly to make safer navigation routing. Even though this sounds like analytics, it can be considered privacy invasion. The current legal framework for privacy protection is obsolescent at the scale of Big Data.
I have meet Kenneth Cukier and read the Big Data special in The Economist but I was swayed to finally read this book after hearing a VMS interview:
For those already versed in BD, I would highly recommend chapters 8 and 9, where they outline the "dark side" of Big Data and some ways to mitigate it.