I think the author forgot his concept halfway through the book. That’s a pity, because I think I’d have liked the book the author was going to write. This felt an awful lot like a collection of interesting bits of data that the author attempted to will into a book, because the stated objection – to demonstrate via data mining sites like Facebook and Google that our public preferences don’t always align with our true beliefs – falls by the wayside midway through. It was interesting reading about how the most recent triple crown winner was saved from sale because data mining contradicted conventional assessment methods and showed that his internal organs predicted he would be an excellent racehorse, but it doesn’t fit the objective laid out by the author.
The diversion might not have rankled if the author delved more deeply into each of his examples; the superficial exploration of the data coupled with stories that only skirted his larger point led me to feel like he read a few examples of data mining, thought “hey, that was cool,” and tried to find a way to link them together into a book, rather than finding a true pattern. For instance, in illustrating a point about whether it was ethical to monitor searches, he mentions a woman who was killed by an ex who had recently and frequently googled murder law, methods, and her name. But he neither spends time on this specific case nor on how many murderers use the internet to plan their crimes and what forms that can take, he simply gives the statistics on how many women are murdered by their partners and how often “kill my girlfriend” is Googled. That’s a pretty shallow take.
I was hoping for something akin to the excellent Dataclysm. This didn’t have nearly the data, nor the interpretation to match that book.