Everybody Lies

It took me about 5 hours on Christmas eve to read the whole of Everybody Lies: What the Internet Can Tell As About Who We Really Are, including Acknowledgments (that is 284 pages out of 338 in the hardback edition). This is worth a mention. In his conclusion, the author Seth Stephens-Davidowitz says he knows very few of us, who started the book, are still reading. Data gleaned from Kindle readers by Amazon says fewer than 7% of those who started reading Thinking Fast, Thinking Slow, a full 499 pages in its hardback edition, actually finished it. Fewer than 3% finished the much-lauded Thomas Piketty book Capital In the 21st Century, which is 704 pages long. Now there is a surprise. We do not need Big Data to tell us that these books were abandoned by vast numbers of those who started reading them. Even those who are unfamiliar with Umberto Eco’s antilibrary effectively maintain one. Books are regularly abandoned, left unread, even unopened. But when they are on your Kindle, Amazon gets to see how much you read of them.

Big Data, as seen in Google searches (and some other online sources), is the digital truth serum, Seth S-D writes. We go online to bare our souls about parents, children, significant others, our bodies, our fears etc, to find answers to dilemmas, personal quandaries, weird curiosities and awkward sense of humour, without fear of losing social approval or our coolness quotient. That idea is the crux of this book.

The other key insights — or powers of Big Data — Seth S-D demonstrates are: everything is data, that Big Data can enable us to zoom in meaningfully into smaller datasets to gain new insights, and that Big Data allows us randomised experiments at scale to find true causal effects in phenomena and observations such as how do our preferences and opinions form.

Big Data for all its potential misuses — Minority Report gets a namecheck — has its uses in helping us feel less alone in our quest and our weirdness. That list of drop down options to complete our search from the Google home page subtly tells us about those others like us, as well as those who started a query the same way but were seeking something else.

Towards the end, chapter 7 of 8, the author contends with limitations of Big Data: the curse of dimensionality (essentially a cognitive limitation), the risk of forgetting the meaning we were seeking through it, the risk of missing what cannot be measured, and the fixation with what can be measured aka the importance of perspective.

The book is quite readable, weaving stories on sex, politics, baseball and other common human pursuits, and often funny. The author’s granny, brother and family find a mention. There are 30 pages of references — Notes — which are oddly not cross-referenced in the main text. The book is also a tad too long, especially the Conclusion. I also have a few other nerdy nitpicks. For instance, in opening with the limits of intuition, Malcolm Gladwell is name-checked but not Gerd Gigerenzer whose work Gladwell paraphrased. Why the nitpick? Because Gigerenzer has made the point about how intuition or gut feel is a sum total of all the “data” we have garnered through life experience, compressed into a heuristic that we draw upon in System 1 thinking (that Kahneman mentions in his largely-left-unread tome). It would have been a neat segue into the author’s messages “everything is data” and the rest of the limitations. But as I said it is a nerdy nitpick.

And one last thing. I bought the Kindle edition of the book first. It was just over $2. About 8% in, I bought the physical book because unsurprisingly, I read faster on paper and retain more, and my eyes do not get tired. Amazon data gleaning may suggest I abandoned the Kindle book. Unless it correlates that abandonment with the data point that I also bought the hardback from another Amazon store (making it easier for them to track me instead of going to the physical store where I buy most of my physical books), and that this review is appearing on this blog.

Everything is data. Big Data however is not everything.

Star rating: 4 out of 5 

Usefulness note: The book will interest all those who enjoyed Freakonomics, More Sex Is Safer Sex, and similar books.