The Problem with Big Data

Today, Netopia contributor Jane Whyatt portrays Viktor Mayer-Schönberger, Oxford University’s professor of internet governance. Mayer-Schönberger is one of the most quoted thinkers in our humble web publication, and I am of course proud to be able to post the portrait. His most recent work, Big Data: A Revolution that Will Transform How We Live, Work, and Think (John Murray, 2013, co-authored with Kenneth Cukier), became the defining book for the venture capital hype around big data business. It’s a good read; it deserves all the credit. It only has one problem: it was released too soon.

Mayer-Schönberger and Cukier are enthusiastic about the opportunities of Big Data, but they also talk about the negative aspects, like the perils of probability in profiling individuals (higher insurance fees for those with increased chances for illnesses, probation decisions based on statistical models of return-felons, even preemptive arrests in the nightmare scenario). The writers say this is the opposite of free will and responsibility, the philosophical underpinnings of democracy and a free market economy. For good measure, there is even a chapter on suggested solutions to these challenges. A new class of professionals—”algorithmists”—should ” – should be brought on to evaluate the consequences of big data applications, shifting the focus of privacy from the collection of personal data to accountability for how it is (misused).

All this is good and thought-worthy, but only a couple of months after the book was published, the greatest big data risk of all was revealed by NSA whistle-blower Edward Snowden: the fact that personal data, once collected, can not only be sold but will leak to secret agencies with computers big enough to find any needle in a petabyte haystack. If only the book had come six months later, it would have taken that aspect into account! How about a couple of new chapters for the next reprint, Professor Mayer-Schönberger?