Soul of a New Data Machine

In another life, I write about comedy performance—reviewing shows and specials, and interviewing comics and comedic actors. As part of that, I'm an avid podcast listener. None of that ever seemed likely to overlap with what I normally cover here, until I caught up to some remarks from author Douglas Rushkoff in a recent appearance on comedian Marc Maron's podcast, where Rushkoff spurred an exchange with Maron about big data.

Rushkoff writes mostly about online media, but his latest book, Present Shock, contains a few short segments relevant to financial services operations, on how collection of big data operates, as well as how using algorithms can wreak havoc, as they did in the trading space when BATS Global Markets debuted its IPO.

Responding to an analogy from Maron about "big data" and "Big Brother," Rushkoff said, "We would like to believe big data is personified and there's a guy at the top of the corporation collecting it, but there isn't. What we're really doing is programming our technologies to extract more value from us. We're the shareholders on the other end."

Months ago, David Saul, chief scientist for State Street, proposed the concept of "smart data" as preferable to "big data." Rushkoff echoes what Saul was driving for. Although Rushkoff was talking about how "big data" can be used in retail, to get more intelligence on customer buying trends, he still was emphasizing how "big data" can be established and applied intelligently, as financial services firms are trying to do. From Saul's perspective, data management programming ought to extract value from data by adding the results of risk calculations to the data.

Nearly a year ago, BNY Mellon's Dennis Smith described the challenges inherent in efforts to draw insight from high volumes of data. Foremost of those challenges was getting high enough data quality when validating collected data to conduct the kind of risk calculations and risk management State Street's Saul aspires to support.

The challenge of harnessing big data continues for the financial industry. Will the industry learn what to do with it just by deploying available technology to find and deliver the data's value? Or does big data need to be "personified," as Rushkoff puts it, with someone at the top of the organization, or at least close enough to the top, like a chief data officer, collecting it and making use of it?

Rushkoff's intent in his remarks likely was to express caution about who makes use of personal data and how, but in financial industry data management, his questions suggest a different problem—that computing power by itself isn't enough to make big data valuable for the industry's purposes.