CU Law professor Paul Ohm. (Photo by Glenn Asakawa/University of Colorado) THE UNDERWHELMING BENEFITS OF BIG DATA (image via

Paul Ohm – The underwhelming Benefits of Big Data

“The cloud is a hodgepodge, and Paul Schwartz, in his rich Article, Information Privacy in the Cloud, tackles many different parts of the confusing combination, giving meaning to mush in his characteristically careful style. Consider his thoughts on the changes being wrought to information privacy law by the move to “networked intelligence in the cloud.” This expression refers, at least in part, to what others have been calling “Big Data,” the trendy moniker for powerful new forms of data analytics. Professor Schwartz weighs the benefits of Big Data techniques against the risks they pose to privacy. Better than some others, he takes care to point to the benefits that truly matter. Too many commentators have too often overstated the benefits of Big Data, inflating studies and praising the merely trivial.

If I do not acknowledge this near the outset, some will misinterpret or misrepresent the point of my Response, claiming falsely that I do not agree with the following patently true prediction: Big Data will lead to important benefits. Whether applied to crises in medicine, in climate, in food safety, or in some other arena, Big Data techniques will lead to significant, new, life-enhancing (even life-saving) benefits that we would be ill advised to

electively forego. This prediction is both obvious and somewhat uninteresting. It flows from the definitional squishiness surrounding the phrase: “Big Data” has become nearly synonymous with “data analysis,” and data analysis is a lynchpin of modern science. To argue against Big Data is to argue against science. That is not my brief.

But some Big Data projects will also lead to bad outcomes, like invasions of privacy and hard-to-detect invidious discrimination. Big Data techniques can help governments spy on their citizens and criminals prey on their victims. As we worry about these negative consequences, and particularly as we consider whether we might forego or shape some forms of Big Data so as to limit their negative effects, we must weigh the associated costs and benefits. In doing so, we should scrutinize carefully claims that the benefits of Big Data outweigh the costs to individuals and society. Too often, when Big Data’s cheerleaders talk about its benefits, they blur the significant with the trivial, the important with the frivolous. Big Data’s benefits are real and important; we should give less attention to those benefits that are not.

Big Data’s touted benefits are often less significant than claimed and less necessary than assumed. Professor Schwartz, however, is not really guilty of overstating Big Data’s virtues. The benefits he touts focus in particular on “information-based forms of health research [that] ‘have led to significant discoveries, the development of new therapies, and a remarkable improve- ment in health care and public health.’” He cites sources that herald new discoveries resulting from advanced data analysis in the treatment of breast cancer, Alzheimer’s disease, and thrombosis. These benefits seem unim- peachably important, even if the specifics remain somewhat underdeveloped in Schwartz’s Article.

Consider in stark contrast a project Schwartz does not mention, but one that has often served as the poster child for the positive benefits of Big Data: Google Flu Trends. Flu Trends is a project that Google’s philanthropic arm,, launched in 2008.8 To test the theory that one might predict the parts of the world suffering from flu outbreaks by watching the symp- toms people type into the Google search engine, Google gave its internal researchers access to its users’ search queries. It turns out the theory works, and Google reports that it can detect likely flu outbreaks a week or two faster than the physician-reporting surveillance efforts of the Centers for Disease Control and Prevention (CDC). To showcase the project, Google publishes an interactive website displaying maps that reveal flu outbreaks around the world, color coding cities, states, and nations according to the estimated prevalence of the virus, in hues ranging from reassuring greens to ominous reds. Almost nobody has anything bad to say about Flu Trends. It represents the triumph of Big Data over illness and potential death, abetted by pervasive surveillance.”


Paul Ohm

The underwhelming Benefits of Big Data


Respond to Paul Ohm – The underwhelming Benefits of Big Data

What do you think?