Big Data Can't Solve Big Problems—Yet: Sean Gourley at The GigaOM Conference
Currently, big data measurements are only capable of solving simple problems. That is to say, big data is not adept at solving big problems. "We're sitting here with one of the biggest technologies we've ever invented, and we're using it to solve trivialities." We have complex data-driven models that teach us the precise color combinations of cereal boxes that will sell the best, and where we have to place them on the shelves to increase profits, Gourley says. However, this data science is nowhere to be found when it comes to solving the problem of childhood obesity driven by the consumption of sugary foods. Not only that, but he says the future of data science, if it progresses in the same way it has thus far, is likely to only offer findings that are largely irrelevant in the scheme of larger social problems. If data scientists were to run programming to find a correlation between Facebook likes and intelligence, he quips that they could conclude that people who like curly fries are among the smartest on the social media site. While the audience laughed at his joke, he says this is a prime example of why data science is not being harnessed to its full potential.
Data is incredibly powerful, he adds. That's why he believes that we need to re-imagine the way we look at data to harness that information. Instead of focusing on data science, Gourley suggests we place our focus on data intelligence. To analyze the issues he does—modern war and conflict—you need to put humans at the center of the data analysis, instead of computers. To solve the complex problems that affect our society, you must look at the messy components of human interaction that isn't necessarily understood through the narrow focus of data science. This, he says, is why he formed Quid. The organization provides strategies for augmenting our ability to understand the complex world around us. Right now, big data cannot solve big problems. But eventually it can, he believes, and it must. And to do so, we must inject the human element back into the method and focus on data intelligence—not just data science.