DATA SCIENCE FICTION

Just like science-fiction is not about an alternative dystopian reality, predictive analytics is not really about the future.

The genre of science-fiction is often associated with constructs of our society taken to an extreme. A despotic regime or no form of governing rules; extreme stratification of society or complete physical and cognitive monogamy. Although in some cases it is also a fully functioning society with one key social construct eliminated or a synthetically engineered scientific product inserted. Due to this conjured image of an extrapolated story, science-fiction readers (or writers) are often believed to be escapists or daydreamers who do not enjoy being involved in the complex moral and social constructs of our world.

However, I argue that this is quite the contrary. As suggested by Le Guin, a science-fiction novel can be interpreted as thought experiment. By taking an existing society and making a level of modifications to it, the goal is to understand how individuals would behave in that system, what the immediate results would be, how it would reach a state of equilibrium, what problems might arise, and, most importantly, how the human beings of today would interpret such a world given the prevailing paradigms and biases that exist in our time. Through the thoughts and actions of realistic, yet fallible characters readers can construe and hypothesize whether the dysfunctions and prejudices that plague our world can in fact be amended by those same flawed human beings. Or, pose an scenario wherein the continual effects of our actions today could result in unsurmountable predicaments.

Most key decisions of the world today are based purely, and sometimes blindly, off a sequence of binary digits, depicted by certain numbers that are representative a best-case scenario which would not be as convincing if it were put forth semiotically or simply based off human-generated thoughts. Although it physically remains the same in size, the world continuously expands as a macrocosm with complex and often unseemly connections across its entities, all centered around an irrational anomaly—human behavior—making any realistic prediction of a future scenario an unrealistic hope.

So the answer to why we value the outputs of machine learning and predictive analytics that possess clairvoyant abilities lies in the very question itself. To understand the far-reaching impact of a change in our large interconnected society, we need to devise our own microcosm where we have control over the variables holding this web together, to understand how they behave. We need to be the authors of this miniature universe or alternative reality, as you may, to create rules, stratify society and inject new synthetic products all just to understand the intricate ways of its very creator. Through the construction of these controlled environments we don’t just get a glimpse into the possible futures, but the very nature of our present, and how we must equip ourselves to navigate this irrational world that we habitat.

Sarah Doctor is a data analytics graduate student at Columbia University. While she is pursuing a technical degree, she is a lover of the arts and loves to draw parallels between the disciplines. She is particularly fascinated by linguistics and using literary constraints to enhance creativity.