I find big data to be a fascinating subject. There are quite a few benefits that it offers scholars in all fields, especially insofar as algorithmically searching large, unwieldy data sources is concerned. Some scholars would approach the interaction and interrogation of big data with some hesitation and would remind of us of the multiple shortcomings that it offers as a research tool. Of course, they still praise its benefits – and rightly so, these scholars aren’t scared to use big data, they just want to feel out the limitations that it presents and to ensure that those limitations are recognized as such. Lev Manovich wrote about some of these problems, and one concern that he noted in particular struck me as being somewhat interesting because it seems to go against a lot of what post-modern scholars would stipulate is a very normal means of interacting with the world.
In particular Manovich cautions the user of big data research to be wary of taking the views and ideas expressed via social media at face value. His concern is that there is potential lack of authenticity at work in a socially constructed environment. I find the assertion to be intriguing, because it almost seems to necessitate that the private sphere is more authentic than the public sphere, disregarding the conception of the multiplicity of Self, or the notion that people might be an amalgamation of varied Selves rather than one consistent Self. In any event, I think that this is an interesting conception that might problematize “deep data” as Manovich calls it (because sometimes it lies to you, or tells a truth that you simply don’t understand in the moment).
Do we attempt to discount these “inauthentic data”? Perhaps we simply let our algorithms gather all the data that they can (and if they gather too much data, maybe we put some algorithms in those algorithms to sort those data) and assume that brute forcing extraordinary amounts of data relegates inauthenticity to the realm of statistical irrelevancy? In that light do we think that the collection of enough “surface data” can lead to “deep data” or that it might simply inform avenues of exploration that we might not have thought of previously? Are the limitations of big data only limitations until our technological prowess is advanced enough to construct an AI that’s so intelligent that it can become self-aware, ushering in the era of our new computer overlords?