Remember that time when Target sent baby equipment coupons to a Minnesotan teenager because they deduced that she was pregnant from data mining? Although many people found this incident creepy, Jordan Ellenberg examines it in a new light and instead focuses on how algorithms still have a lot of room for improvement. He notes that while many algorithms seem to get more and more advanced each year as we feed them an exponential amount of data nuggets, predicting human behavior is more nuanced and precision can change easily.
In a great association, Ellenberg compares humans to weather, but notes the extra difficulty: “In at least one respect, human behavior ought to be even harder to predict than the weather. We have a very good mathematical model for weather, which allows us at least to get better at short-range predictions when given access to more data, even if the inherent chaos of the system inevitably wins out. For human action we have no such model, and may never have one.”
The notion of an algorithm incorrectly predicting what grocery item you would want a discount on does not seem like a big deal, however there are much bigger, more significant situations where correct data predictions are imperative (Ellenberg points to terrorism, for example). We live in a world of big data, and the technologies and tools used to assess and analyze it should aim to be as correct as possible.