Economics has been long striving to predict our behaviors. Unlike natural science, social science deals with far more noise. There’s always noise and error present, not only from the technical errors that come from sampling and other sources, but also from the inherent inconsistencies and occational irrationality of people. This is perhaps why we need an economist, to predict the future that is created by the aggregate of these unpredictability, in order to better prepare ourselves for the uncertainties.

The question is though, what if it were possible to perfectly predict the future? What would that mean? Imagine a situation where we could perfectly foresee the future given certain conditions. In fact, the prediction could be done at such a micro level that we could actually predict each individual’s behavior based on their historical data as well as the data they are producing at any given moment. What would that entail? Maybe this is more than just an empty question but involves a much deeper insight into philosophy as well. The question then becomes: Is this reallly what we want?

Free will is one of the factors that define humans. However, artificial intelligence, deep learning, machine learning, and other modern technologies imply that our behaviors are somewhat predetermined, that if we are given certain conditions and those conditions are previously known, we could calculate exactly what we are going to be doing next. This would be directly counter to the notion of people’s free will. And even worse, by manipulating those factors that influence our behavior, our behaviors can be controlled with extreme subtlety.

I used to think the world described in The matrix was fictional but it seems more and more like our world is becoming something like it.