Everything we’ve done up till now is about passive characterization of the world. But what if you actually want to DO something?
The field of causal inference—the practice of trying to empirically measure causal relationships—has its origins in philosophy and social science. And much of that is concerned primarily with doing causal inference for the sake of understanding phenomena in the world. For many data scientists, however, the field of causal inference is not of interest primarily because it always us to seek understanding in its own right, but rather because we wish to better predict the consequences of actions we wish to take in the world.