# The Book of Why Author: [[Judea Pearl]] ## Review Great initial introduction to the 'Causal Revolution' explaining why we need it, the history of causality, and what the causal methods are. It surpassed my expectations and I've already started digging into more of the details. Would recommend this to anyone interested in the philosophy and history of science, or into epistemology. He frames our ability to understand the world by how well we can mathematically or algorithmically express our thoughts. Words like should, regret, blame, reasons are part of causal language we use to describe the world. We didn't have a way of expressing those ideas mathematically before. If we have knowledge or intuition of how the world (or some subset of it) works then we should be able to encode that knowledge (even partial knowledge) and pair it with data to understand effects through causal inference of our model. Pearl is also the inventor of Bayesian networks so it's not surprising to me that everything is represented graphically with networks. That also intuitively is how I think about cause and effect relationships. The tools and jargon of 'Causality' are in my opinion a superior way of organizing and describing how we think about the world. From personal experience, I think many people don't think hard enough about the causal model that generates data and they resort to thinking superficially about the associations between data. I try to think about the mechanisms that generate the data, and the way he frames his tools and the language he uses is very helpful for thinking clearly about cause and effect. The most important aspect of 'Causality', in my opinion, is how to initially come up with a model of cause and effect. This aspect is called causal discovery and seems to be a developing part of the field. ## Key Ideas ## Related