At its core, Bayes’ theorem is not just a mathematical equation, a formal method for making predictions; it is also a probabilistic and dynamic way of looking at the world. The essence of this approach is the process of Bayesian updating–simply put, the ability to update one’s priors about how the world works whenever one receives new evidence.
Babies and small children are thus Bayesian creatures in many ways. Take for, example, the latest addition to prior probability‘s family, Baby Adys Ann. It would be safe to say that she had no priors or dogmas about the world when she was born on 6 August (Jamaica Independence Day). But over the first few months of her life, might she be developing a set of priors about her little world, such as the causal relation between her squeals of hunger and breast milk? For example, she likes to make a small fuss whenever she is hungry (and for other reasons as well), and each time she begins to fuss, someone who loves her will immediately attend to her needs. Her mother, for example, will nurse her at regular intervals, especially when Baby Adys begins to send out the appropriate signals. So doesn’t Baby Adys, then, update her priors in the sense that she expects, with a high degree of probability, to be nursed whenever she is hungry?
prior probability 