Tom Chivers' book Everything is Predictable is a fantastic dive into a formula known as Bayes' theorem. A book about a formula sounds extremely boring, but its application is extremely useful in everyday life.
Humans appear to have a built-in mechanism for employing something like Bayes' theorem when we encounter new information, even when we don't realize it. The author cites a number of examples. It also helps explain why we have different reactions to new info: our different priors.
But when we use the theorem explictly, we can really harness its power. The authors pits Bayes' type thinking against other probabilistic schools (like stats), but I think they can all be very useful depending on the circumstance.
The most interesting part of the book is the real-world applications that Bayes can be applied to. For example, I've previously believed that medical pre-screens should be how we conduct health-care, in order to find little problems before they become big problems. With the help of some calculations as illustrated by Chivers', I now understand this presents its own set of problems that may cost more than the benefits!
If you think probabilistically, I think you will enjoy this book.
No comments:
Post a Comment