Noise: A Flaw in Human Judgment

"The central theme of this book is super cool for anyone who has learned about bias. Much of Daniel Kahneman’s work, along with that of Amos Tversky and the start of behavioral economics, is centered around the idea of cognitive bias — systemic errors in our thinking that change our judgment in a specific way, usually caused by incorrect applications of heuristics. His most well-known book, Thinking, Fast and Slow, as well as his seminal work in the field , explore these ideas and how we can recognize our biases and go from there. Noise is a really cool sequel. It’s what the second book of Dune ( Dune Messiah ) was to the first book of Dune. The first book of Dune depicts an omniscient, omnipotent figure becoming a leader, and the second book of Dune is about how that’s not a good system of governance for anyone involved. Noise is about the fact that while we have a lot of cognitive biases, in many cases the world is random. It’s what gamers call RNG or random number generator — a fun shorthand way to describe all the randomness in a game’s world, randomness that’s mirrored in ours. It creates this variability in our observation of the world and makes it even harder to make good decisions, because not only do we have to fix for all our biases and get all the systematic priors out of the way, but even accounting for all of that, testing the same samples six times may sometimes yield very different results. The book has two particularly useful sections. The first describes what noise is, and the second does a noise audit to reflect upon that and figure it out. I would recommend the book just from the noise audit perspective. The way I’ve done it, which I find easiest, is to spend a lot of time upfront writing down my priors. ‘Priors’ is a term in probability and Bayesian inference, which can be thought of as our background knowledge or assumptions about a subject before we gather any new information. I write down my priors and why they are my priors. Then, when I observe the outcome of a decision, I don’t ever grade it at 100 or 0. I don’t guarantee that it’s right or that it’s wrong. I’m thinking probabilistically: ‘We can assume this is 70 per cent right. There’s probably enough noise that I could be wrong.’ When making decisions in my personal life, in my professional life, and in investing and business, I echo Paul Saffo’s philosophy of having a lot of strong opinions, weakly held. In part, that’s because I don’t know what noise was involved in that decision-making."