"Noise - A Flaw in Human Judgement" by Daniel Kahneman, Olivier Sibony and Cass R. Sunstein

Book review "Noise"

Nobel prize winner Daniel Kahneman did it again. This time together with Olivier Sibony and Cass R. Sunstein, he manage to write another groundbreaking book, this time: "Noise - A Flaw in Human Judgement".

Well, I did it again too... Manage to read Noise from cover to cover, like it did with his book "Thinking, Fast and Slow". I have to admit it took me quite some time to reach each book and every time I have read (and finished books in between). The material is very academic and dry. But, very much worth the effort.

In a nutshell, Noise uncovers and empirically proves the flaws in human judgement which is a big societal problem. For those think noise refers to what we hear, that's not the noise they talk about. It's the noise in data and in judgement. So margins of error. Esther Duflo, the 2019 Noble prize winner for Economics summarised this book well: "The greatest source of ineffective policies is often not biases, corruption or ill-will, but the three "I"s: Intuition, Ignorance and Inertia. This book masterfully demonstrates why the three are so pervasive, and what we can do to fight them. An essential, eye-opening read."

The writers created six principles to define decision hygiene and how to address the psychological mechanisms that cause noise:

Given the multiple sources of individual differences we should not expect even the best judges to be in perfect agreement on complex judgement problems. The infinite variety of backgrounds, personalities, and experiences that make each of us unique is also what makes noise inevitable.

One strategy for error reduction is debiasing. The writers main suggestion for reducing noise in judgement is 'decision hygiene'. The writers created six principles to define decision hygiene and how to address the psychological mechanisms that cause noise:

• The goal of judgement is accuracy, not individual expression.
• Think statistically, and take the outside view of the case.
• Structure judgements into several independent tasks.
• Resist premature intuitions.
• Obtain independent judgments from multiple judges, then consider aggregating those judgements.
• Favor relative judgments and relative scales.…

You might have already concluded that a solution could (or should) be to use more machine learning and artificial intelligence to eliminate bias and noise... That's what is definitely philosophised in the book to. But they do propose a human fail save...

Long story short, this book is very interesting for any academic researcher, psychologist, AI programmer and every one just interest in how the psyche works.        

Long story short, this book is very interesting for any academic researcher, psychologist, AI programmer and every one just interest in how the psyche works.

#Noise #NoiseBook #DanielKahneman #OlivierSibony #CassRSunstein #MustRead #Research #Bias #Objective #Subjective