The 20-Year-Old Point-of-View: Noise by Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein
By Whom: A team of top professors from Princeton, Oxford, and Harvard whose specialties include psychology, behavioral economics, business and decision making.
For Whom: Judges, Educators, Admissions Officers, Supervisors, Leaders, and Anyone Who Makes Important Judgements
Noise may have the potential to be the next big breakthrough in the study of decision making, and for good reason. One of the authors, Kahneman, is the author of the landmark title Thinking: Fast and Slow, and this book, combined with the help of Sibony and Sunstein, strives to perform a similar task that Kahneman’s first book did: completely revolutionize the way we think about a certain aspect of ourselves, this time in the realm of judgement and decision making.
What the authors of this book have essentially done is take the past 100 years of thought and study on Noise, the variability with which people make judgements. Anyone who’s taken an introductory statistics course will be familiar with the concept, under the names variability or consistency. In a little under 400 pages, they’ve managed to define the concept, point out its impacts, give reasons why it exists and prescribe methods to prevent its most damaging effects. They provide vivid examples of its implications from the fields of artificial intelligence, justice, politics, business, college admissions, grading, employee evaluations, hiring, and more. And yes, there’s a mention of the Oakland A’s.
While I’ll save the technical details for the reader, what Noise was to me was an exploration of the neglected half of error in human judgement. Bias gets lots of press and attention, but noise has stayed relatively in the shadows. The authors take on the task of righting this imbalance, with historical anecdotes, formalized studies, and example implementations all included. If there’s anything that Kahneman has learned from writing his first novel, it’s to sometimes let the details of the situation fall away to give space to the more important messages.
That’s all fine and good, but it’s still 400 pages. While I usually put a criticism here, I can’t really make a comment here that isn’t already addressed by the authors. If one isn’t interested in reading on exactly how they discovered noise or its effects, they suggest skipping specific chapters to get to the applications. While there aren’t many concrete examples of denoising in the real world, they admit that it’s a fairly new concept that not many real companies have been willing to commit to. So all that I can say is that it’s long, but one need not read it in entirety to understand its potential.
Kahneman, Sibony, and Sunstein have created a book that may well change how we view the reliability of human judgements well into the future. There’s so much good stuff in here for everyone that I think I’ll base a lecture off it sometime soon.
5/5. Extremely important.
