143 of 165 people found the following review helpful
Mixed Feelings About This Book,
Verified Purchase(What's this?)
This review is from: Thinking, Fast and Slow (Hardcover)
As one who was brought up with Herbert Simon and"satisficing," I have mixed feelings about this book. As an intelligence professional I know for a fact that corrupt politicians have zero interest in the facts, only in what will profit them personally in the short-term. As much as I would like to see integrity restored as the core value of government, economy, and society, in the larger context in which we live this book is a curiosity.
There are gems and it is certainly worth reading, but as one other reviewer points out, it is not the easiest reading nor the most delightful. Here is what I got out of it (my summary notes, I donate all books right after I read them, to a nearby university).
For those instances when BOTH intelligence (decision-support) officers and their clients (politicians, policy makers, acquisition managers, operational commanders) have integrity--a condition that does not exist today, this book is very useful as a training aid.
01 It strives to provide a deeper understanding of judgments and choices by humans.
02 It fully documents the biases of intuition (judgment informed by past cases)
03 It documents the fact that decision making under uncertainty leads to humans being too prone to believe findings based on inadequate evidence, and too prone to avoid collecting a sufficiency of observations or research findings by others.
The essence of the book is the author's distinction between System 1 and System 2.
System 1 is automatic, fact, and falls prey to illusions.
System 2 is controlled, slow, requires attention, and is easily distracted.
Conclusions about judgment heuristics (rules of thumb):
HARD to think statistically
EASY to think associatively
We have EXCESS CONFIDENCE in what we think we know, and a deep, deep, deep inability to acknowledge our ignorance.
Humans DEVIATE from rational model with two major CORRUPTIONS:
01 Treat problems in isolation instead of as part of a systemic whole
02 Treat problems in relation to framing effects that distort perceptions with inconsequential trivia
QUOTE (34): We found that people, when engaged in a mental sprint, may become effectively blind."
That one sentence made the book worthwhile to me. I have long been a fan of Red Cells and walk-abouts and other forms of being forced to engage outside the box, this one sentence reminded me of my now firm view that all analytic teams need an independent Yoda to challenge them.
The author surprises me with a substantive discourse on how money has caused people to collaborate less--it makes them more independent of one another and displaces the social value of collaboration.
I am fascinated by the author's focus on surprise as a litmus test for the extent to which we are open.
He emphasizes that hypotheses should be confirmed by trying to REFUTE the hypothesis rather than by searching for additional supporting evidence. Having the hypothesis is enough. If it cannot be refutes, THAT is worth much more than a documented but not seriously challenged hypothesis.
QUOTE (117): The tendency to see patterns in randomness is overwhelming.
This in the context of the anchoring effect, and the stark strong impact of preconceived notions that shape perception of inclusive into conclusive, contradictory into confirming.
The discussion of risk, for an intelligence professional, is very very interesting. The author focuses on how critical it is to actually have a measure of risk--to know with some precision what you are defining as risk, why.
I am blown away by a discussion that makes it clear that praise or punishment are generally irrelevant for professionals. They tend to do the best they can, and the odds are such that praise or punishment have no effect but appear to have effect because they zig zag along a mean. "It is what it is." I connect this with the National Football League, and how calm most team players are when they miss a catch or a block. "It is what it is."
I'm not sure this is the book I would want for advanced intelligence courses, but I really like chapter 19 on the illusion of understanding and chapter 20 on the illusion of validity. Even if our political officials are corrupt, we intelligence professionals should at least strive to get it right.
The author is a believer in algorithms over experts. I have mixed feelings about that. Certainly I agree that most experts are wrong and much narrower in their understanding than common competence requires, but I am also very skeptical of algorithms, witness Google's math hacks against digital garbage. As a believer in collective human intelligence (citizen wisdom councils, etcetera), I accept the importance of taking algorithms as far as they can go, but algorithms are no better than the humans who constructed them and the data known to the humans at that time.
I give the author great credit for providing a superb overview across the book of stars in this field. This is not a selfish or self-centered book--it appears to do full justice to all others.
Great thoughts in this book:
01 Capitalists--both inventors and entrepreneurs--overestimate their success rate by two times.
02 Illusion of control is increased by a failure to seek out data from others [this is one reason I champion M4IS2--Multinational, Multiagency, Multidisciplinary, Multidomain Information-Sharing and Sense-Making) and public intelligence in the public interest).
QUOTE (262): Organizations that take the word of overconfident experts can expect costly consequences.
NEW TO ME: Psycho-physics--relation of mind and matter. This may be a different way of talking about quantum physics, but it is one more indication that mind-matter interfaces are going to be a huge area of study in the future.
The author confirms Machiavelli 101--defenders of the status quo are always stronger than reformers seeking change [he does not say this but Kuhn and others do: UNTIL the status quo self-destructs from its own corruption, and the reformers are free to build on its ashes].
On page 411 he provides a very serious critique of libertarianism, pointing out that libertarians assume all individuals are rational and see no value in aggregate services.
QUOTE (417): Observers are less cognitively bvusy and more open to information than actors.
The more I think about this, the more I think that we need a new class of intelligence professionals who are neither collectors nor analysts, but observers "in situ" with decision-makers or "in situ" with crisis situations, and they provide the "third eye". I am writing the chapter on "The Craft of Intelligence" for the next Routledge Handbook of Intelligence Studies, and this is one new idea that I credit to this book and author, that I plan to integrate into my thinking about the future of intelligence as a discipline.
There are two appendices and an excellent index.
As is my custom, I always use Amazon's link feature to point to other related books. Here are ten in the decision-making arena that I consider especially valuable.
Radical Man: The Process of Psycho-Social Development
The Knowledge Executive
Tools for Thought: The History and Future of Mind-Expanding Technology
Thinking in Time: The Uses of History for Decision-Makers
Planning with Complexity: An Introduction to Collaborative Rationality for Public Policy
Business War Games: How Large, Small, and New Companies Can Vastly Improve Their Strategies and Outmaneuver the Competition
Reflexive Practice: Professional Thinking for a Turbulent World
Open Space Technology: A User's Guide
Atlas of Science: Visualizing What We Know
Holistic Darwinism: Synergy, Cybernetics, and the Bioeconomics of Evolution
Tracked by 2 customers
Sort: Oldest first | Newest first
Showing 1-8 of 8 posts in this discussion
Initial post: Jan 21, 2012 6:49:47 AM PST
John Harllee says:
"algorithms are no better than the humans who constructed them and the data known to the humans at that time."
Kahneman would say that the algorithm IS better than the human who constructs it, because it is not subject to the cognitive failures that a human is. But regardless of the reason, Kahneman marshals hard evidence in Chapter 21 that algorithms are in fact superior to human experts in many areas. You shouldn't let your intuition lead you to disregard the evidence!
Thank you for an excellent and useful review.
In reply to an earlier post on Jan 21, 2012 7:04:03 AM PST
Last edited by the author on Jan 21, 2012 7:04:26 AM PST
I'll meet you halfway. I spent a tour at CIA in the advanced information technology arena, specifically focused on expert systems and heuristics. Experts are not really expert. At the same time, computational mathematics is in its infancy--Google evil and Wolfram good. So both are needed, but so also (this is going into my next version of The Craft of Intelligence--are citizen wisdom councils, red teams, independent observers etc.) See my review of Jim Bamford's book on NSA, Body of Secrets: Anatomy of the Ultra-Secret National Security Agency and the last sentence in the book.
Posted on Jul 25, 2013 1:38:48 AM PDT
Amazon Customer says:
I really like the manner and structure of your review of this book. A difficult book to properly characterize I think. You did a fine job of it however. Thanks.
Posted on Aug 1, 2013 3:53:39 PM PDT
Last edited by the author on Aug 1, 2013 3:56:03 PM PDT
"Kahneman would say that the algorithm IS better than the human who constructs it, because it is not subject to the cognitive failures that a human is. "
If so, Kahneman would be wildly wrong. How can you write this with any degree of intellectual honesty or a straight face? Have your biases gotten the better of you?
Have you noticed the Financial Crisis that brought us the Great Depression II? The fall of the House of Mortages was brought on by the very best, the very brightest writing risk assessment programs (algorithms) that were completely flawed by assumptions. All of it managed by people who saw what they wanted to see, and who wanted to see what they saw. Absolutely zero attempt to formulate or disprove a null hypothesis.
Bad algorithms written by flawed people get you this. The algorithm always reflects any cognitive failures of its human creator.
In reply to an earlier post on Aug 1, 2013 4:05:17 PM PDT
Agree. And the evidence is in across many fronts, networked humans with agility and integrity beat computers with old algorithms every time.
Posted on Aug 5, 2013 1:26:30 PM PDT
Scott D. Barton says:
Unusually thoughtful review. It has inspired me to buy the book.
In reply to an earlier post on Aug 5, 2013 1:36:42 PM PDT
Kind of you to comment. All my other reviews are best accessed via Phi Beta Iota so you can browse by category, all reviews lead back to their Amazon page.
Posted on Sep 5, 2013 4:47:44 PM PDT
"QUOTE (417): Observers are less cognitively bvusy and more open to information than actors."
As in, you learn more by listening than by talking? :)
Sorry, but I can't help smile when I see folk wisdom dressed up in 'serious' words.
‹ Previous 1 Next ›