The Big Book Review: "Thinking, Fast and Slow" by Daniel Kahneman (Pt.3)

Read Parts 1 and 2 here:

Welcome back to this series on Thinking, Fast and Slow by Daniel Kahneman, part of the 'Big Book Review' in which we look at sections of a book every month in extreme detail, focusing on what they have to teach us about their topics. Part 3, entitled Overconfidence looks at what businesses, experts and individuals may overlook or misread due to their own faith in themselves. A more extreme and intricate form of the 'Dunning-Kruger Effect'. If you haven't read parts 1 and 2 then I suggest you read those before continuing, of course these articles will cross-reference each other and ideas from previous chapters are not going to be re-explained unfortunately (for the sake of length and word count).

"Thinking, Fast and Slow" by Daniel Kahneman (Pt.3)

Chapter 19 begins with The Illusion of Understanding in which Kahneman states that we believe that our analysis of how the past unfolds is an understanding or an interpretation of understanding, but we are very wrong about this. He focuses on what is called the "narrative fallacy" in which "flawed stories of the past shape our world and our expectations for the future" (p.199). He looks at the difference between different types of stories we tell ourselves. Here's what he has to explain to us about the way in which we interact psychologically with these stories:

"Good stories provide a simple and coherent account of people's actions and intentions. You are always ready to interpret behaviour as a manifestation of general propensities and personal traits - causes that you can readily match effects. The halo effect discussed earlier contributes to coherence, because it inclines us to match our view of all the qualities of a person to our judgement of one attribute that is particularly significant" (p.199).

Our 'matches' and 'assumptions' therefore, are things we do with confidence, but they are nothing more than 'assumptions'. A good story is nothing more than a confidence trickster and thus, your interpretation of behaviour at that one given time is something that is probably not very accurate. Kahneman has given us an anecdote to explain this idea of 'knowing' in which it is clear that there is no 'knowing' at all. He uses the 2008 financial crisis to show us that he has found many people claiming that they knew it was going to happen, often berating the very word 'knew' in order to get to the bottom of all of this:

"In everyday language, we apply the word 'know' only when what was known is true and can be shown to be true. We can know something only if it is both true and knowable. But the people who thought there would be a crisis (and there are fewer of them than now remember thinking it) could not conclusively show it at the time" (p.201).

The idea that optimists who are known to also have some form of expertise would know a catastrophe was imminent does not align with the ideas we will learn about optimism later in this section and therefore, we can only assume that Kahneman is correct in his analysis. Premonition may be reserved for things that have been thought to be true, but there is nothing past that - no understanding, no new knowledge and definitely no such thing as 'intuition'.

This brings us on to hindsight and how bad it cane be for thought. Our own minds make up narratives to better understand things. But Kahneman says that even though this is perhaps a better idea than not understand at all, there is still something wrong, something that will cost us our rationality:

"A general limitation of the human mind is its imperfect ability to reconstruct past states of knowledge, or beliefs that have changed. Once you adopt a new view of the world...you immediately lose much of your ability to recall what you used to believe before your mind changed" (p.202).

And judging by the amount of times a human being changes their mind and world views throughout their lifetime, we can assume that narratives are probably not the best way to practice looking at hindsight or any other form of 'intuitive belief' system. Hindsight itself is a bias, Kahneman states in his analysis. It assesses not the process, but the outcome (p.203) and is therefore completely irrelevant when it comes to studying statistics or the probabilities of whether an outcome will be good or bad. Kahneman looks at this regarding his statement on risk aversion vs. risk taking:

"Although hindsight and the outcome bias generally foster risk aversion, they also bring undeserved rewards to irresponsible risk seekers, such as a general or an entrepreneur who took a crazy gamble and won" (p.204).

This basically shows us that most of all, hindsight deals with outcome bias, we are biased to like the outcome of good, hate the outcome of bad and ignore the process it took to get there entirely. We are therefore doomed to repeat the whole thing without a greater understanding of how we got to the end result.

System 1 is to make sure we have something tidy and coherent but the illusions it brings us are, as Kahneman states, "comforting" (p.204-5). Kahneman also talks about the 'halo effect' as being one of the reasons why books about booming business (and also adversely) bankrupt businesses do so well. They seem to put so much of their focus on the leadership qualities of the people involved and puts nothing down to luck or randomness, or even a regression back to the mean (p.206). There is technically nothing mathematical or statistical about these stories, they are purely coherent and plausible stories that the mind, especially that of system 1, can understand and manipulate. Kahneman explains this:

"Stories of how businesses rise and fall strike a chord with readers by offering what the human mind needs: a simple message of triumph and failure that identifies clear causes and ignores the determinative power of luck and the inevitability of regression" (p.207).

So if system 1 hates randomness, and hindsight makes events feel predictable, our memory seems to get rewritten to fit a story of inevitability. Thus, we think we understand the past far more than we actually do, leading us to believe we can predict the future.

Chapter 20 deals with another illusion, this time about validity. This is basically the fact that we often feel confident in our judgments even when the underlying evidence is weak, ambiguous, or completely useless. Again, we deal with system 1 which Kahneman states is "designed to jump to conclusions from little evidence - and is not designed to know the size of its jumps" (p.209). This makes the statement Kahneman writes on confidence all the more important and true:

"Subjective confidence in a judgement is not a reasoned evaluation of the probability that this judgement is correct. Confidence is a feeling, which reflects the coherence of the information and the cognitive ease of processing it" (p.212).

If system 1 tries to make narratives coherent for the mind to process, and if cognitive ease allows us to process these quicker then overconfidence can be a problem attributed to a lack of understanding and a literal jumping to conclusions. And of course, personal experiences prove to be more solid in the human mind than facts which means that even though facts may "challenge" our need for narratives, it does not necessarily change our reaction towards it. Therefore, without the facts but with the ability for pattern recognition confidently, we often assume that consistency means validity without any real conclusive evidence.

Kahneman counteracts our wants for confidence in consistency by stating that "the world is unpredictable" (p.220), a claim he has made before and continues to make in the text to make sure we understand the randomness of events contrary to how our 'storytelling' minds work. And thus, the feeling of 'I know this' often reflects narrative satisfaction, not real predictive power. Confidence does not equal accuracy.

In Chapter 21, we look at intuition and formulae and learn an uncomfortable truth that I didn't particularly want to learn either, but here we it is: statistical algorithms consistently outperform human judgment. This is even the judgment of experienced experts, in almost every field, where both have been compared. The question Kahneman asks is "why are experts inferior to algorithms?" (p.224). The mixture between confidence and the inability to see past emotion is the bare basics of that comparison between human and machine. These emotions are backed by Kahneman's claim that they can, in fact, be changed so easily that it seems ridiculous:

"The widespread inconsistency is probably due to the extreme context dependency of system 1. We know that from studies of priming that unnoticed stimuli in our environment have a substantial influence on our thoughts and actions. These influences can fluctuate from moment to moment" (p.225).

He explains that even the smallest cool breeze which makes us slightly more positive can have a huge impact on our choices. This does not happen with machines. But when the machines are making decisions about human beings, Kahneman states that the prejudice against them increases by quite a bit even though their accuracy doesn't quite change (p.229).

Formulas win for the following reasons: they weigh each variable consistently, they don’t get tired or biased, they aren’t influenced by irrelevant context and they don’t overinterpret random variation. Kahneman's one rule for this is to replace human intuition wherever possible, with algorithms. Expert intuition is real but rare.

But when, if ever, can we trust intuition? Chapter 22 deals with that issue. This chapter deals with the idea that not all intuition is equal. Kahneman sets strict conditions for when expert intuition can be trusted. We get taken through why exactly it exists in the first place:

"Certain types of intuition are acquired very quickly. We have inherited from our ancestors a great facility to learn when to be afraid. Indeed, one experience is often sufficient to establish a long-term aversion and fear" (p.237).

This probably explains why perhaps it is better to leave certain decisions to machines as everyone has learnt through fear, their own intuitions and have inherited others. Machines have never experienced fear and can't, so the fluctuation between people and their different intuitions don't tend to happen. This is about where Kahneman has a disagreement with a fellow psychologist (who states that human intuition should be trusted when it comes to experts) but, explains both sides equally without judgement. However, he does ask of himself: "when do judgements reflect true expertise?" and comes to the conclusion that whatever is being judged must be regular and predictable, but also they should be learned through sufficient practice (p.240).

This expertise though is "not a single skill" but instead a set of skills, multiple ones. This means one person may have multiple expertise to suit a particular field but be a novice in things that may suit another field and thus, will not have the expertise to be able to commit to accurate judgements (p.241).

If system 1 creates coherency where there is none, does that not mean that it would be able to create judgements where the predictability is forseeable for an expert and thus, pass by the more critical system 2? There's another problem with human judgement (p.243). It's only when the environment is regular enough for the system 1 mechanism in the expert to do a pattern matching of the events can the skill be considered almost as viable as a machine committing to it. So, our conclusion here is definitely that we should only trust intuitions that are shaped by frequent, accurate feedback in stable environments.

Chapter 23 deals with how humans often look at the inside view rather than the outside view of a project due to the fact they have their own biases. The “inside view” focuses on the specifics of a project. Whereas, the “outside view” considers statistical realities of similar projects. The "inside view" occurs more spontaneously as it assesses the "future of our projects" - it is something that also occurs out of personal experience, almost immediately producing biases (p.247). The "outside view", Kahneman states, is much more stable and reliable - but it should only be considered when done properly and scientifically (p.248).

Even if you're well prepared for a project, you must consider the "outside view" which gives us the following pieces of reasoning: most projects take longer than expected, most operations go over budget, most forecasts are too optimistic and unfortunaltely, most businesses fail. This then steps into Kahneman's theory called the 'planning fallacy' (p.250).

In his efforts to overcome the planning fallacy and to teach others how to, he states that people often "overestimate the benefits and underestimate the costs" (p.525). Here are some others ways Kahneman states to overcome this fallacy and they include but aren't limited to: identifying similar past cases, determining their typical outcomes, predicting your case in that context and finally, adjusting modestly for specific factors.

This is where we move into Chapter 24, entitled The Engine of Capitalism in which we take a good look at the way in which confidence plays a major role in the 'venture capitalists' of our society and how optimism fosters this kind of confident behaviour. The main idea is that optimism is psychologically essential for risk-taking. However it also blinds entrepreneurs, investors, and business leaders to the true odds of failure. Kahneman actually states that:

"One of the benefits of optimistic temperament is that it encourages persistence in the face of obstacles" (p.257).

This is not only a benefit though, this is also a huge risk to investors who have put their money behind these projects. From being hit with the "hubris hypothesis" in which a belief in superiority can have serious consequences to the wishful thinking that there will be no competition able to compete (which fosters ignorance) (p.258-61).

Kahneman admits that though capitalism goes around with optimism and that it is "highly valued", it poses the same dangers of having a good feeling about something but not judging it based on any evidence (p.262). The "main benefit" Kahneman states in the midst of optimism is the fact that the person remains resilient. People who score higher on depression tests and have lower optimism also usually have lower resilience as well based on differing life experiences. However, the same optimistic people though taking credit for successes also take very little credit for things that fail (p.263).

But "can overconfident optimism be overcome with training?" Kahneman is "not optimistic" about it (p.264). The system 1 of overconfidence can be tamed but it cannot be rid of entirely. So, when it comes to capitalism, these overconfident venture capitalists may want to stop and kick their system 2 into gear before making decisions. The main takeaway from this chapter is basically that optimism is both productive and dangerous: it increases economic growth while simultaneously causing costly misjudgments.

Conclusion

This one is perhaps a bit shorter than the two previous essays because of the length of the chapters involved. But when we return next time for part 4, there will be more to be made up for as there are far more chapters in that section that continue with the more in-depth analysis of how we think.