Thinking, Fast and Slow D.Kahneman February 12-13, 2014

1 thinking , fast and 12-13, 2014 Human beings are hardly rational economical agents, at least they do not complywith the classical economical theories of economic transactions. They do not formulateutility functions, the pursuit of which they are expected to conduct with relentlessnessand consistency. This is hardly a surprising insight but one which most people, exceptpossibly economical model builders, instinctively hold. On the other hand every model bynecessity involves drastic simplifications, in order to be manageable and clearly show thesalient features. To abandon a simple model which has shown itself useful only makes sensewhen its predictions start to diverge too much from real life. This is a book addressedto a general public, and thus the author, a psychologist turned economist, does not makeany attempt to show how classical economical models fail and how they can be suitablymodified, but constricts himself to elaborate on the obvious shortcomings, which to mostreaders must be intuitively obvious and hence gratifying, as man merely as a rationalagentis to most of us seen as unnecessarily restrictive, with the implications that any theoriesthat neglect those, must by necessity be radically flawed.

2 The author is careful to pointout that he does not make a claim for the irrationality of man, on the contrary, withoutsome rationality assumptions, predictions become impossible, only that he wants to makethe notion of human rationality more nuanced, pointing out its limits. The author shareswith most academic economists a sanguine belief in the possibility of quantization in thesocial world, thus not doubting its ontological aspect, only its present starting point, which provides the title of the book, is the distinction betweenfast and slow thinking , which roughly corresponds to the basic division between the sub-conscious and conscious pioneered by the nowadays discounted Freud1. We all have anuncanny ability to recognize faces, something which so far has not been possible to simu-late on a computer in spite of its superior computing power.

3 We do so very quickly andwith no apparent effort in spite of all the computation that must be involved. Furthermorethe act is not deliberate but involuntary and automatic. We cannot will ourselves not torecognize a well-known face. Those cognitive gifts are given to us by God to serve us wellin our daily life. Although nowadays we prefer to formulate it in terms of natural selectionand survival (and hence reproductive) advantages. Unless pursued, which it seldom is,the latter formulation does not really carry much more illumination than the to those are our abilities to effortlessly retrieve a lot of different facts, such as thatParis is the capital of France, or vocabularies of languages we speak with mindless to the automatic and almost instantaneous thinking is the slow and laborious1It is noteworthy that one of the blurbs printed on the back cover As Copernicus removed the Earthfrom the centre of the universe and Darwin knocked humans off their biological perch, kahneman has shownthat we are not the paragons of reason we assume ourselves to be which is a conscious or subconsciousparaphrase of how Freud himself characterized his , which involves a lot of effort.

4 Which actually shows definite physiological symptomsin terms of dilations or pupils, increases in heart rates and blood pressures, and henceactually allows some objective quantitative measures. This kind of thinking is consciousand craves attention and deliberation and excludes other kinds of thinking . The typicalexample is making a standardized computation in your head, involving keeping items inyour working memory. When standardized you can easily gauge limits to your cognitiveability. Most people can with effort multiply two two-digit numbers but are unable tomultiply two three digit of our lives are spent in a default mood, all the decisions that are required aresupplied by habits, once in a while there is an emergency, the senses realize it quicklyand the fast thinking supplies an automated and instant response which in most cases isfully adequate.

5 In this sense we do not depart from animals in general. However, ourfast thinking is not always appropriate, and then the slow thinking takes over, but onlyreluctantly so, as it is always associated with a large effort (and, one surmises, depletionof energy). If a ready answer is supplied it often takes an effort to question it3 Now ourGod-given intuitions are not always perfect. And here, we do encounter the first differencebetween God and Natural Selection. We are not supposed to question the perfection ofGod, but we are encouraged to explore the failings of Natural first thing to remember that there are too many contingencies in life, especiallythe social life of humans, to have precedents in the development of the human race, orfor that matter to be encodifiable in genes.

6 What may be pre-programmed, so to speak,are certain tendencies to come to conclusions of a speedy judging processes, known asthe gut feeling. The fallacies of our fast thinking , according to the author, is that thefast survival process only looks at the information at hand and does not actively look foradditional information4. Furthermore it weighs the different components, not accordingto their objective importance, but how vividly they appear to the mind. Finally there is apremium on presenting a coherent and logically structured narrative, because as humanswe have a propensity for stories, and are very good at remembering them, as the differentcomponents of it are tied together by logic and association5. What comes through to themind by such natural intuition is compelling.

7 But as Popper reminds us, conviction isa subjective feeling and should not be confused with truth. It takes a lot to challenge2 The author provides as a generic example17 24when I attacked it I thought at first of doublingthe first factor successively, then I realized that I should rewriteit as(17 3) 8and the task becomestrivial. I guess this was not intended by the author. To what kind of thinking does this exploration belongto? It was not painful, yet clearly it usurped my an example. Two people split $ , one picks a dollar more than the other, how do they split instantaneous answer presents itself, namely $1 and $ those are obviously false, yet I fell into thetrap. Why? Lazy, and the question did not interest me in the least. Had it been A is ten years older thanB, their total age is 44, I would not have to as WYSIATI meaning what you see is all there is 5 This is why a mathematician can recall a complicated proof, not verbatim of course, but as a why a genial musician like Mozart, can recall, verbatim in fact, a piece of music, only heard once.

8 Orone, supposes, a chess master, recalling a position by a mere glance, seeing in it a story poignant withpossibilities2such an emotionally charged conviction, and nothing short of a brutal disappointment ofexpectation will do it. Those propensities of the human intuition are responsible for theegregious errors of applying and estimating probabilities which people, even those suitablyeducated, come prey to. To that we will counter the failings of human judgement it is, the author points out, importantto have some mechanical algorithm that does the job. Such a stratagem goes against thegrain of most people. How can a mindless algorithm which often is crude and based oninaccurate assumptions, seriously compete with human experience and that extra-sensorysense that intuition is supposed to represent.

9 (In another age one could have referred toit as divine inspiration, without blushing. Now people would blush at such terminology,but what they are suggesting does not really differ.) The advantages of the algorithmicapproach is that 1) it works against the bias of readily available information and forcesyou to collect more of what is relevant to the case. 2) By using some mindless formulayou check against the tendency to rely on subjective weights and finally 3) by gettinganumber or some such thing, you guard against the seduction of constructing an attractivenarrative. There are also of course obvious disadvantages, but those are so obvious thatthey need not be spelled out. As an example the author cites three components whichcome into play when you are going to predict the quality of a given vintage of wine, allsimply referring to the weather.

10 The amount of precipitation in the fall and spring andthe average temperature of the latter. In particular tasting the wine does not enter atall into consideration, something which sticks in the eye to the classical wine , this simple scheme, seems superior to the regular methods. Whether this is trueor not, we simply have to take the word of the author, there being no other alternative,save an investigation of your own. In particular the stratagem is very controversial in theclinical setting. But once again we have to that the authors word for it, in studies made,thediagnostic precision of clinicians (as testified by subsequent autopsies?) is discouraginglylow, and simple algorithms do better. In fact in modern medicine quantitated test play agreater and greater role, and the old wisdom of the family doctor with his holistic approachof taking in the whole person, is fading into the remembrance of quaint , the disparagement of intuition caused a lot of controversy, and rather to engage inpungent but eventually fruitless polemics the author chose to invite his antagonist in a jointexploration that eventually led to some joint papers and, one assumes, increased compromise they hammered out, was that in some instances there is something to becalled professional intuition, but it needs to be honed during a long time, with constantopportunity to more or less instantaneous feedback to small piecemeal changes.