Daniel Kahneman’s “Thinking, Fast and Slow”

The Kindle ‘locations’ go up to 9363 for this book. The end-notes begin at 7482. Page numbers may be proportional.

This is a famous thought provoking cartoon with a serious and profound point. It asks why we use big words when we don’t need to. It turns out that there is a reason. We need words around which to organize information. German has taken to agglutinating germanic words to make larger germanic words and that may be a better plan than the English proclivity to agglutinate Latin or ancient Greek words. In either case we learn these new words in the optimistic expectation that we will have much new information to accrete to them. The book makes this as their first point in the introduction.

(@373) Before you go beyond this point you absolutely must watch the video here! Count carefully I was way off.

Here is a very premature hypothesis. Jane Jacobs proposes two sorts of institution “guardians” and “merchants”. I propose roughly that System 1 (Kahneman’s fast — ancient) defines the nature of guardians and System 2 (Kahneman’s slow — new) defines the nature of merchants, who must contrive win-win situations. Military strategy often demands a high degree of logic, but its impetus is System 1. System 1 is what we could do 2 million years ago.

Another hypothesis is that system 2 is as described here.

Kahneman’s reports on experience and memories thereof should inform theories of valence wherein we seek or avoid repeats of such experience.

(@415)I am surprised that System 1 reads words. I suppose that simple experiments prove this.

(@705) Consuming glucose during system 2 work: I presume that this is accompanied with actual energy expenditure such as might be detected in a calorimeter. This is the only obvious case but not the only possible case. This bears on the appellation “hothead”.

(@849) “associative activation”, “spreading cascade of activity”—Computer like I say. An explicit test of Kanerva’s ideas comes into view here. Following a linked list monopolizes some specific mechanism necessary to hold a Kanerva address. It can move from node to node along links only sequentially and probably at a rate of several clock cycles per link for some basic concept of clock cycle. There might be several of these but that causes problems; ask any memory engineer. Naïvely this suggests sequential spreading whereas the subjective sense is indeed a branching spread. Perhaps this is an illusion. Can a testable hypotheses be generated here?

I do think that System 2 makes heavy use of System 1 in order to do its System 2 thing. System 1 has only a light context to disrupted and restored.

(@881) The priming effect is described here simply. It is a key phenomenon to be accounted for.

(@1123) Zajonc: The mere exposure effect. This is clearly an evolutionary adaptation to be comfortable with the familiar, especially if it has not bitten you.
Beware the Novel! (Point made at @1142)

(@1390) I must tell a tale about myself and the ‘Halo effect’. When I was perhaps 6, I met a fellow named John and surprised that he was nice. I wondered why I was surprised and then realized that I had had a strong negative association with that name; I had known a few disagreeable Johns. I knew that my bias was illogical but I was reluctant to abandon it as I knew that I would have to undertake other such examinations of opinions.

(@1545) Correlation between snap judgement of photo of politician, and electoral success: There is a flaw: some politicians have more money and other sorts of support. They will have both more campaign action and better photographers—thus the positive correlation.

(@1715) The Perspective interpretation. I wonder whether this effect would have been observed before the invention of perspective in 1413? One must learn to see perspective, as well as learn to produce it. This is curious for seeing perspective is what we evolved to do when there was no flat surface that we were looking at. It took some sort of suppression of that fact to see perspective in a drawing.

(@2964) This (good) chapter feels like it should have been in “The Signal and The Noise”.


Better: I think that it is true that regression toward the mean is not intuitive. I wonder why we have not evolved to do better here; it does not sound hard.

(@3442) “Leaders who have been lucky are never punished for having taken too much risk. Instread, they are believed to have had the flair and foresight to anticipate success … ”. The flip side is that a priori probability is no easier to know a posteriori.

(@3454) “… a correlation coefficient, which varies between 0 and 1.” No, it varies between −1 and 1.

(@3556) Quote:

From an evolutionary perspective, or from the perspective actually at hand, what is the use of an uncertain prediction?

(@3701) Quote

I wonder about the methodology of evaluating the predictions. Indeed what is the required syntax of a prediction? Probability of rain tomorrow is a poor form of prediction. (more)

(@3745) Here is Meehl’s marvelous book.

(@3780) Quote regarding wine:

This sounds like a pleasant business opportunity! It takes a long planning horizon however and thus it takes a long time after the publication of Ashenfelter’s notes for price equilibration to arrive.

(@3953) To his chapter end aphorisms he might add the motto on Harold Brown’s (Carter’s secretary of defense) desk:

I enjoyed this chapter (21) the most.

(@3983) “… they knew the sculpture was a fake without knowing how they knew …”
This reminds me of recognizing a number that is a square.

(@3988) I feel vindicated in Kahneman’s and Klein’s rejection of Gladwell’s fake statue analysis in “Blink”.

(@4009) Kahneman quotes Simon’s definition of intuition: “… Intuition is nothing more and nothing less than recognition.”. I think I agree. Then Kahneman continues “That strong statement reduces the apparent magic of intuition to the everyday experience of memory.”. Well ‘of recognition’, but that is still pretty magical. Yet the reduction is useful.

(@4106) Speaking of therapists Kahneman says: “The feedback they receive from their patients’ long term outcomes is sparse, delayed, or (usually) nonexistent, and in any case too ambiguous to support learning from experience.”. The efficient market hypothesis also relies on such feedback and the wine price conundrum can be described as the failure of the market’s intuition due to insufficient training.

(@4179) The story is of planning a project with several people and after the norma (optimistic) estimate of two realized that such projects took about seven years—thus discouragement. I recall after Tymnet was pretty much completed and in broad service, saying that if we had known that the project would take that long, roughly 7 years, we would not have started and top management would not have let us start. The project was seen as a great success.

(@4301) This chapter makes no allowance, so far, for unplanned excess gains, such as experienced in Tymnet. The company grew faster than we had foreseen, partly due to Tymnet and partly due to unanticipated raw demand, and the benefit of Tymnet was proportional to this growth. Tymnet was highly synergistic with acquisitions. Consider the unplanned downsides, but also the unplanned upsides.

I have often thought that considering the success rates for new restaurants that if those hewing to Kahneman’s teachings would open no restaurants. That would unfortunate for those who eat out. This does not mean that Kahneman is wrong, but perhaps that restaurant patrons exploit restaurant owners. More or less this point is made in the next chapter 24. (@4327)

(@4349) The notion is that the economy runs mainly on those who are:

This is not surprising from a Darwinian vantage point. With two populations of enterprises differing only in optimism of the leaders, the realists will tend to shut down the enterprise which is good for the investors, but bad, on the average, for the economy. The remaining visible enterprises will be run by optimists.

(@4423) Quoting a question Kahneman asks of entrepreneurs:

I claim that this is a malformed question. It is presumably clear how to drive a firm into the ground and refraining from doing so takes little talent. In that sense the entrepreneur has much to do with the success of the company. That is not the sense of the question but it is not clear what the sense is.

(@4578) I am getting anxious to hear about ‘nonlinear utility’ or ‘concave utility functions’. A rational person will not bet his entire worth on a double or nothing proposal even with a 0.51 probability of winning, unless he can off-lay some risk. This is not an entirely psychological bias; or at least it has roots in survival theory. These issues are raised by @4590.

(@4588) Fechner’s logarithmic suggestion is interesting. Lest it be considered vacuous I note that sensing electricity does not follow that pattern. The ratio between an insensible voltage and a painful one is only about a factor of two. (40:80 volts) This is perhaps because it was not adaptive to sense small voltages.

(@4648) Kahneman may be right about Betty’s choice, but that would not be my choice. I think it would not have been my choice even before I learned about concave utilities. I do not buy ‘Bernoulli’s error’ on the basis of my subjective feelings.

(@4725) Continuing the above issues, I still am not ‘risk-seeking’ in case of bad options; and I think it is my System 1 at work.

(@4864) I am sad that there is no talk here of evolutionary theory.

(@4896) The twins on the indifference curve are a good gedanken experiment. It explains much of wage stickiness.

(@5280) Regarding the choice presented in B: I indeed chose the 2nd option—as a $9600 insurance policy. That is perhaps too strong a convex utility curve. I did the arithmetic before I decided. It is a millionaire, after-all, that is paying the premium! I think Kahnemanis trying to rescue his ‘reference point’ notion. In the choice at hand my reference point is having the money.

(@5636) Regarding insurance: “Here again, people buy more than protection against an unlikey disaster; they eliminate a worry and purchase peace of mind.” This can be explained by simple reference to a convex utility curve, àla Bernoulli.


Another example is climbing a mountain on a hot day. On the way up you swear you will never climb this mountain again. Then you recall a previous such oath. -----5877