Tuesday, December 6, 2011

"Thinking Fast and Slow" by Daniel Kahneman

I've just watched the mutual dick sucking festival between Newt Gingrich and Donald Trump and tried to keep from retching.

Newt Gingrich: nothing a little spontaneous human combustion wouldn't cure.

Although, not really wishing much harm on my fellow humans, I'd just as soon put the entire Republican clown car - including the sticky wet turd Donald Trump - onto a ice flow and shove them off into the Bering Sea. You know, an experiment,  just to see who gets eaten last, or which orca finally gets tired of playing with which bloated corpse. Nothing bizarre. Nothing grotesque. Just some good old-fashioned natural justice.

Speaking of Newt, I suspect he wouldn't get the fast-or-slow solution to the following problem, and it is mainly because Mr. Six Sigma is, despite conservative misconceptions, intellectually a fucking retard, and ethically an empty shit bag.

Ready? Think Fast! Think Slow! Problem: A bat and a ball together cost $1.10. The bat is $1.00 more than the ball. How much does the ball cost? Answer at the end.

This is one many examples of figuring out how people think, and that's the cool thing about, not just Daniel Kahneman's papers, but Kahneman and Amos Tversky's papers. They have fun puzzles in them.

In case you haven't figured it out, this is a mini-review of Kahneman's book "Thinking Fast and Slow". I call this a mini-review because I obtained the book through the public library's new purchases reservation system, and so have had only two weeks to read it. So, I skimmed. Just as well, as probably 80% of the material I was already familiar with. As is Kahneman, for this book is primarily, a summary of not only his lifelong research, but other psychologists findings as well. As such, I would classify this book as a worthwhile introductory textbook, and leave it at that. If you are unfamiliar with this area, then I recommend reading it. If, like me, you have some familiarity with the materials, then I suggest you skim and not worry overmuch on missing the juicy stuff. The salient points will catch your eye, and you will be rewarded. 

I first came across Kahneman in an article in "The Economist" magazine, which was devoted to demolishing the Chicago School's neo-classical version of economic models. I'm also ashamed to say that I also came Kahneman and Tversky through Malcolm Gladwell's book "Blink". (The nicest thing I can say about that book is that it provided a fun bibliography to work through). Back to the Chicago School, not being an economist or a psychologist, I still, through casual observation of human beings and occasional chance readings, came to the conclusion long ago that human beings are not rationally self-interested agents. Or rather, that a rational description served poorly to describe the human animal. In a more succinct manner, identical to my critique of the works of Ayn Rand, would be put thus: "What kind of a fucking retard would buy into this bullshit?"

Kahneman devotes a large amount of the book to developing two fictional characters, the two functional personae that make up your mind: System 1 and System 2.

System 1 is, for lack of a better term, your intuitive self, your associational engine. System 2, again for lack of a better term, is your rational self, your logical computer. System 1 has been honed by millions of years of evolution to be the very clever, very spontaneous, animal mind that each and every one of us relies upon to make it through the day. System 1 is the fast thinker.  System 2 is the slow thinker. System 2 allocates attention to effortful activities such as complex computations, problems of agency, choice, judgement, and concentration. It is also extremely lazy. If System 2 can get away with not having to do work, and rely upon the mental activities of System 1, it will. This lies at the heart of practically every cognitive illusion and fallacy we operate under. Not all. We have to keep in mind that logic is stupid. That rationality is NOT the same as intelligence. But still, it requires work to be logical, and if we can get away with the minimum amount of effort, we will.

From my own personal introspections, I offer an indefensible allegory. We've all heard of the false myth that we use only ten percent of our brains. I submit that only ten percent of my brain is used to produce the conscious me, the "ego". The other ninety percent is used to produce the subconscious me, the myriad associational, emotional, instinctive, embodied processes that make up the majority of me. (In some sense, this part is an alien other, a portion that, failing to be replicated in an artificial intelligence, pretty dooms the whole project from the get-go. Then again, a realistic replication of this alien other, in an attempt to simulate human intelligence, may also doom an AI to automatic insanity).

So, in summary, I recommend the book to any reader. There's both insight for the beginner, and clarified summaries for the old hand.

Answer: Fast think: the ball costs ten cents. Slow think: if x is the amount of the ball, then x + (x + $1.00) = $1.10, then 2x = $1.10 - $1.00, then x =  five cents.

Don't feel bad if you quickly said ten cents. Even Nobel Prize winning economists and people from Harvard business school and MIT get this one wrong. That's fast thinking for you.

Better to be quick and wrong than right and dead.


  1. That's one helluva lede, John!

    I must admit even though I've read your next-to-last paragraph 3 times, I still can't get even my slowest thinking to grasp it. But I will persist!

  2. What bugs me is that I used to be able to do algebra like in my head and now I have to write it down on a scrap of paper!

    Maybe I've been indoctrinated but I believe the linguists when they say we have a language machine in our heads. It implies that what Kahneman is saying boils down to the absence of an arithmetic instinct.

    Except for a few people like John von Neumann, who could do long square roots in his head, but was amused by dropping pencils on the floor so he could look at girls' legs.

  3. OK, Barry explained it to me, I now know algebra!

  4. Zina, Don't feel bad, I had to look up "lede".


    Kahneman say the most effortful thing you can do is the Add 1 exercise. Take any four digit number and add 1 to each digit in your head. (I'm assuming it's modulo 10, in other words, 9 + 1 = 0 instead of 10, so that 4729 becomes 5830.

    Perhaps it is, but what about the savant capabilities of people such as Johnny vN and kids within the (autistic) spectrum? Having a rich family history of people being slightly in the spectrum but not so much that you could notice (one brother can give you day of the week for a date, father and maternal grandfather had eidetic memory, all brothers pretty much walking encyclopedias in at least one subject, and yours truly can hold 3D dynamic models in his head), I've got to think that some people have a numeric instinct... although Kahneman called it a lack of statistical instinct (bad at probabilities and judging risk) and I would say that is probably correct.

  5. I found the observations in the book about human behavior interesting (the raw data). But not the author's interpretations, models or conclusions about the behavior. The models overreach too much in their conclusions and are too simplisitic (the curse of the so-called social sciences). To me, understanding that people will usually react a particular way in a given situation (observation) is far more interesting than coming up with a half-baked theory explaining "why".

    Marlene Detierro (CashFund)