Transcript for Daniel Kahneman uncut

Steve Paulson: I want to start with the back story to this book actually. I read that writing the book was quite an ordeal for you. Can you explain why?

Daniel Kahneman: Oh, yes. It was really- It's a very difficult thing for an academic to write to the general public. It's just challenging, and you're trying to satisfy two audiences and that is very difficult. And then, you're an old man, and that makes things a lot more difficult, because you think things are harder than they used to be. So, altogether, this wasn't fun.

PAULSON: (laughs) But from what I heard, this goes beyond the usual self doubt that an academic has about writing a popular book. You actually went to the trouble of actually hiring various psychologists to evaluate what you wrote, right?

KAHNEMAN: Well, yes, I did. And Michael Lewis is making sure that he has written a story in which he tells this incident and I think it's going to accompany me for the rest of my life. But what I wanted to make sure of actually, I was quite concerned about what my colleagues would think of the work. I mean my main audience is the public at large, but I didn't want it to destroy my academic reputation. And I make some decisions in the presentation of the material that-, I deviated from sort of the standard way of speaking in my profession. So, I did ask people- and it had to be anonymous and I paid them, all that is true, to tell me whether I should go ahead with the book, or whether my reputation would be served by simply waking away from it. And they told me I should go ahead, and I did.

PAULSON: Well, it sounds like they actually gave you glowing reviews.

KAHNEMAN: Well, no, actually. I think, they said, “It's OK.”Michael Lewismade it sound a little bit warmer than, I forget what adjective he used, and then the New York Times said yesterday that the reviews were rapturous, and I actually have a note to write, an email to these people that the word did not come from me. Because they were not rapturous, they just told me, “It's OK. You might as well go ahead and write it. It won't destroy your reputation.

PAULSON: Well, there's more than a little irony here, in this story. Here you are a Nobel prize winning psychologist who is expressing all kinds of self-doubt about what you wrote. I guess I'm wondering is there something particularly about the way you approach the world that naturally lends itself towards doubting what you've done.

KAHNEMAN: No. It was a particular thing towards what I choose to do. And, I use the language of two systems in the mind, System One and System Two. And this language is questionable with psychology. I mean, we've always been warned never to try to explain psychology by invoking little people inside the head, because it's too easy. And here I am, invoking in effect, little people within the head. And I think I have an excellent reason for doing that, and I think this is absolutely essential to communicating these ideas to the public, and indeed it helps me think about them. But that was the crutch of what was worrying me. Whether in trying to communicate with the public I was simplifying things too much, and whether I was being selfish in the way that I described the work. Whether I was sufficiently aware of what is going on in the field that young people are doing these days. So, I've always thought that an old man writing a book should be very careful, because you tend to know best the thing that you knew a long time ago. And you learn more slowly than you used to. So I was worried about the impact of that on my work.

PAULSON: Well, let's talk about these two ways that our mind works. What you call System One and System Two. First of all, explain what System One is.

KAHNEMAN: Well, System One- Well, let me first apologize and say that there are no systems really, in the sense of interactive parts. But what there is, is, there is a set of mental operations that occur automatically. It's when you see something. When a two-year-old child points and says, “Doggie,” something is happening that seems to be experienced passively. It's something that happens to you. As we see the world, we don't decide to see the word. When we see a word on a poster, we read that word and we don't decide to read it. So there are many activities of that kind. When I say, “Capital of Russia,” a word came to your mind. When I say, “two plus two,”a word came to yourmind. All of these things happen by themselves, and feel as if they were happening by themselves.

PAULSON: So there's something sort of instantaneous about it. It sounds like it has something to do with pattern recognition.

KAHNEMAN: It is pattern recognition, a lot of it. It is recognition. It is instantaneous; at least it is very quick. And it happens automatically. It is not deliberate. It doesn't feel like something we do. In contrast there are other mental activities that feel like something we do. We are the author of it. So, if you try to compute the product of 17 by 24, well in the first place you've got to decide to do it. And, then when you do it, you're following a program and you're doing that deliberately. And you feel that you are the author of your own actions. So subjectively, it's quite different. And, it those two classes of operations that I describe as System One does this and System Two does that.

PAULSON: So System Two is, I guess what we would call conscious reasoning?

KAHNEMAN: Yes, and System Two is something else, actually. System Two is, I lumped with System Two another set of operations, which are control operations. We don't say anything that comes to mind. We direct our attention to somethings and we direct it away from other things. So, we are controlling ourselves. And that control operation is also joined with System Two and the characteristic they share in common is that they are deliberate and also they're effort-full. They can be impaired when you're busy. So my rule about System Two, is that, the definition of an operation of System Two, is that it's something you cannot do when you're making a left turn into traffic.

PAULSON: Hmmm.

KAHNEMAN: Because we are limited in our ability to exert effort. And both deliberate computation and self-control demand effort.

PAULSON: Well, you have a very curious comment in your book. You say although System Two believes itself to be where the action is, the automatic System One is the hero of your book.

KAHNEMAN: It certainly is, and I think it is the hero of our mental life too. In the sense that, most of the day, what we do is very largely skilled and automatic, and most of what we do all day is quite good. Because we're skilled at what we do, and skilled activities below with System One. Now, when we think of who we are and of what we experience, we experience our self thinking deliberately. Or we experience our self doing things voluntarily. And so our experience is very closely linked t System Two. Much of what happens that I call System One activities, we're really unconscious. We're not aware of what's going on. We're not aware of what's going on within our memory, as a word is said and all the associations of that word are activated in memory. We're not aware of that process at all. And so we are biased to think that we are our System Two, but in fact our System Two, most of the time, does what is suggested by System One. It just endorses.

PAULSON: Now the obvious system is whether we should basically trust System One, these instinctive, intuitive responses or whether we should try to check them more often because they're actually not very reliable.

KAHNEMAN: Well, the theory clearly-, we have very little option in this matter. System Two could not carry us through the day. Deliberation and thinking are costly and ineffective, and mostly we have to trust our habits and we have to trust our perceptions and we have to trust our impressions.

PAULSON: But much of your research has shown that a lot of those natural tenancies to trust our observations, they feed biases that are often, actually, very inaccurate.

KAHNEMAN: Absolutely correct.  I mean System One, it has a model of the world. But the model of the world that it operates in is very simplified and System One operates on-, is a machine for jumping to conclusions, so it operates on very little information as if it had more information. We're very rarely sensitive to what we don't know. We operate on whatever we know. We make the best story possible. And then there is a very striking characteristic of System One. It is rarely stumped. That is, when System One faces a question that it cannot answer, it spontaneously will come up with the answer to another question, to a simpler one, that it does have an answer to. And it is this process of answering the wrong question that very frequently leads to the mistakes that I've been studying.

PAULSON: Give me one or two examples of this. So how System One comes to the wrong conclusion.

KAHNEMAN: Well, System One, for example is going to be sensitive- Well, no, here is an example. Suppose you ask people that are about to travel abroad, and you offer insurance and ask, some of them would be willing to pay for insurance. And in one case the insurance is for something that happens during the trip, is for death for any reason. And in the other case you offer them a policy that covers death in a terrorist incident. That study was done when there was a fair amount of terrorism in Europe, and it was done in the United States. People pay more for the second policy than for the first.

PAULSON: Even thought the first presumably would include the second, right?

KAHNEMAN: Now, this is absurd. It would include the second, and it would include many other sources, forms of death, which are a lot more likely than dying in an incident. But, what happens is this, what we do when we do when we buy insurance is we measure how afraid we are. That's a simpler question than, how much should I by willing to buy insurance. So the first (?), what comes to mind immediately is the reaction which is basically, the intensity of the fear that I experience. And then you map that intensity of fear into a number, and amount of dollars that you are willing to pay for insurance. That is answering the wrong question.

PAULSON: So you're saying that we often think we are making a rational decision, but actually it's not rational at all.

KAHNEMAN: Well, you know, the word rationality is a complex word, and I try to use it as little as I can, really. Rationality's defined within economics and decision theory in particular ways. It's defined by a set of axioms but a set of rules, and those rules are not for a human mind. So, the human mind is simply not built to be rational, as economists define rationality. And they have to define rationality that way for mathematical reasons. In my way of thinking, people are quite reasonable and they make mistakes. And I don't like to hear them described as irrational because I think that's much too extreme.

PAULSON: When you say the human mind is not built to be rational, what do you mean?

KAHNEMAN: Well, what I have in mind is that, in the concept of rationality, as it is embodied within economic theory, all beliefs are internally consistent, and all preferences are internally consistent. And preferences are only affected by things that matter, they are never affected by whether you describe a cold-cut as 10% fat or 90% fat free. A rational economic agent, according to economics, would be willing to pay the same amount to both cold-cuts. In fact most people will pay more for the one that's 90% fat free.

PAULSON: Really. That's fascinating.

KAHNEMAN: Oh, yes.

PAULSON: So even though they say the same thing, just the way you word something, I guess advertisers know this (laughs.)

KAHNEMAN: Of course.

PAULSON: It has tremendous impact.

KAHNEMAN: And the reason is straight forward. You have an emotion to the word “fat-free.”And you have another emotion to the word “fat.” And thatis a System One reaction. It's an immediate, intuitive emotion. The emotion is aroused. And then, you are more attracted to the 90% fat-free than to the 10% fat and you pay more because you are more attracted. It's very much the same story as deciding how much you'll pay for insurance by how afraid you are. And you are more afraid of dying in a terrorist incident than you are afraid of dying. It's absurd, but that's the way the mind works.

PAULSON: Now, you're book has been described in some quarters as kind of the anti- Malcolm Gladwell (laughs) treatise on how the mind works. Of course Gladwell is the person who really popularized the idea that we should really go with our gut. We should trust our instincts because that's often, sort of, the best response. It sounds like you're saying our instincts are quite often misleading?

KAHNEMAN: Well, the word “quite often”is beautiful because it can be quietoften this way, and quite often the opposite way, and that is truth with respect to Malcolm Gladwell, he popularized an ocean of blink, and he clearly emphasized the very interesting stories about the marvels of intuition. He didn't deny that intuition can lead you to mistakes, and there are several chapters in his book in which he elaborates on them. So there is that debate and the truth is intuition, quite often, is wonderful and quite often makes mistakes.

PAULSON: You have a number of fascinating stories that you tell in your book and one of them is about an experience you had when you were doing national service in the Israeli army. It was a test called the leaderless group challenge. Can you describe this?

KAHNEMAN: Yes, I can. Well, I used to be, as part of my service, I was involved in the selection of candidates for officer training school. And one of the tests that we have was the leaderless challenge. And the way that it was run, it was run in an obstacle course and you would bring a group of eight soldiers. You removed all insignia of ranks, so they didn't know what rank they had. You faced them with a wall, there was a telephone pole right next to them. You told them to pick up the telephone, which they did, and then you gave them their instructions, which were to get to the other side of the wall, without the pole touching the wall or the ground, and without anybody touching the wall. If either of these things happened they had to start over. Very stressful exercise. There is no leader, it's a difficult task. It quite often was very hot, and dusty, and sweaty. And they've got to operate as a group and to decide what to do. And then, you have the observers, and I was one of the observers. And, as an observer, the thing that is very striking is that you see people natures revealed. Because they're under intense stress, and some of them appear to be leaders and strong and other appear to be wimps and others are just followers and other are actually nasty. And all of this you can see, because you feel that under the stress of the moment, truth is revealed. So that's one experience that we had there. There was another experience that occasionally, every few weeks, we would have a statistics day. And on that day we would be told how well we were doing in predicting the achievements of those people that we did send to officer training school. And the answer was always the same. We had no idea. We couldn't do it.

PAULSON: (laughs) So you thought you had picked the good leaders and actually you had not.

KAHNEMAN: We had not. But, the third part of the story is the interesting one. Is this was the army, and so, whether we believed or not that we're doing something useful, they would bring the next bunch of recruits the next day and we'd take them to the obstacle field, we'd face them to the all, pick up the phone pole. Five minutes later, we would see their true natures revealed. So with the same confidence that we had before we knew the statistics. So it's that. You can know that you don't know. You can have the statistics about your own performance, but the emotions, the sense of a compelling sense that you understand something, that we have watching these people struggling with the task was not effected at all by learning the statistics. That to some extent set me off on my career. I coined a term for it, I called it the illusion of validity. And many years later, I studied it.

PAULSON: Which is to say that we are often confident, or over confident when we really shouldn't be (laughs) based on the way the world really works.

KAHNEMAN: Oh, absolutely. In almost every way, we tend to be over confident, and the reason is that System One, or associative memory, generates interpretations of situations that tend to be much more coherent that reality. So it neglects ambiguities, it suppresses ambiguities. So, we have over confidence that way. We have a sense that the world is understandable and predictable, because anything that happens we have a story about it. And so, there is a financial crisis, and then there's a pretty clear therefore that we're inclined to believe in about how the financial crisis came about. That gives us a sense that the financial crisis was knowable in advanced. And it wasn't.

PAULSON: So we always have to fit everything that happens into some narrative, some story, even when the details really don't fit in reality.

KAHNEMAN: Yes. It happens automatically. Our mind is a sense making organ. It is a story telling organ, and it will assemble whatever it knows into the best story possible. Even when the quality of the information is poor, even when there is little information, you'll come up with a story and you'll tend to believe in it. And your confidence in the story depends on how coherent it is. It doesn't really depend on the quality of the evidence on which the story is based.

PAULSON: Now is this a polite way of saying that we are always deceiving ourselves. Well, we are deceiving ourselves in the sense that we're blind to our own blindness. That I think is generally true. We have a great deal of difficulty acknowledging the role of what we don't know. So we know what we know, and we tell a story about what we know. If I ask you is this national leader and I tell you, well, she is intelligent and strong. Now you already have a sense that she's a good leader. Now, I could continue, you know nothing about her character. I could tell you she's corrupt and cruel. But you have already jumped to a conclusion on the basis of very little information. You have a story. If I'd stopped there, you would feel that you know that here is a strong leader. That's the way the mind works. It's intended to jump to conclusion. It's designed that way, and that is a very good thing, because it enables us to act and react without stopping to deliberate too much.

PAULSON: You also write in your book about our two selves, the experiencing self and the remembering self. Can you explain the difference there?

KAHNEMAN: Yes. As you go through life, you have experiences, and you live the present and moments of the present are very short, and then they turn out into episodes. So there is one thing, one way of thinking about life, or thinking well-being or happiness is the quality of your experience as you go through life. Occasionally, you stop to evaluate yourself, or you evaluate an episode or a meeting after it ended. You say, “Was this a good episode or a bad episode?”This is your remembering self. So, the experiencing self isthe one that answers the question, “Does it hurt now?”That's theexperiencing self. The remembering self answers the question, “How was your trip to Albania?”And that's the remembering self. Oddly enough, theremembering self doesn't always know what the experiencing self went through, and that we can demonstrate experimentally. So that people, their remembering self tells itself a story about what happened, and that story doesn't always reflect the true experiences that went on in the past.

PAULSON: Well, you have some fascinating experiments that you did. And I have to say, you're a genius in constructing some of these experiments (laughs). One of them was something called the cold hand situation. Can you explain what you did there?

KAHNEMAN: What we did there, we asked people to immerse their hand up to the wrist in cold water, in water at 14 degrees Celsius. And on one of the occasions, everybody had, actually, two occasions. They thought they were going to have three, but they had two. On one occasion they had one minute at 14 degrees Celsius.

PAULSON: Which we should say is extremely cold.

KAHNEMAN: It's extremely cold. It's tolerable, that is, we didn't have actually causalities, or people leaving, but it hurts. Then seven minutes later, we bring them in and we give them another experience. And the other experience could be one minute at 14 degrees, just the same as before, but instead of telling people as we had on the other occasion, “Take your hand out, here is a warm towel, dry yourself,”we did something else. We opened a valve,and during the next 30 seconds the water became very slightly warmer, in fact it got to 15 degrees Celsius.

PAULSON: (laughs)

KAHNEMAN: So they had an experience that was strictly worse than the other experience. One is 60 seconds of pain and then 30 seconds of diminishing pain, but still painful, not as good as warming you hand with a towel. Then you wait another seven minutes and you come to those people and you tell them, “Well, you can choose to repeat either the experience you had with your left hand or the experience you had with your right hand. What do you choose?”And the large majority of them choose to expose themselves to thelonger experience. That is they choose to expose themselves to more pain. Why? It's because it ended well, and when we tell ourselves stories, how they end is extremely important. So you can take a bad story and make it better by ending suffering, so long as the suffering that you're adding is diminishing, so that it ends as an improvement.

PAULSON: (laughs) So you get to play with the way people respond to suffering, or the way they think about suffering?

KAHNEMAN: Well, this is the way that people think about lives. Just think about the role of endings in life. So you have the mother and her daughter and they haven't been speaking to each other for years and the mother is dying and now they get reconciled. And they have an hour together in which they're at peace. This is enormously important to us. The face that the mother had, the experience before she died of meeting her daughter again and loving her daughter again. Now, if you stop to think about it, the emphasis that we put on this is ridiculous. It's one hour and they were estranged for years and years. But what we think about, what our remembering self thinks about, it thinks in stories. And the story of the mother is very different when it ends that way.

PAULSON: Hmmm….

KAHNEMAN: We're story tellers.

PAULSON: Uh-huh. Well, I've often wondered whether we can ever really call up the experience itself or whether every original experience is tainted by memory. That we're always sort of reconstructing it in probably a somewhat fictional way.

KAHNEMAN: Well, that certainly is the case. You have theories about what happened to you, so. But in addition there is a major difference between the experiencing self and the remembering self in how, how you weigh episodes in your life. I'll give you an example. Suppose you have a vacation that is two weeks long, or the same vacation that is one week long. And suppose the second week is just as good as the first. In some real sense in terms of your experiencing self, the longer vacation is twice as good as the first one, because the second week is as good as the first one, and you have two weeks. In terms of the remembering self, there will barely be a difference between a one week vacation and a two weeks vacation if you stayed in the same place.

PAULSON: Hmmm…

KAHNEMAN: Because you are not adding many new memories. So the evaluation of life and of episodes in life is very different if you take the perspective of maximizing the quality of experience. Or if you take the perspective of memory. Very important when we think of people with Alzheimer or of elderly people with deficient memories. They still have an experiencing self. They may not remember the experience that they're having, but they're still having experiences. And you can make those better or worse.

PAULSON: Getting back to the vacation example, I've sometimes wondered whether the most important thing about the vacation is the actual experience of it, hopefully it being enjoyable, or whether it's how we think about it later on. Because the actual experience may only go a week or two weeks, but there's a lifetime of what we make of that experience.

KAHNEMAN: Yeah, but there is actually an interesting question, which is how much do you consume, how much during the rest of your life do you consume your memories. And I gave a Ted Talk about this last year, and I told the story of the best vacation of my life, which was in Antarctica for two or three weeks. And I know it's the best memory of my life, and I think about it occasionally as the best experience of my life. Now if I added up how long I have thought about it, in the last four years, I estimated it wildly as 25 minutes. In short very, very little.

PAULSON: Hmmm….

KAHNEMAN: So the time that we spend remembering and the time that we spend living are on different scales. And yet we give enormous importance to the time we spend remembering, and to our memories, and to the stories we tell.

PAULSON: Which raises a question of is there a way to construct experiences so they become memories, because then that would presumably enhance our life?

KAHNEMAN: Oh, this is something that certainly other people do for you. So when you go to Disney Land, they construct your experiences. And basically, they're not so much constructing your experiences as they're constructing your memories of the experiences. Because this is what you are going to keep. You're going to keep the memories, you're going to keep the pictures, and that is what this is all about. And there are actually different industries that cater to giving you experiences, or that cater to giving you memories. So, adventure tourism is to a very large extent for good memories, rather than for an experience as such. Now, whereas resorts most cater to your experience and not to your memory.

PAULSON: The question I supposed that comes out of this research is, is there anything that we can do? Based on your findings that somehow improves our judgments, our sense of how to tap into our better selves, and to put aside those ways that we delude ourselves.

KAHNEMAN: Well, you know, there is, if my description of things is correct, more or less, there is essentially one thing that you can do, which is to slow yourself and involve, bring in the reinforcement of System Two. And we can't do that a lot because System Two is just too slow. We must trust ourselves most of the time. When the stakes are high, it may be useful to slow yourself down. And there is something else that we can do, is we can help others make better decisions. I describe my book as an attempt to educate gossip. That is, it's an attempt to allow people to think more intelligently about the judgments and experiences of others. Because gossip is a pleasant activity. It's painless and you can make it interesting. And, I think a world in which people understand how other people make judgments and decisions, is ultimately going to be a world in which better judgment and decisions are made. Because we do anticipate the gossip of others, and if we anticipate intelligent gossip, we are more likely to make intelligent decisions.

PAULSON: So that's the take away here. (laughs) We should have more intelligent gossip?

KAHNEMAN: I think that's the take away in part, because I really do not believe that good intentions to make good judgments are going to be terribly useful.

PAULSON: One final question, you have a fascinating and difficult life story going back to your childhood. You were in France, a Jew, in a Jewish family, during the Nazi occupation there. And I know your father was at one point picked up and carted off to a camp. I'm sure very traumatic. Has that influenced your later work as an academic? These kinds of questions that you consider?

KAHNEMAN: I can't trace much of what I've done to those early years. People do that occasionally. I think I was interested in people as a very small child. It's true that because I was interested in people I observed very keenly what was happening to us and what was happening to others as a child during the war years, but I don't think that I'm a psychologist because of what happened to me in my youth.

PAULSON: OK. Well, thank you. This has been fascinating.

KAHNEMAN: Thank you.

Comments for this interview

reputation and little things (charlie, 05/31/2013 - 11:04am)

Perhaps Daniel's declining memory has failed to recollect the fact that Bertrand Russell wrote books most academics can't understand and others that out of discipline academics, and others, easily can.
Perhaps his Lordship was savvy enough to even more value out of his by the pound hired chaperones of reputation than has Daniel to almost extinct date-but that, since it can't be a personal memory of the past -or even wavering reasonable- must be altogether rationalistic-however a-System1(or is it not quite 2?) it thunders to the economic story tellers.