Making Sense - Sam Harris - #119 — Hidden Motives

00:00:00.000 Today, I am presenting the audio from the event I did in Denver with Robin Hanson.

00:00:22.740 Robin's a professor of economics at George Mason University, and he's the author with

00:00:26.860 Kevin Simler of a very interesting book titled The Elephant in the Brain, Hidden Motives in

00:00:32.340 Everyday Life. I give more of his bio from the stage, but I really enjoyed this conversation

00:00:38.440 with Robin. We spoke about all the related issues here of selfishness and hypocrisy and norms and

00:00:45.300 norm violations, cheating, deception, self-deception, the evolutionary logic of conversation,

00:00:53.340 social status, signaling and counter-signaling, common knowledge. There's many interesting

00:00:58.180 topics here. I enjoyed the event. Unfortunately, the audio is a little wonky. We are completely

00:01:03.880 at the mercy of whatever recording we get from these venues, and there are a few moments where

00:01:11.080 things cut out. It's a little echoey. It's not that bad. Once you start listening, you will acclimate

00:01:15.920 to it, but it was a good conversation. And so now I bring you Robin Hanson.

00:01:20.920 Ladies and gentlemen, please welcome Sam Harris.

00:01:29.640 Well, thank you all for coming out. Really, it's amazing to see you all or see some fraction

00:01:54.280 of you. I'm going to jump right into this. We have a very interesting conversation ahead

00:01:59.300 of us because I have a great guest. He is a professor of economics at George Mason University.

00:02:05.260 He is also a research associate with the Future of Humanity Institute, which you might know focuses

00:02:11.700 on existential risk and other big topics of ethical importance. He has a PhD in social science

00:02:18.860 science from Caltech, a master's in physics and the philosophy of science. He did nine years of

00:02:25.120 research with Lockheed and NASA, studying mostly artificial intelligence and also Bayesian statistics.

00:02:33.280 And he's recognized for his contributions in economics and especially in prediction markets,

00:02:39.420 but he's made contributions in many other fields. And he has written a fascinating book,

00:02:44.600 which unfortunately is not for sale here today, but you should all buy this book because it was

00:02:49.680 really, it's amazingly accessible and he just touches so many interesting topics. That book is

00:02:54.960 The Elephant in the Brain, Hidden Motives in Everyday Life. Please welcome Robin Hanson.

00:03:00.020 Thanks for coming.

00:03:12.380 So your reputation for being interesting precedes you.

00:03:16.380 I deny it all.

00:03:19.600 So I want to, there are many things we can talk about and as you know, but I want to focus

00:03:25.820 on your book and I want to move in a kind of a linear way through your book because your book is

00:03:29.900 just, is so rich and I don't think we will do the book justice, but we will try. The book is really

00:03:36.300 a kind of a sustained meditation on selfishness and hypocrisy. We have these ideas about why we do

00:03:45.900 things and then we have all of the evidence accruing for the real reasons why we do these things and the

00:03:51.940 mismatch there is, is rather harrowing to consider. And your book is just an unvarnished look at that.

00:03:58.060 So I want to tour through this, but perhaps before we get to some of these specific topics,

00:04:03.300 how do you view the project of your book? What were you up to in writing? I should say that you have a

00:04:07.700 coauthor on the book, Kevin Simler, who's not here tonight, but what were you doing writing this book?

00:04:13.940 This was what I wish I would have known when I started my social science career

00:04:17.880 many years ago. I started out in physics and then went into computer science. And in those areas,

00:04:25.320 I noticed that people were really eager for innovation. And then I seemed to see that in

00:04:29.720 social science, there were even bigger innovations possible. And so I moved there and then I was

00:04:34.200 puzzled to find that people were not very interested in innovations compared to the other areas.

00:04:39.800 And I kept also finding other puzzles in social science, ways in which our usual theories don't

00:04:44.600 make sense of what's going on. And my book, our book is an attempt to explain a lot of the major

00:04:51.640 puzzles in social science and a lack of interest in innovation. And one of the conclusions is that

00:04:57.640 we're just doing policy wrong, policy analysis wrong. But first we have to get into the basics here.

00:05:04.040 Well, so, and there's really two levels of it. There's, there's how you as a person might think

00:05:10.360 about these things, but it's the level of personal hypocrisy and then the mismatch between what your

00:05:14.840 motives actually are and what you may think they are. And then there's this, the fact that institutions

00:05:20.760 have this kind of structure or this, this blindness where the institutions think they're about something

00:05:25.560 and they're, they seem not to be upon analysis, an institution like medicine or, or a university or, or,

00:05:31.000 or, um, so what's the basic problem? Why is there this, this mismatch between what we think we're doing

00:05:37.560 and what we're actually doing? So if you've read many psychology books, you're familiar with the

00:05:42.280 idea that people are not always honest about what they're doing and why. And, uh, you might find that

00:05:48.120 trite and kind of boring by now, because of course we all know that, but so far people haven't taken

00:05:53.800 that insight to our major social institutions. And that's what we think is new and original about our

00:05:59.000 book. We say that, uh, not only are you not always honest about whether you like to go to the opera

00:06:04.120 with your spouse or whether you enjoy playing and cleaning up after your kids, you're also not

00:06:09.400 honest with yourself about why you go to school and why do you go to the doctor and why you vote,

00:06:14.200 uh, and why you do art. That is these, uh, deviations between what we think we're doing

00:06:22.120 and our actual motives, uh, infect many major social institutions and they therefore, you know,

00:06:29.880 should make us reconsider the basics of what these institutions are for and therefore why we

00:06:35.240 support them and whether we should subsidize them and how we should structure them and everything.

00:06:39.320 Right. So unlike many conversations I have here, I have a, a list of nouns that I just, that are kind

00:06:47.480 of a ladder through which we could, we could walk. Uh, let's start with norms and, and what you call

00:06:54.760 meta norms. So what is a norm and what, and why do we have them and what, and what does it mean to,

00:07:01.080 to protect them or, or to fail to protect them? So animals like chimpanzees and most other social

00:07:07.400 animals, they have a complicated social world and they pay attention to each other and they reward

00:07:11.800 others for helping them and punish others if they hurt them. So they, they have many regular behaviors,

00:07:16.840 but humans uniquely have norms in the sense that we have a rule of what you're supposed to do or not

00:07:21.800 supposed to do. And if somebody else sees you breaking the norm, it's a rule that they're

00:07:26.200 supposed to do something about it. They're supposed to tell other people that you've broken a norm and

00:07:30.600 then try to find a way to make you stop breaking the norms. And so humans have these norms about what

00:07:36.440 we're supposed to do or not supposed to do. And many of these norms are quite common around the world.

00:07:41.000 We're supposed to help each other. We're supposed to not brag, not, um, be violent to each other.

00:07:47.240 We're supposed to make group decisions together by consensus. And we're not even supposed to have

00:07:52.120 subgroup coalitions, uh, people who are aligned against the others. These are just common human

00:07:55.800 norms. And many of these norms are expressed in terms of motives. So there's a rule that we're not

00:08:01.720 supposed to hit each other on purpose. It's okay to hit accidentally, but not on purpose. And so,

00:08:08.680 we are, because our ancestors had these norms and they were so important, their social world was their

00:08:14.040 main world. Uh, we developed these big brains that we have mainly, apparently, for social reasons.

00:08:18.920 We developed these big brains to deal with our social world. And we have the biggest brains of all,

00:08:23.320 so our social world must have been the most complicated. Uh, but norms were a big part of

00:08:27.960 this world. And so we have this part of our brain that's all the time thinking about what we're doing

00:08:33.960 and trying to explain why we're following good motives. We're, and that's, in a sense, the conscious part

00:08:40.280 of your mind. You are the conscious part of your mind and you aren't necessarily the one in charge

00:08:45.640 of your mind. There's this idea that instead of, say, being the president or the king, you're the

00:08:50.200 press secretary. You don't actually know why you do things, but you're supposed to make up a good

00:08:55.800 excuse. And you do that. You're constantly looking at what you're doing and asking yourself, what would

00:09:02.600 be a good explanation for this thing I'm doing? And you're good at that. You're good at coming up with

00:09:07.800 excuses for what you're doing. But you don't actually know what you're actually doing. You don't realize

00:09:12.360 that you don't know. Yeah. And this is a very robust but not really celebrated neurological

00:09:21.320 finding. And it becomes horribly elaborated in people who have what's called a split-brain procedure,

00:09:27.640 where as a remedy for grand mal seizures, you can cut the corpus callosum, which connects the two

00:09:34.840 hemispheres of the brain. And that prevents the seizure activity from moving from one hemisphere

00:09:38.920 to the other. And what people have found going back now many decades is that the left, most people,

00:09:46.760 the left linguistically agile hemisphere confabulates reasons for doing things when those reasons are

00:09:54.040 brought out in an experimental paradigm. Those reasons are just manifestly not so. So you can present

00:09:59.320 the right hemisphere of the brain with a demand, like, you know, get up and walk toward the door.

00:10:05.320 And then you can ask the linguistically competent left hemisphere, why are you walking toward the door?

00:10:11.720 And it will confabulate a reason, like, I want to get a Coke. This is a result from a classic

00:10:16.600 experiment, which I think you cite in your book. These experiments were done by Roger Sperry and Michael

00:10:22.280 Kazanaga and Iran Zeidel. And the left hemisphere just continually completes the picture linguistically

00:10:31.240 without any apparent awareness that those claims are out of register. They're based on nothing.

00:10:37.160 They're just, I mean, this is what the word confabulate means, just to just make up this reason out of

00:10:43.080 full cloth. And it seems that, I mean, though most of us have not had our brains split, we have an

00:10:50.200 ability to give a post-hoc rationalization for why we did things, which certainly in an experimental

00:10:57.880 paradigm can be shown to really have no relationship to the proximate cause of our actions. And it is

00:11:03.320 embarrassing if caught on video. So we're living with this fact that we are our own press secretary,

00:11:13.000 giving the, at minimum, the most benign, but often just the most grandiose and apparently noble

00:11:20.440 rationale for why we're doing what we're doing. And yet evolution and other modes of logic suggest that

00:11:26.520 that isn't the reason for why we do much of what we do. Well, since you are in the habit of just making

00:11:31.640 up excuses, that means you could be wrong a lot, but doesn't mean you are wrong a lot. Maybe you

00:11:36.840 are mostly right, even though you would be wrong if you didn't know. So we have to go further than

00:11:41.640 just the possibility that you're being wrong to decide you're wrong. So we have to wonder, well,

00:11:47.560 how sure can you be about most of your activity, whether it's the real reason you have? Now,

00:11:53.480 one thing to be clear about is almost any area of life, like going to school or going to the doctor,

00:11:58.120 is big and complicated, the world's complicated, so a great many motives are relevant. And if we

00:12:03.080 average over people, surely thousands of different motives are relevant for almost everything we're

00:12:07.480 doing. And so what we have to be asking here is what is the main motive? What's the most dominant

00:12:12.360 motive? Not what's the only motive? So, I mean, just as an example, if you say the dog ate my homework

00:12:18.920 as an excuse, that only works because sometimes dogs eat homework. If dogs didn't exist, 0.64

00:12:24.040 it wouldn't make much sense. Dragon ate my homework doesn't work. So these things that we come up

00:12:29.880 with as excuses for our behavior, they only work as excuses because sometimes they're true. They

00:12:34.200 have an element of truth. So we're not going to say that your usual motive isn't at all applicable.

00:12:39.640 The claim is just, it's not as true as you think. And you're not saying that no one has noble motives.

00:12:45.240 Exactly. So there is real altruism, there's real nobility. There's real all of these things,

00:12:50.200 exactly. Sometimes people get up to get a Coke. Yes. But in addition, there are evolutionary

00:12:57.880 reasons why we would be self-deceived about our motives. We're actually, and this is based

00:13:04.040 often on the work of Robert Trivers, who's done a lot of work on self-deception and the evolutionary

00:13:08.280 logic there, we are better at deceiving others. We're better at getting away with norm violations

00:13:15.960 if we, in fact, are not aware that our press secretary is not telling the truth,

00:13:21.640 which is to say that if we, in fact, are self-deceived, we are better at deceivers.

00:13:25.400 So if we want to lie, it's better not to know we're lying because then it seems sincere.

00:13:31.240 Right. Well, you can be sincere. Right.

00:13:35.240 The easy way to seem sincere is to be sincere, even if you're wrong.

00:13:38.680 It's a famous Seinfeld episode, I believe. You're not lying if you believe it.

00:13:49.000 I should say that basically, this is something you and I should probably talk about, but

00:13:54.040 the jury's out as to whether or not knowing any of what we're about to talk about is good for you.

00:13:59.400 Sorry.

00:14:01.240 There's a psychological experiment being performed on you and you have not consented.

00:14:04.680 Memory wipe pills will be available after the session.

00:14:11.640 So how do you think about, let's take cheating as, cheating is a classic norm violation.

00:14:17.720 There's reason to think that our brains have evolved in large measure to both to cheat and to detect

00:14:23.240 cheating in others. How do you think about cheating in your line of work?

00:14:28.120 Well, cheating is again violating norms. And so we want to live in a community where the

00:14:34.600 norms are enforced and we also want ourselves to be exceptions to the rules.

00:14:40.120 So, for example, you know, most criminals actually think crime is a bad thing.

00:14:45.320 They just think that their particular acts don't quite count as crimes. So we all basically would

00:14:52.120 like to make exceptions for ourselves. So the question is how and one of the ways we can do it

00:14:55.880 is to not be very honest about what we're doing with ourselves.

00:14:58.280 This, this may not be relevant, but I just put me in the mind of it. I've never understood why

00:15:06.360 no one remarks on the fact that when we think of like just reducing our speed limit laws,

00:15:11.480 what that would do in terms of saving lives. And we could save tens of thousands of lives a year.

00:15:16.680 But if we, if we made cars that could not exceed the speed limit,

00:15:20.440 that would guarantee that no one would exceed the speed limit. But no one would want that. No,

00:15:25.640 no one, no one who thinks that we should have speed limits would want a car that it would

00:15:31.640 slavishly follow the speed limit. How do you make that? We all, is that just synonymous with wanting

00:15:38.200 the right to cheat on the speed limit? I mean, are we all imagining some emergency where you have to

00:15:43.080 speed past the speed limit? So the whole theme here is that in your head, you think you want

00:15:49.160 things. So in your head, you think you want to enforce speed limits. With your actual actions,

00:15:54.360 you don't. You want to speed. And there's a contradiction there. And you don't want to look

00:15:59.240 at that contradiction. So you look away. And that's the elephant in your brain. As you know,

00:16:03.720 the elephant in the room is the thing we all know is there that you don't want to look at. And the

00:16:06.920 the elephant in your brain is this contradiction between what you say you want and what you actually

00:16:11.480 do. So let's, let's actually raise this issue now, whether this is this line of thinking or this

00:16:18.600 analysis has a downside. So if in fact, it's true that we are better fit to our social environment

00:16:27.240 with a certain amount of ignorance with respect to our own motives, so that we, that it's optimal,

00:16:33.240 there's like an attractor of optimal fitness, which is, which entails some measure of self-deception.

00:16:39.880 And we are in the process, you in the process of writing this book, all of us in the process

00:16:43.880 of talking about it, are to some degree undeceiving ourselves about these things. Why isn't that bad

00:16:51.800 for us? And is that, is it worth worrying whether it's bad for us? So apparently evolution constructed

00:16:58.200 you to be ignorant about why you do things. It thought, yes, it might be useful if you know why

00:17:02.680 you do things, but that's to be traded off against all these other benefits of not knowing why you do

00:17:06.760 things. So you were constructed not to know. If the situation you're in in the modern world is

00:17:12.120 much like the situation evolution anticipated for you, that's probably better in your personal interest,

00:17:17.800 not knowing. You're probably better off going on with the usual sort of ignorance that the rest of

00:17:22.280 us have had, and acting that way, because you'll get along that way, and that's what evolution

00:17:26.760 anticipated for you. Now, evolution couldn't think of everything, so you could be in an environment

00:17:31.320 today which is not something evolution might have participated in, or you might be in an unusual

00:17:35.480 situation. For example, you might be a salesperson or a manager, the sort of person for whom it's really

00:17:40.760 important to understand people's motives and to be able to read them and understand what's going on.

00:17:44.920 You also might be a nerd, like myself, that as most people can just intuitively read the social

00:17:51.000 world around them and do the right thing. Some of us can't, and some of us need more conscious

00:17:54.920 analysis in the world to figure out what's going on, and so you may appreciate this more cynical

00:17:59.400 conscious analysis, even if it has some disadvantages. But most importantly... As a self-help seminar,

00:18:06.120 I think that's not going to sell tickets. Not that you'd be nerds or anything, but some of us are.

00:18:14.520 But I also just think if you're going to be a specialist in policy analysis, if you're going to stand up and

00:18:19.240 say, I have studied education or medicine and I have thought about what changes would be better,

00:18:24.680 it's more your responsibility to know what's actually going on in those worlds, even if it costs

00:18:29.320 you some degree of social awkwardness to know that. I think at least social analysts and policy analysts

00:18:35.720 should understand these things. Yeah, so let's take an institutional example. Take education. What is it

00:18:41.640 that we're deceived about with respect to education? So again, just to be clear, just because you might be

00:18:47.240 deceived about many things doesn't mean you are. So I need to walk you through arguments to convince you

00:18:51.560 that in fact, in each area, your motives isn't what you think it is. Now, my colleague, a beloved

00:18:57.480 colleague, Ryan Kaplan, has a book just out called The Case Against Education, and he goes through a whole

00:19:02.840 book like The Treatment of This. Our chapter is just a summary of that, but a summary is sufficient.

00:19:08.440 The summary is when you ask people why do you go to school, if they are answering in front of a public

00:19:13.640 speech or in a letter of application, say, they will tell you the usual story is to learn the material

00:19:18.520 so that you can become a more useful person later. That's our standard story about school.

00:19:23.800 And there are a number of puzzles in education that just don't make sense of that theory, and I'm going

00:19:28.120 to offer another theory that makes more sense of it. Some of these puzzles are you don't actually

00:19:33.000 learn very much at school. Most of the stuff you learn isn't very useful.

00:19:37.080 Yet, people who don't learn useful things are paid more. So, bartenders who go to college make

00:19:44.840 more than bartenders who go to high school. You do make more for more years of school in terms of

00:19:50.680 your wages, but the last year of high school and the last year of college is worth as much as the

00:19:54.680 other three years combined. But you don't learn more in the last year of high school or college.

00:20:00.600 I went to Stanford for a while for free without registering or applying simply by walking in and

00:20:05.560 sitting on classes. One of the professors gave me a letter of recommendation based on my performance.

00:20:11.640 Nobody tries to stop you from doing that. Why? You can get the very best education for free if you

00:20:17.320 don't want a credential. That calls into question the idea that you're there for the learning

00:20:21.800 as opposed to something else. So, the alternative theory is that you're there to show off and to

00:20:29.080 gain a credential that shows that you are worthy of showing off. That is, you are smart, conscientious,

00:20:33.960 conformist. You're willing to do the sorts of things that they ask you to do. You take ambiguous

00:20:39.400 instructions with long deadlines and consistently over several decades, over several years, complete

00:20:45.800 mildly boring assignments. Great preparation for future workplaces.

00:20:52.920 And by the end, you have shown that and that's something employers value. And that's a standard

00:20:57.880 plausible explanation for education. Most of you will find that plausible unless you are an education

00:21:03.080 policy expert, in case you will be offended, and search for another explanation. So, in most of

00:21:10.280 these areas, most of you will nod and say, yeah, that makes sense, unless this is your precious area.

00:21:14.840 For all of us, there is something precious in our lives, something sacred. And for that,

00:21:18.680 we will be more reluctant to accept one of these more cynical explanations of what's going on there.

00:21:23.320 But as long as education isn't sacred for you, you'll probably nod and say, yeah, you don't learn

00:21:27.560 much in school. But so now, what is signal and what is noise there? Are employers wrong to value those

00:21:33.640 things? What should people do differently as a result of understanding this about the status quo?

00:21:40.200 Individually, you shouldn't do different. Individually, if you want to convince an

00:21:42.920 employer in our world that you have what it takes, you do need to go to school, jump through the hoops,

00:21:48.040 and perform well. And in fact, you might do that better if you aren't aware that you're just doing

00:21:52.280 arbitrary things to show off to an employer. That may be demotivating for you. You might be better off

00:21:56.680 pretending to yourself and believing that you're learning. It would be useful.

00:22:01.960 But the point is, you are showing that you have a characteristic, not creating a characteristic.

00:22:06.120 The school isn't changing you. It's distinguishing you. It's like certifying you as different. 0.65

00:22:11.320 So now, what's the role of common knowledge in some of these situations? You should define what

00:22:18.040 common knowledge is. It's not common knowledge. What common knowledge is?

00:22:21.320 So think about cheating. He asked about cheating. And think of the rule that you're not supposed

00:22:26.520 to drink alcohol in public. This is a rule. And there are people who are supposed to enforce

00:22:31.320 this rule, the police. And you might think this, of course, is relatively easy to enforce.

00:22:38.520 But think of the example of people putting an alcoholic beverage inside a paper bag and drinking

00:22:43.800 it outside. This happens. Now, ask yourself, how hard could it be for the police to know that you're

00:22:50.760 drinking alcohol if you're drinking some bottle of a paper bag out of stores? Of course they know.

00:22:56.440 But you're giving them an excuse to look the other way. That is, that's not common knowledge. We

00:23:02.760 don't know that we all know that we all know that it's alcohol. Somebody could be fooled and that's

00:23:08.600 enough to pretend that you don't know. So this is why it's actually much easier to cheat in many ways

00:23:15.080 than you might have thought. We have all these rules and we're supposed to enforce them, but we're

00:23:19.000 not very eager to enforce them. We'd rather go about our business and ignore the rule violations. And so

00:23:24.600 a rule violation needs to be kind of blatant. And other people need to see us see the rule violation,

00:23:29.880 and then we kind of feel forced to do something about it. But if it's not blatant, it's not something

00:23:35.720 we all can see and know that we know, then you might prefer to pretend you didn't see. And many of you

00:23:41.320 probably have seen things that are not supposed to happen as you walk by the street. And you just

00:23:45.480 keep walking, hoping that nobody saw you saw it, because then you could pretend you didn't see it

00:23:50.360 and go about your business, because it would be a pain and trouble to stop and try to enforce the rules.

00:23:54.440 Yeah. Well, also, there's so much about our social lives where we know there's a subtext to what's

00:24:03.240 going on. But if that subtext ever became explicit, it would destroy the basis of trust or a good

00:24:10.280 feeling. Or like if you said to someone, I'm only inviting you over to dinner tonight because you

00:24:16.200 invited me last time and I needed to reciprocate. Right, exactly. You know, that's why we're having

00:24:21.400 this dinner. That, on some level, that we all know that's going on, but to make it explicit is sort

00:24:29.240 of antithetical to being friends with people. Right. So there's often many levels of what's going on.

00:24:38.200 And in fact, we expect to see that in movies and stories. So if somebody, as an actor, was given a

00:24:46.760 script and the script said, you're at a romantic dinner with somebody else, and the two of you are

00:24:51.880 there talking to each other, and what you're saying to each other is, I love you, I love you too,

00:24:55.960 this is great, we're having a wonderful relationship, this is a wonderful restaurant,

00:24:59.320 isn't this a great night? The actor will tell you, I can't act that. Because there's just one level

00:25:05.560 there, and that doesn't seem plausible at all. We expect in a scene like that there to be multiple

00:25:10.600 levels. That is, there's the surface level of I love you, isn't this great, and something else

00:25:14.920 must be going on. And the actor will actually look for another level so they can act the scene.

00:25:19.800 I'm afraid you'll leave me, so I'm trying to make sure you don't. Or I'm thinking of leaving you,

00:25:23.560 and so I'm trying to let you off nice. Something to make there be two levels of motives, because that's

00:25:28.360 what we expect to see out of actors and scenes. So we are really, at some level, we kind of know

00:25:33.720 that people are quite often pretending one motive and really acting another motive.

00:25:37.640 There's one thing, one chapter in your book on conversation, which I found fascinating,

00:25:42.920 because conversation is fairly mysterious in terms of the mismatch between what we think is going on

00:25:49.160 and what is actually going on, and why it would be valued in an evolutionary sense. So let's talk about

00:25:53.880 what most people think is going on during a conversation, and what seems to actually be

00:25:58.200 going on. So we're going meta here, because of course this is a conversation here, and we will

00:26:02.600 try to pretend that this isn't true about our conversation, because that's the social exception.

00:26:09.080 The jig is up. Exactly. So the usual story, if you ask why are you talking to your friend,

00:26:15.800 why did you spend an hour talking, why didn't you do the dishes or something useful,

00:26:18.920 you might say, well, we're exchanging information. We each have information, the other person doesn't,

00:26:26.920 and by talking and exchanging information, we can all know more. And this is the standard rationale

00:26:31.800 for most of our conversations. This, what I'm about to tell you, applies not just to personal

00:26:35.880 conversation, but also applies to our news media conversations, to academic conversations and

00:26:40.600 journals. In all of them, the standard rationale is information. That's why you read the newspaper,

00:26:45.160 of course right, to get more information. Now, there are many features of our conversations

00:26:49.880 that don't fit very well with this explanation. That's, again, my main argument here is to show

00:26:54.920 you the detailed puzzles that don't fit with the explanation, then offer you another explanation

00:26:59.400 that fits better. So some of the puzzles here are, if it was about exchanging information,

00:27:05.240 we would keep track of debts. I might say, well, I've told you three useful things so far,

00:27:09.720 you haven't told me any of the useful things. It's your turn. We would be more eager to listen

00:27:15.160 than to talk. It would be our turn to talk and then sigh. Okay, fine, I'll tell you something.

00:27:19.960 We would be searching for the most valuable things to tell each other, the things that mattered most

00:27:23.480 to each other, and we would talk about important things instead of the trivialities that we usually

00:27:27.240 fill our conversations with. And it would be fine to jump from topic to topic as long as we were

00:27:32.040 saying something valuable and important, because the point is to correct to communicate information.

00:27:36.280 But as you know, the usual norm of conversation is to slowly drift from topic to topic, none of

00:27:40.760 which need to be very important, but each time we should say something relevant to that topic.

00:27:45.960 Now, an alternative explanation in sharing information with this theory is that we are

00:27:49.880 showing off our backpack of tools and resources that we can show we can bring to bear to any topic

00:27:56.600 you dare to offer. So it's important that the conversation meander in a way no one of us can control,

00:28:04.520 so that we are each challenged to come up with something relevant to whatever that is.

00:28:10.520 And by impressing you with knowing something, having a friend or a resource, having a tip,

00:28:15.160 having some experience that's relevant to whatever you bring up, I show you that if you and I stay

00:28:19.800 allies and associates in the future, whatever problems you have, I'll have something relevant.

00:28:24.520 I'm ready for you with resources that would be useful to you, because look what I can do no matter

00:28:29.240 what conversation topic comes up. Yeah, well, the mismatch between desire to listen and desire to

00:28:36.200 talk is pretty, I mean, I think that's the one that people will find very salient, because if it was

00:28:41.400 really about just getting information, we would be massively biased toward listening. We would be

00:28:47.880 stingy with, I mean, we would be pricing out all of our disclosures. We'd have much bigger ears and

00:28:53.560 smaller mouths. Yeah. So then how do you think about gossip and reputation management and what's

00:29:02.200 happening in that space? We do, in fact, exchange information. Yeah. So again, it works as an excuse

00:29:08.440 because it's partly true. We do exchange information and it is somewhat valuable. It's just not the main

00:29:13.400 thing we're doing. But often, well, the information we're exchanging is meta to the actual apparent

00:29:19.560 topic. As you may know, indirectly through what people say, they tell you other things like bragging

00:29:27.000 about themselves indirectly by telling you about their great vacation in some expensive rare place.

00:29:33.720 And they talk about each other, often in the guise of, you know, saying what's been happening. But we are

00:29:40.120 very interested in knowing about each other and evaluating each other. And so part of what's going

00:29:45.800 on when we're impressing people is we're not only impressing the people who immediately see us, we're

00:29:49.560 impressing the other people who will hear about it indirectly. And so it's important that we impress

00:29:55.000 other people in ways that can transfer through that gossip to the other people who will hear about it.

00:30:00.280 And we are trying to avoid negative gossip or negative reputation of things that would make us look

00:30:05.240 back. And this is, you know, a basic explanation for why a lot of decisions in the world are really

00:30:10.200 quite shallow. So for example, as an employer, you might look at an employee and say, this potential

00:30:17.320 employee looks really good. Yes, they don't have a college degree, but they don't need a college

00:30:21.480 degree for this. And I can tell they could do the job. But then you might think to yourself,

00:30:24.760 yes, but other people will hear that I hired this person. And they will notice that this person

00:30:28.920 doesn't have a college degree, and they will gossip about it. And then I might look bad for having

00:30:33.560 hired someone with a college degree, and maybe I just don't want to take that chance. So even if I

00:30:37.480 know that this person could do the job, I still, because I'm trying to impress this wider audience

00:30:43.080 who will gossip about it, I am pushed to make shallower choices based on less than I know.

00:30:49.800 Is there anything that you do differently in this area based on having thought about this? Do you view

00:30:55.160 gossip as a negative character trait that should be discouraged in yourself, or do you just see it as

00:31:03.560 inevitable or socially useful as a way of correcting misaligned reputations?

00:31:10.120 I understand and appreciate gossip as an important human role as a natural nerd.

00:31:15.880 I'm not as inclined or interested in it personally, but that's my failing, not the world's.

00:31:23.720 So is social status the main metric to which all of this is pegged? Is that what we're

00:31:32.760 concerned about as subtext virtually all the time? It's one of the things, but it's actually less than

00:31:38.840 people might think. So if you're forced to admit you're showing off, often the thing you want to

00:31:47.960 admit to showing off is how great you are. That is how smart or conscientious or careful, how knowledgeable,

00:31:57.160 but plausibly at least half of what you're doing in showing off is showing off loyalty,

00:32:02.920 not ability. And so perhaps we push medicine in order to show that we care about people. We

00:32:09.640 participate in politics to show that we're loyal to our side. We do a lot of things to show loyalty,

00:32:15.000 and that is not something we're as eager to admit. Because of course, by trying to be loyal, we are

00:32:21.320 showing some degree of submission to those we are trying to...

00:32:24.360 Yeah, so that is a somewhat craven motive to sign on to. I'm being loyal, right?

00:32:31.080 Right.

00:32:33.320 But why is that? In fact, loyalty is a virtue that we acknowledge.

00:32:37.320 So humans actually have two different kinds of status, and it's suspicious and noticeable that we

00:32:42.200 don't make the distinction very often and we merge them together. There's dominance and prestige.

00:32:46.920 Dominance is more having power over someone, and prestige is earning respect. And the difference

00:32:53.800 of these actually show up in where your eyes go and how you look. When somebody has dominance

00:32:58.760 over them, you are not supposed to look them in the eye. Looking them in the eye shows defiance.

00:33:02.680 But if somebody has prestige, you are supposed to look at them. They are like, we presumably up here,

00:33:07.960 you are looking at us, we are claiming we have prestige, and you're not supposed to look away at them.

00:33:12.760 I wish you wouldn't put it that way. Yes, well... How embarrassing.

00:33:17.720 And so people want to get prestige and they don't want to admit to accepting dominance or submitting

00:33:25.320 to dominance, but of course we do. And so... But do people admit to wanting prestige?

00:33:30.920 More so. They might admit to accepting prestige, although not to seeking it, of course. Now,

00:33:37.720 in ancient history, most societies had kings and their neighbors had tyrants.

00:33:45.320 Tyrants dominated because they, bad guys over there, had dominance, and those people were

00:33:49.400 submitting to dominance and what a terrible thing they had to suffer. But we have a king who has

00:33:53.160 prestige and it's okay for us to look up to and obey our king because he's worthy of the status.

00:33:58.040 And so, this is often how people come to terms with their bosses. So, from a distance,

00:34:05.640 people say how terrible it is that we all obey our bosses at work. But each of us at work often

00:34:10.920 makes peace with that by saying, well, my boss is okay. He's earned that right to be in that role. And

00:34:16.360 I'm okay with doing what he says. Right. So now, I don't want to spend a lot of time on politics,

00:34:22.440 but obviously, everything you've written about is relevant to politics. And as I was reading the

00:34:28.840 book, it seems somewhat mysterious to me that in the current moment, someone like Trump seems to

00:34:35.320 violate more or less every rule you mention in your book. I mean, the things we've evolved not to do,

00:34:41.560 or not to do brazenly, like brag, right? Or lie without any hope of being believed.

00:34:46.760 Or advertise our most crass motives in place of more noble ones that could have been plausible,

00:34:59.080 right? He seems to get away with all of this. So, how do you explain the success of what's

00:35:06.600 essentially the anti-evolutionary algorithm? Sure. So, let's start with something called

00:35:14.680 countersignaling. So, ordinarily, if you have an acquaintance and you are trying to show that you

00:35:22.360 like an acquaintance, you will do things like, you know, smiling at them, being polite, flattering them,

00:35:28.520 opening the door for them, offering them some food. Those are ways we ordinarily show someone

00:35:33.640 that we are trying to be friendly. When you have really close friends, however, often you go out of

00:35:40.680 your way to do the opposite. You insult them. You trip them. You don't show up for some meeting.

00:35:47.560 Why do you do the opposite for a close friend? So, that's in part to show that you are more than an

00:35:53.080 acquaintance. Once it's clear that you are at least an acquaintance, people might wonder,

00:35:58.360 how close are we? And doing the opposite can show that you are more than an acquaintance. A paper that

00:36:06.200 discussed this was called too cool for school. As you know, many students try to show how studious,

00:36:11.480 how good they are at school by studying hard and doing well. And then some students try to show

00:36:15.880 that they can ace everything without trying. And that's, again, countersignaling. So he's managed

00:36:22.440 to convince half the country that he's their best friend by revealing all of these. But remember,

00:36:27.800 politics is about loyalty signaling. And at one level, we might all want politicians who are high

00:36:35.000 status. We might all want politicians who are articulate and tall and went to a good school

00:36:39.560 and smart and say all the right polite things and have salmon, etc. And so, in general, we would all

00:36:45.560 want the same thing there. But if you want to show that your side is different and you are being

00:36:51.480 especially loyal to your side, you may have to go against the norm. So, as you may know, when

00:36:57.880 the election of Trump, there was a subset of our society who felt neglected, who felt that their

00:37:04.440 voice was not being heard and that the politician establishment was not catering to them. And so,

00:37:10.920 Trump stood up and said, I will cater to you. And he went out of his way to show loyalty to that group

00:37:18.840 by counter-signaling in many ways, by doing the opposite of what an ordinary politician might do

00:37:24.520 to appeal to everyone, to show I really am appealing to you and you in particular. And I'm going out of

00:37:30.280 my way to raise the cost of appealing to other people to appeal more to you to show that I really

00:37:35.560 am loyal to you. And he did convince that group that he was unusually loyal to them and they voted for

00:37:41.720 him on that basis. And he has successfully counter-signaled his way into the presidency. The rest of the

00:37:47.720 world and other people are saying, but this is not the usual leader. And of course, the people who

00:37:51.400 voted for him said, yes, that's exactly how I knew he was trustworthy on my side, is that he

00:37:57.000 counter-signaled the usual signals of overall political competence. But we often do that to signal

00:38:07.080 loyalty. We often go out of our way to pay costs to signal loyalty. So, one of our chapters is on

00:38:11.240 religion, a topic that I know my guest host up here has written a lot about. And one of the standard

00:38:17.080 stories about religion is, you may agree to unusual rituals and to believe strange things

00:38:24.600 in order to convince people that the people you share those rituals and strange beliefs with, that

00:38:29.320 you are tied to them. And that it will be expensive for you to leave them and that they

00:38:33.560 therefore rely on you. Actually, viewing back to Trump, just for a second, viewing a lot of this

00:38:39.960 through the lens of loyalty explains a few other things. Because when you look at how

00:38:45.400 people in his inner circle, or people who have to function like his press secretary,

00:38:51.400 try to make, with as brave a face as possible, try to put some positive construal on his lying or his

00:38:59.800 mistakes or his misrepresentations of fact. That does function as a kind of loyalty test.

00:39:06.680 I mean, whenever you're just, when people with real reputations have to get out there and,

00:39:11.480 you know, put both feet in their mouths, so as to pretend that the president didn't do likewise,

00:39:17.640 it is a kind of, it is a kind of, I mean, it looks like a bizarre hazing ritual from here, but

00:39:23.080 it does signal loyalty. But again, those people across the border, they have tyrants and we have kings.

00:39:29.800 It's easy to criticize the other side for being excessively loyal and submissive, but

00:39:34.440 this happens all across the political spectrum. It's not just on the Trump side.

00:39:38.200 Yeah.

00:39:43.160 I don't know what that yeah meant. I wasn't accepting that at all.

00:39:49.080 That was confabulation, because you were wondering. So, I know you're an economist by day.

00:39:55.800 Let's spend a few minutes on incentives, because it seems to me that

00:39:58.520 many of the problems in our lives are the result not of bad people doing bad things, because they're

00:40:06.200 bad, but all of us good people, or more or less good people, struggling to function in systems

00:40:15.400 where the incentives are not aligned so as to get us to do the right thing most of the time,

00:40:20.520 or make it easy enough to do the right thing. I think you have a line in your book about

00:40:24.120 incentives being like the wind. I mean, you can decide to row into it, or you can tack against it,

00:40:30.520 but it's better to have the wind at your back. And so, how do you think about incentives, and what's the

00:40:35.720 low-hanging fruit here? But what is it that we could be doing differently in any area of consequence?

00:40:41.080 So, our book is about motives, and money, and power, and respect are things we have as motives,

00:40:50.280 and incentives are often aligned with those things. So, we often do the things that give us more of

00:40:57.640 these things we want. But we'd rather not admit that those are our highest priorities, and so we're

00:41:05.400 usually reluctant to overtly just do what it takes to get the money or respect. So, in most areas of

00:41:12.360 life, we have to put some sort of gloss of some higher motive that we have to pretend we're trying

00:41:17.640 to do. And that means often very direct, simple incentive schemes don't work. They're too obvious.

00:41:23.720 Right.

00:41:24.200 You know, just like your incentive to reciprocate the dinner,

00:41:29.080 that's an incentive you have, but you have also an incentive to not admit that too directly.

00:41:35.400 Right.

00:41:36.040 Because otherwise, you would force them to admit that they mainly wanted the reciprocation. And so,

00:41:43.160 this is an issue with incentives that many of the problems we have in the world happen because we

00:41:49.800 have insufficient incentives to do the right thing, but often that's because we don't want to admit how

00:41:55.880 important incentives are so we don't want to admit that we need incentives so we don't restructure

00:42:00.200 things to produce the incentives we need. And because we want to pretend we don't need the

00:42:05.400 incentives. So, for example, your doctor's incentive to give you the best treatment can often be

00:42:11.000 compromised by the fact that they might, under one incentive system, just want to treat you more

00:42:15.480 just because they get paid every time they treat you, or another incentive system might need to treat

00:42:20.040 you less because they have to pay out of their pocket every time they treat you. Under either case,

00:42:24.280 their incentives might not be well aligned with you, but you could have set up some sort of more

00:42:30.200 direct incentive system where, you know, they had a stake in your health, but you might not be

00:42:35.320 comfortable with asking for that because that might show that you didn't trust your doctor. You might

00:42:40.920 rather on the surface pretend like you trust your doctor and they trust you and you have a nice,

00:42:45.400 comfortable relationship. This is also a problem in financial investment, actually. An awful lot of

00:42:51.880 people invest an awful lot in intermediaries who take a lot of money, but don't offer that much

00:42:57.800 in return. And people often just like the relationship they have with the intermediary,

00:43:02.840 and they don't want a distrusting relationship that would have some explicit stronger financial

00:43:07.240 incentives, so they accept a weak relationship. People often want to, like, feel like you had a

00:43:13.080 relationship and that relationship is degraded by the idea that you might not have trusted them.

00:43:20.920 I'm a researcher in academia and most money comes in the form of grants where they say apply for the

00:43:27.160 grant and then they might give you the grant. We've long known that prizes are often more effective. A

00:43:31.720 prize is where they say if you do the following thing then we'll give you this much money.

00:43:35.880 And a prize can give stronger incentives for people to do things, but a prize is less trusting.

00:43:40.200 And you as the granting agency often want to just form a relationship with someone and then take

00:43:45.560 credit for them as if we were buddies. And this prize sort of makes an arms-like distance where

00:43:51.720 clearly I don't trust you if I'm going to only pay you if you do this measurable thing. And so we'd

00:43:57.160 rather have this closer relationship than to have a stronger incentive. Is there a meta level to many

00:44:03.000 of these considerations where it can be reasonable to not follow the purely rational line through all

00:44:10.680 of these problems? It sounds like what would happen is if we took all of this to heart we would try to

00:44:18.680 bootstrap ourselves to some new norms that that paid better dividends by or that seem more rational

00:44:25.240 economically or otherwise, or in terms of like health outcomes. And yet given human nature we might

00:44:32.680 find the price of anchoring ourselves to those new norms to be unacceptable for one reason or another.

00:44:37.320 So the way I would summarize this is to say our usual institutions let us pretend to be trying to get

00:44:46.680 the thing we pretend to want while actually under the surface giving us the things we actually want.

00:44:55.000 Policy analysts typically try to analyze how to give policy reforms that would give us more of the things

00:45:00.760 we pretend to want. And we're usually uninterested in that because we know we don't actually want more of the things

00:45:06.680 we pretend we want. If you could design a policy reform that let us continue to pretend to get the things

00:45:14.680 we pretend to want while actually getting more of what we actually want. We'd like that, but we can't admit it.

00:45:20.360 If we stumble into it we'll stay there. But if the policy analysts were just to out loud say,

00:45:26.520 well this is a system that will give you more of this thing is what you actually want, but admit it,

00:45:30.440 don't you? We don't want to admit it and then we won't want to embrace that. So yes, what we want to do

00:45:36.200 is pay for the appearance of the thing we're pretending to want and we're often paying a lot for that appearance.

00:45:41.960 Hmm. I would love to see a transcript of what you've just said there.

00:45:49.640 So I'm going to ask you some rapid fire kind of bonus questions here. I want to leave a lot of

00:45:56.280 time for Q&A because though conversation isn't about just exchanging information, you have a lot of

00:46:02.520 information to exchange and I want to get the audience involved. But if you had one piece of advice for a

00:46:10.200 person who wanted to succeed in your area of work, what would that be?

00:46:15.720 I am an intellectual and my measure of success would be insight. There are other measures of

00:46:23.400 success. You could have a prestigious position, you could make a lot of money, you could get a lot of

00:46:28.200 friends. But if the measure of success is insight, then a number of strategies, one of which is just to

00:46:35.400 look for neglected areas. So as we talked about in conversation, there's a strong norm in ordinary

00:46:40.600 conversation to follow the conversation, to talk about what everybody else is talking about. And

00:46:45.400 academics do that, news media does that, and we do that in ordinary conversations with people. But

00:46:50.680 for intellectual contribution, if you jump right in on what everybody else is talking about, your

00:46:55.960 chances of making a large impact are pretty small. You're adding a small amount to what everybody else is

00:47:00.280 talking about. If you go talk about what somebody else isn't talking about, find something important

00:47:04.440 but neglected, your contribution can be quite large, even if you're not especially brilliant

00:47:09.320 or well-tooled. And so one very simple heuristic, if you want to produce intellectual insight, is just

00:47:14.600 to look at what other people aren't looking at that seems important and hope that later on they'll

00:47:19.320 come around to your topic and realize that you did make a contribution. But how long would you stay in

00:47:24.200 that important area, waiting for people to come around? You don't have to stay, you have to stay

00:47:29.480 long enough to make a contribution, and then you can go off looking for another area to make a

00:47:33.160 contribution to. What, if anything, do you wish you had done differently in your 20s, 30s, or 40s? You

00:47:39.320 can pick the relevant decade. Well, I wandered around a bit, much like Sam, in that I started my PhD

00:47:48.920 program at the age of 34 with two kids, eight zero and two. It's a relatively late start. That was,

00:47:54.680 in some sense, the price for continuing to switch because other areas seem to actually be more

00:47:59.400 objectively important and have more promise. But as I said before, this book that I'm out with here

00:48:08.920 is summarizing the thing I wish I would have known at the beginning of that social science career, which

00:48:12.920 is that we are just often not honest with ourselves about our motives. So the thing I'm most known for

00:48:19.480 actually is something called prediction markets, betting markets on important topics. And they do

00:48:23.960 work well, and they give people something they say they want, which is more accurate estimates and

00:48:28.120 information on important topics. And it turns out people are usually not very interested in them,

00:48:33.720 even though you can show over and over again in many ways that they work and they're cheap, etc.

00:48:38.600 And part of why I didn't realize that that would happen is I took people at the word for what they

00:48:44.680 want. So you wish you hadn't spent so much time on prediction markets? Well, I wish I would have

00:48:49.000 understood the constraint that people are not honest about what they want and thought about that

00:48:52.760 constraint when I was initially trying to design institutions. So I've read many other ideas and

00:48:58.840 worked on ideas for reforming politics and medical purchasing and information aggregation, etc. And in

00:49:04.920 each case, I assumed the usual story about what we're trying to do and worked out a better answer.

00:49:11.960 And we actually can not only work out better answers, we can show them and not only in math,

00:49:16.040 but in lab experiments and field experiments, we do actually know many ways to make the world better

00:49:19.880 substantially. And the world's not interested, most of them, because we know how to make the world

00:49:24.120 better according to the thing that people say they want, to learn more at school, to get healthier at

00:49:28.760 the hospital, to get more policy and politics. But in fact, emotionally at people's heart, they kind of know that's not what they

00:49:34.920 want. And so they're not interested. So I wish I would have known that 20 years ago. And this book is

00:49:40.280 hopefully to somebody at a younger career, somebody can pick this up. You might know a 20-year-old who's

00:49:45.320 been saying for a while, everybody's bullshitting, nobody's telling the truth. Where can I find out

00:49:50.280 what's really going on? I'm hoping our book can be that book.

00:49:53.240 So 10 years from now, what do you think you'll regret doing too much of or too little of at this point in your life?

00:50:01.640 I mean, if I knew that, I would presumably be doing something different.

00:50:07.000 Do you actually think that's true? Isn't that just one of the problems? For instance, you know

00:50:14.440 you want to lose weight, you know how to lose weight, but you still can't get the ding-dong out of your

00:50:18.200 hand? The major issue would be if I'm neglecting in the long run for the short run, right? I don't

00:50:22.200 know if I am. But yes, if I am neglecting in the long run, then I would regret not investing more in

00:50:27.480 the long run. But I am primarily investing in this long run effort to produce intellectual insight.

00:50:33.640 And I actually think there are scale economies in that. So the more fields you learn, the more

00:50:39.880 mental models and tools you have is to learn new fields. So you can actually learn new fields faster

00:50:45.480 the more fields you have. So if your intellectual project is to learn many fields and then find ways

00:50:49.320 to combine the insights from them together, that's something you continue to do more and better

00:50:53.960 as you get older. And so I'm enjoying that wave and not thinking I'm over at all.

00:51:00.200 What negative experience, one that you would not wish to repeat, has been most valuable to you?

00:51:07.240 Most valuable to me, negative experience? Or changed you for the better.

00:51:13.080 But it's got to be negative. You wouldn't want to repeat it.

00:51:15.400 Well, so early in my academic career, I sort of really just failed to do the simple standard

00:51:26.440 thing of applying to the best colleges. I'm not sure what went wrong, but somehow my family or me

00:51:31.480 somehow just did not go through the process of applying to good colleges far away. We just

00:51:35.640 sent me to the local college, which was easy for me, too easy compared to my colleagues. So I had lots

00:51:42.760 of free time. So perhaps I might have thought I should have gone to a more challenging college,

00:51:47.960 and then people would have challenged me. But that made me who I am in the sense that with all that

00:51:52.600 free time, I just started studying stuff on my own. I sort of made up my own topics and made up my own

00:51:58.360 questions and just started going in and working on things. And so actually, I was a physics

00:52:03.480 undergraduate major, and the first two years of physics classes are going over all the major topics.

00:52:08.040 And then the second, the last two years are going off over all the major topics again with more math.

00:52:13.320 And I had all these questions that the math was not answering. And so what I did in the last two

00:52:18.200 years of college was to just play with the equations, just rearrange them, try different ways. And by

00:52:24.280 spending the semester rearranging the equations, I could ace the exams. But I didn't do any of the homework.

00:52:30.920 And so professors who had a formula, like so much percentage homework, so much percentage exams,

00:52:34.920 they, you know, didn't know what to do with me exactly. And so I got low grades in some classes,

00:52:39.640 although I was, you know, people were willing to give me letters and recommendations. But basically,

00:52:43.560 that formed me. That is, I became the person who didn't like do what I was told. I wasn't following

00:52:50.440 a path that people had led for me. And I wasn't like going down, learning the things I was supposed

00:52:54.280 to learn. I was just making up my own problems and my own questions and working them out for myself.

00:52:59.320 And in the end, that has some advantages, but I'm not sure that was best overall.

00:53:05.160 I'm going to put that in the bragging category.

00:53:17.080 What worries you most about our collective future? 0.96

00:53:20.520 We are collectively ignorant compared to what we could be. We are a vast population, a vast world,

00:53:29.800 a lot of smart people, very capable people. We have many great tools. And we just don't pull that

00:53:35.240 together into a consensus that we can use very well. We fail to do something we could do quite easily.

00:53:41.960 My work on prediction markets was one attempt to try to create an institution which would allow us to

00:53:46.840 collect what we know together effectively and efficiently. And it would work if anybody was

00:53:50.200 interested. But we're not very interested. And so part of my intellectual work is just try to

00:53:55.080 diagnose why aren't we interested as part of the understanding how could we do better. And I think

00:53:59.320 this fact that we're all trying to show off to each other is part of it. So if I ask, well, what's

00:54:03.160 going wrong with our showing off? I would say the problem is we are showing off to audiences that are

00:54:08.200 too ignorant. That is, if we focused on a really smart audience, a really knowledgeable audience,

00:54:15.080 you're trying to show off to them, then we would be forced to show off in better ways.

00:54:20.200 So for example, we haven't talked much about it, but basically I've said medicine is mostly about

00:54:24.680 showing that we care rather than helping people to get healthy. So when grandma's sick, you make sure

00:54:30.360 she gets expensive medical treatment, the sort that everybody would say is the reasonable thing to do, 0.63

00:54:34.600 even if it's not actually very effective. But as long as your audience doesn't know it's not very

00:54:40.200 effective, they will still give you credit for being caring about grandma. If your audience knew that the

00:54:44.920 medicine you were pushing hurt her instead of helping her, they would not consider you as such

00:54:49.080 a caring person. So the more that our audience knows about what actually works and has what effects,

00:54:54.600 the more we would all be pushed to do things that actually had good effects as part of the process

00:54:59.960 of trying to show off and show that we care. Similarly in politics. Actually, before we move on to that,

00:55:04.840 say more about the mismatch in medicine. How is it that we know, or how is it that you think you know

00:55:10.440 that it's more about caring than about results? So again, the structure is a set of puzzles that

00:55:15.880 don't make sense from the usual point of view. So it turns out we have data on variations in health

00:55:22.680 and variations in medicine and there's almost no relationship. That is geographic areas that spend

00:55:28.360 more on medicine or have people do more doctor visits, those areas are not healthier. We also even

00:55:33.640 have randomized experiments where some people have been given randomly a low price of medicine and they

00:55:37.800 consume more and other people have a high price and they consume less and then there's no difference

00:55:43.480 in health between these groups. So at a very basic level, there's very little if not any correlation

00:55:49.240 between health and medicine. Not only that, there are other things that correlate strongly with health

00:55:53.960 that people show very little interest in. Well, there must be a lower bound to that though,

00:55:58.680 because some medicine is life-saving clearly, right? Where are you putting the line between... Well,

00:56:04.040 it's not a line. That is, there's a whole mix of medicine and some of the stuff helps and that

00:56:07.640 means other stuff hurts. So if you could just get the stuff that helps and avoid the stuff that hurts,

00:56:13.880 why then you could do better? But people show relatively interest in doing that. And so some

00:56:19.000 medicine hurts, not only does it do zero, it on average hurts. We are not interested in exercise, air

00:56:27.160 quality... But what's the measure of people not being interested in the information that would allow

00:56:31.800 them to get better medicine? We have experiments and studies where people have been given access

00:56:39.560 to information and asked if they would be willing to pay much for it and even just give it and seeing

00:56:43.560 if it affects the behavior. And consistently, if you give people privately information about the quality

00:56:48.840 of medicine, they just aren't interested and don't act on it. And they won't pay for it, right?

00:56:53.880 They certainly won't pay for it, exactly. So there was a study of people about to undergo heart surgery,

00:56:59.880 where a few percent of people undergoing heart surgery die. So that means you face a few percent

00:57:04.920 risk of death. That should be a serious situation. They said, we have statistics on the local surgeons

00:57:09.800 and the local hospitals in terms of what the percentage is of those patients dying there. And

00:57:14.200 it varies by quite a bit, twice as much in some places than other places. Would you like this information?

00:57:21.000 Only eight percent were willing to pay 50 bucks.

00:57:23.160 And those who were just given the information didn't act on it.

00:57:28.360 Why is it that I think that everyone I know is in the eight percent?

00:57:32.040 Well, that's what they're pretending. A way to understand this is to think about

00:57:37.640 Valentine's, which happened recently. On Valentine's, it's traditional to try to show that you care

00:57:42.280 about someone by, say, buying them a box of chocolates. Now, when you do this, do you ask how hungry they are

00:57:47.560 when you think about how large a box to buy? No, plausible. You need to buy as much chocolate as

00:57:54.040 it takes to show you care more than somebody else, regardless of how hungry they are, which is like

00:57:58.680 medicine. We just give people a lot of medicine, even though the extra medicine isn't very useful.

00:58:03.160 And if you ask, well, how do I know which quality of chocolate to get? You know that you need to give

00:58:07.240 a quality of chocolate, a signal that's a common signal of quality. If you happen to privately know that

00:58:12.600 this is a great kind of chocolate, or they happen to privately know a certain kind of thing is a

00:58:16.840 great kind of chocolate, that won't actually affect whether you interpret this as a generous

00:58:20.920 act. The interpretation of generosity is based on a common signal of quality. So if medicine is

00:58:26.440 a way to show that we care, then similarly, what we want is common signals of quality. We aren't

00:58:31.320 actually very interested in private signals of quality of medicine, which is what we actually see.

00:58:36.680 All right, back to rapid-fire questions.

00:58:39.800 I'm taking too long.

00:58:40.920 No, no, I've been asking follow-ups as well.

00:58:44.040 If you could solve just one mystery or problem...

00:58:49.640 If you'd like to continue listening to this podcast, you'll need to subscribe at

00:58:53.160 samharris.org. You'll get access to all full-length episodes of the Making Sense podcast,

00:58:58.200 and to other subscriber-only content, including bonus episodes and AMAs,

00:59:03.000 and the conversations I've been having on the Waking Up app. The Making Sense podcast is ad-free,

00:59:08.040 and relies entirely on listener support. And you can subscribe now at samharris.org.

00:59:12.280 samharris.org.

00:59:22.920 Thank you.

Making Sense - Sam Harris - March 12, 2018

#119 — Hidden Motives

Episode Stats

Summary

Transcript