AI transcript
0:00:02 (upbeat music)
0:00:07 Pushkin.
0:00:11 Hey, it’s Jacob, and I want to tell you
0:00:14 about a podcast called What Could Go Right?
0:00:18 The show is hosted by Zachary Carabell and Emma Varvalukas.
0:00:20 They are the founder and executive director
0:00:22 of The Progress Network.
0:00:25 And on the show, they sit down with expert guests
0:00:28 to discuss the world’s most pressing issues.
0:00:32 And they do not resort to pessimism or despair.
0:00:35 Plus, every Friday, the hosts share reports
0:00:38 highlighting the latest progress from across the globe.
0:00:40 For a dose of optimistic ideas from smart people,
0:00:43 listen to What Could Go Right?
0:00:44 It’s a podcast.
0:00:47 You can get it where you get podcasts.
0:00:50 Hey, happy New Year.
0:00:52 We are very happy to be back.
0:00:55 And I have one request before we start the show.
0:00:57 I’m asking you a favor.
0:00:59 And the favor is this,
0:01:04 would you please send us an email to problem@pushkin.fm
0:01:09 and tell us what you like about the show
0:01:10 and what you don’t like about the show.
0:01:12 And specifically, what kinds of things
0:01:14 you want to hear more of
0:01:16 and perhaps what kinds of things you don’t want to hear.
0:01:21 Again, it’s problem@pushkin.fm.
0:01:22 I’m going to read all the emails.
0:01:24 So thank you in advance for sending them.
0:01:29 Claude Shannon is this huge figure
0:01:31 in the history of technology.
0:01:34 He’s one of the key people who worked at Bell Labs
0:01:36 in the middle of the 20th century
0:01:38 and really came up with the ideas
0:01:41 that made modern technology possible.
0:01:44 But I’m going to be honest with you.
0:01:48 I never really understood what Claude Shannon figured out
0:01:50 that was such a big deal.
0:01:52 But the people who know about technology,
0:01:53 who know about the history of ideas,
0:01:56 they say Shannon’s a giant.
0:01:58 Claude Shannon is like the nerd’s nerd.
0:02:02 He’s the techno intellectuals, techno intellectual.
0:02:06 And so for today’s show, I wanted to understand
0:02:08 what did Claude Shannon figure out
0:02:11 and why is it so important for the modern world?
0:02:20 I’m Jacob Goldstein and this is What’s Your Problem.
0:02:22 My guest today is David Shea.
0:02:27 David is a professor of electrical engineering at Stanford.
0:02:29 He’s studied Shannon for decades.
0:02:32 He teaches Shannon’s work to his students.
0:02:35 And David used Shannon’s work
0:02:38 to make a breakthrough in cell phone technology.
0:02:40 And that breakthrough, that breakthrough
0:02:42 that came to us via Shannon and Shea,
0:02:45 it affects every phone call we make.
0:02:48 David and I talked about Shannon’s key insights
0:02:51 and about how David’s own work built on Shannon.
0:02:54 And we also talked about the big chunk of Shannon’s life
0:02:58 that was taken up with juggling and riding unicycles
0:03:00 and building mechanical toys.
0:03:03 But to start, we talked about how in the middle
0:03:06 of the 20th century, Bell Labs wound up driving
0:03:09 so much technological innovation.
0:03:16 Yeah, so Bell Labs was the research lab of AT&T.
0:03:21 AT&T at that time was the phone company, okay?
0:03:23 Nowadays, we have many phone companies, right?
0:03:28 We have Verizon, we have T-Mobile, et cetera.
0:03:31 But those days, there was only one phone company
0:03:33 and that’s a monopoly.
0:03:36 So a monopoly needs to justify its existence.
0:03:39 – So it doesn’t get broken up by the government.
0:03:40 – It doesn’t get broken up.
0:03:42 Of course, it eventually got broken up.
0:03:45 But at that time, it was a monopoly.
0:03:49 And so one way of justifying its existence
0:03:53 is to say that, okay, it says to the American people,
0:03:57 to the government that we will always spend
0:04:01 a certain percent of our revenue on this research lab,
0:04:04 called Bell Labs.
0:04:08 And whatever Bell Labs come up with is kind of our
0:04:11 contribution not only to our bottom line,
0:04:16 but also to technology of the country.
0:04:18 – So they have this sort of public mission
0:04:21 to prevent the government from breaking them up.
0:04:25 – Yeah, and so therefore it also allows researchers
0:04:29 a very free reign to do research
0:04:31 that not necessarily tied to, like say,
0:04:35 a particular business unit, okay?
0:04:36 So they can be very creative.
0:04:38 And that’s the atmosphere of Bell Labs.
0:04:42 So Bell Labs attracted a bunch of very smart people
0:04:44 because smart people wants to work on their own problem,
0:04:47 not the problem that the manager gives them.
0:04:51 – Okay, that’s one characteristic of smart people.
0:04:54 And so, yeah, that was the heydays of Bell Labs.
0:04:57 Lots of smart people inventing amazing stuff.
0:04:59 Laser was invented there.
0:05:03 Information theory, the transistor was invented there.
0:05:08 Sort of almost all the foundation of the information age,
0:05:13 where there’s hardware, algorithm, software,
0:05:16 is in some sense all have the roots at Bell Labs.
0:05:19 So that was the contribution to mankind, actually,
0:05:21 I should say, not only to America.
0:05:24 – So Shannon gets there at this time, right?
0:05:28 He’s there when they’re inventing,
0:05:31 certainly the transistor, what’s he do?
0:05:34 Tell me about his work there when he gets there.
0:05:35 What’s he working on?
0:05:41 – Yeah, so I think Shannon always have his own agenda, right?
0:05:46 We know for a fact that he has been interested
0:05:49 in the problem of communication,
0:05:54 that idea of having a grand theory of communication,
0:05:57 even back in 1938, I think ’37, ’38,
0:06:00 because he wrote a letter at that time
0:06:03 to a very famous person named Venera Bush.
0:06:04 – Yeah.
0:06:06 – Venera Bush is very famous.
0:06:10 It was, I think, president of MIT or dean of MIT,
0:06:12 and then he became sort of a scientific advisor
0:06:14 to the president.
0:06:17 And so he wrote a letter to Venera Bush in 1938
0:06:19 and said, “Hey, you know what?
0:06:20 I’m really interested in this question
0:06:22 of how to find one theory
0:06:25 that unifies all possible communication systems.
0:06:27 There’s so many different communication systems out there,
0:06:30 but I think there’s something at the heart of every system.
0:06:31 And I’m trying to get to the heart.
0:06:34 – And nobody had thought of it in that way, right?
0:06:38 It seems like part of his, part of why Shannon
0:06:39 is such a big deal is like,
0:06:41 as I understand it, people is like, you know,
0:06:43 people understood, like, they were trying to figure out
0:06:45 how to make the phone work better.
0:06:47 And they were trying to, you know,
0:06:49 make movies be clearer or whatever.
0:06:52 But there wasn’t this idea that you could abstract it
0:06:54 until Shannon came along.
0:06:56 – And the reason is very simple, actually,
0:06:59 because if you have a physical system
0:07:01 that you want to build, right?
0:07:02 What do you see, right?
0:07:04 You say, “Hey, man, the video, for example,
0:07:06 I’m seeing you right now.
0:07:08 I’m not seeing you very clearly, I have to say.”
0:07:11 – Yes, I made a closet, I made a closet.
0:07:13 – Right, yeah.
0:07:15 – Then I would say, hey, how to try to improve the image?
0:07:18 Maybe I can try to, you know, fix this pixel
0:07:21 or do some filtering of your noise.
0:07:25 So I’m very tied to the very specific details
0:07:28 of the specific problem, because why I’m the engineer?
0:07:30 I need to improve the system, not in 10 years,
0:07:32 but tomorrow, you know, tomorrow,
0:07:33 I need to get a better system.
0:07:34 – You don’t need a theory of the system.
0:07:37 You just want a clearer picture, yeah.
0:07:39 – Yeah, I’m in the wits, right?
0:07:40 I’m in the wits.
0:07:44 And Shannon, because of his training,
0:07:47 and also because of the embassy of a place like Bell Labs,
0:07:50 could afford to sort of step back
0:07:52 and just look at the broader forest
0:07:55 as opposed to the details of specific trees.
0:07:59 – So, okay, Shannon’s big idea
0:08:02 comes out in this paper he publishes in 1948.
0:08:05 The paper’s called “A Mathematical Theory of Communication.”
0:08:07 It’s like his great work.
0:08:09 Tell me about that paper.
0:08:12 – So that paper is actually a very interesting paper.
0:08:15 In fact, when I teach information theory,
0:08:18 I teach from the paper itself,
0:08:21 because I thought it’s an amazing way,
0:08:23 not only of learning information theory,
0:08:26 but learning how to write a scientific paper properly.
0:08:28 – Uh-huh.
0:08:29 – Okay, and you know,
0:08:32 not everyone does research and information theory,
0:08:33 but everybody has to write.
0:08:34 – Uh-huh.
0:08:36 – Okay, every researcher has to write
0:08:40 to express their ideas to the peers and to the audience.
0:08:43 So in that paper, very interesting.
0:08:44 The first paragraph of the paper, okay,
0:08:46 is already very interesting.
0:08:48 Because typically when people write a paper nowadays,
0:08:51 they tell you, oh, how great my invention is.
0:08:52 It’s gonna change the world.
0:08:55 Every paper is gonna change the world.
0:08:57 But in fact, his first paragraph
0:09:00 focused on telling you what his paper is not achieving.
0:09:02 – Uh-huh.
0:09:03 – I mean, that’s a master.
0:09:05 That’s a master’s, right?
0:09:08 I mean, how many papers that you read nowadays
0:09:10 tells you in the beginning,
0:09:13 hey, you know what guys, expectation management here,
0:09:15 this paper is not about this, not about that.
0:09:16 – Don’t get your hopes up.
0:09:18 Hey, don’t get your hopes up, yeah.
0:09:19 – Exactly, that’s exactly what he did.
0:09:21 Expectation management.
0:09:22 – Okay.
0:09:25 – Nowadays, today we will call it expectation management.
0:09:28 And now those days, I guess he just calls it honesty.
0:09:31 And his whole point was,
0:09:35 often people associate information with meaning.
0:09:36 – Uh-huh.
0:09:37 – Okay?
0:09:39 And then he said in this paper, we ignore meaning.
0:09:42 We ignore meaning, okay?
0:09:46 So that was the first thing he did.
0:09:47 Which is brilliant,
0:09:50 because once you tie information with meaning,
0:09:53 then he will never be able to make any progress.
0:09:56 It’s just too difficult and too broad and too vague a problem.
0:09:59 – Everybody gets stuck on this idea of meaning
0:10:00 and what is meaning.
0:10:02 And he’s like, forget about meaning.
0:10:05 So if we’re gonna forget about meaning, what is left?
0:10:08 – Yeah, so actually the biggest,
0:10:10 I think breakthrough of that paper
0:10:15 is to really focus on the thing that matters.
0:10:19 And cut away a lot of stuff that really doesn’t,
0:10:21 not that it doesn’t matter,
0:10:22 but it doesn’t matter in terms of solving
0:10:24 the communication problem.
0:10:27 The communication, so then he said,
0:10:29 okay, what is the communication problem?
0:10:31 The communication problem is the following,
0:10:36 is that there are multiple possibilities of a word.
0:10:40 And my goal is to tell the receiver, destination,
0:10:42 which of the multiple possibilities
0:10:44 is the correct possibility.
0:10:47 – Yeah, and so in language, it’s basically,
0:10:49 it’s a finite set, language is a finite set.
0:10:51 It’s very large, but if we’re speaking,
0:10:54 and we both know that we’re speaking English,
0:10:56 then essentially you are hearing the words
0:10:57 and decoding them.
0:10:59 And you know that it is a series of words
0:11:02 and you just have to figure out which words.
0:11:04 I mean, like that, for example?
0:11:05 – Yes, like that.
0:11:08 – Okay, so that’s the frame he builds, then what?
0:11:14 – Okay, all right, then once you have this framing, right,
0:11:18 then you can ask the question, okay,
0:11:20 what is the goal of communication?
0:11:22 The goal of communication is to communicate
0:11:27 as fast as I can, right?
0:11:30 And the natural question is, why is there a limit
0:11:34 on how fast I can communicate to you?
0:11:38 Because if there’s no limit, then amazing world, right?
0:11:40 We can communicate so fast.
0:11:42 – It’s like instant telepathy.
0:11:45 It’s like you instantly beam me every thought in your head.
0:11:45 Yeah, okay.
0:11:47 – Exactly, the natural question he has,
0:11:50 once you set up this finite set, as you mentioned,
0:11:52 is okay, given these finite sets,
0:11:55 is there a limit on how fast I can communicate to you?
0:11:59 And so that was the question that was the heart of the paper,
0:12:04 which is to, so he formulated this notion
0:12:09 of a capacity that communication system is like a pipe.
0:12:14 It’s like you’re pushing water through this pipe
0:12:16 and the size of the pipe limits
0:12:19 of how fast you can push water through it.
0:12:22 And analogously, in communication,
0:12:24 there’s this notion of a size of a pipe,
0:12:25 which is called a capacity.
0:12:30 And you figure a way of computing this capacity
0:12:32 for different communication medium.
0:12:35 Any communication medium,
0:12:37 you can actually compute a capacity
0:12:38 for that communication medium,
0:12:42 and that limits how fast you can communicate information
0:12:43 over that medium.
0:12:46 Whether that medium is wireless, over the air,
0:12:49 or over the wire line, like I’m talking to you,
0:12:52 I communicate over the air, I talk to my wifi,
0:12:54 the wifi goes through some copper cables,
0:12:57 some optical fiber, all these are physical medium,
0:13:00 but you can compute a capacity
0:13:02 for each of these different mediums.
0:13:08 – And I know that part of the paper looks at,
0:13:17 say redundancy in various modes of communication.
0:13:21 And on a related note, patterns, right?
0:13:23 There’s this whole section of the paper
0:13:25 where he looks at the frequency
0:13:28 with which letters occur in English,
0:13:31 and kind of builds an idea around that.
0:13:34 Tell me about those pieces of the paper.
0:13:37 – Yeah, so let’s talk about the word redundancy.
0:13:38 – Yeah, was that the wrong word?
0:13:39 – Was that the right word?
0:13:40 – No, no, no, no, no, no.
0:13:41 That’s not only not the wrong word,
0:13:44 but it’s actually the most important word, I would say.
0:13:45 – Okay. – Almost.
0:13:49 Because you go back to the question,
0:13:50 to the thing I was talking about,
0:13:52 which is how fast you can communicate, right?
0:13:53 – Yeah.
0:13:55 – So what he discovered was actually,
0:13:57 there’s no limit on how fast you can communicate.
0:13:59 You can always communicate very fast,
0:14:02 but what the guy can hear is gibberish,
0:14:05 and he cannot really distinguish what you’re trying to say.
0:14:07 It’s like so much noise in the system,
0:14:08 that he cannot really figure out what to say.
0:14:10 – Even if you’re face-to-face, right?
0:14:11 Even if you’re face-to-face,
0:14:13 you’re not going over the phone or whatever.
0:14:16 If you talk too fast, the listener won’t understand
0:14:16 because you’re going too fast.
0:14:18 – Yeah, and anybody who goes
0:14:21 to a crazy professor’s lecture would know about this,
0:14:23 where the professor just keeps on talking
0:14:26 at a million miles per hour,
0:14:28 and the student just sits there,
0:14:29 and nobody understood a thing,
0:14:31 and the professor calls the day when it’s finished.
0:14:36 So basically, what he’s saying is that,
0:14:38 hey, you know what, to make sure
0:14:41 that the information goes through reliably,
0:14:43 reliably, that’s the first word,
0:14:48 you need to introduce redundancy in your message.
0:14:53 And what he figured out is, in some sense,
0:14:56 the optimal way of adding redundancy,
0:14:59 because you can always be stupid in adding redundancy.
0:15:03 For example, I can keep on repeating the same word
0:15:06 100 times to you, and then you’ll probably get it,
0:15:07 and then I move on to the next word.
0:15:09 I cannot move on the next word,
0:15:14 but that would take me 100 times slow, right?
0:15:18 And so that’s not a very smart way of adding redundancy.
0:15:20 So what he figured out is an optimal way
0:15:24 of adding redundancy, so that you can communicate reliably,
0:15:29 and yet, at the maximum, what he calls, capacity limit.
0:15:31 And that was a totally amazing,
0:15:34 actually formulation of the problem,
0:15:37 and highly non-obvious.
0:15:41 And I think that is sort of the amazing contribution
0:15:43 of this guy, Shannon, you know.
0:15:45 – It’s optimization.
0:15:49 He optimizes communication across any channel,
0:15:53 where you’re balancing efficiency,
0:15:55 or speed, and reliability.
0:15:56 That is the trade-off.
0:15:59 And he figures out how to optimize for that trade-off.
0:16:01 – Yes, yes.
0:16:05 He figured out how to optimize that trade-off,
0:16:11 but that trade-off turns out to be very interesting.
0:16:14 It’s a very interesting trade-off.
0:16:16 So typically, when we think about trade-off,
0:16:19 we think about like a smooth curve, right?
0:16:21 As when you’re doing something,
0:16:23 then you can get better performance.
0:16:26 But what he showed was that there’s kind of
0:16:28 like a cliff effect.
0:16:30 – Okay.
0:16:33 – And the cliff effect is that if you communicate
0:16:35 below this number called capacity,
0:16:39 then you can always engineer a system
0:16:42 to make the signal, the communication,
0:16:44 as reliable as you want.
0:16:46 So reliable, that’s completely clean.
0:16:47 – Wow.
0:16:51 – Whereas if you communicate above this number of capacity,
0:16:54 then there’s nothing you can do to make it signal clean.
0:16:55 It’s just completely gibberish.
0:17:00 So it’s a very sharp trade-off that he identified.
0:17:02 It’s not a smooth trade-off.
0:17:04 – And if you’re running the phone company,
0:17:05 that’s exactly what you want to know, right?
0:17:08 So then you can tune it all the way to capacity,
0:17:11 and then not try and tune it any more after that
0:17:13 because it’s not going to get any better.
0:17:14 – Correct.
0:17:15 And that’s the goal of 60 years of engineering
0:17:18 to achieve his goal, his vision.
0:17:22 His vision in 1948, it took people around 60 years
0:17:24 to get to that, implement his vision.
0:17:28 – Well, so you are part of that story, right?
0:17:31 Let’s let you walk onto the story now.
0:17:35 So you tell me about your work and how Shannon’s work,
0:17:38 you know, how you built on Shannon’s work.
0:17:41 Tell me about how you built on Shannon’s work.
0:17:47 – Yeah, so I did my PhD in the ’90s, in the ’90s.
0:17:51 My advisor was a Shannon’s student,
0:17:54 and so I learned information theory, okay?
0:17:57 Now, at that time, information theory
0:17:59 was almost a dead subject, okay?
0:18:03 When I was a PhD student, the first thing my advisor
0:18:05 told me, maybe you’re following Shannon,
0:18:07 is, hey, don’t work in information theory.
0:18:09 – Wow.
0:18:10 – You’ll never find a job.
0:18:12 You’ll never find a job with this stuff, okay?
0:18:14 – That’s a tough moment.
0:18:15 That must be a tough moment for you.
0:18:16 – It’s pretty tough, yeah.
0:18:20 Because at that time, there’s not much progress
0:18:24 made in the theory, and there’s no killer applications either.
0:18:26 There’s no very killer applications
0:18:30 that need all this sophisticated information theory, okay?
0:18:31 So it’s like a dead field.
0:18:34 – Was there a while when people used it to, like,
0:18:36 whatever, make landline phones work better,
0:18:38 like in the ’50s or something, where people were like,
0:18:40 oh, great, now we’ve got this theory
0:18:42 and we can make the phone work better?
0:18:48 – Yeah, so the thing is that the solutions
0:18:51 that people come up with to achieve these capacity limits
0:18:53 is very complicated, okay?
0:18:56 And the electronics, the technology’s just not enough
0:18:58 to build these complicated circuits.
0:19:01 So information theory have had not a very significant
0:19:05 impact in the ’50s, ’60s, or even ’70s.
0:19:06 – So it’s like one of those cases
0:19:09 where the theory is just too far ahead
0:19:12 of the technology to be useful.
0:19:15 – Yeah, and so people kind of start losing interest
0:19:18 in the theory, they say, oh, this is a bunch of math,
0:19:19 it’s not impacting the real world,
0:19:22 and so students are drifting away from the field,
0:19:25 but there are still always a few students, okay,
0:19:27 who are just so enumerated by the theory
0:19:30 that they keep on pursuing it.
0:19:33 And my advisor’s one of the leading professors in this area,
0:19:36 and he would have, like, one student
0:19:40 every decade, every decade, to do research in information.
0:19:42 – You were that student?
0:19:43 You were that student? – And I was not that student.
0:19:44 – Oh, okay. – And I was not
0:19:46 that student, okay?
0:19:48 At that time, that slot was already taken
0:19:50 by an earlier student. – Okay.
0:19:53 – Who was way smarter than me, who was way smarter than me,
0:19:56 and that said, he was a student of the decade
0:19:58 in information theory, okay?
0:20:01 Now, so I was assigned to work on some other problems,
0:20:03 okay, completely unrelated, okay?
0:20:05 But anyway, the point, though, is that when I graduated,
0:20:08 something happened, okay?
0:20:11 And that was the beginning of the wireless revolution.
0:20:12 – Uh-huh.
0:20:16 – That was the time when only a million people
0:20:19 have cell phones, and those cell phones,
0:20:21 I don’t even remember, it’s like, gigantic brick.
0:20:24 – Yeah, like, there’s that famous scene
0:20:25 from the movie “Wall Street,” right?
0:20:27 That’s the one that everybody talks about,
0:20:28 where it’s like, bigger than a brick.
0:20:31 People say brick, but it’s actually bigger than a brick.
0:20:33 It’s like a big, hardback book or something.
0:20:36 – Yeah, and actually, those days,
0:20:37 because there’s so few of these phones,
0:20:39 it’s like a prestige.
0:20:41 It’s like a prestige to have this brick.
0:20:42 – Yeah. – Okay.
0:20:44 – Yeah, you couldn’t get that brick.
0:20:47 You had to be rich to get that brick, yeah.
0:20:50 – Yeah, and so the wireless revolution was happening
0:20:52 because people realized that, hey, you know what?
0:20:55 Be able to communicate anytime, anywhere is really valuable.
0:20:58 And so people are now getting interested.
0:21:01 And at that time, what people realized is that, whoa,
0:21:05 this wireless physical media
0:21:07 is really tough to communicate over
0:21:09 because the bandwidth is so limited
0:21:11 and the noise is so much, right?
0:21:14 FCC was limiting the bandwidth allocation
0:21:16 to these applications a lot.
0:21:19 – Uh-huh, the Federal Communications Commission,
0:21:22 the government wasn’t letting wireless companies
0:21:23 use much bandwidth for themselves.
0:21:24 – Yeah, because all the bandwidth,
0:21:27 most of them are allocated for military purposes.
0:21:29 And there’s only very little bandwidth allocated
0:21:30 at that time for civilians.
0:21:32 And so those bandwidths were auctioned out
0:21:34 to companies with a very high price.
0:21:35 – Yeah.
0:21:39 – And so it became very important to be very efficient
0:21:43 in using this very expensive property, okay?
0:21:45 And then people realized, hey,
0:21:47 if we want to be really efficient,
0:21:51 then we need a theory which is about efficiency.
0:21:53 So people start thinking, okay, all right,
0:21:55 so information theory was dead,
0:21:56 but now it’s gonna come back to life
0:21:59 because we have this really important problem,
0:22:02 a really expensive spectrum that was allocated by FCC,
0:22:04 and we want to squeeze as much of it as possible.
0:22:05 – As much communication,
0:22:07 we need a sort of mathematical theory
0:22:09 of communication, if you will.
0:22:12 – And that was the renaissance of information theory
0:22:16 spurred by this amazing technology of wireless,
0:22:19 which took us from one million phones
0:22:22 to 10 billion phones today.
0:22:26 – Everybody has 1.1 phones.
0:22:30 And information theory play a big role in that revolution.
0:22:40 – In a minute, how David used Claude Shannon’s 1948 paper
0:22:42 to come up with an idea that we all use
0:22:44 every time we make a phone call.
0:22:52 – Hey, it’s Jacob.
0:22:54 I’m here with Rachel Botsman.
0:22:58 Rachel lectures on trust at Oxford University,
0:23:01 and she is the author of a new Pushkin audiobook
0:23:04 called How to Trust and Be Trusted.
0:23:05 Hi, Rachel.
0:23:06 – Hi, Jacob.
0:23:09 – Rachel Botsman, tell me three things
0:23:11 I need to know about trust.
0:23:16 – Number one, do not mistake confidence for competence.
0:23:17 Big trust mistakes.
0:23:19 So when people are making trust decisions,
0:23:23 they often look for confidence versus competence.
0:23:27 Number two, transparency doesn’t equal more trust.
0:23:29 Big myth and misconception.
0:23:32 And a real problem, actually in the tech world.
0:23:35 The reason why is because trust
0:23:38 is a confident relationship with the unknown.
0:23:41 So what are you doing if you make things more transparent?
0:23:44 You’re reducing the need for trust.
0:23:49 And number three, become a stellar expectation setter.
0:23:54 Inconsistency with expectations really damages trust.
0:23:55 – I love it.
0:23:57 Say the name of the book again
0:23:59 and why everybody should listen to it.
0:24:02 – So it’s called How to Trust and Be Trusted.
0:24:04 Intentionally, it’s a two-way title
0:24:08 because we have to give trust and we have to earn trust.
0:24:10 And the reason why I wrote it is because
0:24:13 we often hear about how trust is in a state of crisis
0:24:15 or how it’s in a state of decline.
0:24:18 But there’s lots of things that you can do
0:24:20 to improve trust in your own lives,
0:24:22 to improve trust in your teams,
0:24:25 trusting yourself to take more risks
0:24:28 or even making smarter trust decisions.
0:24:29 – Rachel Botsman, the new audio book
0:24:32 is called How to Trust and Be Trusted.
0:24:33 Great to talk with you.
0:24:34 – It’s so good to talk with you
0:24:36 and I really hope listeners listen to it
0:24:39 because it can change people’s lives.
0:24:44 – Let’s talk for a moment about your role, right?
0:24:47 Like you actually played an important role there.
0:24:50 – Yeah, so I was at Bell Labs.
0:24:54 – Uh-huh, just like Claude, just like Claude Giada.
0:24:56 – Yeah, so I spent one year at Bell Labs
0:25:00 as a so-called postdoc right after my PhD.
0:25:03 Before I moved to Berkeley to become a professor there,
0:25:04 I spent one year there.
0:25:06 And that’s what people were talking about
0:25:07 at that time at Bell Labs.
0:25:10 Hey, this new thing, wireless information theories
0:25:12 come back to life.
0:25:14 We can try to use information theory
0:25:16 and adapt it and extend it
0:25:18 to this wireless communication problem.
0:25:20 And so that’s when I said,
0:25:23 “Whoa, this information theory I learned from Bob Gallagher.
0:25:25 Finally, there’s a place to use it.
0:25:28 Finally, I can actually make a living.
0:25:31 Make a living out of it.”
0:25:33 Unlike what my advisor told me, it’s not dead.
0:25:34 It’s coming back to life.
0:25:35 – Yeah.
0:25:38 – And so that’s sort of my start in the field.
0:25:43 And yeah, so I invented a bunch of stuff
0:25:45 and actually applied this,
0:25:47 connected information theory to the real world.
0:25:50 And every time you use a phone,
0:25:52 you’re using my algorithm,
0:25:55 which is based on the theory of information.
0:25:58 – Huh, and so you’re,
0:26:00 that’s a cool thing to be able to say, first of all,
0:26:02 that’s a very good flex.
0:26:04 Your algorithm,
0:26:08 it’s the proportional fair scheduling algorithm, right?
0:26:09 – Yes, yes.
0:26:09 – What is that?
0:26:10 What’s it do?
0:26:13 – All right, so I should tell you a little bit of story.
0:26:15 I think a story is,
0:26:17 and then I’ll tell you what it does, okay?
0:26:19 So I went to,
0:26:22 so that was the end of 1999, around 1999.
0:26:26 So I was doing all this information theory stuff
0:26:29 at Berkeley, writing many papers.
0:26:32 But then I always have a thought back on my,
0:26:34 which is, “Hey, is this stuff going to be useful?”
0:26:36 And so I went to a,
0:26:37 I decided to go to a company, a wireless company,
0:26:38 who actually built these things
0:26:40 and see whether this theory can be used.
0:26:43 And the company I went to is called Qualcomm.
0:26:44 – Okay.
0:26:45 – I’ve heard of Qualcomm.
0:26:46 – No, you’ve heard of Qualcomm,
0:26:48 but at that time it was a small company.
0:26:49 It was not very big, okay?
0:26:51 And at that time,
0:26:54 they have this problem they’re working on, okay?
0:26:56 Which is the following, all right?
0:26:57 So in wireless communication,
0:27:00 there’s a concept called base station, okay?
0:27:05 And the base station serves many cell phones
0:27:06 in the vicinity of the base station.
0:27:08 It’s called cell, okay?
0:27:09 – Is it like a tower?
0:27:09 Is it what we would call it?
0:27:11 – Yes, like a tower, that’s right.
0:27:12 It’s always on the tower.
0:27:14 There’s some electronics there.
0:27:17 And that’s how the base station
0:27:20 is supposed to beam information to many phones.
0:27:20 To many phones.
0:27:22 – You still see them, you see them, whatever,
0:27:23 on top of a big building
0:27:25 or when you’re driving down the freeway, right?
0:27:26 That’s what you’re talking about, yeah.
0:27:27 – That’s right.
0:27:29 And sometimes on fake trees.
0:27:30 – Yeah, I love the fake trees.
0:27:32 In New Jersey, they love the fake trees, yeah.
0:27:34 – New Jersey, that’s right.
0:27:37 New Jersey fake trees, yes.
0:27:40 So at that time, they were looking at this problem,
0:27:44 which is, hey, okay, my bandwidth is limited,
0:27:47 but I have many users to serve, okay?
0:27:49 How do I schedule my limited resource
0:27:51 among all these users, right?
0:27:54 Because I only have one total bandwidth.
0:27:55 And so at that time, people were saying,
0:27:57 okay, maybe something simple.
0:28:00 I give one nth of the time to the end user, right?
0:28:02 So the boost of five users,
0:28:04 I serve this user for a little bit
0:28:05 and I serve the second user for a little bit
0:28:07 and third user, fourth user, fifth user.
0:28:08 – I mean, the idea is you’re switching really fast.
0:28:09 You’re just like switching kind of.
0:28:12 – Yeah, switching really fast, yeah, exactly.
0:28:14 And then when I went there, I said, okay,
0:28:16 good, this is a problem, it’s a good problem.
0:28:19 And I said, hey, instead of fixating
0:28:21 on this particular scheduling policy,
0:28:24 why don’t we do a Shannon thing?
0:28:29 – A Claude Shannon thing, you thought of, yeah, okay.
0:28:31 – The Claude Shannon thing is what?
0:28:34 Is to look at the problem from first principle,
0:28:39 not pre-assume a particular solution
0:28:41 or a particular class solution even,
0:28:44 and ask ourselves, what is the capacity
0:28:45 of this whole system?
0:28:51 And how do I engineer the system to achieve that capacity?
0:28:52 Okay? – Uh-huh.
0:28:55 – And it turns out that if you look at the problem this way,
0:28:58 then it turns out that the optimal way of scheduling
0:29:01 is not the one that they were trying to design.
0:29:05 And the reason is because in wireless communication,
0:29:07 there’s a very interesting characteristic,
0:29:10 which is called fading.
0:29:12 – Okay. – Okay.
0:29:15 When I talk to you over the air,
0:29:17 the channel actually goes up and down
0:29:20 strong and weak, strong and weak very rapidly.
0:29:24 What I mean is when I send an electromagnetic signal
0:29:27 from the base station to the phone,
0:29:32 that signal get amplified and attenuated very rapidly.
0:29:35 – It goes up and down. – It goes up and down.
0:29:36 – Okay. – Okay?
0:29:38 – Can we say it gets stronger and weaker?
0:29:39 Can we say it gets stronger and weaker?
0:29:40 – Stronger and weaker. – Okay.
0:29:42 – Yes.
0:29:44 And so the optimal way that information
0:29:47 so it has to do is actually not divide the time
0:29:52 into slots blindly, but really try to schedule a user
0:29:57 when the channel is strong.
0:29:59 – Uh-huh, uh-huh.
0:30:02 – And then from that on, I designed a scheduling algorithm
0:30:05 which is more practical by sort of leverage
0:30:08 of this basic idea from information theory.
0:30:11 – And so the base station is basically monitoring
0:30:13 the strength of the incoming signals
0:30:16 from all the different phones.
0:30:17 – Correct, correct. – And saying,
0:30:18 “Oh, that one’s strong, I’m gonna grab that one.
0:30:20 “Oh, that one’s strong, I’m gonna grab that one.”
0:30:21 That’s what’s happening.
0:30:22 – Correct, correct.
0:30:25 – And how does that, I mean, I get in a kind of
0:30:27 big first principles way sort of analogously,
0:30:29 it follows from Shannon, but is there anything
0:30:34 sort of specific in Shannon that leads you
0:30:36 to this algorithm?
0:30:42 – So remember, Shannon is a very general theory.
0:30:44 – Yeah. – Okay?
0:30:47 It basically says that given any communication medium
0:30:52 or any communication setting, you can try to calculate
0:30:55 this notion of a capacity.
0:30:57 So the very general theory.
0:31:01 What I did was to apply it to a very specific context,
0:31:05 which is this base station serving multiple user setting.
0:31:06 – Yeah.
0:31:10 – And then apply his framework to analyze
0:31:12 the capacity of that system.
0:31:13 – Uh-huh.
0:31:16 – And in the process of analyzing the capacity,
0:31:19 you can also figure out what is the optimal way
0:31:22 of achieving that capacity.
0:31:24 Remember you mentioned capacity is really
0:31:28 an optimization problem, and Shannon was able
0:31:30 to solve this optimization problem in general,
0:31:32 but now I specialize it in some sense
0:31:35 to this pretty specific setting,
0:31:38 except that the setting is used by everybody.
0:31:39 – Yes, yes.
0:31:42 – At that time, it was like, research is about timing.
0:31:45 And I was there at the right place at the right time
0:31:50 because Qualcomm turns out to completely dominate
0:31:52 the entire third generation technology.
0:31:53 – Yeah, 3G.
0:31:56 – So when I was able to convince them that,
0:31:59 hey, your way of doing things is no good,
0:32:02 this way suggested by Shannon is actually far better.
0:32:05 Please use this way.
0:32:06 It took me a few months,
0:32:08 but I was able to persuade them to implement it.
0:32:12 And then it got into the standard through the domination.
0:32:14 And then every standard after that
0:32:16 uses the same, based on the same algorithm.
0:32:19 So it was good because as I said,
0:32:21 I’m at the right place at the right time.
0:32:24 You know, when you try to contribute to engineering,
0:32:26 it’s too late if the system is built already,
0:32:28 because people don’t want to change the whole system
0:32:30 to accommodate your new idea.
0:32:34 But it was very early in the design phase.
0:32:37 – So, okay, so you made this breakthrough
0:32:41 in wireless communications using Shannon’s work.
0:32:46 Were there similar breakthroughs in other domains?
0:32:47 – Any communication medium, right?
0:32:50 It could be optical fiber.
0:32:54 It could be DSL modem, DSL modem.
0:32:56 Underwater communication.
0:32:58 Almost all these communication systems
0:33:01 are now designed based on his principle.
0:33:07 So his impact of this theory is kind of global.
0:33:09 It’s the entire communication landscape.
0:33:15 – There’s a story I read about Shannon,
0:33:20 when he is developing information theory.
0:33:23 He takes a book off the shelf
0:33:25 and he reads a sentence to actually his wife.
0:33:29 And it’s something like the lamp was sitting on the,
0:33:30 and she says, table.
0:33:32 And he says, no, I’ll give you a clue.
0:33:35 The first letter is D and she says desk.
0:33:38 And when I heard that story,
0:33:41 what I thought of was large language models.
0:33:43 Like that sounds exactly like a large language model.
0:33:47 And so I’m just fishing, I’m just curious,
0:33:52 like does his work matter for machine learning,
0:33:55 large language models, et cetera, or no?
0:34:00 – Yeah, so that’s a very interesting point.
0:34:03 Now I’m not an expert by any means in AI
0:34:04 or large language models.
0:34:08 I’m not a professional researcher in that area.
0:34:11 But I think you can actually see some commonality, right?
0:34:15 Is that, these models in some sense,
0:34:17 they don’t care about meaning either.
0:34:20 – Yeah, very good, very good, yeah.
0:34:21 – Right?
0:34:24 Actually, I just came to my,
0:34:26 this discussion is very interesting.
0:34:28 Because it’s really just patterns.
0:34:31 It’s just which patterns are more likely
0:34:32 than other patterns, right?
0:34:34 The example you gave about desk and lamp
0:34:36 is basically about patterns.
0:34:38 And information theory is really analyzing
0:34:42 sort of the number of possible patterns in some sense.
0:34:47 So there is definitely a philosophical connection,
0:34:50 I believe, starting from Shannon
0:34:52 to these large language models.
0:34:53 – So let me ask you about one other,
0:34:56 and this is one that you are professionally involved in.
0:35:03 Cryptocurrency and blockchain, you have studied it
0:35:06 and you started a company, right?
0:35:08 Is there a connection between Shannon’s work
0:35:09 and cryptocurrency?
0:35:12 – Yeah, so what attracts me
0:35:16 to work in this area of blockchain
0:35:21 is that blockchain actually has one very common
0:35:23 philosophical connection to information theory,
0:35:25 which is the following.
0:35:29 In blockchain, the problem is not communication per se,
0:35:31 it’s called consensus, okay?
0:35:33 It’s a different problem,
0:35:36 but it’s essentially allow a bunch of users
0:35:40 at different places to come to an agreement on something, okay?
0:35:41 – Yes.
0:35:45 – Now, the goal of designing blockchain
0:35:49 is really to be so-called fault-tolerant.
0:35:49 – Fault-tolerant.
0:35:51 – Which means fault-tolerant,
0:35:54 which means that even if, say,
0:35:57 one-third of the users are bad guys
0:35:59 and send you some gibberish message,
0:36:00 – Yeah.
0:36:04 – You can still, the rest two-third people
0:36:06 can still come to an agreement, okay?
0:36:08 All right?
0:36:09 So you look at this problem,
0:36:12 it’s actually not that different from communication
0:36:14 in information theory because it’s kind of combating.
0:36:16 – The bad guys are the noise,
0:36:18 or the good guys are the signal, yeah.
0:36:19 – And the good guys are the signal
0:36:22 and they try to introduce redundancy, okay?
0:36:25 To help them to fight against these bad guys.
0:36:28 – Yes, and there’s an optimization problem
0:36:30 where the more redundancy you have,
0:36:33 the sort of slower the system is, the more ponderous.
0:36:36 And so you try, the optimization problem is
0:36:39 you try to figure out what is the optimal number
0:36:41 of bad guys that you can tolerate
0:36:43 and your system still works.
0:36:46 That is the analogous to the capacity problem.
0:36:51 So I find the philosophical connection very appealing.
0:36:52 And so that’s sort of one reason
0:36:55 why I got attracted to working to this area.
0:36:59 – Why do you think more people don’t know about Shannon?
0:37:04 Like all of the sort of intellectuals in technology say
0:37:10 he’s like one of the great thinkers of the 20th century.
0:37:14 But most people have never heard of him.
0:37:16 Why do you think that is?
0:37:19 – So Shannon was actually a very shy person.
0:37:24 Very shy person, he hates publicity.
0:37:29 He hated when people interview him.
0:37:30 You remember, right?
0:37:32 He’s basically a very modest person.
0:37:34 Remember the first paragraph I talked to you about.
0:37:37 He tells you what he’s, that he’s not accomplishing.
0:37:40 And so he’s a very modest, very shy person,
0:37:42 not into publicity.
0:37:47 And I think that sort of impact not only himself,
0:37:49 but also everybody who works in that field.
0:37:54 Adopt this as kind of like a metric, right?
0:37:56 That, hey, we should all be modest because why?
0:37:59 Look at this guy who accomplished so much
0:38:00 and he’s still so modest.
0:38:01 Who are we?
0:38:03 Who are we, right?
0:38:05 So as a result,
0:38:08 the field doesn’t really sell themselves very well.
0:38:11 The marketing engine, the marketing DNA is not there.
0:38:15 And so people don’t know about him.
0:38:16 – So I want to talk for a minute
0:38:18 about the rest of Shannon’s life.
0:38:22 He writes this huge paper when he’s in his early 30s,
0:38:26 eventually goes on to be a professor at MIT.
0:38:30 And he seems to spend a lot of his career
0:38:33 juggling, riding a unicycle,
0:38:35 building mechanical toys, building games.
0:38:40 And he never does sort of great influential work again.
0:38:44 And I’m curious, what do you make of that?
0:38:46 How do you sort of fit his whole career together?
0:38:50 – So there’s a single, there’s a theme
0:38:53 that unifies all this in my mind, which is playfulness.
0:38:59 Because in his mind, research is really about puzzles.
0:39:03 He doesn’t understand something.
0:39:05 It’s like a puzzle to him.
0:39:08 And he’s trying to figure out the pieces of the puzzle.
0:39:11 Information theory was like that.
0:39:13 The puzzles, he sees all these real-world systems.
0:39:15 They seem to all share some commodity,
0:39:16 but nobody understood it.
0:39:17 So there’s a puzzle
0:39:19 and he’s always thinking about the puzzle.
0:39:22 And finally his paper basically solved that puzzle.
0:39:25 So everything to him is playfulness.
0:39:27 I think it’s plain, it’s a game.
0:39:29 Puzzle needs to solve the puzzle.
0:39:30 And that’s his mind.
0:39:31 That’s how his mind works.
0:39:33 So although it seems very different,
0:39:37 things that he did pre and post information theory,
0:39:41 but it’s actually in my mind, quite swanky monotonous.
0:39:44 (upbeat music)
0:39:49 – We’ll be back in a minute with the lightning round.
0:39:51 (upbeat music)
0:39:55 – Hey, it’s Jacob.
0:39:57 I’m here with Rachel Botsman.
0:40:01 Rachel lectures on trust at Oxford University.
0:40:04 And she is the author of a new Pushkin audiobook
0:40:08 called How to Trust and Be Trusted.
0:40:09 Hi, Rachel.
0:40:10 – Hi, Jacob.
0:40:13 – Rachel Botsman, tell me three things
0:40:15 I need to know about trust.
0:40:19 Number one, do not mistake confidence for competence.
0:40:21 Big trust mistakes.
0:40:23 So when people are making trust decisions,
0:40:27 they often look for confidence versus competence.
0:40:31 Number two, transparency doesn’t equal more trust.
0:40:33 Big myth and misconception.
0:40:35 And a real problem, actually in the tech world,
0:40:38 the reason why is because trust
0:40:41 is a confident relationship with the unknown.
0:40:43 So what are you doing?
0:40:45 If you make things more transparent,
0:40:47 you’re reducing the need for trust.
0:40:52 And number three, become a stellar expectation setter.
0:40:58 Inconsistency with expectations really damages trust.
0:40:59 – I love it.
0:41:00 Say the name of the book again
0:41:02 and why everybody should listen to it.
0:41:05 – So it’s called How to Trust and Be Trusted.
0:41:07 Intentionally, it’s a two-way title
0:41:11 because we have to give trust and we have to earn trust.
0:41:13 And the reason why I wrote it
0:41:15 is because we often hear about
0:41:17 how trust is in a state of crisis
0:41:19 or how it’s in a state of decline.
0:41:21 But there’s lots of things that you can do
0:41:24 to improve trust in your own lives,
0:41:26 to improve trust in your teams,
0:41:28 trusting yourself to take more risks,
0:41:31 or even making smarter trust decisions.
0:41:33 – Rachel Botsman, the new audio book
0:41:35 is called How to Trust and Be Trusted.
0:41:36 Great to talk with you.
0:41:37 – It’s so good to talk with you.
0:41:40 And I really hope listeners listen to it
0:41:42 because it can change people’s lives.
0:41:49 – So I read that you recently asked people
0:41:52 at your company to give five-minute talks.
0:41:55 I’m curious why you did that.
0:41:56 That’s interesting to me.
0:41:56 Why’d you do that?
0:42:02 – So the shorter the talk, the harder it is to give.
0:42:03 – Yeah.
0:42:08 – So you can explain an idea in five minutes.
0:42:10 Then I think your idea is actually not very good.
0:42:13 – Aha, that’s good.
0:42:16 – Most good ideas you can get the point to
0:42:18 across in five minutes.
0:42:20 Remember, I’m an information theorist by training.
0:42:23 So communication to the limit
0:42:26 is what I’m passionate about.
0:42:28 – If you had to give a five-minute talk,
0:42:30 what would it be about?
0:42:33 – About Shannon, I guess.
0:42:38 He’s my hero, he’s my hero.
0:42:43 – So one, you talked about the importance of timing
0:42:48 in research of not only finding the right problem,
0:42:50 but finding the right problem at the right time, right?
0:42:54 Both in terms of Shannon’s work and in terms of your work.
0:42:59 You’re also a professor and a manager,
0:43:01 like how do you help other people
0:43:03 find the right problem at the right time?
0:43:09 – Yeah, finding the right problem at the right time
0:43:11 is probably the most difficult
0:43:15 because, you know, timing is everything.
0:43:18 However, this is hard to teach.
0:43:21 What you try to do is to be ready.
0:43:27 So one very famous information theorist told me this.
0:43:30 He said, you know, everybody will get lucky
0:43:32 at some point in time in their career.
0:43:37 However, most people, when they get lucky,
0:43:38 they’re not ready.
0:43:41 So they don’t realize that they get lucky.
0:43:43 And so they missed the opportunity.
0:43:44 They went a different direction.
0:43:46 Luck tells you you should go this way,
0:43:48 but you went the other way, lost it.
0:43:50 – That makes me so scared.
0:43:54 – And so what I teach my students is always be ready.
0:43:55 It’s like your muscles.
0:43:57 You have to be always train your muscles
0:43:59 so that when you are lucky,
0:44:00 you can capitalize on the luck.
0:44:06 – So you talked about Shannon’s playful nature.
0:44:09 Like he was a juggler, he rode a unicycle.
0:44:10 You do anything like that?
0:44:13 Do you have any weird hobbies?
0:44:17 – No, no.
0:44:20 The only weird hobby is I love to talk to people like you.
0:44:21 – Fair.
0:44:24 You love going on podcasts, that’s the juggling
0:44:26 of the 21st century.
0:44:30 Who’s your second favorite underrated thinker?
0:44:33 – My advisor.
0:44:35 – Ah, Bob Gallagher.
0:44:37 – My advisor, Gallagher, yeah, Bob Gallagher.
0:44:40 He taught me how to think about research
0:44:44 because he learned from Shannon and I learned from him.
0:44:47 – And if you boil down what–
0:44:49 – Your advisor learned from Shannon
0:44:51 and what you learned from your advisor,
0:44:52 what would it be?
0:44:53 What did you learn?
0:44:58 – Yeah, I learned about taking a very complicated problem
0:45:01 and strip it down to the essential.
0:45:02 – Uh-huh.
0:45:05 – And then formulate a problem around that and solve it.
0:45:07 That’s an art.
0:45:11 It’s not something you can convert it
0:45:14 into a mathematical formula and teach students.
0:45:18 It’s just based on intuition, experience.
0:45:21 And that’s what Shannon taught my advisor.
0:45:24 And that’s what my advisor taught me.
0:45:26 And that’s what I try to teach my students.
0:45:29 Really, teaching is not really about giving the formula.
0:45:31 It’s really just learning by examples.
0:45:34 I observe what he does.
0:45:36 And then my students observe what I do
0:45:37 as I interact with them.
0:45:40 And hopefully this art will carry on
0:45:43 from generation to generation.
0:45:46 – Finding the essence of the problem.
0:45:47 – Yeah.
0:45:50 (upbeat music)
0:45:57 – David Shea is a professor at Stanford.
0:46:01 Today’s show was produced by Gabriel Hunter Chang.
0:46:03 It was edited by Lydia Jean Kott
0:46:05 and engineered by Sarah Brugier.
0:46:09 You can email us at problem@pushkin.fm.
0:46:11 I’m Jacob Goldstein and we’ll be back next week
0:46:13 with another episode of “What’s Your Problem?”
0:46:20 – Hey, it’s Jacob.
0:46:22 I’m here with Rachel Botsman.
0:46:25 Rachel lectures on trust at Oxford University
0:46:29 and she is the author of a new Pushkin audiobook
0:46:32 called “How to Trust and Be Trusted.”
0:46:33 Hi, Rachel.
0:46:34 – Hi, Jacob.
0:46:35 – Rachel Botsman.
0:46:39 Tell me three things I need to know about trust.
0:46:44 – Number one, do not mistake confidence for competence.
0:46:45 Big trust mistakes.
0:46:47 So when people are making trust decisions,
0:46:51 they often look for confidence versus competence.
0:46:55 Number two, transparency doesn’t equal more trust.
0:46:57 Big myth and misconception
0:47:00 and a real problem actually in the tech world.
0:47:02 The reason why is because trust
0:47:06 is a confident relationship with the unknown.
0:47:09 So what are you doing if you make things more transparent?
0:47:12 You’re reducing the need for trust.
0:47:17 And number three, become a stellar expectation setter.
0:47:22 Inconsistency with expectations really damages trust.
0:47:23 – I love it.
0:47:25 Say the name of the book again
0:47:27 and why everybody should listen to it.
0:47:30 – So it’s called “How to Trust and Be Trusted.”
0:47:32 Intentionally, it’s a two-way title
0:47:36 because we have to give trust and we have to earn trust.
0:47:39 And the reason why I wrote it is because we often hear
0:47:41 about how trust is in a state of crisis
0:47:43 or how it’s in a state of decline.
0:47:45 But there’s lots of things that you can do
0:47:48 to improve trust in your own lives,
0:47:50 to improve trust in your teams,
0:47:52 trusting yourself to take more risks
0:47:55 or even making smarter trust decisions.
0:47:57 – Rachel Botsman, the new audio book
0:48:00 is called “How to Trust and Be Trusted.”
0:48:01 Great to talk with you.
0:48:02 – It’s so good to talk with you
0:48:04 and I really hope listeners listen to it
0:48:06 because it can change people’s lives.
0:00:07 Pushkin.
0:00:11 Hey, it’s Jacob, and I want to tell you
0:00:14 about a podcast called What Could Go Right?
0:00:18 The show is hosted by Zachary Carabell and Emma Varvalukas.
0:00:20 They are the founder and executive director
0:00:22 of The Progress Network.
0:00:25 And on the show, they sit down with expert guests
0:00:28 to discuss the world’s most pressing issues.
0:00:32 And they do not resort to pessimism or despair.
0:00:35 Plus, every Friday, the hosts share reports
0:00:38 highlighting the latest progress from across the globe.
0:00:40 For a dose of optimistic ideas from smart people,
0:00:43 listen to What Could Go Right?
0:00:44 It’s a podcast.
0:00:47 You can get it where you get podcasts.
0:00:50 Hey, happy New Year.
0:00:52 We are very happy to be back.
0:00:55 And I have one request before we start the show.
0:00:57 I’m asking you a favor.
0:00:59 And the favor is this,
0:01:04 would you please send us an email to problem@pushkin.fm
0:01:09 and tell us what you like about the show
0:01:10 and what you don’t like about the show.
0:01:12 And specifically, what kinds of things
0:01:14 you want to hear more of
0:01:16 and perhaps what kinds of things you don’t want to hear.
0:01:21 Again, it’s problem@pushkin.fm.
0:01:22 I’m going to read all the emails.
0:01:24 So thank you in advance for sending them.
0:01:29 Claude Shannon is this huge figure
0:01:31 in the history of technology.
0:01:34 He’s one of the key people who worked at Bell Labs
0:01:36 in the middle of the 20th century
0:01:38 and really came up with the ideas
0:01:41 that made modern technology possible.
0:01:44 But I’m going to be honest with you.
0:01:48 I never really understood what Claude Shannon figured out
0:01:50 that was such a big deal.
0:01:52 But the people who know about technology,
0:01:53 who know about the history of ideas,
0:01:56 they say Shannon’s a giant.
0:01:58 Claude Shannon is like the nerd’s nerd.
0:02:02 He’s the techno intellectuals, techno intellectual.
0:02:06 And so for today’s show, I wanted to understand
0:02:08 what did Claude Shannon figure out
0:02:11 and why is it so important for the modern world?
0:02:20 I’m Jacob Goldstein and this is What’s Your Problem.
0:02:22 My guest today is David Shea.
0:02:27 David is a professor of electrical engineering at Stanford.
0:02:29 He’s studied Shannon for decades.
0:02:32 He teaches Shannon’s work to his students.
0:02:35 And David used Shannon’s work
0:02:38 to make a breakthrough in cell phone technology.
0:02:40 And that breakthrough, that breakthrough
0:02:42 that came to us via Shannon and Shea,
0:02:45 it affects every phone call we make.
0:02:48 David and I talked about Shannon’s key insights
0:02:51 and about how David’s own work built on Shannon.
0:02:54 And we also talked about the big chunk of Shannon’s life
0:02:58 that was taken up with juggling and riding unicycles
0:03:00 and building mechanical toys.
0:03:03 But to start, we talked about how in the middle
0:03:06 of the 20th century, Bell Labs wound up driving
0:03:09 so much technological innovation.
0:03:16 Yeah, so Bell Labs was the research lab of AT&T.
0:03:21 AT&T at that time was the phone company, okay?
0:03:23 Nowadays, we have many phone companies, right?
0:03:28 We have Verizon, we have T-Mobile, et cetera.
0:03:31 But those days, there was only one phone company
0:03:33 and that’s a monopoly.
0:03:36 So a monopoly needs to justify its existence.
0:03:39 – So it doesn’t get broken up by the government.
0:03:40 – It doesn’t get broken up.
0:03:42 Of course, it eventually got broken up.
0:03:45 But at that time, it was a monopoly.
0:03:49 And so one way of justifying its existence
0:03:53 is to say that, okay, it says to the American people,
0:03:57 to the government that we will always spend
0:04:01 a certain percent of our revenue on this research lab,
0:04:04 called Bell Labs.
0:04:08 And whatever Bell Labs come up with is kind of our
0:04:11 contribution not only to our bottom line,
0:04:16 but also to technology of the country.
0:04:18 – So they have this sort of public mission
0:04:21 to prevent the government from breaking them up.
0:04:25 – Yeah, and so therefore it also allows researchers
0:04:29 a very free reign to do research
0:04:31 that not necessarily tied to, like say,
0:04:35 a particular business unit, okay?
0:04:36 So they can be very creative.
0:04:38 And that’s the atmosphere of Bell Labs.
0:04:42 So Bell Labs attracted a bunch of very smart people
0:04:44 because smart people wants to work on their own problem,
0:04:47 not the problem that the manager gives them.
0:04:51 – Okay, that’s one characteristic of smart people.
0:04:54 And so, yeah, that was the heydays of Bell Labs.
0:04:57 Lots of smart people inventing amazing stuff.
0:04:59 Laser was invented there.
0:05:03 Information theory, the transistor was invented there.
0:05:08 Sort of almost all the foundation of the information age,
0:05:13 where there’s hardware, algorithm, software,
0:05:16 is in some sense all have the roots at Bell Labs.
0:05:19 So that was the contribution to mankind, actually,
0:05:21 I should say, not only to America.
0:05:24 – So Shannon gets there at this time, right?
0:05:28 He’s there when they’re inventing,
0:05:31 certainly the transistor, what’s he do?
0:05:34 Tell me about his work there when he gets there.
0:05:35 What’s he working on?
0:05:41 – Yeah, so I think Shannon always have his own agenda, right?
0:05:46 We know for a fact that he has been interested
0:05:49 in the problem of communication,
0:05:54 that idea of having a grand theory of communication,
0:05:57 even back in 1938, I think ’37, ’38,
0:06:00 because he wrote a letter at that time
0:06:03 to a very famous person named Venera Bush.
0:06:04 – Yeah.
0:06:06 – Venera Bush is very famous.
0:06:10 It was, I think, president of MIT or dean of MIT,
0:06:12 and then he became sort of a scientific advisor
0:06:14 to the president.
0:06:17 And so he wrote a letter to Venera Bush in 1938
0:06:19 and said, “Hey, you know what?
0:06:20 I’m really interested in this question
0:06:22 of how to find one theory
0:06:25 that unifies all possible communication systems.
0:06:27 There’s so many different communication systems out there,
0:06:30 but I think there’s something at the heart of every system.
0:06:31 And I’m trying to get to the heart.
0:06:34 – And nobody had thought of it in that way, right?
0:06:38 It seems like part of his, part of why Shannon
0:06:39 is such a big deal is like,
0:06:41 as I understand it, people is like, you know,
0:06:43 people understood, like, they were trying to figure out
0:06:45 how to make the phone work better.
0:06:47 And they were trying to, you know,
0:06:49 make movies be clearer or whatever.
0:06:52 But there wasn’t this idea that you could abstract it
0:06:54 until Shannon came along.
0:06:56 – And the reason is very simple, actually,
0:06:59 because if you have a physical system
0:07:01 that you want to build, right?
0:07:02 What do you see, right?
0:07:04 You say, “Hey, man, the video, for example,
0:07:06 I’m seeing you right now.
0:07:08 I’m not seeing you very clearly, I have to say.”
0:07:11 – Yes, I made a closet, I made a closet.
0:07:13 – Right, yeah.
0:07:15 – Then I would say, hey, how to try to improve the image?
0:07:18 Maybe I can try to, you know, fix this pixel
0:07:21 or do some filtering of your noise.
0:07:25 So I’m very tied to the very specific details
0:07:28 of the specific problem, because why I’m the engineer?
0:07:30 I need to improve the system, not in 10 years,
0:07:32 but tomorrow, you know, tomorrow,
0:07:33 I need to get a better system.
0:07:34 – You don’t need a theory of the system.
0:07:37 You just want a clearer picture, yeah.
0:07:39 – Yeah, I’m in the wits, right?
0:07:40 I’m in the wits.
0:07:44 And Shannon, because of his training,
0:07:47 and also because of the embassy of a place like Bell Labs,
0:07:50 could afford to sort of step back
0:07:52 and just look at the broader forest
0:07:55 as opposed to the details of specific trees.
0:07:59 – So, okay, Shannon’s big idea
0:08:02 comes out in this paper he publishes in 1948.
0:08:05 The paper’s called “A Mathematical Theory of Communication.”
0:08:07 It’s like his great work.
0:08:09 Tell me about that paper.
0:08:12 – So that paper is actually a very interesting paper.
0:08:15 In fact, when I teach information theory,
0:08:18 I teach from the paper itself,
0:08:21 because I thought it’s an amazing way,
0:08:23 not only of learning information theory,
0:08:26 but learning how to write a scientific paper properly.
0:08:28 – Uh-huh.
0:08:29 – Okay, and you know,
0:08:32 not everyone does research and information theory,
0:08:33 but everybody has to write.
0:08:34 – Uh-huh.
0:08:36 – Okay, every researcher has to write
0:08:40 to express their ideas to the peers and to the audience.
0:08:43 So in that paper, very interesting.
0:08:44 The first paragraph of the paper, okay,
0:08:46 is already very interesting.
0:08:48 Because typically when people write a paper nowadays,
0:08:51 they tell you, oh, how great my invention is.
0:08:52 It’s gonna change the world.
0:08:55 Every paper is gonna change the world.
0:08:57 But in fact, his first paragraph
0:09:00 focused on telling you what his paper is not achieving.
0:09:02 – Uh-huh.
0:09:03 – I mean, that’s a master.
0:09:05 That’s a master’s, right?
0:09:08 I mean, how many papers that you read nowadays
0:09:10 tells you in the beginning,
0:09:13 hey, you know what guys, expectation management here,
0:09:15 this paper is not about this, not about that.
0:09:16 – Don’t get your hopes up.
0:09:18 Hey, don’t get your hopes up, yeah.
0:09:19 – Exactly, that’s exactly what he did.
0:09:21 Expectation management.
0:09:22 – Okay.
0:09:25 – Nowadays, today we will call it expectation management.
0:09:28 And now those days, I guess he just calls it honesty.
0:09:31 And his whole point was,
0:09:35 often people associate information with meaning.
0:09:36 – Uh-huh.
0:09:37 – Okay?
0:09:39 And then he said in this paper, we ignore meaning.
0:09:42 We ignore meaning, okay?
0:09:46 So that was the first thing he did.
0:09:47 Which is brilliant,
0:09:50 because once you tie information with meaning,
0:09:53 then he will never be able to make any progress.
0:09:56 It’s just too difficult and too broad and too vague a problem.
0:09:59 – Everybody gets stuck on this idea of meaning
0:10:00 and what is meaning.
0:10:02 And he’s like, forget about meaning.
0:10:05 So if we’re gonna forget about meaning, what is left?
0:10:08 – Yeah, so actually the biggest,
0:10:10 I think breakthrough of that paper
0:10:15 is to really focus on the thing that matters.
0:10:19 And cut away a lot of stuff that really doesn’t,
0:10:21 not that it doesn’t matter,
0:10:22 but it doesn’t matter in terms of solving
0:10:24 the communication problem.
0:10:27 The communication, so then he said,
0:10:29 okay, what is the communication problem?
0:10:31 The communication problem is the following,
0:10:36 is that there are multiple possibilities of a word.
0:10:40 And my goal is to tell the receiver, destination,
0:10:42 which of the multiple possibilities
0:10:44 is the correct possibility.
0:10:47 – Yeah, and so in language, it’s basically,
0:10:49 it’s a finite set, language is a finite set.
0:10:51 It’s very large, but if we’re speaking,
0:10:54 and we both know that we’re speaking English,
0:10:56 then essentially you are hearing the words
0:10:57 and decoding them.
0:10:59 And you know that it is a series of words
0:11:02 and you just have to figure out which words.
0:11:04 I mean, like that, for example?
0:11:05 – Yes, like that.
0:11:08 – Okay, so that’s the frame he builds, then what?
0:11:14 – Okay, all right, then once you have this framing, right,
0:11:18 then you can ask the question, okay,
0:11:20 what is the goal of communication?
0:11:22 The goal of communication is to communicate
0:11:27 as fast as I can, right?
0:11:30 And the natural question is, why is there a limit
0:11:34 on how fast I can communicate to you?
0:11:38 Because if there’s no limit, then amazing world, right?
0:11:40 We can communicate so fast.
0:11:42 – It’s like instant telepathy.
0:11:45 It’s like you instantly beam me every thought in your head.
0:11:45 Yeah, okay.
0:11:47 – Exactly, the natural question he has,
0:11:50 once you set up this finite set, as you mentioned,
0:11:52 is okay, given these finite sets,
0:11:55 is there a limit on how fast I can communicate to you?
0:11:59 And so that was the question that was the heart of the paper,
0:12:04 which is to, so he formulated this notion
0:12:09 of a capacity that communication system is like a pipe.
0:12:14 It’s like you’re pushing water through this pipe
0:12:16 and the size of the pipe limits
0:12:19 of how fast you can push water through it.
0:12:22 And analogously, in communication,
0:12:24 there’s this notion of a size of a pipe,
0:12:25 which is called a capacity.
0:12:30 And you figure a way of computing this capacity
0:12:32 for different communication medium.
0:12:35 Any communication medium,
0:12:37 you can actually compute a capacity
0:12:38 for that communication medium,
0:12:42 and that limits how fast you can communicate information
0:12:43 over that medium.
0:12:46 Whether that medium is wireless, over the air,
0:12:49 or over the wire line, like I’m talking to you,
0:12:52 I communicate over the air, I talk to my wifi,
0:12:54 the wifi goes through some copper cables,
0:12:57 some optical fiber, all these are physical medium,
0:13:00 but you can compute a capacity
0:13:02 for each of these different mediums.
0:13:08 – And I know that part of the paper looks at,
0:13:17 say redundancy in various modes of communication.
0:13:21 And on a related note, patterns, right?
0:13:23 There’s this whole section of the paper
0:13:25 where he looks at the frequency
0:13:28 with which letters occur in English,
0:13:31 and kind of builds an idea around that.
0:13:34 Tell me about those pieces of the paper.
0:13:37 – Yeah, so let’s talk about the word redundancy.
0:13:38 – Yeah, was that the wrong word?
0:13:39 – Was that the right word?
0:13:40 – No, no, no, no, no, no.
0:13:41 That’s not only not the wrong word,
0:13:44 but it’s actually the most important word, I would say.
0:13:45 – Okay. – Almost.
0:13:49 Because you go back to the question,
0:13:50 to the thing I was talking about,
0:13:52 which is how fast you can communicate, right?
0:13:53 – Yeah.
0:13:55 – So what he discovered was actually,
0:13:57 there’s no limit on how fast you can communicate.
0:13:59 You can always communicate very fast,
0:14:02 but what the guy can hear is gibberish,
0:14:05 and he cannot really distinguish what you’re trying to say.
0:14:07 It’s like so much noise in the system,
0:14:08 that he cannot really figure out what to say.
0:14:10 – Even if you’re face-to-face, right?
0:14:11 Even if you’re face-to-face,
0:14:13 you’re not going over the phone or whatever.
0:14:16 If you talk too fast, the listener won’t understand
0:14:16 because you’re going too fast.
0:14:18 – Yeah, and anybody who goes
0:14:21 to a crazy professor’s lecture would know about this,
0:14:23 where the professor just keeps on talking
0:14:26 at a million miles per hour,
0:14:28 and the student just sits there,
0:14:29 and nobody understood a thing,
0:14:31 and the professor calls the day when it’s finished.
0:14:36 So basically, what he’s saying is that,
0:14:38 hey, you know what, to make sure
0:14:41 that the information goes through reliably,
0:14:43 reliably, that’s the first word,
0:14:48 you need to introduce redundancy in your message.
0:14:53 And what he figured out is, in some sense,
0:14:56 the optimal way of adding redundancy,
0:14:59 because you can always be stupid in adding redundancy.
0:15:03 For example, I can keep on repeating the same word
0:15:06 100 times to you, and then you’ll probably get it,
0:15:07 and then I move on to the next word.
0:15:09 I cannot move on the next word,
0:15:14 but that would take me 100 times slow, right?
0:15:18 And so that’s not a very smart way of adding redundancy.
0:15:20 So what he figured out is an optimal way
0:15:24 of adding redundancy, so that you can communicate reliably,
0:15:29 and yet, at the maximum, what he calls, capacity limit.
0:15:31 And that was a totally amazing,
0:15:34 actually formulation of the problem,
0:15:37 and highly non-obvious.
0:15:41 And I think that is sort of the amazing contribution
0:15:43 of this guy, Shannon, you know.
0:15:45 – It’s optimization.
0:15:49 He optimizes communication across any channel,
0:15:53 where you’re balancing efficiency,
0:15:55 or speed, and reliability.
0:15:56 That is the trade-off.
0:15:59 And he figures out how to optimize for that trade-off.
0:16:01 – Yes, yes.
0:16:05 He figured out how to optimize that trade-off,
0:16:11 but that trade-off turns out to be very interesting.
0:16:14 It’s a very interesting trade-off.
0:16:16 So typically, when we think about trade-off,
0:16:19 we think about like a smooth curve, right?
0:16:21 As when you’re doing something,
0:16:23 then you can get better performance.
0:16:26 But what he showed was that there’s kind of
0:16:28 like a cliff effect.
0:16:30 – Okay.
0:16:33 – And the cliff effect is that if you communicate
0:16:35 below this number called capacity,
0:16:39 then you can always engineer a system
0:16:42 to make the signal, the communication,
0:16:44 as reliable as you want.
0:16:46 So reliable, that’s completely clean.
0:16:47 – Wow.
0:16:51 – Whereas if you communicate above this number of capacity,
0:16:54 then there’s nothing you can do to make it signal clean.
0:16:55 It’s just completely gibberish.
0:17:00 So it’s a very sharp trade-off that he identified.
0:17:02 It’s not a smooth trade-off.
0:17:04 – And if you’re running the phone company,
0:17:05 that’s exactly what you want to know, right?
0:17:08 So then you can tune it all the way to capacity,
0:17:11 and then not try and tune it any more after that
0:17:13 because it’s not going to get any better.
0:17:14 – Correct.
0:17:15 And that’s the goal of 60 years of engineering
0:17:18 to achieve his goal, his vision.
0:17:22 His vision in 1948, it took people around 60 years
0:17:24 to get to that, implement his vision.
0:17:28 – Well, so you are part of that story, right?
0:17:31 Let’s let you walk onto the story now.
0:17:35 So you tell me about your work and how Shannon’s work,
0:17:38 you know, how you built on Shannon’s work.
0:17:41 Tell me about how you built on Shannon’s work.
0:17:47 – Yeah, so I did my PhD in the ’90s, in the ’90s.
0:17:51 My advisor was a Shannon’s student,
0:17:54 and so I learned information theory, okay?
0:17:57 Now, at that time, information theory
0:17:59 was almost a dead subject, okay?
0:18:03 When I was a PhD student, the first thing my advisor
0:18:05 told me, maybe you’re following Shannon,
0:18:07 is, hey, don’t work in information theory.
0:18:09 – Wow.
0:18:10 – You’ll never find a job.
0:18:12 You’ll never find a job with this stuff, okay?
0:18:14 – That’s a tough moment.
0:18:15 That must be a tough moment for you.
0:18:16 – It’s pretty tough, yeah.
0:18:20 Because at that time, there’s not much progress
0:18:24 made in the theory, and there’s no killer applications either.
0:18:26 There’s no very killer applications
0:18:30 that need all this sophisticated information theory, okay?
0:18:31 So it’s like a dead field.
0:18:34 – Was there a while when people used it to, like,
0:18:36 whatever, make landline phones work better,
0:18:38 like in the ’50s or something, where people were like,
0:18:40 oh, great, now we’ve got this theory
0:18:42 and we can make the phone work better?
0:18:48 – Yeah, so the thing is that the solutions
0:18:51 that people come up with to achieve these capacity limits
0:18:53 is very complicated, okay?
0:18:56 And the electronics, the technology’s just not enough
0:18:58 to build these complicated circuits.
0:19:01 So information theory have had not a very significant
0:19:05 impact in the ’50s, ’60s, or even ’70s.
0:19:06 – So it’s like one of those cases
0:19:09 where the theory is just too far ahead
0:19:12 of the technology to be useful.
0:19:15 – Yeah, and so people kind of start losing interest
0:19:18 in the theory, they say, oh, this is a bunch of math,
0:19:19 it’s not impacting the real world,
0:19:22 and so students are drifting away from the field,
0:19:25 but there are still always a few students, okay,
0:19:27 who are just so enumerated by the theory
0:19:30 that they keep on pursuing it.
0:19:33 And my advisor’s one of the leading professors in this area,
0:19:36 and he would have, like, one student
0:19:40 every decade, every decade, to do research in information.
0:19:42 – You were that student?
0:19:43 You were that student? – And I was not that student.
0:19:44 – Oh, okay. – And I was not
0:19:46 that student, okay?
0:19:48 At that time, that slot was already taken
0:19:50 by an earlier student. – Okay.
0:19:53 – Who was way smarter than me, who was way smarter than me,
0:19:56 and that said, he was a student of the decade
0:19:58 in information theory, okay?
0:20:01 Now, so I was assigned to work on some other problems,
0:20:03 okay, completely unrelated, okay?
0:20:05 But anyway, the point, though, is that when I graduated,
0:20:08 something happened, okay?
0:20:11 And that was the beginning of the wireless revolution.
0:20:12 – Uh-huh.
0:20:16 – That was the time when only a million people
0:20:19 have cell phones, and those cell phones,
0:20:21 I don’t even remember, it’s like, gigantic brick.
0:20:24 – Yeah, like, there’s that famous scene
0:20:25 from the movie “Wall Street,” right?
0:20:27 That’s the one that everybody talks about,
0:20:28 where it’s like, bigger than a brick.
0:20:31 People say brick, but it’s actually bigger than a brick.
0:20:33 It’s like a big, hardback book or something.
0:20:36 – Yeah, and actually, those days,
0:20:37 because there’s so few of these phones,
0:20:39 it’s like a prestige.
0:20:41 It’s like a prestige to have this brick.
0:20:42 – Yeah. – Okay.
0:20:44 – Yeah, you couldn’t get that brick.
0:20:47 You had to be rich to get that brick, yeah.
0:20:50 – Yeah, and so the wireless revolution was happening
0:20:52 because people realized that, hey, you know what?
0:20:55 Be able to communicate anytime, anywhere is really valuable.
0:20:58 And so people are now getting interested.
0:21:01 And at that time, what people realized is that, whoa,
0:21:05 this wireless physical media
0:21:07 is really tough to communicate over
0:21:09 because the bandwidth is so limited
0:21:11 and the noise is so much, right?
0:21:14 FCC was limiting the bandwidth allocation
0:21:16 to these applications a lot.
0:21:19 – Uh-huh, the Federal Communications Commission,
0:21:22 the government wasn’t letting wireless companies
0:21:23 use much bandwidth for themselves.
0:21:24 – Yeah, because all the bandwidth,
0:21:27 most of them are allocated for military purposes.
0:21:29 And there’s only very little bandwidth allocated
0:21:30 at that time for civilians.
0:21:32 And so those bandwidths were auctioned out
0:21:34 to companies with a very high price.
0:21:35 – Yeah.
0:21:39 – And so it became very important to be very efficient
0:21:43 in using this very expensive property, okay?
0:21:45 And then people realized, hey,
0:21:47 if we want to be really efficient,
0:21:51 then we need a theory which is about efficiency.
0:21:53 So people start thinking, okay, all right,
0:21:55 so information theory was dead,
0:21:56 but now it’s gonna come back to life
0:21:59 because we have this really important problem,
0:22:02 a really expensive spectrum that was allocated by FCC,
0:22:04 and we want to squeeze as much of it as possible.
0:22:05 – As much communication,
0:22:07 we need a sort of mathematical theory
0:22:09 of communication, if you will.
0:22:12 – And that was the renaissance of information theory
0:22:16 spurred by this amazing technology of wireless,
0:22:19 which took us from one million phones
0:22:22 to 10 billion phones today.
0:22:26 – Everybody has 1.1 phones.
0:22:30 And information theory play a big role in that revolution.
0:22:40 – In a minute, how David used Claude Shannon’s 1948 paper
0:22:42 to come up with an idea that we all use
0:22:44 every time we make a phone call.
0:22:52 – Hey, it’s Jacob.
0:22:54 I’m here with Rachel Botsman.
0:22:58 Rachel lectures on trust at Oxford University,
0:23:01 and she is the author of a new Pushkin audiobook
0:23:04 called How to Trust and Be Trusted.
0:23:05 Hi, Rachel.
0:23:06 – Hi, Jacob.
0:23:09 – Rachel Botsman, tell me three things
0:23:11 I need to know about trust.
0:23:16 – Number one, do not mistake confidence for competence.
0:23:17 Big trust mistakes.
0:23:19 So when people are making trust decisions,
0:23:23 they often look for confidence versus competence.
0:23:27 Number two, transparency doesn’t equal more trust.
0:23:29 Big myth and misconception.
0:23:32 And a real problem, actually in the tech world.
0:23:35 The reason why is because trust
0:23:38 is a confident relationship with the unknown.
0:23:41 So what are you doing if you make things more transparent?
0:23:44 You’re reducing the need for trust.
0:23:49 And number three, become a stellar expectation setter.
0:23:54 Inconsistency with expectations really damages trust.
0:23:55 – I love it.
0:23:57 Say the name of the book again
0:23:59 and why everybody should listen to it.
0:24:02 – So it’s called How to Trust and Be Trusted.
0:24:04 Intentionally, it’s a two-way title
0:24:08 because we have to give trust and we have to earn trust.
0:24:10 And the reason why I wrote it is because
0:24:13 we often hear about how trust is in a state of crisis
0:24:15 or how it’s in a state of decline.
0:24:18 But there’s lots of things that you can do
0:24:20 to improve trust in your own lives,
0:24:22 to improve trust in your teams,
0:24:25 trusting yourself to take more risks
0:24:28 or even making smarter trust decisions.
0:24:29 – Rachel Botsman, the new audio book
0:24:32 is called How to Trust and Be Trusted.
0:24:33 Great to talk with you.
0:24:34 – It’s so good to talk with you
0:24:36 and I really hope listeners listen to it
0:24:39 because it can change people’s lives.
0:24:44 – Let’s talk for a moment about your role, right?
0:24:47 Like you actually played an important role there.
0:24:50 – Yeah, so I was at Bell Labs.
0:24:54 – Uh-huh, just like Claude, just like Claude Giada.
0:24:56 – Yeah, so I spent one year at Bell Labs
0:25:00 as a so-called postdoc right after my PhD.
0:25:03 Before I moved to Berkeley to become a professor there,
0:25:04 I spent one year there.
0:25:06 And that’s what people were talking about
0:25:07 at that time at Bell Labs.
0:25:10 Hey, this new thing, wireless information theories
0:25:12 come back to life.
0:25:14 We can try to use information theory
0:25:16 and adapt it and extend it
0:25:18 to this wireless communication problem.
0:25:20 And so that’s when I said,
0:25:23 “Whoa, this information theory I learned from Bob Gallagher.
0:25:25 Finally, there’s a place to use it.
0:25:28 Finally, I can actually make a living.
0:25:31 Make a living out of it.”
0:25:33 Unlike what my advisor told me, it’s not dead.
0:25:34 It’s coming back to life.
0:25:35 – Yeah.
0:25:38 – And so that’s sort of my start in the field.
0:25:43 And yeah, so I invented a bunch of stuff
0:25:45 and actually applied this,
0:25:47 connected information theory to the real world.
0:25:50 And every time you use a phone,
0:25:52 you’re using my algorithm,
0:25:55 which is based on the theory of information.
0:25:58 – Huh, and so you’re,
0:26:00 that’s a cool thing to be able to say, first of all,
0:26:02 that’s a very good flex.
0:26:04 Your algorithm,
0:26:08 it’s the proportional fair scheduling algorithm, right?
0:26:09 – Yes, yes.
0:26:09 – What is that?
0:26:10 What’s it do?
0:26:13 – All right, so I should tell you a little bit of story.
0:26:15 I think a story is,
0:26:17 and then I’ll tell you what it does, okay?
0:26:19 So I went to,
0:26:22 so that was the end of 1999, around 1999.
0:26:26 So I was doing all this information theory stuff
0:26:29 at Berkeley, writing many papers.
0:26:32 But then I always have a thought back on my,
0:26:34 which is, “Hey, is this stuff going to be useful?”
0:26:36 And so I went to a,
0:26:37 I decided to go to a company, a wireless company,
0:26:38 who actually built these things
0:26:40 and see whether this theory can be used.
0:26:43 And the company I went to is called Qualcomm.
0:26:44 – Okay.
0:26:45 – I’ve heard of Qualcomm.
0:26:46 – No, you’ve heard of Qualcomm,
0:26:48 but at that time it was a small company.
0:26:49 It was not very big, okay?
0:26:51 And at that time,
0:26:54 they have this problem they’re working on, okay?
0:26:56 Which is the following, all right?
0:26:57 So in wireless communication,
0:27:00 there’s a concept called base station, okay?
0:27:05 And the base station serves many cell phones
0:27:06 in the vicinity of the base station.
0:27:08 It’s called cell, okay?
0:27:09 – Is it like a tower?
0:27:09 Is it what we would call it?
0:27:11 – Yes, like a tower, that’s right.
0:27:12 It’s always on the tower.
0:27:14 There’s some electronics there.
0:27:17 And that’s how the base station
0:27:20 is supposed to beam information to many phones.
0:27:20 To many phones.
0:27:22 – You still see them, you see them, whatever,
0:27:23 on top of a big building
0:27:25 or when you’re driving down the freeway, right?
0:27:26 That’s what you’re talking about, yeah.
0:27:27 – That’s right.
0:27:29 And sometimes on fake trees.
0:27:30 – Yeah, I love the fake trees.
0:27:32 In New Jersey, they love the fake trees, yeah.
0:27:34 – New Jersey, that’s right.
0:27:37 New Jersey fake trees, yes.
0:27:40 So at that time, they were looking at this problem,
0:27:44 which is, hey, okay, my bandwidth is limited,
0:27:47 but I have many users to serve, okay?
0:27:49 How do I schedule my limited resource
0:27:51 among all these users, right?
0:27:54 Because I only have one total bandwidth.
0:27:55 And so at that time, people were saying,
0:27:57 okay, maybe something simple.
0:28:00 I give one nth of the time to the end user, right?
0:28:02 So the boost of five users,
0:28:04 I serve this user for a little bit
0:28:05 and I serve the second user for a little bit
0:28:07 and third user, fourth user, fifth user.
0:28:08 – I mean, the idea is you’re switching really fast.
0:28:09 You’re just like switching kind of.
0:28:12 – Yeah, switching really fast, yeah, exactly.
0:28:14 And then when I went there, I said, okay,
0:28:16 good, this is a problem, it’s a good problem.
0:28:19 And I said, hey, instead of fixating
0:28:21 on this particular scheduling policy,
0:28:24 why don’t we do a Shannon thing?
0:28:29 – A Claude Shannon thing, you thought of, yeah, okay.
0:28:31 – The Claude Shannon thing is what?
0:28:34 Is to look at the problem from first principle,
0:28:39 not pre-assume a particular solution
0:28:41 or a particular class solution even,
0:28:44 and ask ourselves, what is the capacity
0:28:45 of this whole system?
0:28:51 And how do I engineer the system to achieve that capacity?
0:28:52 Okay? – Uh-huh.
0:28:55 – And it turns out that if you look at the problem this way,
0:28:58 then it turns out that the optimal way of scheduling
0:29:01 is not the one that they were trying to design.
0:29:05 And the reason is because in wireless communication,
0:29:07 there’s a very interesting characteristic,
0:29:10 which is called fading.
0:29:12 – Okay. – Okay.
0:29:15 When I talk to you over the air,
0:29:17 the channel actually goes up and down
0:29:20 strong and weak, strong and weak very rapidly.
0:29:24 What I mean is when I send an electromagnetic signal
0:29:27 from the base station to the phone,
0:29:32 that signal get amplified and attenuated very rapidly.
0:29:35 – It goes up and down. – It goes up and down.
0:29:36 – Okay. – Okay?
0:29:38 – Can we say it gets stronger and weaker?
0:29:39 Can we say it gets stronger and weaker?
0:29:40 – Stronger and weaker. – Okay.
0:29:42 – Yes.
0:29:44 And so the optimal way that information
0:29:47 so it has to do is actually not divide the time
0:29:52 into slots blindly, but really try to schedule a user
0:29:57 when the channel is strong.
0:29:59 – Uh-huh, uh-huh.
0:30:02 – And then from that on, I designed a scheduling algorithm
0:30:05 which is more practical by sort of leverage
0:30:08 of this basic idea from information theory.
0:30:11 – And so the base station is basically monitoring
0:30:13 the strength of the incoming signals
0:30:16 from all the different phones.
0:30:17 – Correct, correct. – And saying,
0:30:18 “Oh, that one’s strong, I’m gonna grab that one.
0:30:20 “Oh, that one’s strong, I’m gonna grab that one.”
0:30:21 That’s what’s happening.
0:30:22 – Correct, correct.
0:30:25 – And how does that, I mean, I get in a kind of
0:30:27 big first principles way sort of analogously,
0:30:29 it follows from Shannon, but is there anything
0:30:34 sort of specific in Shannon that leads you
0:30:36 to this algorithm?
0:30:42 – So remember, Shannon is a very general theory.
0:30:44 – Yeah. – Okay?
0:30:47 It basically says that given any communication medium
0:30:52 or any communication setting, you can try to calculate
0:30:55 this notion of a capacity.
0:30:57 So the very general theory.
0:31:01 What I did was to apply it to a very specific context,
0:31:05 which is this base station serving multiple user setting.
0:31:06 – Yeah.
0:31:10 – And then apply his framework to analyze
0:31:12 the capacity of that system.
0:31:13 – Uh-huh.
0:31:16 – And in the process of analyzing the capacity,
0:31:19 you can also figure out what is the optimal way
0:31:22 of achieving that capacity.
0:31:24 Remember you mentioned capacity is really
0:31:28 an optimization problem, and Shannon was able
0:31:30 to solve this optimization problem in general,
0:31:32 but now I specialize it in some sense
0:31:35 to this pretty specific setting,
0:31:38 except that the setting is used by everybody.
0:31:39 – Yes, yes.
0:31:42 – At that time, it was like, research is about timing.
0:31:45 And I was there at the right place at the right time
0:31:50 because Qualcomm turns out to completely dominate
0:31:52 the entire third generation technology.
0:31:53 – Yeah, 3G.
0:31:56 – So when I was able to convince them that,
0:31:59 hey, your way of doing things is no good,
0:32:02 this way suggested by Shannon is actually far better.
0:32:05 Please use this way.
0:32:06 It took me a few months,
0:32:08 but I was able to persuade them to implement it.
0:32:12 And then it got into the standard through the domination.
0:32:14 And then every standard after that
0:32:16 uses the same, based on the same algorithm.
0:32:19 So it was good because as I said,
0:32:21 I’m at the right place at the right time.
0:32:24 You know, when you try to contribute to engineering,
0:32:26 it’s too late if the system is built already,
0:32:28 because people don’t want to change the whole system
0:32:30 to accommodate your new idea.
0:32:34 But it was very early in the design phase.
0:32:37 – So, okay, so you made this breakthrough
0:32:41 in wireless communications using Shannon’s work.
0:32:46 Were there similar breakthroughs in other domains?
0:32:47 – Any communication medium, right?
0:32:50 It could be optical fiber.
0:32:54 It could be DSL modem, DSL modem.
0:32:56 Underwater communication.
0:32:58 Almost all these communication systems
0:33:01 are now designed based on his principle.
0:33:07 So his impact of this theory is kind of global.
0:33:09 It’s the entire communication landscape.
0:33:15 – There’s a story I read about Shannon,
0:33:20 when he is developing information theory.
0:33:23 He takes a book off the shelf
0:33:25 and he reads a sentence to actually his wife.
0:33:29 And it’s something like the lamp was sitting on the,
0:33:30 and she says, table.
0:33:32 And he says, no, I’ll give you a clue.
0:33:35 The first letter is D and she says desk.
0:33:38 And when I heard that story,
0:33:41 what I thought of was large language models.
0:33:43 Like that sounds exactly like a large language model.
0:33:47 And so I’m just fishing, I’m just curious,
0:33:52 like does his work matter for machine learning,
0:33:55 large language models, et cetera, or no?
0:34:00 – Yeah, so that’s a very interesting point.
0:34:03 Now I’m not an expert by any means in AI
0:34:04 or large language models.
0:34:08 I’m not a professional researcher in that area.
0:34:11 But I think you can actually see some commonality, right?
0:34:15 Is that, these models in some sense,
0:34:17 they don’t care about meaning either.
0:34:20 – Yeah, very good, very good, yeah.
0:34:21 – Right?
0:34:24 Actually, I just came to my,
0:34:26 this discussion is very interesting.
0:34:28 Because it’s really just patterns.
0:34:31 It’s just which patterns are more likely
0:34:32 than other patterns, right?
0:34:34 The example you gave about desk and lamp
0:34:36 is basically about patterns.
0:34:38 And information theory is really analyzing
0:34:42 sort of the number of possible patterns in some sense.
0:34:47 So there is definitely a philosophical connection,
0:34:50 I believe, starting from Shannon
0:34:52 to these large language models.
0:34:53 – So let me ask you about one other,
0:34:56 and this is one that you are professionally involved in.
0:35:03 Cryptocurrency and blockchain, you have studied it
0:35:06 and you started a company, right?
0:35:08 Is there a connection between Shannon’s work
0:35:09 and cryptocurrency?
0:35:12 – Yeah, so what attracts me
0:35:16 to work in this area of blockchain
0:35:21 is that blockchain actually has one very common
0:35:23 philosophical connection to information theory,
0:35:25 which is the following.
0:35:29 In blockchain, the problem is not communication per se,
0:35:31 it’s called consensus, okay?
0:35:33 It’s a different problem,
0:35:36 but it’s essentially allow a bunch of users
0:35:40 at different places to come to an agreement on something, okay?
0:35:41 – Yes.
0:35:45 – Now, the goal of designing blockchain
0:35:49 is really to be so-called fault-tolerant.
0:35:49 – Fault-tolerant.
0:35:51 – Which means fault-tolerant,
0:35:54 which means that even if, say,
0:35:57 one-third of the users are bad guys
0:35:59 and send you some gibberish message,
0:36:00 – Yeah.
0:36:04 – You can still, the rest two-third people
0:36:06 can still come to an agreement, okay?
0:36:08 All right?
0:36:09 So you look at this problem,
0:36:12 it’s actually not that different from communication
0:36:14 in information theory because it’s kind of combating.
0:36:16 – The bad guys are the noise,
0:36:18 or the good guys are the signal, yeah.
0:36:19 – And the good guys are the signal
0:36:22 and they try to introduce redundancy, okay?
0:36:25 To help them to fight against these bad guys.
0:36:28 – Yes, and there’s an optimization problem
0:36:30 where the more redundancy you have,
0:36:33 the sort of slower the system is, the more ponderous.
0:36:36 And so you try, the optimization problem is
0:36:39 you try to figure out what is the optimal number
0:36:41 of bad guys that you can tolerate
0:36:43 and your system still works.
0:36:46 That is the analogous to the capacity problem.
0:36:51 So I find the philosophical connection very appealing.
0:36:52 And so that’s sort of one reason
0:36:55 why I got attracted to working to this area.
0:36:59 – Why do you think more people don’t know about Shannon?
0:37:04 Like all of the sort of intellectuals in technology say
0:37:10 he’s like one of the great thinkers of the 20th century.
0:37:14 But most people have never heard of him.
0:37:16 Why do you think that is?
0:37:19 – So Shannon was actually a very shy person.
0:37:24 Very shy person, he hates publicity.
0:37:29 He hated when people interview him.
0:37:30 You remember, right?
0:37:32 He’s basically a very modest person.
0:37:34 Remember the first paragraph I talked to you about.
0:37:37 He tells you what he’s, that he’s not accomplishing.
0:37:40 And so he’s a very modest, very shy person,
0:37:42 not into publicity.
0:37:47 And I think that sort of impact not only himself,
0:37:49 but also everybody who works in that field.
0:37:54 Adopt this as kind of like a metric, right?
0:37:56 That, hey, we should all be modest because why?
0:37:59 Look at this guy who accomplished so much
0:38:00 and he’s still so modest.
0:38:01 Who are we?
0:38:03 Who are we, right?
0:38:05 So as a result,
0:38:08 the field doesn’t really sell themselves very well.
0:38:11 The marketing engine, the marketing DNA is not there.
0:38:15 And so people don’t know about him.
0:38:16 – So I want to talk for a minute
0:38:18 about the rest of Shannon’s life.
0:38:22 He writes this huge paper when he’s in his early 30s,
0:38:26 eventually goes on to be a professor at MIT.
0:38:30 And he seems to spend a lot of his career
0:38:33 juggling, riding a unicycle,
0:38:35 building mechanical toys, building games.
0:38:40 And he never does sort of great influential work again.
0:38:44 And I’m curious, what do you make of that?
0:38:46 How do you sort of fit his whole career together?
0:38:50 – So there’s a single, there’s a theme
0:38:53 that unifies all this in my mind, which is playfulness.
0:38:59 Because in his mind, research is really about puzzles.
0:39:03 He doesn’t understand something.
0:39:05 It’s like a puzzle to him.
0:39:08 And he’s trying to figure out the pieces of the puzzle.
0:39:11 Information theory was like that.
0:39:13 The puzzles, he sees all these real-world systems.
0:39:15 They seem to all share some commodity,
0:39:16 but nobody understood it.
0:39:17 So there’s a puzzle
0:39:19 and he’s always thinking about the puzzle.
0:39:22 And finally his paper basically solved that puzzle.
0:39:25 So everything to him is playfulness.
0:39:27 I think it’s plain, it’s a game.
0:39:29 Puzzle needs to solve the puzzle.
0:39:30 And that’s his mind.
0:39:31 That’s how his mind works.
0:39:33 So although it seems very different,
0:39:37 things that he did pre and post information theory,
0:39:41 but it’s actually in my mind, quite swanky monotonous.
0:39:44 (upbeat music)
0:39:49 – We’ll be back in a minute with the lightning round.
0:39:51 (upbeat music)
0:39:55 – Hey, it’s Jacob.
0:39:57 I’m here with Rachel Botsman.
0:40:01 Rachel lectures on trust at Oxford University.
0:40:04 And she is the author of a new Pushkin audiobook
0:40:08 called How to Trust and Be Trusted.
0:40:09 Hi, Rachel.
0:40:10 – Hi, Jacob.
0:40:13 – Rachel Botsman, tell me three things
0:40:15 I need to know about trust.
0:40:19 Number one, do not mistake confidence for competence.
0:40:21 Big trust mistakes.
0:40:23 So when people are making trust decisions,
0:40:27 they often look for confidence versus competence.
0:40:31 Number two, transparency doesn’t equal more trust.
0:40:33 Big myth and misconception.
0:40:35 And a real problem, actually in the tech world,
0:40:38 the reason why is because trust
0:40:41 is a confident relationship with the unknown.
0:40:43 So what are you doing?
0:40:45 If you make things more transparent,
0:40:47 you’re reducing the need for trust.
0:40:52 And number three, become a stellar expectation setter.
0:40:58 Inconsistency with expectations really damages trust.
0:40:59 – I love it.
0:41:00 Say the name of the book again
0:41:02 and why everybody should listen to it.
0:41:05 – So it’s called How to Trust and Be Trusted.
0:41:07 Intentionally, it’s a two-way title
0:41:11 because we have to give trust and we have to earn trust.
0:41:13 And the reason why I wrote it
0:41:15 is because we often hear about
0:41:17 how trust is in a state of crisis
0:41:19 or how it’s in a state of decline.
0:41:21 But there’s lots of things that you can do
0:41:24 to improve trust in your own lives,
0:41:26 to improve trust in your teams,
0:41:28 trusting yourself to take more risks,
0:41:31 or even making smarter trust decisions.
0:41:33 – Rachel Botsman, the new audio book
0:41:35 is called How to Trust and Be Trusted.
0:41:36 Great to talk with you.
0:41:37 – It’s so good to talk with you.
0:41:40 And I really hope listeners listen to it
0:41:42 because it can change people’s lives.
0:41:49 – So I read that you recently asked people
0:41:52 at your company to give five-minute talks.
0:41:55 I’m curious why you did that.
0:41:56 That’s interesting to me.
0:41:56 Why’d you do that?
0:42:02 – So the shorter the talk, the harder it is to give.
0:42:03 – Yeah.
0:42:08 – So you can explain an idea in five minutes.
0:42:10 Then I think your idea is actually not very good.
0:42:13 – Aha, that’s good.
0:42:16 – Most good ideas you can get the point to
0:42:18 across in five minutes.
0:42:20 Remember, I’m an information theorist by training.
0:42:23 So communication to the limit
0:42:26 is what I’m passionate about.
0:42:28 – If you had to give a five-minute talk,
0:42:30 what would it be about?
0:42:33 – About Shannon, I guess.
0:42:38 He’s my hero, he’s my hero.
0:42:43 – So one, you talked about the importance of timing
0:42:48 in research of not only finding the right problem,
0:42:50 but finding the right problem at the right time, right?
0:42:54 Both in terms of Shannon’s work and in terms of your work.
0:42:59 You’re also a professor and a manager,
0:43:01 like how do you help other people
0:43:03 find the right problem at the right time?
0:43:09 – Yeah, finding the right problem at the right time
0:43:11 is probably the most difficult
0:43:15 because, you know, timing is everything.
0:43:18 However, this is hard to teach.
0:43:21 What you try to do is to be ready.
0:43:27 So one very famous information theorist told me this.
0:43:30 He said, you know, everybody will get lucky
0:43:32 at some point in time in their career.
0:43:37 However, most people, when they get lucky,
0:43:38 they’re not ready.
0:43:41 So they don’t realize that they get lucky.
0:43:43 And so they missed the opportunity.
0:43:44 They went a different direction.
0:43:46 Luck tells you you should go this way,
0:43:48 but you went the other way, lost it.
0:43:50 – That makes me so scared.
0:43:54 – And so what I teach my students is always be ready.
0:43:55 It’s like your muscles.
0:43:57 You have to be always train your muscles
0:43:59 so that when you are lucky,
0:44:00 you can capitalize on the luck.
0:44:06 – So you talked about Shannon’s playful nature.
0:44:09 Like he was a juggler, he rode a unicycle.
0:44:10 You do anything like that?
0:44:13 Do you have any weird hobbies?
0:44:17 – No, no.
0:44:20 The only weird hobby is I love to talk to people like you.
0:44:21 – Fair.
0:44:24 You love going on podcasts, that’s the juggling
0:44:26 of the 21st century.
0:44:30 Who’s your second favorite underrated thinker?
0:44:33 – My advisor.
0:44:35 – Ah, Bob Gallagher.
0:44:37 – My advisor, Gallagher, yeah, Bob Gallagher.
0:44:40 He taught me how to think about research
0:44:44 because he learned from Shannon and I learned from him.
0:44:47 – And if you boil down what–
0:44:49 – Your advisor learned from Shannon
0:44:51 and what you learned from your advisor,
0:44:52 what would it be?
0:44:53 What did you learn?
0:44:58 – Yeah, I learned about taking a very complicated problem
0:45:01 and strip it down to the essential.
0:45:02 – Uh-huh.
0:45:05 – And then formulate a problem around that and solve it.
0:45:07 That’s an art.
0:45:11 It’s not something you can convert it
0:45:14 into a mathematical formula and teach students.
0:45:18 It’s just based on intuition, experience.
0:45:21 And that’s what Shannon taught my advisor.
0:45:24 And that’s what my advisor taught me.
0:45:26 And that’s what I try to teach my students.
0:45:29 Really, teaching is not really about giving the formula.
0:45:31 It’s really just learning by examples.
0:45:34 I observe what he does.
0:45:36 And then my students observe what I do
0:45:37 as I interact with them.
0:45:40 And hopefully this art will carry on
0:45:43 from generation to generation.
0:45:46 – Finding the essence of the problem.
0:45:47 – Yeah.
0:45:50 (upbeat music)
0:45:57 – David Shea is a professor at Stanford.
0:46:01 Today’s show was produced by Gabriel Hunter Chang.
0:46:03 It was edited by Lydia Jean Kott
0:46:05 and engineered by Sarah Brugier.
0:46:09 You can email us at problem@pushkin.fm.
0:46:11 I’m Jacob Goldstein and we’ll be back next week
0:46:13 with another episode of “What’s Your Problem?”
0:46:20 – Hey, it’s Jacob.
0:46:22 I’m here with Rachel Botsman.
0:46:25 Rachel lectures on trust at Oxford University
0:46:29 and she is the author of a new Pushkin audiobook
0:46:32 called “How to Trust and Be Trusted.”
0:46:33 Hi, Rachel.
0:46:34 – Hi, Jacob.
0:46:35 – Rachel Botsman.
0:46:39 Tell me three things I need to know about trust.
0:46:44 – Number one, do not mistake confidence for competence.
0:46:45 Big trust mistakes.
0:46:47 So when people are making trust decisions,
0:46:51 they often look for confidence versus competence.
0:46:55 Number two, transparency doesn’t equal more trust.
0:46:57 Big myth and misconception
0:47:00 and a real problem actually in the tech world.
0:47:02 The reason why is because trust
0:47:06 is a confident relationship with the unknown.
0:47:09 So what are you doing if you make things more transparent?
0:47:12 You’re reducing the need for trust.
0:47:17 And number three, become a stellar expectation setter.
0:47:22 Inconsistency with expectations really damages trust.
0:47:23 – I love it.
0:47:25 Say the name of the book again
0:47:27 and why everybody should listen to it.
0:47:30 – So it’s called “How to Trust and Be Trusted.”
0:47:32 Intentionally, it’s a two-way title
0:47:36 because we have to give trust and we have to earn trust.
0:47:39 And the reason why I wrote it is because we often hear
0:47:41 about how trust is in a state of crisis
0:47:43 or how it’s in a state of decline.
0:47:45 But there’s lots of things that you can do
0:47:48 to improve trust in your own lives,
0:47:50 to improve trust in your teams,
0:47:52 trusting yourself to take more risks
0:47:55 or even making smarter trust decisions.
0:47:57 – Rachel Botsman, the new audio book
0:48:00 is called “How to Trust and Be Trusted.”
0:48:01 Great to talk with you.
0:48:02 – It’s so good to talk with you
0:48:04 and I really hope listeners listen to it
0:48:06 because it can change people’s lives.
Claude Shannon is a major figure in the history of technology. Known as the father of information theory, Shannon spent decades at Bell Labs and MIT. But what exactly did Claude Shannon figure out, and why is it so important?
To answer that question, Jacob talked with David Tse, a professor of electrical engineering at Stanford who studied under one of Shannon’s students, and who teaches Shannon to his own students today. David used Shannon’s work to make a breakthrough in wireless communication that underpins every phone call we make today.
See omnystudio.com/listener for privacy information.