Pod Insights a16z Podcast a16z Podcast | From Mind at Play to Making the Information Age About this episode Modern technology owes much to the introduction of the binary digit or "bit", first proposed by Claude Shannon in "A Mathematical Theory of Communication”, a paper published in 1948. The bit would go on to transform analog to digital, making Shannon the father of the information age. His contemporaries (and collaborators) included Vannevar Bush, Alan Turing, and other architects of the digital era.
In this podcast, moderated by a16z board partner Steven Sinofsky, the authors of the new book about Shannon, A Mind at Play -- Jimmy Soni and Rob Goodman -- discuss the life and mind of the mathematician, engineer, and cryptographer from his roots as a precocious tinkerer in Gaylord, Michigan to the halls of MIT and Bell Labs. But this conversation is also, more broadly, about how genius and innovation happens... beginning with play.
Transcript Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Speaker 00 Welcome to the A16Z Podcast. Today, we're doing one of our book episodes, and we're talking about genius and the process of innovation through the life of Claude Shannon, the father of information theory. He was also an architect of the digital age, who, among other things, worked with Venever Bush and befriended Alan Turing. This conversation is moderated by A16Z board partner Steven Sinofsky, with special guests Jimmy Soni and Rob Goodman, authors of a new biography of Shannon, just out, called A Mind App Play. Rob's voice is the first you'll hear right after Stephen's.
0:30 Steven Sinofsky? I want to start off by just setting some context. Back in, I think, around 1990, Scientific American said, decades after this paper was published in 1948, that Shannon created, quote, "the Magna Carta of the information age." What did they even mean by that? That's a pretty big statement.
0:46 Speaker 02 Shannon's paper, A Mathematical Theory of Communication, is something like a founding document. It laid out the principles that make the digital transmission of information possible. Shannon, in this paper, does things like introduce the concept of the bit, explain how you can quantify information, explain how you can use digital codes to compress information and to send it with arbitrarily perfect accuracy. So all these things that are foundational to digital communications in the present, Shannon lays them out, and that's Magna Carta scale achievement.
1:14 Steven Sinofsky? Let's kind of go back in time and go back to, you know, his earliest years. And he was born in 1916, more than 100 years ago. He probably didn't have much in the way of electricity, of indoor plumbing, of all of those kind of things. It was very early in the 20th century, in industrialization, in the Midwest. What was he born into?
1:35 Speaker 03 So he was born in Upper Michigan, and his childhood is the childhood of a sort of boy tinkerer, right? He plays with broken radios and takes them apart and puts them back together. There was a line that all the broken radios in Gaylord passed through Claude Shannon's hands. He builds a makeshift elevator in the back of a barn with a friend.
1:57 Steven Sinofsky? So Gaylord is his hometown. How big is that town?
2:00 Speaker 03 Two or 3,000 people.
2:02 Steven Sinofsky? Tiny.
2:02 Speaker 03 It was like a small village, but it had, because it was close to railroads, it had some commercial importance to the region. But his dad is a probate judge, mom is a teacher, and he is a boy who's constantly playing, building things, always trying to sort of figure out how to rig things up. So he rigs up a barbed wire telegraph between his house and a friend's house.
2:25 Steven Sinofsky? So he uses the barbed wire as the transmission?
2:26 Speaker 03 Yeah, exactly. Claude is a boy. I mean, he's just doing this for fun.
2:29 Steven Sinofsky? So that's sort of the rough equivalent of like hacking away at a Raspberry Pi today?
2:33 Speaker 03 Basically, yes, basically. And then he's also, you know, he's like, he has inspiration from his family. His grandfather was an inventor. He actually filed a patent to improve on the washing machine. And so he's inspired by that, and he's a distant cousin of Thomas Edison. So that's kind of in the family lore as well. But he has a very normal childhood compared to other geniuses. I mean, we are not talking about somebody who, whose parents are drilling him in the finer points of advanced mathematics at a young age. They allow him to play. They allow him to do, do his, live his own life.
3:03 Steven Sinofsky? But also compared to many of the other scientists that emerged in, that would also later become his contemporaries, he also didn't face like oppression.
3:12 Speaker 02 Yeah, compared to all the people who made such a difference in technology in the 20th century, Claude Shannon had just a pretty idyllic childhood. And one of my favorite parts of researching this book was reading some newspaper headlines in this tiny town of Gaylord. Some of our favorites were "meeting held to discuss artichokes" and "Vern Matz loses finger." So those are the kinds of things that made the front page of the newspaper in Gaylord, Michigan. And that's the kind of childhood Shannon had.
3:34 Steven Sinofsky? One of the things that fascinates me about whenever I get to read about the inventors from that, that started in that era is just how versatile they are. Like, it's rather incredible that, that you could become essentially as expert as you can become in so many different things.
3:53 Speaker 03 His mom is a musician, so he ends up playing the French horn and later picks up the jazz clarinet. He does reasonably well in school. It turns out that part of the reason he decided to go whole hog on mathematics is because his sister was good at math and there was a little bit of sibling rivalry. He finds his way to the University of Michigan. Michigan and he comes at a really interesting time because the engineering school has just gone through this massive expansion.
4:16 Steven Sinofsky? And I find it relevant to what's going on today with the discussion about coding and the importance of sort of transforming education to the modern era. He seemed to be fortunate enough that the University of Michigan was busy transforming itself from sort of a normal, as you would, like liberal arts school. Into like, hey, we need to be an engineering school.
4:35 Speaker 02 There's this quote from, I think, the dean of engineering who's almost excited because the engineering department is about to pass the liberal arts department registration. And he says, by God, we'll pass them yet. So it's just this sort of idea that the economy is changing around the school. And the school is really investing a lot in its, its, its capital, both physical capital and human capital to keep up with that.
4:55 Steven Sinofsky? What, what else happened to him at Michigan that was, was such an interesting part of his?
5:00 Speaker 03 Yeah, he started publishing answers to these mathematical puzzles that came at the very back of academic journals. By the way, we actually published the puzzles in the book. So for anybody who wants to try to solve what Claude Shannon was solving, you're welcome to try.
5:13 Steven Sinofsky? Yeah, I'm good.
5:14 Speaker 03 Rob and I couldn't, but imagine that you're a college junior or senior. You're like picking up these academic journals, flipping to the back, looking at these puzzles, working out long solutions, sending them in for publication. What it suggested to us was this was a guy who wasn't going to go back and run the family furniture business. He was actually going to try to make it as an academician and as somebody who was going to get some advanced training. And it's a pretty incredible thing when you think about a kid like Claude Shannon with not particular, no particular means. From a reasonably modest family, a small town, going to the University of Michigan and managing, you get two pieces published in these journals later, later in his collegiate career. That's, that's a pretty extraordinary thing when you think about where he is coming from.
5:57 Steven Sinofsky? Yeah. I mean, especially because most of the people reading those and answering them were probably on the East Coast at Harvard and MIT. And then also in Michigan, he was either being pulled to the mechanical world of farming or to the soon to be created auto industry.
6:12 Speaker 02 Right. And he was someone who made a point of studying engineering and mathematics at the same time. And I think that was relatively rare to double major in those two things. Shannon said he just did it because he was just a few courses away from getting a double major. So why not?
6:23 Steven Sinofsky? So he finished up at Michigan and then like did this awesome thing where he's just like, hey, I think I'm going to go to MIT.
6:28 Speaker 02 So again, it kind of testifies a little bit to Shannon's ambition. He sees this job application invitation on something the size of a postcard that's posted up in the engineering building at Michigan. And it says, come to MIT and work as a graduate student with Veneva Bush and the Differential Analyzer, which is one of the leading computing machines of the day. It's an analog computer. So Shannon sends off his application. So it testifies, one, to the fact that he had this decent publication record for an undergrad, but also to the fact that Veneva Bush, who was one of the great sort of scientific networkers and organizers in 20th century America, he had a real eye for talent. He was the first person really well up in the scientific hierarchy to spot Claude Shannon's talent and to sort of invite him into the big leagues, in a sense.
7:13 Steven Sinofsky? So Bush wrote this very famous article called As We May Think.
7:17 Speaker 02 Right.
7:17 Steven Sinofsky? Which is sort of the history of the iPhone, or a tablet, or a whole bunch, and the web, and a whole bunch of other stuff, all rolled into a single paper, which itself is phenomenal.
7:27 Speaker 02 Right.
7:27 Steven Sinofsky? But he does, so he ends up getting a job with Bush. I was fascinated, you know, by this description because it goes back to being talented in many things. Like, Bush had this whole philosophy of engineering that was deeply, and he was running, he was like a dean at MIT, so he was in charge of a lot of stuff. And so he was, he didn't believe in like the pure theory, but he also didn't believe in sort of the pure mechanical. He had this bizarre view of, at the time, of like, sort of, how do you think with your hands?
7:55 Speaker 02 Yeah, and he had a great example when he was constructing this differential analyzer. He said he was working with a pretty, not very well schooled mechanic to actually build this thing. He said by the end of the process of putting this thing together, this mechanic had pretty much learned the basic concept of calculus. He didn't really get it on an intellectual level, but he knew it with his hands. He got it in his bones because Bush and the machines that he built were all about analog processes, about acting out differential equations, about thinking about how to make things through the act of building. Bush said that he was never really more than a, you know, second-rate mathematical brain, but he was a great builder and a great organizer. And he was really someone who put those skills to use for doing math, for acting out math.
8:38 Speaker 02 And then that was something that I think was really key to what made him such an important figure in the field of computing.
8:44 Steven Sinofsky? I don't know that there's a figure in science at that time who was better as a mentor for Claude Shannon than Veneva Bush. Just building on this, Bush was building essentially these analog purpose-built machines to solve problems. And it's really hard for us to wrap our heads around what was going on. But it was sort of like, if you want to track the trajectory of a missile, then you build like a bunch of metal that operates in a certain way, and you spin wheels, and the answer to the missile's trajectory pops out the other end. But if you wanted to then do some weather forecast, the machine was irrelevant.
9:17 Speaker 02 You'd have to completely rebuild it from scratch. So it really wasn't the most practical machine, but these are room-sized computers, huge, they call them fearsome things of gears and shafts. The problem solving that an analog computer was doing was actually replicating what the problem looked like, and then figuring out the solution. So it did have to be rebuilt. It broke constantly. It was really frustrating. People had to watch it 24 hours a day because the bad things would happen if you didn't.
9:40 Steven Sinofsky? And like debugging it involved like, you know, filing more off of some part, like adding a tooth in a gear or kind of. And what I think is actually really neat in hindsight is Bush was attempting to push his students, including Shannon, to build like the general purpose version of this machine that could like solve any differential equation. And if you kind of do an analogy today, that's a lot like, like people saying, hey, let's go solve general AI, when all the grad students are using machine learning to pick out kitten videos. And so they understand how to use machine learning for kitten videos, but the idea of like, you know, figure out what videos to go find and look at and classify them and understand, just seems really far off. It gives much more optimism that general AI
10:22 Steven Sinofsky? might be solved, because if you were Vannevar Bush, you were just not getting closer to your general purpose differential engine until Shannon comes along.
10:30 Speaker 02 And the really interesting part about that is that when you set people onto these general problems, you can't necessarily predict where the solution is going to come from or what's going to be productive. So what Bush is interested in is configuring, like you said, a general purpose analog computer that can reassemble itself on the fly and can use electrical relays to change the quantities of the various variables that it shafts and gears represent. And Shannon takes this in a very different direction through his study of the electrical relays in the switching system, when he realizes that this can really be combined with a And logic, and this is something that Chris Dixon wrote in his great article about Shannon in the Atlantic, where he said that Shannon figured out how to map logic, Boolean logic, onto the physical world.
11:11 Speaker 02 And he did this because Bush sort of set him to deal with this problem in general computing. And it turned out to be hugely productive because Shannon, along with Turing's paper in the same year, is really laying the foundations for all the digital computers that come afterwards.
11:24 Steven Sinofsky? Well, it sounds like also that that was another example of, you know, one person who was skilled in many disciplines, applying the different disciplines across them. I mean, he understood Boolean logic, he understood math and calculus, and he was a tinkerer.
11:40 Speaker 03 And Andy had actually worked at the phone company as well, so he understood switches. He studies logic as an undergrad, he manages to work at the phone company, he gets Vannevar Bush as a mentor, and he works on the differential analyzer. There is a bit of this that you feel like is almost kismet. And these things, logic and the switches and the analyzer, had been in the ether. It took Claude Shannon to fuse them together.
12:00 Steven Sinofsky? So he comes up with this sort of breakthrough notion of bringing together, you know, logic gates, Boolean logic, circuits. And it seems as amazing as this was, it wasn't quite a leap to like the computer.
12:13 Speaker 02 It was recognized almost immediately as a really important piece of work. It won the Nobel Prize, which is different than the Nobel Prize. We had to point that out in the book, which is an award for engineering papers. So after he's done this amazing piece of work in the area of switching and logic, Bush says to him, why don't you go write your dissertation on theoretical genetics now? Because why not? Once you're Claude Shannon. So Claude Shannon says, okay, and he goes off and does it. And it's possible that it took him out of direct contact with that field at least for a temporary amount of time. And then the war happens.
12:40 Steven Sinofsky? Yeah, it's really interesting because at this point, everyone is taken to focusing on the needs of the war, the war department. And he's decidedly apolitical and doesn't appear to be particularly religious or even dogmatic in anything other than his beliefs about math and engineering.
12:55 Speaker 03 Sorry, one other thing, he's very dogmatic about jazz music.
12:58 Steven Sinofsky? Oh, okay, aren't we all? How did that play in the environment he's in, where, you know, people had fled from Europe because of the war? What was on his mind? Did he care about the repercussions of technology? Or did he? Put aside the beliefs.
13:13 Speaker 03 It absolutely played a role. So this is actually a very hard time in Claude Shannon's life. It's the cusp of the American entry into World War II. He himself admits, he doesn't want to do the draft. He's a frail guy. He likes to keep his own, you know, counsel. His first marriage is collapsing and he has gone from MIT to Princeton's Institute of Advanced Studies where he's on fellowship and he doesn't quite know what's going to come next. And there's the very real risk that he gets drafted and sent overseas to fight. And what he has now, you know, over the course of his undergrad and his graduate studies, acquired some impressive mentors. Those mentors get him a wartime contract working at Bell Laboratories that spares him from the draft. More importantly, it puts
13:55 Speaker 03 him working on practical applications of mathematics and technology, and not just practical, we're talking the most practical. His first project is on fire control, which basically is how do you shoot things down from the sky?
14:07 Steven Sinofsky? Yeah, it's like an anti-aircraft, it's like a control unit for, for a really fast shooting anti-aircraft gun, not, not like fire or flames.
14:13 Speaker 03 Right. But there's complicated mathematics that has to go into figuring out how you do that and do it at scale. And it leads him to connect with many of the, the senior figures at Bell Laboratories who are so impressed by his work that they are then able to pull him into the laboratories permanently. The war is a, I mean, it changes the lives of everyone in that generation. For Claude Shannon, it leads him from fire control to cryptography, which is an important development in his life. But I, I do think that in a way, without the war, I'm not sure that you get to the 1948 paper, to the, you get to the theory of communication, because he, he could well have gone in a completely different direction.
14:52 Steven Sinofsky? He was at Princeton and it's, it's like good grief, the guy's in his twenties and he's hanging out with von Neumann, with Morris, with, with Einstein.
14:59 Speaker 03 We haven't filled our favorite Einstein story. Claude Shannon is at the IAS in Princeton, giving a lecture on something or other. And halfway through the talk, Einstein, Einstein walks into the room and he sits for a couple of seconds and he leans over and whispers something to someone in the back row and then he walks out again. And then Shannon immediately after the lecture's done, like runs up the top, said, oh my God, what does Einstein think about my lecture? He said, oh, he just wanted to know where the men's room was.
15:23 Steven Sinofsky? Or the tea and cookies.
15:23 Speaker 03 Or the tea and cookies. We've got two versions of the story. I like the men's room one better because it's just that extra level.
15:28 Steven Sinofsky? Then crypto comes up. And so, you know, in hindsight, he seemed again fortuitously or by some higher power uniquely qualified to go after cryptography. You know, the field was completely different when he started it back to analog differential machines and stuff like that. In fact, that like he worked on one of the early realtime systems, you know, SigSally, and that did not look like any computer that we ever would think of. What did he do to change cryptography?
15:55 Speaker 03 So there's a number of things. I think it's worth also being, you know, it's worth sort of level setting where he's at in his life. So he's just finished his graduate studies. His first marriage has collapsed. It was a really emotionally difficult event. He moves to the West Village in New York and he starts going to Bell Labs every day. Bell Labs has gone from, I think, 3,000 employees to 9,000 employees and a lot of the employees in the office are wearing military uniforms. And so this is a really tense time. There's a lot of work to be done. There are basically, everybody is working a six or seven day work week just until, you know, until the war ends. And into this mix steps Claude Shannon with his knack for math, with his boyhood fascination
16:38 Speaker 03 with codes and a kind of facility for code breaking and for code making. And that's what he does for a little while. He focuses on how the US can better encrypt the messages that it's sending to the Brits and he focuses on kind of understanding the fundamentals of cryptography. He writes a famous, a now famous paper that is classified for years. I believe it's called a Mathematical Theory of Cryptography and he proves the existence of a one-time pad, the existence of an unbreakable code. One of the more interesting elements of this work is that it puts him in touch with Alan Turing in what is probably honestly my favorite chapter in the book. He and Alan Turing are having tea every day.
17:16 Steven Sinofsky? Alan Turing's a little older than him at this point.
17:18 Speaker 03 Yeah, but Alan, so Alan Turing's on a billet from the British government to make sure that what the US is doing in terms of the messages it's sending is secure. The Brits were very suspicious that the US just wasn't going to get it right. And so Alan Turing's at Bell Labs and Claude Shannon's at Bell Labs and these sort of two giants of computing meet and these are guys who don't make new friends easily, become friendly and have tea every day. It's just an incredible story.
17:44 Steven Sinofsky? It's amazing to me that all of this is happening at what is effectively a corporate lab.
17:48 Speaker 03 And it is, it's actually a testament to Bell Laboratories that somebody like Shannon is, A, invited to be there in the first place, because it's not like he has a specific job title. He joins the mathematical research group, and they basically go around cherry picking the problems that they would like to solve, and they don't have to do anything they don't want to do.
18:05 Steven Sinofsky? Well, they're the phone company.
18:06 Speaker 03 That's true. It helps to have a government monopoly, a government-backed monopoly. But the truth is, that he has this extraordinary mentor, and the head of that department, in Thornton Frye, who realizes that there are a bunch of academic mathematicians who don't want to stay in academia, and you don't really know what to do with them. And he sort of says, "Well, if you invite them in, and attach them to engineers, attach them to physicists, they can help. They can amplify and help solve problems." And so Claude Shannon is one of these sort of flexible problem solvers. So Bell brings people like him in. The second thing that Bell does, it's like, you know, companies have newsletters today, they have blogs today. Bell is publishing a full academic journal for most of the 20th century, and
18:48 Speaker 03 these are rigorous papers distributed around the country, university researchers read them. And it really is a hallmark of that era, that someone like Claude Shannon, who by day is working on cryptography, on the coloration of wires for the phone service, is also in a place where he can write a 77-page paper that developed an entire field from scratch.
19:09 Steven Sinofsky? So now we're post-war, and we're back to Shannon going on to solve even bigger problems. And the notion of communication comes up, and it's, at this point, it was still rather primitive. I think like, the idea of like, wow, the signal doesn't make it from point A to point B, means like, increase the amplifier, make it louder. And if we all know from the dinner table, screaming doesn't make your point get across. But even the very notions of signal noise, all of these haven't really been formed yet.
19:36 Speaker 02 Right. The solution to noise, the solution to a noisy channel, or distortion, was just to talk louder, brute forcing the problem. And Shannon discusses ways to get around that by talking smarter and in code.
19:48 Steven Sinofsky? But this breakthrough seems, it's not like others, because the problem goes back to the 1890s and the telegraph. And he himself had sort of been formulating it over this 10-year journey of...
20:00 Steven Sinofsky? Of thinking about it.
20:01 Speaker 03 Yeah, he actually wrote a note to to Venever Bush first suggesting that he was working on this, the theory that all messages or all communications were essentially the same. This is 10 years before he ever publishes the paper. It is kind of interesting to think of this idea like marinating in his brain as he's traveling through different parts of his life. The other important point is that he takes, it takes 10 years for these things to crystallize. We tend to want very quick reactions to things. We think, you know, the moment our tweet goes up, if it doesn't get responded to, oh, God, what have I done? For Claude Shannon, this was 10 years, often working at night and on the weekends, thinking, pondering, writing things down, scribbling, and then eventually coming back and dropping
20:46 Speaker 03 a theory that when it was announced, people said it came like a bomb.
20:50 Steven Sinofsky? Yeah, that's actually, to me, it's just very refreshing to hear, like, there's a moment, but it also took many, many years. Sometimes history has a tendency to tell everything. Like, it happened on Tuesday and then the paper gets published. But what was the big assumption that he made in his theory of communication that that really sort of changed everybody's mind?
21:09 Speaker 02 I'd say there are a few things. A lot of the history of information up to Shannon was this question of abstraction. How can we get away from the meaning that any message has and think about messages in a more objective way? How can you measure the information content of a message? And Shannon's predecessors, people like Nyquist and Hartley, had been sort of groping to a solution to this problem, as had many others in the field. But it's Shannon who really comes up with the final formulation of how do you quantify information? What does it even mean to say, how much information is in a book? How much information is in a song? How much information is in a video? Or so on. And then what Shannon does in introducing the bit, which he starts off calling the binary digit, before one of his colleagues comes up with bit as a good abbreviation.
21:50 Speaker 02 What Shannon does is he talks about how we can think about information as a resolved uncertainty, and how we can think about information probabilistically, and how we can use these tools to actually calculate information in a really objective way. And then, once we do that, once we can actually do hard science with our messages, that enormously simplifies the problems of compression and accurate communication. But this first step, getting past that semantic level and getting to the objective quality of information, I think that's a key in the whole paper.
22:23 Steven Sinofsky? You know, we do look back and go like the bit came out of the paper. At the time, was that like viewed as sort of a key innovation? Or when people had the elevator conversation about the paper, what was the elevator conversation?
22:36 Speaker 03 That he had set some outer limits for what engineers would would try to do. That the paper was a model of clarity and concision, that he had explained all possible forms of communication in a sort of single diagram. And that he had done it all without anyone knowing, that he had no collaborators of any of any kind. And that it was published in two sections within the Bell Technical Journal, the Bell Systems Technical Journal. What we take from the paper only starts to matter in the 80s. And so at the time, this was still a discussion very much in theory, but the power and force of his theory was was immediately seen.
23:17 Steven Sinofsky? One of my, the things that I loved that I didn't know was that the first two papers were called A Mathematical Theory of Communication. And then when he went to write his own book, he changed it to The Mathematical Theory of Communication. What actually transpired in the interim there?
23:33 Speaker 03 One, he must have gotten some pretty good feedback on it. You know, this isn't just a theory, it is the theory, that he had stumbled on the big one. The other thing is the, I think also the intervention of Warren Weaver, who comes on as Shannon's co-author for the book, who is sort of a science popularizer. And he's also someone who's very, like Bush, he probably doesn't describe himself as having a first-rate scientific mind, but he's someone who loves literature. He collects translations of Alice in Wonderland. He supposedly can identify wine varietals by tasting. So he's just a sort of Renaissance man. And he comes on encouraging Shannon to take this to press and writing a section of the book that's sort of a layperson's explanation of what information theory is. People think that Weaver, in some ways, is a co-originator of the theory.
24:16 Speaker 03 And he was always in a hurry to downplay that and to say, no, I just was the popularizer.
24:20 Steven Sinofsky? One of the things that's interesting, too, is that, you know, he's 32 when he writes the the paper in 1948. And but all along, people like Weaver played an important role in helping him like basically finish his work.
24:32 Speaker 03 He was very careful about the people that he let into his orbit. So part of that is just he was a natural introvert and he kind of kept his own, he sort of kept to himself. But his friends were just brilliant people. In 1948, after the paper is published, Probably the most significant person in Shannon's life enters his life. That is Betty Moore, who is actually, her title is computer. She is a computer at Bell Labs. And what that meant was that she was helping engineers do math. She herself was a Phi Beta Kappa graduate of what is, what was at Douglas Women's College, what is now Rutgers. She's got a lot of talent. She publishes herself. She's a musician. And Shannon, shy though he is, like starts talking to her. They start dating and they're, they're a match right away.
25:15 Speaker 03 People just understand that when they're together, there's a different kind of connection. And they connect interpersonally, but they also connect mathematically. She does help him complete his work. Shannon was the kind of person who would see solutions in his head. And so he would think about problems and then see the end state. And he wasn't actually that interested in like explaining to other people how he got to that end state. And I guess when you're as smart as Claude Shannon, you're kind of allowed to get away with that. But Betty Shannon understood that in order for Claude Shannon to have the kind of impact he was going to have, the work would need to get finished. So she would actually sit with him. A lot of his earliest papers are in her handwriting. And she would, she would do the math, do the intervening math. She would challenge him. She would include historical references, literary references in papers.
25:58 Speaker 03 And she never got any credit for this. And hopefully, you know, our work starts to restore that a bit. But, but Betty Shannon is one half of what I think of as like one of the great creative partnerships of the 20th century.
26:10 Steven Sinofsky? Yeah, I think that's, I mean, I think that that's just a fantastic point is certainly part of the times we learned, we know the same thing about Pierre Marie Curie and the same thing about Einstein and Mileva and so on. Amazing foundation gets laid. And this career, he won many awards that he'd ever seemed to seek out in his later years. He ended up meeting like many of the progeny of the computing era. At one point he ends up meeting Steve Jobs. How did that come about?
26:38 Speaker 03 So this is a story that was relayed to us by, by Claude Shannon's daughter. And it's, it's an extraordinary moment. They're both Steve Jobs and Claude are the recipients of honorary degrees from the University of Pennsylvania. And after the-
26:53 Steven Sinofsky? When is this?
26:54 Speaker 03 I believe it's the 1980s. After the ceremony is over. People are sort of milling about the quad. And if you can imagine Claude Shannon at this point by the eighties, because his work has started to actually be implemented. I mean, he's won a national medal. He's a revered figure. So there's a crowd around him. People want to shake his hand. People want to be around him. And Steve Jobs is well known, but not as well known as he is now. And so Steve Jobs actually has to, he goes into this throng of people and elbows his way into this audience with Claude Shannon. And he has to, he has to try to meet Claude, not the other way around. So he gets up to him and he shakes his hand and he says, you know, Dr. Shannon, it's a real, it's a real honor to meet you. You know, my, my name is Steve Jobs. I work at, at Apple Computer. And Shannon looks at him and says, oh, that's great, Steve. It's nice to meet you.
27:39 Speaker 03 What do you do at Apple? And it's just incredible moment. Steve Jobs actually sends the Shannon's, he assembles an Apple II and sends it to Claude. So they have one of the only Apple II's, I guess, that was assembled by Steve Jobs himself. And it's a real point of pride later for them. It's a kind of funny moment in, in computing history that these two giants met.
27:58 Steven Sinofsky? I really wanted to pull a couple of quotes about him from the book because I, I found them just so telling. And also to help us to reflect on our own culture. You know, one person said, he never argued his ideas. If people didn't believe in them, he just ignored those people. So that has that, that sort of Silicon Valley feel to it.
28:17 Speaker 02 Right. He, as Jimmy was saying, that Shannon was the sort of person who was really occupied with doing interesting things, whether it was tinkering in his workshop or thinking through interesting math problems. And if you weren't going to support that, he would very politely excuse himself and he'd go back to his office and work on whatever he was working on. Or he'd start unicycling down the hallway to get away from you.
28:36 Steven Sinofsky? Yeah, I love the unicycling only because that seems to have roots in PC era as well.
28:40 Speaker 03 Claude Shannon, to my mind, one of the most interesting things about him is he just spends his entire life pursuing the problems that interest him most. And then the moment that he's taken them, as far as he'd like to take them, he goes on and chases a different problem. So it's interesting, right? Because he could have continued to trade on information theory for decades. He had the opportunity to be a scientific celebrity. He was in, he was profiled in Vogue magazine. I mean, he had a dapper suit and everything, the cigarette, the whole deal. And he just, he sort of walks off the stage, but pursues artificial intelligence, then pursues robotics, goes and builds a chess playing machine, builds an artificially intelligent mouse that can navigate a maze. I, what I find inspiring about him is this is someone who had lucrative.
29:26 Speaker 03 Prestigious options and almost always went for the problem that interested him most.
29:31 Steven Sinofsky? Well, I really want to thank you so much for joining us. And so this was Jimmy Soni and Rob Goodman, co-authors of Mind at Play: How Claude Shannon Invented the Information Age. Thank you very much for being here.
29:41 Speaker 03 Thanks, this is great. Yeah, thank you for having us.
Previous Company Building in Crypto: A Conversation between Brian Armstrong and Chris Dixon Next What's Next for Education Startups Page updated November 13, 2026 · transcript and summary are machine-generated