mary fallon's must-read political books, former senator david bourne on his a letter to america, also rare books from gal lay key and others from the history of science collection at ou. and sunday at 5 p.m., oklahoma history on c-span3. a look into african-american life in 920s oklahoma and native american artifacts from the special collections at the oklahoma history center. once a month c-span's local content vehicles explore cities across america. this weekend from oklahoma city on c-span2 and 3. >> george dyson talks about the team led by john von neumann at princeton who built the first computer and discusses the impact of that work on other fields of science. this is about an hour, 20 minutes. >> finally, before we get started i do have one very, very special introduction tonight. i'd like to recognize on behalf of george akrevoe emmanouilides, hired as a secretary at the age of 16 to work for herman gold stein when he was working at the moore school, and when herman gold stein moved from the project to ias to work directly with john von neumann and that incredible team there, the akrevoe went with him and had a very distinguished career doing far, far more than simply being a secretary. she's traveled to be here for this program. akrevoe, would you, please, stand up? you're so tiny, akrevoe. i saw people in the back sort of craning to see you. we're delighted you're here, and we may even take a few questions from her, so if you'd like to ask questions of someone who worked directly with the legends, please, feel free to pass those up. and now to tonight's program. more than seven decades ago seems distant and inaccessible in many ways, and yet you are the's cathedral: the origins of the digital universe, makes it as vivid and relevant as today. indeed, it could hardly be more relevant. the cathedral is governed, powered and driven by mere variations on the early code in the machines they envisioned and built. it would be simplistic to say that in von neumann's case the stories are important because as dyson writes, the digital universe and the hydrogen bomb were brought into existence by the same team, von neumann's team at the institute for advanced study at princeton. what is far more to found, however, and what we will also deal with tonight is the legacy and the implications of those fateful days, as if hydrogeneral bomb itself is not fateful enough. and at the center of it, the man: john von neumann, inventer, towering intellect. the physicist edward teller described him in this way: its members will resemble johnny von neumann. this is not george dyson's first attempt to help us understand at both a technical and human level the way we coexist and cocreate with computing. he is the author of darwin among the machines as well, two other books also, and he writes and speaks frequently on this subject, and this is not totally his day job because as you may know, in addition to being a science historian and author, he is a boat builder and designer from his home in bellingham, washington. ladies and gentlemen, please, join me in welcoming george dyson. [applause] >> thank you very much. >> welcome. >> thank you. >> so glad to have you here, george. >> great to be here. fantastic exhibits. they're just incomprehensibly good. >> well, thank you. thank you very much. we're delighted you're here, and you're among friends. these are your people, george. [laughter] so we're going to have some fun tonight. let's talk about, first, the process of writing the book. now, you have an intensely personal connection to princeton and the advanced, the institute of advanced study because of your father, freeman dyson. >> and my mother. >> and your mother. >> my mother was there first. >> absolutely. so talk a little bit about that, what it was like to grow up and be among people at isa? >> well, i've got to be careful what i say, but for a child, for a young boy it was not that interesting a place because it was populated mainly by theoreticians who, you know, would work on pencil and paper all day and go off in the woods and talk about physics, and maybe the most exciting thing were the chalkboards, they still used chalkboards that the time, no powerpoint. but this was there out build anything the back where julian big lowe was building this machine, so i spent a lot of my time poking around and taking things apart that was discarded as scrap from the project. so that was -- >> there's a famous story about your babysitter being einstein's secretary. [laughter] >> yeah. people, behind every great man, this is somebody who keeps track of things, and that was helen who was a fantastically intelligent woman. she was sort of, she was einstein's search engine -- [laughter] when einstein needed something, it was helen who knew where everything was. but she didn't have her own children. she grew up in a family, i can't remember, it was 11 children? a huge number of children in her family. and she missed that, so she adopted our family. i had four younger systems, she really was their babysitter, and my job was to make her life difficult while she was trying to watch my sisters. [laughter] but i owe her a tremendous lot because i was really being difficult one day, and she said why don't you stop, settle down and read a book. and i said there's no books, i've read all the books. and she went to the shelf and gave it to me and said read this. and it was con tick key, and that was the first adult book that i read and, you know, that changed my life. >> that really did change your life. >> and she had the perception to see that. i don't think that was an accident. i think she was, i'm going to give this kid this book. >> so hanging out with this crowd as a kid, did it become a natural thing for you to be curious about who they were and about von neumann in particular and the meaning of computing? >> yeah, that came later really thanks to esther, my sister who had such an influence on the community here. thanks to esther, i started going to her technology conferences in the early '80s, and i saw this whole world, you know, the world of the personal computer was flourishing. and then i realized that that came from this, you know, this outbuilding behind where i had grown up. and i wanted to understand that, so that's when i became interested in sort of going back and finding out what really happened. and i have to say outright that i'm less interested in who was -- not trying to find out who was first than trying to find out what really happened. >> in fact, you specifically say this is not a book about firsts. >> not the fest electronic computer, it's not the first program computer. you can almost say maybe it's the first computer with a fully random access memory, but even then there was a couple that were first. so it's the first of nothing, but it was the ancestor in the sense that it was the one that got copied. just like, you know, newton was the first iphone but didn't become the iphone. >> we've learned the hard way that between the word fist and the word computer -- first and the word computer, there are about 19 adjectives. >> yeah. don't say first and don't say invent, and you're safe. >> you told me a fascinating story about the treasure-trove of papers that you were allowed to get access to, some that hadn't seen the light of day since the mid 1940s. can you talk about that? >> yeah. there were a number -- i mean, the reason the book exists were people kind enough to let me into their garages, their basements, in the case of the institute, sort of their own archives. the institute for advanced study had been very protective of their, essentially their privacy, they're a private organization. and thanks to charles who's one of your benefactors, it was really charles who pushed the door open, said let george in to have access to this stuff, and i found unbelievably amazing things. i mean, just -- you know, for someone like me it was just unbelievable to go there. my daughter who came with me, you know, she knows every day i would poke around these documents all day and then come home and go back in the morning. because i normally work in archives where i've got 48 hours, and i'm sleeping on somebody's couch, and i have to beg a photocopier, and it was amazing to have that access. >> you write about an amazing piece of paper that you found which looks like it was torn from a note pad and crumpled up and thrown away, but then somehow retrieved and uncrumpled. >> where yes. >> talk about that. >> this wasn't -- in the institute of archives, this was in the papers that julian bigelow who was an engineer who, like most, saved things. when he died, his family allowed me to sort of go through his papers. and in there was a scrap of lined paper that on the top said let a word be 40 bd, so that's like 1946, 40 binary digits and a command and an address. ten bits for the command, and so to me, that's like the tablet of moses saying let there be the command line. [laughter] and from the rest of the world, you know, came from there. you know, where did these things start? and tomorrow someone may say, no, i've got a piece of paper in my basement that's 1943. which is quite probable. eckhardt and lockly may well have -- i mean, certainly did produce the same idea probably earlier. >> lots of amazing things come out of basements and garages. >> yes. >> have you found some -- this is the last question about the process. akrevoe is here tonight in part because you encountered a number of amazing people who were there at the time and who you could talk to as original sources. >> yeah. i was a little late. i should have done this ten years earlier. but there still were enough people left for me to get some firsthand -- and i relied a lot on the oral histories that were done by the babbich institute. >> you are the's ca -- touring's -- you are turing's c, how did you come to call this book? >> for context, it's 100 years ago, sara turing is five month bees pregnant. young alan is just about to come on the scene. so i, you know, one of the good things about turing is he left very few papers, so you can read everything he wrote. with von neumann it's hopeless to try to read everything he wrote. so i read everything that turing wrote, and in 1950 he wrote this, you know, tremendously famous paper, as famous as his 1936 paper on universal computation, and this was a paper about artificial intelligence. and he could see the critics coming, it's like intelligent design that, you know, if you're going to create intelligent machines, you're playing god, and we shouldn't go there. so he made the statement that when we create these intelligent machines, we are no more creating souls than we are in the process of creating children. we are simply creating manages for the -- mansions for the souls that only he can create. and i love that sentence, that phrase. and when i went -- 2005 i went to google, i got invited in because that was 60 years since this project began, so they commemorated that. and the engineers there gave me a very deep inside tour of what was going on. and when i walked out of there, i just, i was stunned. they were really, truly doing everything that turing had imagined, building a large -- a machine that would answer all questions they could ask in a nondetermine fistic way. and i thought, this is not turing's mansion, this is turing's cathedral. and, boy, that phrase, that became the title of the book. >> yeah. >> and the second level is simply that the ca three drag is built by large numbers of anonymous people whose names are not remembered, but the cathedral remains. >> over many years. >> and those are real people, like akrevoe, who did the work. and, you know, the cathedral wouldn't be there without every stone being there. >> you are thing and von neumann did overlap at princeton. >> yeah, for two years. >> how much is known about the interaction between turing and von neumann? >> we know between 1936 and 1938, and at that time the institute has nothing to do with princeton university. people say at princeton, you think it's the university like the hoover institute is at stanford. but at that time they didn't have a building, so they made a deal with a math department, so the institute actually was sort of parasitic on the math department. so turing and von neumann, even though turing was at the university, von neumann was at the institute, they had a lot of contact. and turing, that's where he corrected the final proof of his great paper. so he certainly had an influence on von neumann. what's sort of dark is what happened during the war when von neumann went off to england to work with the british, and turing came to america to work with the americans, and that part is still sort of a black hole. but like a lot of the stuff from plexly park, it just may take a long time and eventually come out. >> do you have any suppositions about any of that? >> i think von neumann was in england working on the nuclear work because the british made a lot of contributions to los alamos, a lot of the ideas came from britain, and i think von neumann went over this to sort of jump-start some of that, but i think von neumann was so good at everything that i can't believe they didn't bring him into the cryptography question, and he came back -- one of the forgotten things von neumann says explicitly is that he got, he actually credits his ideas about programming to a visit he took to a computational laboratory in england. he says that in writing. which lab -- you know, it was, he doesn't mention any of the e enigma work, but, of course, you wouldn't mention that. >> there are a couple of exchanges in the book, maybe many, where you cite discussions that von neumann is a part of with colleagues there at the ias where von neumann expolice sitly says, yes, that's one of the things turing is working on. it seems he was aware of turing's theories as he was working on his seminal paper, and it was already -- the implication is it was already having some influence on his thinking about this whole area. >> yes. yeah, it was, i mean, von neumann was a mathematical lo jig, and he followed that stuff very closely. i decided to do a little physical research rather than just speculation, so i went and found von neumann's copy of turing's paper. and it's in the institute library now in one of those shelves which you have to turn the cranks to open them up because nobody goes in there, and there are all the volumes of the proceedings of the london mathematical society, and they're all there with perfect bindings, all intact, and then there's one volume, volume 42 with turing's paper in it, and if you take it out, all the pages fall out. it's complete hi disintegrated from being read so many times. so i think that's pretty good evidence they read that paper. [laughter] >> so let's talk about von neumann. well educated in budapest, goes to berlin, joins the academy, is a professor. the nazis begin to dismiss jews from german academies. he resigns, leaves. and he's appoint today the faculty at princeton. and as he goes to the ias, he encounters really a remarkable group who were already there and really a remarkable intellectual atmosphere. oppenheimer called it an intellectual hotel. >> right. >> so talk about the ias as von neumann would have experienced it. >> well, the thing we forget, most people remember the institute for advanced study because of einstein and the nuclear physicists and the string theorists. people forget about the institute also had a very strong school in the history of art, a school of classical -- i don't know how to pronounce it, studying greek epigraphs and a school of archaeologists. so there was all this other culture there. and, of course, akrevoe can tell you, people like homer thompson really great, he was the model for raiders of the lost ark. and the art historians. so all these people were there. it wasn't just math and physics, it was a very rich place. and oppenheimer comes in -- someone invited t.s. eliot. in fact, the position i had there was created for t.s. eliot there of sort of the strange outside artist who's allowed to do, you know, comes in, does something that doesn't belong to one of the schools. now they have a school of biology, so it was a rich place. about von neumann's coming in the first place, he didn't come alone, he came with eugene rigner. and at that time princeton university was not hiring jewish professors, so they couldn't really hire von neumann flat out, but they found a loophole. there was no problem, they could hire two hungarians halftime. [laughter] they couldn't hire one hungarian full time, but they could hire two halftime. so they offered johnny von neumann and eugene rigner which to them was ten times what you could make in europe, so they said, yes. that's how they got both those guys at once. >> then we're going to skip ahead just slightly because i want you to talk a bit about how he wound up in los alamos. >> yes. sort of everybody wound up in los alamos. [laughter] >> yeah. >> there was a, you know, almost a special train from new jersey to loss hall most. los alamos. von neumann didn't stay there, he was trance gent. he had so much going on. he went to los alamos, you didn't leave until the war was over, they couldn't have people go and leave. but von newman had this sort of special pass where he could go in and out, and he had a tremendous influence. los alamos was the prototype of all these great laboratories, and there was a deal made, though, that was very explicit that oppenheimer made, sort of a deal with the devil. we will build you this bomb, but you have to let us -- you know, we won't tell you how to use the bomb if you don't tell us how to do science, and we'll be free to do all the science we want in our spare time. and that's why so much good science came out of los alamos. >> and he was both incredibly challenged and invigorated by the intellectual process of trying to think through the very complicated problem. but he was also increasingly as the project moved forward deeply troubled from an ethical standpoint as well, wasn't he? >> von neumann? >> yes. >> yeah. again, that's what i discovered from the most remarkable body of documents i found was in marina von newman's basement. -- von neumann's basement, it's always next to the water heater, filing cabinet. known papers went to the library of congress, but this filing cabinet didn't go, and in the bottom drawer was all the handwritten correspondence between johnny and his wife from 1937 to 1960, so that gives you a day by day, firsthand picture of what people were really thinking at the time. >> because he was using that as just a means of getting this out, right? be his corps -- his correspondence with his wife? >> it was like e-mail is to us. but he would have a full day of meetings and solving problems and still write 16 pages in fountain pen to his wife, and they had a difficult marriage. they were always in different places because she was doing the coding for these early bomb calculations, so she might be working in aberdeen, and he would be in los alamos or oak ridge, and these letters went back and forth. and they were half in hungarian, but my friends in cambridge translated all the interesting hungarian. >> did you know that this secret trove of letters -- >> i had no idea. i couldn't have imagined it. to me, the most interesting period of american history that period from just before world war ii til put snuck. -- sputnik. after sputnik we have a very good record, but there's a period there it's really not clear what people said -- the oppenheimer trial is pretty good because they've got everybody under oath, but clary describes the day they sort of closed the mousetrap on oppenheimer and what everybody's reactions were that day. you're not going to get that anywhere else. >> how did you come apart that stash? >> thanks to marina von neumann who knew that i was doing this project, and she finally said, well, you know, maybe you should come to ann arbor and look at this stuff. to her, it was awkward because this is the woman that her father left her mother for. she didn't really want to look. very, very personal letters. but she trusted me to go through them and take out what was, you know, what was useful for the history of computing. and it brings the book to life. i don't think there would be a book here that's alive without her voice. it's something about having english as a second language, she just -- you know, they drive, 1940 von neumann gets invited to give lectures in seattle, so they driver in 190 across the -- 1940 across the united states. she records that. it's route 66 and stopping at gas stations, it's all there. >> her notes, her personal accounts are just riveting. >> and she committed suicide at the end, so she kept a journal right to the very end. so she deserves a book of her own, but it's not for me to write but, hopefully, someone else. >> so in this incredibly pair pa tettic life that von neumann is leading in 1944, he encounters ec hard and mockly and their work on the inyak for the first time. talk about that encounter and then what happened. >> yeah. eckhardt and mockly were way ahead. they had built the inniac. there's no doubt they built it, no one argues whether they built it or not. it's a clear case of something that was first even though it can be traced to other people and to, you know, have jane smiley tracing it back to -- [inaudible] but the iniac, and von neumann always said it, it was just a pioneering thing. because he was scientific adviser to the ballistic research board, he got to see it. when he saw it, he just immediately saw what it could do. and i think part of that's visual. and iowa yes slow can tell you. when it's run, you're seeing the bits, you're seeing the lights moving around. you can see the numbers and, you know, you're within the computation. and he had that kind of mind, you know, visually-thinking mind. so i think the moment he saw that, he could see all this other stuff coming. >> so in the way it was physically built, you're literally standing inside be it. >> and the thing we forget forget is the iniac was very advanced, it was a multiple core processer, it was like 16 core on a chip where you divide the computation up into a parallel path. >> the, herman goldstein who was instrumental working with them says that when he saw it for the first time, when he saw the iniac, it changed his life forever. >> forever, certainly, yes. >> are there von neumann's own words that express what he was feeling? because suddenly everything that had been theoretical for him now becomes physical and real. >> yes. he -- there are words. and like, for instance, in these letters to his wife, i mean, he fell in love with the iniac. he saw that machine, he knew what it could do, and then it was this very wonderful thing where he taught his wife how to program it. because she saw -- well, it's a debate who saw what first, but somebody saw that -- they were thinking, they were designing the institute machine and were writing code for it, but the machine wasn't ready, and then somebody said, well, we could rewire the iniac and run these stored programs on the iniac, and clay went down and spent six weeks rewriting it, and that was such a fruitful time. so there was a mixture of their marriage and the coding and the bombs, it's an amazing period. >> and she had no, no training whatsoever for any of this. >> no. well, no, she did have training. that's the interesting thing. during the war when johnny went off to england to do what, we don't know, clary was left behind in princeton, and everybody was trying to help with the war. she got a wartime job. women could just apply, and her wartime job was with the center for population research at princeton university. and her job was modeling populations of people. what would happen if you created, you know, a new state in the middle east or something, you know, so they modeled these growths of populations. and that's what she needed to solve these early problems. early monte carlo is all studying the population of neutrons, and phishing is like having two children, and escape is like emigrating, so she learned, she taught herself the math mathematics of population statistics, and that ended up being what they needed. it was a strange accident. >> there's a point of controversy in history about von neumann's relationship to the intellectual work that eckhardt and mockly were doing, and the paper that they produced which was the design paper and whether von neumann wanted to circulate that, was more of an open source tide to use -- guide to use a modern phrase, didn't believe that this secret because he understood the power it could unleash should truly be secret or whether he played by the rules, so to speak. what do you think about that? >> well, he certainly broke the rules. von neumann broke the rules, and this is very controversial. there is no doubt that, you know, von neumann did not write that whole paper. they certainly weren't all his ideas. he put them together. there's some debate what parts of that paper goldstein wrote and what parts von neumann wrote. there's no question that a lot of the ideas came from eckhardt and mockly, and they should have been co-authors in the modern standard. but the paper was released under john von neumann's name, and he, you know, all i'm trying to do is find out sort of what the truth is, and the truth is that, yeah, it was released, and then it was considered a publication, so it voided the chance of patents on the iniac. all i can tell you, you know, in terms of a smoking gun is that in 1945 von neumann signed a consulting agreement with ibm. so he was already -- so this result was highly favorable for ibm that there were no patent restrictions on these ideas. but i think from von neumann's point of view, i don't think he was out to do anything unscrupulous, he just thought this would be for the good of everybody, and i think you have to remember that they were all coming out of world war ii where all these groups who later -- during the war they were all collaborating, rca and eckhardt-mockly, they were all cooperating together. people from manchester were coming to princeton, people from princeton were going to manchester, the sort of arguing about what should get credit came later. >> and they do all come out of world war ii. in fact, the war ends, von neumann is headed back to ias, wants to transplant the entire team and take them all and continue the work, and eckhardt and mockly decline. gold steven goes, but eckhardt and mockly decide their going to pursue -- >> yes. originally, bigelow was supposed to be eckhardt. eckhardt declined. so there's a lot -- and people went back and forth. there's a great letter from johnny to his wife saying they've shanghaied two of our guys, and we've shanghaied two of theirs. just like what goes on here -- [laughter] they're always stealing -- >> exactly. >> and so in the end it became very bitter. and i, again, you can read some documents in this book that are pretty incriminating that the eckhardt-mockly company was doing quite well p they had a contract for three machines that would have put them firmly on the path that ibm took, and then they had their security questioned, and they lost the contract. there's a disturbing memo describing what had happened because i don't think they were a security risk, but it really put their company off the path forward. >> and von neumann was bitter about their decision, wasn't he? he was very critical of this vector. >> he was annoyed. i mean, he just wanted to go full speed ahead x he felt they were holding things up. and they would feel differently. so i don't -- i try not to take sides, but i think both, there's truth on both sides. >> so in 1946 he puts that aside, he does go back to i ark s, and they -- ias, and they begin working on this highly-improved successor to the iniac called the maniac. >> mathematical and numerical integrator and calculator. that name was then adopted by -- [inaudible] >> yes. the air of the acs. >> yes. >> he says some interesting things. first of all, he's very practical. he says we're not going to originate anything -- >> right. >> we're simply going to with the state of the art as it existed at that moment. and that meant vacuum tubes and crt memory as it existed at the time and 15 tons of air-conditioning. >> yes. and magnetic wire input/output, and they did -- they innovated a lot of things. one was, you have to remember at that time you could not modify ibm equipment. it was like the old telephones, you couldn't change it. but they did modify ibm, they got some punch card machines at that time still read on the 12-bit side, and they converted it so it would read on the 80-bit side, and that's why we have 90-character line -- 80-character lines. >> was that their innovation? >> no, that was a guy named hue let crane who moved to stanford, he just died a few years ago. no, he did that alone and got in real trouble, and then the ibm people came and said, well, wait a minute, we could probably sell this. and they did very well. that's what put them in the data processing business. >> francis -- [inaudible] the british author, had a great review of the book in the guardian of london today, we were talking about it earlier, and he says no other book brings to life anything so vividly or appreciatively like the immense engineering difficulty of creating electronic logic for the first time. talk about the design, the team, the plan, von neumann's drive to do this. he was just tireless in bringing all these theories to life. >> yeah. and he was trying to do the almost impossible. anybody looking at this, what they did was really crazy. they're making this 40-bit, parallel machine where one bit of every word is in a different tube, and these tubes are asill scope tubes that if you walk past, you know, wearing a wool sweater, you might throw of the three bits off, and nothing worked. yet they got it to work in new jersey which is like the least has pit call environment for delicate electronic equipment. >> you tell the story about every time a car went by, it would wipe out the memory and they'd have to start over. >> yeah, thunderstorms -- [laughter] they were so persistent. so the fact they got this thing working as well as they did is quite amazing. >> you use the faze a deal -- phrase a deal with the devil talking about los alamos and the atomic bomb. there was another deal with the devil made in the development of the computer that von knew moon wanted to build, wasn't there? >> yeah. i feel, and i think this is still -- this is sort of a fable for the future -- but the deal was made that, you know, it's a metaphor, but the deal with the devil was made that the devil could have this weapon that could destroy all life on earth, and von neumann and the scientists would get this computer that would reveal all knowledge. the this incredible trade. like give me everlasting life, and i'll give you my firstborn child, sort of thing. and we think that, you know, kind of we won the deal because at that time it's incomprehensible to us today how real the threat of worldwide, global thermonuclear war was at that time n. the 1950s it was a 20-minute launch window to destroying the world, and we survived that. we don't really worry about those bombs like we used to. so it seems like we got the better end of the deal. but i think what you have to remember is that computers could be equally threatening. maybe the devil is out there saying i didn't really want the bombs, i want the computers. and that's what i think we need to be watchful for, that we do not let this global computing network that is so beautiful, it is a cathedral, but make sure that it does not become the tool of, you know, some totalitarian maniac. >> in fact, you talk about that, and i want to get to that in a second, but it was interesting -- interesting's not the right word, it's incredibly compelling that von neumann could see both move anything parallel. >> yes. >> having worked in los alamos, he could see what the net result of the hydrogen bomb was likely to be and at the same time had this premonition of what computing taken to it completely far extent could also turn out to be. >> right. he did not foresee the internet. >> no. >> give him credit for everything, but not the internet. >> no, i understand. [laughter] and then they devise what you call a real problem and the perfect cover. so they're working on thermonuclear explosion models, but they're also working on strategic weather -- >> on meteorology. they did need a cover for this work, and meteorology was the perfect cover, so he brought in real meteorologist, and anytime you check your iphone and get a five-day forecast, it's the same codes they developed right there, just with better input data and infinitely more processing power. >> the two applications that seemed to catch on at exactly the same time is as computing power increased, weapons and weather. >> their both hydrodynamics, that's really what it is. >> so you talked about clary learning to write code for the ini, ac and this, to me s another one of the great innovations of the book. as they're developing the monte carlo algorithm im, von neumann is coaching her further, and be she's getting involved in writing code for monte carlo. >> yeah. >> has that been known before, that clary von neumann was writing code for the monte carlo? >> maybe if you really looked. what's interesting, some of these codes -- like, there's one of them in an envelope. it's an envelope that you could mail with two stamps, and it's a code, you know, what we would call the source code for a hydrogen bomb, so it's the opposite of today. this code would run, and it would run on iniac for six weeks to get, essentially, a one-bit answer, sort of a yes or no answer. whereas now, you know, in a moy cosecond -- microsecond your screen is refreshed, and you've used up a zillion bits. so there are very different kinds of codes, but i think incredibly important. and in monte carlo, i mean, the perfect example of, you know, ulam sort of invented it while he was recovering from are a brain virus, and they told him don't think too much, so he started playing solitary, and he realized, oh, we could do computing this way following random paths. [laughter] >> how did monte carlo, which is an incredibly sophisticated way of going about writing software, how did that happen so early in the evolution of computing when the machines were so primitive and memory was so small? >> well, they needed it is why it happened. they needed to follow these populations of neutrons, and they didn't have the horsepower to do it in an analytical way, so they had to do it in a thattist call way. -- statistical way. the beauty of monte carlo, it's not an approximation. it actually is closer to the way physics really works. physics is, at its essence, not deterministic. >> can you explain a little bit about that? there are a lot of people who will know about monte carlo, but there may be other people watching or listening who don't. >> yeah. it's not that easy to explain, but instead of trying to get an exact answer, you, you sort of develop a game of chance that approximates the problem and run that game of chance. the more you play, the better your answer gets. like if you were gambling in a casino, you know, if you gambled a really long time, you would get an accurate estimate of how, what the take of the other side is, you know? are they taking 3 be % or -- 3% or 2%. and it's what is so beautiful, again, couldn't have imagined it was true, but johnny and clary von neumann, they meet in monte carlo, the casino, right? but johnny has gone there, he has a system for roulette x he's lost all his money -- [laughter] and his first marriage is, you know, he's still married to marina's mother, but he goes over to the bar, and there is clary whose husband is a compulsive, addicted gambler, and she's unhappy. and he knows her from childhood. she was this, you know, very attractive figure skater. and he buys her a drink. or, no, she buys him a drink. he has no money. >> because she had the money. >> yeah, she had the money. [laughter] so they met, they meet in monte carlo. you couldn't make that up. >> yeah, yeah. it's a great story. it absolutely is. so work begins in '46, and then in the summer of 1951, a team from los alamos comes to princeton. they load a very large thermonuclear calculation into the maniac, it runs for 24 hours without interruption -- without interruption. >> for six weeks. >> for six weeks it runs -- >> or 60 days, actually almost more like eight weeks. >> yeah. were they flabber gatted? were they confident? what was the reaction? >> again, nobody was supposed to talk about it. we're not even supposed to know the maniac was working. we're supposed to know it was then publicly dead waited -- dedicated with an ice screw upture and -- sculpture later. that's when we were building ivy mike, the first big hydrogen bomb. and so there's a few people left, harris maher is still alive who was there and -- [inaudible] who died not too long ago. but people were, you know, they were not supposed to talk about what they were doing. they were just testing out the machine, but actually they were running a real problem. you can tell, and akrevoe would probably know, you can tell by the dates at which nick shows up. >> [inaudible] >> yes. they -- >> [inaudible] >> we should get her a mic. >> [inaudible] >> we'll have to get you a mic. akrevoe's telling how nick me -- came, and akrevoe thought that nick had a girlfriend in new mexico. [laughter] >> she just said he was so handsome. [laughter] and then the hydrogen bomb is detonated 60 years ago this year, november 1, 1952, in the south pacific. von neumann writes about knowing, and he uses the phrase that they were creating a monster. but then he also goes on to say that he felt it would be unethical for the scientists not to see through to the end what they knew they were capable of. and i think it's very interesting that he juxtaposed ethics in that way. the monster being created, but the scientific obligation to see it through to it end. is that characteristic of the von neumann that you've uncovered doing research? >> yes. and of stan too. they had this argument repeatedly, and he would -- his answer was, no, it's physics. we have got to do the physics. and we've, you know, we've got to know what happens. if there's a way to know what happens at those energy densities, it's our job to find out. it's not our job to say whether it's good or bad. >> the very next year the soviets detonate their hydrogen bomb. >> well, their -- yeah. it wasn't very successful. so they turned out, just like with the germans, they were not as far along as we were afraid of. but the thing to remember that von neumann at los alamos, he had had been working with claus fuchs, and then he found out that klaus fuchss was a russian spy, so the russians had all, as much knowledge as we did about -- and might well be pushing for the hydrogen bomb. >> many and less than four years later, von neumann dies of cancer. >> yeah, tragically. >> and the team scatters. >> pulls the plug. >> yeah. the computing team at ias, because the plug yets pulled -- gets pulled, it does not last. it, essentially, collapses. >> right. i think it's the setback of the decade. they had this group going that was doing scientific computing for scientific purposes which ibm picked up on, you know, the research center at yorktown heights and stuff started doing that, but there was a gap in between that was lost. but it's understandable, too, why the institute didn't want to become a computing center. it was von neumann who kept it going, and when he was gone, it was over. we've got audience questions? >> we do. i'm going to -- there are a lot of really good questions here. let me ask you just two or three things before we get to these. let's talk about the implications of all of this because you talk a lot about the implications of where computing is today and where it's going in the book. you said about a week ago that the last time you checked the digital universe -- let me just make sure i've got these numbers right -- is expanding by two trillion transistors a second in processing power and five trillion bits per second in storage. >> right. that's like hard disk storage. >> right. von neumann had predicted a universe of 10,000 switches, i think, was what he -- >> well, he said that was all you needed for a computer. they didn't know about transistors, but 10,000 switching units would be enough. >> so with this unleashing, there are three things that i want to coffer. one is artificial intelligence. you write, when von neumann spoke of computers, he never talked about artificial intelligence, and turing talked of little else. >> yes. >> so talk about that dichotomy, the two of them and where you personally believe, and you write about it in the book, ai is headed as the result of all of this. >> yeah. i'm more on the turing side. >> exactly. >> i love speculating about artificial intelligence. he never published anything until it was perfectly proved, you know, he spoke in perfect, complete sentences, you know, and turing was very much the other way. just sort of stuttered and said what he thought. and so they were just very different characters. i think, and that's the tragedy of von neumann's death. he was interested in artificial intelligence, but he didn't want to publish anything until he had a complete theory of it, and he never got there. turing died at age 41 and von neumann at age 53. >> is what we're seeing now the approximation of artificial intelligence as they, turing, might have thought of it? >> i think it's oddly close to what turing was looking at. people remember turing's 1950 paper, the one with the imitation -- and they remember his 1936 paper, it was a universe alma chien. the one i think is equally important but less remembered is his 1938 which was his ph.d. dissertation at princeton that was on nondetermine fistic machines that he called oracle machines. these are machines that are determine fistic, but every wince in a while -- every once in a while they take a jump. and we put it together, and that's intelligence so turing, in fact, he believed that he had proved that a machine that never makes mistakes can never be intelligent, that it's only -- [laughter] it is. and, i mean, girdle proved that as well. it's the -- if it never makes mistakes, it's not ever going to be intelligent. but if you look at like what, for instance, google is doing, you know, you have this enormous deterministic machine, all whatever it is, a million servers now, they're all predictable turing machines in the classical sense, yet they're connected by these done networkistic -- nondeterministic links which are the people. every time you're given ten links and you click on one, a nondeterministic process. and the machine ip corporates the state of that nondeterministic leap into the state of the machine, and that's why google can get you those results in a millisecond, because it knows what other people, where other people have found the meaning. and i think, you know, you can't imagine sort of a more perfect blueprint for an oracle machine than what google is doing right here. and that's not scary or anything else, it's just, it's just they're doing it, and we love it. we couldn'tly without it now. >> the second one i wanted to talk about was the computer as an organism. and you talk a lot about in the book that, again, it's the juxtaposition of you are thing and von neumann d of turing and von neumann. you say many decades later we still face the same questions. turing's question was what it would take for machines to begin to think, von neumann's was what it would take for machines to begin to reproduce. >> right. >> now, the notion of a replicating computer is in your book. you talk about that both logically and practically. what do you think the implications are? >> well, it's what happened. that's why we ended up with silicon valley because these machines became, effectively, self-replicating. nobody's there in a workshop making computers, they're replicating themselves, and i think that's why this von neumann machine is so important. even though there were other machines everywhere else, because it's the one that became the pattern, it's bigelow's machine that the chip factories that used to be right here, you know, make millions of every day. >> so is organism the word? is that the word to apply to it? >> no. i'm more interested in codes than organisms and that's what, you know, if you have to pick the great characters in this book, it's von neumann, it's clary, it's bigelow. codes themselves could be viewed as organisms because they self-reproduce, they replicate, they cross-breed, and he looked at that, you know n1953. so i think in a way the more interesting stuff happened on the coding side, not the hardware side. you know, the chips are just sort of the soup out of which the interest withing stuff happens. >> and the third one is big computers. you've touched on it a little bit with google. von neumann envisioned a world in which, of course, there was no network, but a few big computers would perform all of the world's computations. and you see that vision being realized in some respects now -- >> strange how we're going back to that. >> we're going back to the future. >> von neumann's vision was there would be three or four big computers, and you would dial in and do your computation and get your results over a network, and then we went to this vast, distributed network, and now we're going more to things like google and facebook which, essentially, are large computers in a very broad sense. but we, you know, in a way they are -- of course, people in the industry, gone back and forth many times from sort of the intelligence being in a terminal, being in the server, it goes back and forth. >> okay. let's get to some audience questions now. here's one that says it's not unreasonable to say that theoretical computer science is still dominated by turing's con cements. do you think it's -- concepts. do you think it's possible for that to change in the near future? >> yes. and i think the way it will change is not, not from the bottom up. i mean, we're never going to, i don't think we're ever going to escape the turing machine running on the von neumann two-dimensional matrix. turing had the one-dimensional model, and von neumann did a two-dimensional implementation. but now we're free to build all sorts of different models, so i think the answer is, yes, it is going to change. >> be here's a question from someone who went to the ias in 1955 and saw that the computer that they had looked cluey with some vacuum tubes hanging out in this space with stiff wires. was that a temporary situation, or did it generally look like that? >> sounds to me like they were doing some diagnostic thing or something, they might have been looking for -- i don't know. i'd like to know exactly when that was. >> 1955. >> they had monitor stations, that could have been what it was. they had tubes they would go around and look in the memory from outside. >> by the way, there was a real conflict at the ias, wasn't there, among the physicists and mathematicians and what they called the computer guys, right, the computer people? >> yeah. >> the computer people were relegate today the worst space. >> yeah. they were put in the basement. >> in the basement. [laughter] >> be next to the boiler room. and then they were put in an outbuilding. >> yeah. was that ever reconciled, or did they just have to go on about their business? >> no. what's been reconciled now through charles she moanny who build, you know, the most fabulous, luxurious building at the institute built by a hungarian programmer. so now they have the best quarters. >> so the hungarians had the last laugh. [laughter] there's a suggestion here that von neumann had to, essentially, charm the ias into making a machine. a place that was mostly about theory was now going to get real. is that your -- >> very true, yes. >> talk about that. >> he threaten today leave, he had offers from university of chicago, mit, he could have gone anywhere, but he wanted to stay there so, yeah, they couldn't let him go. >> if you were to describe what von neumann's vision was for the u.s. and society in this country as he found it and chose to make his home, do you