A machine-generated transcription of the video at on youtube
[10:00.380 --> 10:02.620] So it's terrific to be here.
[10:02.620 --> 10:11.700] And of course I want to thank Cheryl and Nicole for inviting me and for organizing this.
[10:11.700 --> 10:18.660] I will get the self-promotion out of the way.
[10:18.660 --> 10:24.180] Just heard, I've just written a book called The People's Network.
[10:24.180 --> 10:29.300] It's about the, it's about the early years of the telephone actually.
[10:29.300 --> 10:36.860] And it's not what I'm going to speak about today, but the argument I want to make today
[10:36.860 --> 10:41.540] is sort of a piece with my work on the early telephone.
[10:41.540 --> 10:52.660] I study the history of information and communication systems and how they shape the way we communicate
[10:52.660 --> 10:57.900] with one another and interact with each other and with the world.
[10:57.900 --> 11:02.740] And in the broadest possible sense, the argument I want to make, the argument sort of the
[11:02.740 --> 11:06.260] argument of the book on the telephone, but also the argument I want to make today about
[11:06.260 --> 11:09.100] Cold War and video games.
[11:09.100 --> 11:14.620] Marshall McCluen said, we shape our tools and then our tools shape us.
[11:14.620 --> 11:19.340] He was actually, I think, borrowing from Winston Churchill who said we shape our buildings
[11:19.340 --> 11:21.300] and then our building shape us.
[11:21.300 --> 11:24.580] Because I'm talking about video games, I'm just going to adjust that to say we shape
[11:24.580 --> 11:27.740] our toys and then our toys shape us.
[11:27.740 --> 11:31.580] And both sides of that equation matter.
[11:31.580 --> 11:37.660] What this tells us when I say we shape our tools, when we build our invent or develop a
[11:37.660 --> 11:43.540] communication and information system like the computers or the internet or 100 years
[11:43.540 --> 11:50.500] before the telephone, the telegraph, when we construct these things, we make choices.
[11:50.500 --> 11:51.500] We make all sorts of choices.
[11:51.500 --> 11:53.340] People building them make all sorts of choices.
[11:53.340 --> 11:56.980] They build the computer to do one task and not to do another.
[11:56.980 --> 12:01.420] They put a telephone line here and not over here.
[12:01.420 --> 12:10.100] And we make these decisions and in making these decisions, ideas get embedded into the technology
[12:10.100 --> 12:11.100] itself.
[12:11.100 --> 12:15.460] And that's what I'm really interested in is how ideas, political ideas, cultural ideas,
[12:15.460 --> 12:22.780] social ideas get built into the tools and the systems we use.
[12:22.780 --> 12:27.140] Which means that the choices we make when developing these systems really matter because
[12:27.140 --> 12:31.620] they do, our tools do shape us down the road.
[12:31.620 --> 12:36.060] Not just ideas get embedded into the things that we make.
[12:36.060 --> 12:42.100] Economic and political interests, whole philosophies can be literally sort of built into our systems
[12:42.100 --> 12:44.060] and our machines.
[12:44.060 --> 12:49.700] And it's tricky because once we've built the machines, those choices that we made in
[12:49.700 --> 12:53.380] building them become invisible to us.
[12:53.380 --> 12:58.460] We forget that the computer could have been like this and could have been like this and
[12:58.460 --> 12:59.860] that we made this choice.
[12:59.860 --> 13:02.020] Our choices come to seem natural.
[13:02.020 --> 13:07.660] The ideas shape the technology and the technology reinforces the ideas and the choices we made
[13:07.660 --> 13:13.300] along the way, the fact that we could have done something different, become invisible to
[13:13.300 --> 13:14.300] us.
[13:14.300 --> 13:20.780] So we forget and we don't notice, we rarely notice how profoundly our communication and
[13:20.780 --> 13:25.860] our information systems, whatever they may be, how profoundly they shape the way we see
[13:25.860 --> 13:28.620] the world.
[13:28.620 --> 13:35.140] As I say, this is the same argument in my book on the telephone, but for some reason my
[13:35.140 --> 13:40.620] 18 year old students seem much more interested when I make this argument about video games
[13:40.620 --> 13:45.940] than when I make this argument about the telephone in the 19th century.
[13:45.940 --> 13:54.060] So Cold War and video games.
[13:54.060 --> 13:57.780] When I give this talk to 18 year olds, I can assume that everybody knows what video games
[13:57.780 --> 14:03.540] are, but I cannot assume that they know what the Cold War was.
[14:03.540 --> 14:09.460] Today's freshmen were born six years after the fall of the Berlin Wall.
[14:09.460 --> 14:12.020] With this audience, hopefully that is reversed.
[14:12.020 --> 14:16.740] I don't need to back you up too much on the Cold War, maybe we'll have some remedial
[14:16.740 --> 14:20.260] information on video games.
[14:20.260 --> 14:25.740] But the Cold War, of course, was an arm showdown between the communist world and the capitalist
[14:25.740 --> 14:33.180] world, led by the Soviet Union and the United States, lasted roughly 50 years from the end
[14:33.180 --> 14:37.540] of the Second World War into the 1980s or 90s.
[14:37.540 --> 14:41.220] We call it the Cold War because it never became an all-out war, but of course there were plenty
[14:41.220 --> 14:44.100] of wars and bloodshed during that period.
[14:44.100 --> 14:51.620] Proxy wars fought all over the world in place of the terrible nuclear conflict that thankfully
[14:51.620 --> 14:54.420] never came.
[14:54.420 --> 15:01.140] I like this map as an image of the Cold War, this polar projection, because this is a
[15:01.140 --> 15:06.540] map that before the Cold War, no one really would have had reason to draw the world this
[15:06.540 --> 15:07.540] way.
[15:07.540 --> 15:09.940] This is a funny way, a different way of looking at the world.
[15:09.940 --> 15:14.940] The North Pole in the center with the United States and the USSR facing each other across
[15:14.940 --> 15:21.260] the Pole, across Canada, Africa and South America pushed right off to the periphery off
[15:21.260 --> 15:26.020] the map, maps are a way of looking at the world, of course.
[15:26.020 --> 15:34.220] This is a particularly Cold War way of looking at the world.
[15:34.220 --> 15:40.620] The talk I'm going to give today will make sense, I think.
[15:40.620 --> 15:47.700] It will make more sense, I think, if you think of the Cold War, not as a war that never came,
[15:47.700 --> 15:52.980] not just as a conflict, but as a way of looking at the world.
[15:52.980 --> 15:59.900] When we say that the Cold War lasted from 1945 to, let's say 1989 or 1992, what we're
[15:59.900 --> 16:09.820] saying is that during that time, every conflict that happened around the world was viewed through
[16:09.820 --> 16:13.620] Cold War lenses, through Cold War lenses, that the Cold War was a way of seeing the world
[16:13.620 --> 16:19.980] and a way of thinking about things and assuming that everything fit into this Manichaean conflict
[16:19.980 --> 16:27.780] between communism and capitalism.
[16:27.780 --> 16:36.900] The late great historian, Eric Hobbesbaum, said he once wrote, he who says industrial revolution
[16:36.900 --> 16:40.980] says cotton, meaning that the history of the industrial revolution and the history of
[16:40.980 --> 16:45.940] cotton are so intertwined that if you're talking about one, you're talking about the other.
[16:45.940 --> 16:54.860] So I'm going to borrow that phrasing from Hobbesbaum and say, she who says Cold War says computers.
[16:54.860 --> 16:59.180] But again, the history of computers and the history of the Cold War are so inextricably
[16:59.180 --> 17:06.860] intertwined that if you're talking about one, you are in fact talking about the other.
[17:06.860 --> 17:12.420] We wouldn't have had the modern computer without Cold War research.
[17:12.420 --> 17:16.980] And we wouldn't have had the Cold War, at least not in the shape that it took, without
[17:16.980 --> 17:18.980] the computer.
[17:18.980 --> 17:25.980] It's not just that Cold War military research developed the computer, although it did.
[17:25.980 --> 17:30.540] It's not just that computers were used by the military, although of course they were.
[17:30.540 --> 17:36.860] But more than that, Cold War politics and a certain Cold War mentality, a Cold War way
[17:36.860 --> 17:43.540] of looking at the world, was embedded in the machines, was built into the computers themselves.
[17:43.540 --> 17:49.060] And our ideas about computers, but also in the actual technical architecture of the machines.
[17:49.060 --> 17:55.100] And then the machines in turn, the computers made possible the politics and the strategy
[17:55.100 --> 18:01.300] of the Cold War.
[18:01.300 --> 18:07.100] My first computer looked like this, not this, but this.
[18:07.100 --> 18:15.900] All too came out in 1977. And again, to my students, this looks hilariously clunky and
[18:15.900 --> 18:18.820] old-fashioned and quaint.
[18:18.820 --> 18:24.860] But really, I like to say that it's not that different from the computers we used today.
[18:24.860 --> 18:25.860] And it's sort of structure.
[18:25.860 --> 18:28.260] If you look at it, it has the same parts.
[18:28.260 --> 18:35.740] It has a screen, it has a keyboard, it has this removable electronic memory.
[18:35.740 --> 18:41.740] You can use it for word processing, you can use it for games, you can use it for communication.
[18:41.740 --> 18:46.860] The first computers, the ones built during and right after the Second World War, looked
[18:46.860 --> 18:47.860] like this.
[18:47.860 --> 18:53.140] It's the Harvard Mark I built in 1944.
[18:53.140 --> 18:58.500] These were the code-breaking machines and the calculating machines.
[18:58.500 --> 19:04.260] Huge, electromechanical machines like the Mark I, like ENIAC.
[19:04.260 --> 19:09.380] And they don't look anything like the Apple II that I got when I was a kid or like the
[19:09.380 --> 19:10.380] computers we used today.
[19:10.380 --> 19:12.500] There's this big gap.
[19:12.500 --> 19:16.260] And there's a big gap in our kind of, in the history of computers and our understanding,
[19:16.260 --> 19:18.740] in the public understanding of the history of computers.
[19:18.740 --> 19:22.620] Most people have heard of these big machines and have a kind of mental image of them.
[19:22.620 --> 19:27.420] And of course, most people know what a personal computer looks like.
[19:27.420 --> 19:33.860] But we don't have a great sense of how we got from here over to there.
[19:33.860 --> 19:38.540] And that's the Cold War is the missing part of that story.
[19:38.540 --> 19:40.140] There were technical changes, of course.
[19:40.140 --> 19:43.180] The computers got smaller and faster and more powerful.
[19:43.180 --> 19:50.140] But there's also, between the Mark I and the Apple II, there's a huge shift in thinking
[19:50.140 --> 19:56.060] about what the computer was and what the computer was for and how people would interact with
[19:56.060 --> 19:59.740] these machines.
[19:59.740 --> 20:08.740] Howard Aiken, the man who designed the Mark I, famously said in 1947, there will never
[20:08.740 --> 20:09.740] be enough problems.
[20:09.740 --> 20:14.180] There will never be enough work, enough calculation for more than one or two of these
[20:14.180 --> 20:17.420] computers.
[20:17.420 --> 20:20.620] And there are a number of quotes like this, that people love to try it out.
[20:20.620 --> 20:25.940] Thomas Watson, who was the president of IBM, also in the 1940s, said, he estimated that
[20:25.940 --> 20:30.340] the world market for computers maybe they could sell six.
[20:30.340 --> 20:35.860] That's the market for computers because how many people are going to need a machine
[20:35.860 --> 20:41.060] that can do thousands of calculations per second?
[20:41.060 --> 20:45.780] So the point of these quotes, and do you see them all all over the place, the point is
[20:45.780 --> 20:49.260] not that Howard Aiken or Thomas Watson were dumb.
[20:49.260 --> 20:55.340] The point is not that they didn't understand the computers they were building.
[20:55.340 --> 21:00.580] The point is that in 1947, even people who knew more about computers than anyone else in
[21:00.580 --> 21:06.580] the world had little idea about how the technology was going to develop, about what the computer
[21:06.580 --> 21:10.940] was going to become, and about really what it was for.
[21:10.940 --> 21:18.260] In the history of computers, the machine, the hardware, often has preceded the application,
[21:18.260 --> 21:19.980] the software, the demand.
[21:19.980 --> 21:26.620] We think that necessity is the mother of invention, but often in computing, the machine came
[21:26.620 --> 21:30.660] first and then people had to figure out, well, what are we going to do with this machine?
[21:30.660 --> 21:35.420] That's why computer people talk about the killer app.
[21:35.420 --> 21:41.540] The killer app is the application, the software that makes the hardware worth why.
[21:41.540 --> 21:48.980] My Apple 2, the Apple 2 came out in 1977, but they sat dusty on shelves until in 1979,
[21:48.980 --> 21:56.100] a company called Visicalc came up with the first easy to use spreadsheet software.
[21:56.100 --> 22:00.060] Suddenly businessmen started coming into computer stores and saying, I want to buy one of
[22:00.060 --> 22:04.140] those Visicalc machines, not one of those Apple 2s, but one of those Visicalc machines.
[22:04.140 --> 22:10.380] Once the use, once the software, the application, came along, suddenly people could see why they
[22:10.380 --> 22:14.780] might want to have one of these funny little machines in their house.
[22:14.780 --> 22:19.100] In the history of computers, it's often a search for the killer app for the application that
[22:19.100 --> 22:20.940] makes the hardware worth while.
[22:20.940 --> 22:25.820] That was true in 1947, too, when Aiken was saying, I don't know why we're going to need
[22:25.820 --> 22:30.660] more than two of these things.
[22:30.660 --> 22:35.700] When Aiken said this, of course he was picturing machines like the Mark 1.
[22:35.700 --> 22:42.500] He was picturing these huge building-sized calculators, but computers, even as he said this,
[22:42.500 --> 22:45.500] computers were about to take a left turn.
[22:45.500 --> 22:49.940] Of course they would get smaller, they would get cheaper, they would get more powerful,
[22:49.940 --> 22:55.020] but there was also, as I said, a deep change in their basic function in what they were
[22:55.020 --> 22:59.060] used for and how human beings interacted with them.
[22:59.060 --> 23:03.300] The turn that Aiken and others did not foresee was called real-time computing.
[23:03.300 --> 23:10.340] In 1947, the phrase real-time had never been used in connection to computers.
[23:10.340 --> 23:12.060] They didn't have that concept.
[23:12.060 --> 23:17.220] In 1947 with Aiken's machine, you have some really big calculation you want to do.
[23:17.220 --> 23:22.660] You would program the machine using switches or punch cards, and then you'd turn it on and
[23:22.660 --> 23:27.140] go away, and it would chunk, chunk, chunk, chunk through the calculations, and later you
[23:27.140 --> 23:31.220] would come back, and there would be your answer in the form of more punch cards or switches
[23:31.220 --> 23:35.100] or something like that or print out.
[23:35.100 --> 23:41.260] In the 1950s, people started imagining and building computers that would respond, and
[23:41.260 --> 23:45.860] would receive input from the real world and respond quickly enough for us to talk about
[23:45.860 --> 23:50.820] real-time, that they could actually interface with the real world in some way, that systems
[23:50.820 --> 23:56.860] where a human operator would put information into the computer and get information back
[23:56.860 --> 24:01.740] almost immediately, creating a feedback loop between the human being and the computer.
[24:01.740 --> 24:05.820] The human being and the computer would be engaged in some sort of activity together,
[24:05.820 --> 24:09.980] sort of like a dance, that the information would be going back and forth from the human
[24:09.980 --> 24:16.380] being to the computer, in a system a loop that operated fast enough that you could simulate
[24:16.380 --> 24:21.740] some real world process, or maybe even intervene in some real world process.
[24:21.740 --> 24:27.380] And that, we're so used to this as being how a computer works, but that was a really
[24:27.380 --> 24:33.100] big change from the Mark I and the big machines that preceded it.
[24:33.100 --> 24:36.820] Big sort of conceptual change in just how, not just in the machine, but in how the human
[24:36.820 --> 24:39.740] being interacts with the machine.
[24:39.740 --> 24:43.900] So I called this on the slide a real-time human machine feedback loop, which is pretty
[24:43.900 --> 24:48.300] jargon-y, but what would a real-time human machine feedback loop look like?
[24:48.300 --> 24:52.380] Well, it looks sort of like this.
[24:52.380 --> 24:57.860] Information passes from the screen into the user and back from the user into the screen
[24:57.860 --> 25:01.460] in a loop fast enough to simulate some real world process.
[25:01.460 --> 25:05.700] This is a real-time human machine feedback loop, probably one of the most sophisticated
[25:05.700 --> 25:08.060] that we have.
[25:08.060 --> 25:12.900] Or stepping back into history a little bit, it might look something like this.
[25:12.900 --> 25:14.940] I'll get to this in a few moments.
[25:14.940 --> 25:17.500] This is arguably the first video game.
[25:17.500 --> 25:21.660] It's called Space War in 1962.
[25:21.660 --> 25:35.180] But I'm getting ahead of myself with that.
[25:35.180 --> 25:40.100] In the period we're talking about, from the 1940s through to the 1960s, nobody spent
[25:40.100 --> 25:45.900] more money and effort on developing computers than the US military.
[25:45.900 --> 25:51.460] So as a result, nobody had more impact than the United States military on what computers
[25:51.460 --> 25:57.220] looked like and what they were for, what they were good at.
[25:57.220 --> 26:01.580] There's kind of a myth that the private sector developed the computer.
[26:01.580 --> 26:05.420] That's true that much of the research took place at civilian universities, places like
[26:05.420 --> 26:11.300] MIT and private companies like IBM, AT&T, General Electric.
[26:11.300 --> 26:13.460] But the military was paying for this research.
[26:13.460 --> 26:19.100] At one point, 85% of MIT's research budget came from the military.
[26:19.100 --> 26:23.740] So the computer industry likes to think of itself as something sort of private, created
[26:23.740 --> 26:24.740] by the private sector.
[26:24.740 --> 26:28.500] But in the 1950s, in computers, there was no private sector.
[26:28.500 --> 26:33.140] It was a military project.
[26:33.140 --> 26:36.620] Why did the military embrace computers in this way?
[26:36.620 --> 26:41.300] Certainly they saw the potential of the technology maybe earlier than anyone else.
[26:41.300 --> 26:43.980] And they were right to do so.
[26:43.980 --> 26:48.180] But it's also true that there are fads and fashions in intellectual life.
[26:48.180 --> 26:49.740] There are trends in intellectual life.
[26:49.740 --> 26:53.540] And things can come in, ideas can come in and out of fashion.
[26:53.540 --> 27:02.820] And in the early Cold War, there was an intellectual fashion trend, if you like, for using statistics
[27:02.820 --> 27:10.700] and calculation to grapple with complex social and organizational problems.
[27:10.700 --> 27:13.700] And they called this social science.
[27:13.700 --> 27:19.500] And social science was never more popular or prestigious than in the early Cold War
[27:19.500 --> 27:21.140] years.
[27:21.140 --> 27:26.900] And the Cold War planners put a great deal of faith into the idea that human social behavior
[27:26.900 --> 27:30.500] really was a science, that you could model it scientifically.
[27:30.500 --> 27:36.060] And they created new kinds of social science disciplines, like game theory, operations
[27:36.060 --> 27:41.140] research, systems theory, poured money into these fields.
[27:41.140 --> 27:47.980] Universities like Western built big, ugly buildings and called them social science centers.
[27:47.980 --> 27:55.700] And all of this activity was driven by a kind of faith in models and simulations.
[27:55.700 --> 28:04.460] In the idea that you could build a miniature mathematical model of the world, like a game.
[28:04.460 --> 28:13.300] And that you could learn about the real world by playing with that little model world.
[28:13.300 --> 28:18.380] One group that was really at the forefront of this and that loved this kind of thinking,
[28:18.380 --> 28:25.820] models and simulations, was the Rand Corporation.
[28:25.820 --> 28:29.620] Rand was one of the most influential Cold War think tanks.
[28:29.620 --> 28:31.780] It was set up by the Air Force.
[28:31.780 --> 28:33.980] It was independent, but it was created by the Air Force.
[28:33.980 --> 28:38.060] Rand just stands for R&D, Research and Development.
[28:38.060 --> 28:42.660] The Research and Development Corporation, the Rand Corporation.
[28:42.660 --> 28:46.980] It was staffed by brilliant minds from a host of disciplines.
[28:46.980 --> 28:50.940] There were mathematicians, there were political scientists, there were physicists, there were
[28:50.940 --> 28:51.940] economists.
[28:51.940 --> 28:57.300] Rand called itself a university without students.
[28:57.300 --> 29:00.300] My colleagues and I always sort of sigh wistfully at that thought.
[29:00.300 --> 29:04.060] A university without students.
[29:04.060 --> 29:11.860] The Air Force gave them pots of money and groovy furniture and couches to lounge on.
[29:11.860 --> 29:20.540] And the freedom to work on anything they wanted as long as it related to the Cold War.
[29:20.540 --> 29:27.500] Rand's specialty became one of their most famous analysts, Herman Khan, called Thinking
[29:27.500 --> 29:29.540] the Unthinkable.
[29:29.540 --> 29:37.020] What Rand was good at was imagining and planning for nuclear Armageddon.
[29:37.020 --> 29:45.180] But Khan and others like him popularized ideas like game theory, zero sum games, mutual
[29:45.180 --> 29:52.620] assured destruction as they planned for and tried to imagine nuclear war.
[29:52.620 --> 29:58.180] This is a famous or really an infamous table in Khan's 1960 book on thermonuclear war.
[29:58.180 --> 30:03.020] If you can see it says tragic but distinguishable post-war states.
[30:03.020 --> 30:10.220] And Khan was talking about, well, what's the difference between a war that kills 20 million
[30:10.220 --> 30:13.620] Americans and a war that kills 40 million Americans?
[30:13.620 --> 30:16.540] And how will we, how long will it take to recover?
[30:16.540 --> 30:22.260] And how can we plan to make it more likely that the World War III will kill 20 million
[30:22.260 --> 30:28.500] Americans and not 40 or not 80 or not 160 million?
[30:28.500 --> 30:30.180] People thought Khan was a monster.
[30:30.180 --> 30:37.980] People were not thrilled to learn that serious top level advisors were having discussions
[30:37.980 --> 30:42.660] about war plans that might leave 40 million people dead.
[30:42.660 --> 30:44.180] Khan always felt that that was unfair.
[30:44.180 --> 30:47.100] He said, look, I said that these states were tragic.
[30:47.100 --> 30:50.900] I just also want to point out that these states are distinguishable.
[30:50.900 --> 30:53.340] And this is this kind of distinction that Khan tried to make.
[30:53.340 --> 30:57.900] And whether you think that that is a fair distinction to make or whether it's all equally monstrous,
[30:57.900 --> 31:06.540] you know, that's a sort of a, how we position ourselves morally to these sorts of questions.
[31:06.540 --> 31:10.020] But certainly, and Khan liked to provoke people.
[31:10.020 --> 31:15.740] And people like Khan and also Henry Kissinger, horrified people by talking about nuclear
[31:15.740 --> 31:18.980] war as if it were a game.
[31:18.980 --> 31:24.020] But they said that's the only way you can think about it.
[31:24.020 --> 31:31.140] There's a telling kind of interchange between Khan and Air Force General Curtis Lamey.
[31:31.140 --> 31:34.740] Now Curtis Lamey was not squeamish about bombs.
[31:34.740 --> 31:36.780] Lamey was the champion of strategic bombing.
[31:36.780 --> 31:40.900] He oversaw the bombing of Japan in the Second World War.
[31:40.900 --> 31:45.980] But he resented the kind of takeover of the Pentagon by civilian eggheads like Herman
[31:45.980 --> 31:47.180] Khan.
[31:47.180 --> 31:51.260] And so he once blasted and he said, why that son of a bitch was in junior high school when
[31:51.260 --> 31:53.020] I was at bombing Japan.
[31:53.020 --> 32:00.300] And Khan's response was, how many thermonuclear wars have you fought, General?
[32:00.300 --> 32:05.820] Khan's point being, or the argument being that Lamey and his generation of military
[32:05.820 --> 32:10.420] generals believed in personal experience.
[32:10.420 --> 32:16.060] The new civilian intellectuals that were taking over the defense department at this time
[32:16.060 --> 32:22.860] said that in the new world of the atomic bomb, personal experience is of no use.
[32:22.860 --> 32:29.980] The only way to know anything about the new world of nuclear war is to simulate it as
[32:29.980 --> 32:33.340] a game or in a computer.
[32:33.340 --> 32:36.500] So that's what they did.
[32:36.500 --> 32:42.500] At places like Rand and also MIT and the Pentagon, they planned for nuclear war and planned for
[32:42.500 --> 32:46.580] all kinds of conflicts by playing games, by doing simulations.
[32:46.580 --> 32:50.660] These were not necessarily computer games.
[32:50.660 --> 32:54.140] These could be sort of like big war games fought on a map.
[32:54.140 --> 32:58.900] The picture is not terrific, but there's a big hex map with lots of little pieces and
[32:58.900 --> 33:01.900] tokens on it that they're moving around.
[33:01.900 --> 33:04.900] They would use dice.
[33:04.900 --> 33:09.140] Military officials would sort of run these scenarios and the Pentagon guys and the
[33:09.140 --> 33:10.940] Rand guys would run through these scenarios.
[33:10.940 --> 33:16.940] The picture is very dark, but the man standing in the middle in the dark shirt is Daniel
[33:16.940 --> 33:24.500] Ellsberg, who would later be famous as the man who leaked to the Pentagon papers, saying
[33:24.500 --> 33:30.780] that evidence that the defense department knew the war in Vietnam was unwinnable and contributed
[33:30.780 --> 33:34.660] in a way to the downfall of Richard Nixon.
[33:34.660 --> 33:42.460] So here's Ellsberg in 1959 playing war games, playing simulations of war.
[33:42.460 --> 33:45.300] And actually this picture is where this story started for me.
[33:45.300 --> 33:49.860] Seeing this picture is where I got sort of interested in this story because if you are
[33:49.860 --> 33:56.220] a nerd of a certain age like myself, this looks very familiar.
[33:56.220 --> 33:58.940] These kinds of hex maps and dice and this sort of thing.
[33:58.940 --> 34:03.420] This is the, in addition to being part of the Cold War story, this is the prehistory
[34:03.420 --> 34:09.020] of war games and role playing games like Dungeons and Dragons and World of Warcraft.
[34:09.020 --> 34:14.900] It all comes from this place.
[34:14.900 --> 34:20.580] SPI, the sort of company that pioneered a lot of the war games that I and people like
[34:20.580 --> 34:23.500] me played when we were kids.
[34:23.500 --> 34:27.860] The founder got the idea for the hex maps when he saw this picture in Life magazine in
[34:27.860 --> 34:28.860] 1959.
[34:28.860 --> 34:31.260] He said, oh, hex is, that's much better than squares.
[34:31.260 --> 34:38.420] So he got the idea from, you know, directly from Rand and the military industrial complex.
[34:38.420 --> 34:39.420] These two things are connected.
[34:39.420 --> 34:44.620] But in the 1950s and 1960s, it wasn't kids in rec rooms and basements playing these games.
[34:44.620 --> 34:51.020] It was the joint chiefs of staff and the Army War College and people at MIT and even in
[34:51.020 --> 34:52.820] the White House.
[34:52.820 --> 34:55.340] Here's another even more elaborate.
[34:55.340 --> 34:57.140] This is a game.
[34:57.140 --> 34:59.860] It's called the Naval Electronic Warfare Simulator.
[34:59.860 --> 35:04.900] But what is happening here is that all these guys are playing an elaborate complex war
[35:04.900 --> 35:11.420] game using both computers and analog sort of tools.
[35:11.420 --> 35:19.740] In September 1961, just after the construction of the Berlin Wall, Rand, Rand, a role-playing
[35:19.740 --> 35:26.980] simulation of the Berlin crisis for a group of Camp David that included Henry Kissinger,
[35:26.980 --> 35:32.860] McGeorge Bundy, Carl Kason, Walt Rostow, all these high level Kennedy advisors, really
[35:32.860 --> 35:38.980] one of the sort of most high powered clusters of geeks ever to gather around a gaming table.
[35:38.980 --> 35:44.860] A year later, during the Cuban Missile Crisis, events in the real world started to look
[35:44.860 --> 35:49.220] scarily like the events of the simulation they had run a year before.
[35:49.220 --> 35:55.580] And Daniel Ellsberg said to Walt Rostow, this shows how realistic that Berlin game was.
[35:55.580 --> 36:01.740] And Rostow said, or how unrealistic all of this is.
[36:01.740 --> 36:07.900] On Halloween 1963, so 50 years ago this week, Rand ran another game for the joint chiefs
[36:07.900 --> 36:14.820] of staff, the director of the US budget, and Attorney General Robert Kennedy.
[36:14.820 --> 36:18.380] And Bobbi was really intrigued by the game, by the whole system, and he proposed he got
[36:18.380 --> 36:19.380] excited.
[36:19.380 --> 36:21.820] He said, oh, we should run one of these about civil rights.
[36:21.820 --> 36:25.940] Run a game about the civil rights legislation.
[36:25.940 --> 36:30.900] That never happened, of course, a few weeks later was the assassination of his brother.
[36:30.900 --> 36:35.780] But if I had asked you, if you remember the Kennedys, if I had asked you, which Kennedy
[36:35.780 --> 36:41.660] brother was most likely to spend his Halloween night playing a war game with a bunch of other
[36:41.660 --> 36:45.260] eggheads, it would have to be Bobbi, clearly.
[36:45.260 --> 36:51.420] The only one you could imagine doing that would be Robert Kennedy.
[36:51.420 --> 36:55.100] Computers were part of these games.
[36:55.100 --> 36:57.140] Computers were made for these kinds of games.
[36:57.140 --> 36:59.500] Literally, this is what the computers were made for.
[36:59.500 --> 37:04.220] They were made to facilitate games and simulations like this.
[37:04.220 --> 37:12.660] Computers aided the games, and in so doing, they encouraged and enabled this dream of shrinking
[37:12.660 --> 37:19.740] the real world down into a little toy world where all the variables could be controlled
[37:19.740 --> 37:24.140] and the outcomes could be predicted.
[37:24.140 --> 37:38.580] In light of the moment we took before this talk for Remembrance Day to remember our veterans,
[37:38.580 --> 37:46.340] I want to make it clear that my point with all this is not to be critical of the military.
[37:46.340 --> 37:49.740] I'm not critical of the military and what they do.
[37:49.740 --> 37:54.180] I think what I want to be critical of and what I hope that we should all be skeptical
[37:54.180 --> 37:56.300] of is this dream.
[37:56.300 --> 38:02.620] This dream that with the right algorithms, with the right machines, that war can be made
[38:02.620 --> 38:08.100] painless, that war can be made bloodless, that there is some sort of a push button solution
[38:08.100 --> 38:13.660] that we can take the world's problems and solve them in a painless, bloodless way using
[38:13.660 --> 38:17.940] these marvelous tools.
[38:17.940 --> 38:24.940] This brings us to, I would say, the most important computer system that most people have never
[38:24.940 --> 38:26.500] heard of.
[38:26.500 --> 38:28.180] It's called Sage.
[38:28.180 --> 38:34.340] Sage stands for semi-automatic ground environment, the Sage computer.
[38:34.340 --> 38:39.620] Sage's story begins at the very start of the Cold War.
[38:39.620 --> 38:45.220] As you all know, the end of the Second World War in 1945, the United States used the atomic
[38:45.220 --> 38:47.780] bomb against Japan.
[38:47.780 --> 38:52.420] In the war ended, the United States was the only country with the bomb.
[38:52.420 --> 38:55.260] Most Americans thought that that was going to be true for a long time.
[38:55.260 --> 38:59.740] For maybe 10 or 20 years, they would have this atomic monopoly.
[38:59.740 --> 39:05.260] During those early years of the US atomic monopoly, the American Air Force told anyone who
[39:05.260 --> 39:09.620] would listen that there is no defense against the atomic bomb.
[39:09.620 --> 39:15.900] People said, should we build bomb shelters, anti-aircraft guns, defense against this weapon,
[39:15.900 --> 39:18.940] and the Air Force said, no, no, no.
[39:18.940 --> 39:20.380] The bomber will always get through.
[39:20.380 --> 39:23.220] There is no defense against this weapon.
[39:23.220 --> 39:30.060] That was fine until 1949 when, of course, the Soviet Union tested their first atomic
[39:30.060 --> 39:31.860] weapon.
[39:31.860 --> 39:39.940] Everyone in America was asking what is the Air Force doing to protect us from atomic attack?
[39:39.940 --> 39:45.540] The Air Force quickly got in gear and sucked up a whole bunch of projects and started throwing
[39:45.540 --> 39:46.540] money at them.
[39:46.540 --> 39:48.900] A bunch of civil defense, air defense projects.
[39:48.900 --> 39:53.740] We have to find ways to defend against air attack.
[39:53.740 --> 39:57.020] One of the projects that they adopted was called World Wind.
[39:57.020 --> 40:04.860] This was a project based at MIT involving real-time computer simulation.
[40:04.860 --> 40:11.860] World Wind had started out in 1944, in the last years of the war, as an analog flight simulator.
[40:11.860 --> 40:16.620] It was partly electronic and partly mechanical.
[40:16.620 --> 40:18.860] This is what a flight simulator looked like in the 1940s.
[40:18.860 --> 40:21.180] This is a link trainer.
[40:21.180 --> 40:26.900] It kind of looks like something at Walmart that you can put your kids in and put a tunnion
[40:26.900 --> 40:33.260] and it will shake them around for a few minutes while you try and get your shopping done.
[40:33.260 --> 40:39.340] An MIT graduate student named Jay Forrester wanted to build a smarter simulator.
[40:39.340 --> 40:45.220] He wanted to build a general simulator or universal simulator by which he meant one that he
[40:45.220 --> 40:49.940] could program to simulate any set of flying conditions.
[40:49.940 --> 40:53.820] A kind of video game.
[40:53.820 --> 41:02.380] Between 1944 and 1949, Forrester's project got more and more elaborate and he got interested
[41:02.380 --> 41:07.740] as it got more complicated, he got interested in digital computers and he speculated he said
[41:07.740 --> 41:14.100] maybe someday computers would be fast enough that we could simulate aerial warfare in something
[41:14.100 --> 41:17.620] approximating real-time.
[41:17.620 --> 41:23.660] In the early 1950s, the Air Force learned of Forrester's project and basically they adopted
[41:23.660 --> 41:27.260] World Wind and they gave him a blank check to work on this.
[41:27.260 --> 41:33.060] So this thing that was designed to simulate aerial combat, sort of a simulation, a game,
[41:33.060 --> 41:38.340] became a thing actually meant to conduct aerial combat.
[41:38.340 --> 41:44.480] And eight years and eight billion dollars later, what Forrester gave them was the Sage
[41:44.480 --> 41:48.260] Air Defense Network, the semi-automatic ground environment.
[41:48.260 --> 41:53.380] And as I said, it's this extremely important and influential computer system but most
[41:53.380 --> 41:55.900] people have not heard of it.
[41:55.900 --> 42:02.420] I'm going to show you, I'll let IBM explain to you how Sage worked.
[42:02.420 --> 42:03.420] If we can show the...
[42:03.420 --> 42:06.260] It's just like a two and a half minute video clip.
[42:06.260 --> 42:11.060] If you can go to full screen.
[42:11.060 --> 42:14.860] This is where America's peace of mind begins.
[42:14.860 --> 42:19.940] Around the clock, radars, electronic eyes, watch the skies and report what they see to
[42:19.940 --> 42:24.300] Sage, defense system of the United States Air Force.
[42:24.300 --> 42:27.980] Here is a Sage Center on 24-hour alert.
[42:27.980 --> 42:33.700] And its heart is a computer developed by a research team from MIT and IBM working with
[42:33.700 --> 42:35.060] the Air Force.
[42:35.060 --> 42:41.220] The Sage Computers speeds the information for decisions by man in our missile age.
[42:41.220 --> 42:47.460] Every scheduled flight across American Frontiers is recorded ahead of time on IBM punch cards,
[42:47.460 --> 42:49.820] then fed into the Sage Computers.
[42:49.820 --> 42:55.380] Now the computer can draw a picture of what is supposed to be in the sky at any moment.
[42:55.380 --> 43:01.500] It continually compares this expected picture with the real picture as seen by radar.
[43:01.500 --> 43:06.980] If a flying object does not belong, it appears on this viewing screen.
[43:06.980 --> 43:09.220] There's one now at the right of the screen.
[43:09.220 --> 43:13.300] They call it a blip, unknown flying object.
[43:13.300 --> 43:14.820] Friend or foe?
[43:14.820 --> 43:17.220] Within seconds the Air Force will know.
[43:17.220 --> 43:19.940] The officer fires a light gun at the target blip.
[43:19.940 --> 43:22.780] This tells the computer to track the object.
[43:22.780 --> 43:26.740] At the launching site a long-range bow-mark missile is ready for firing.
[43:26.740 --> 43:30.420] Now they ask the computer to calculate an intercept point.
[43:30.420 --> 43:36.380] X marks the spot where the bow-mark missile would meet the moving target if fired immediately.
[43:36.380 --> 43:40.620] The officer in charge makes the final decision.
[43:40.620 --> 43:43.780] Fire.
[43:43.780 --> 43:51.340] At the moment of launching the bow-mark missile receives instructions from the IBM computer.
[43:51.340 --> 43:59.220] As the missile screens toward target, radar keeps on tracking.
[43:59.220 --> 44:04.420] With electronic control, the computer automatically adjusts the missile to meet any change in the
[44:04.420 --> 44:05.740] target flight.
[44:05.740 --> 44:09.180] There is no escape.
[44:09.180 --> 44:14.180] Intercept.
[44:14.180 --> 44:15.700] This was a test.
[44:15.700 --> 44:23.460] One of many successful tests of the Sage bow-mark security team, our new system of air defense.
[44:23.460 --> 44:29.260] To be ready for the worst, so that the worst will never happen, America is now armed with
[44:29.260 --> 44:31.820] instant electronic reflexes.
[44:31.820 --> 44:38.100] The Sage computer, made by IBM, is another example of the vast new powers that man can
[44:38.100 --> 44:42.020] achieve through the creative use of his mind.
[44:42.020 --> 44:50.900] Strength for national defense, speed for informed decisions, surface for a growing America.
[44:50.900 --> 45:00.100] This is IBM, freeing man's mind to shape the future.
[45:00.100 --> 45:01.100] Pretty cool, huh?
[45:01.100 --> 45:04.580] Oh, we're on the back to the beginning.
[45:04.580 --> 45:06.180] Here we go.
[45:06.180 --> 45:09.980] We have to catch up.
[45:09.980 --> 45:12.780] There we are.
[45:12.780 --> 45:17.340] Sage was the largest computer ever built.
[45:17.340 --> 45:19.340] I mean, the physically largest computer ever built.
[45:19.340 --> 45:23.620] Since computers are smaller now, it will probably hold that record for a while.
[45:23.620 --> 45:29.780] Sage computers were actually four story buildings, half an acre of floor space, tens of thousands
[45:29.780 --> 45:31.780] of vacuum tubes.
[45:31.780 --> 45:32.780] There were lots of them.
[45:32.780 --> 45:37.940] There were 22 Sage command computers, 22 of these buildings in the United States, one
[45:37.940 --> 45:41.660] in Canada, one in North Bay.
[45:41.660 --> 45:47.380] Sage cost more than the Manhattan Project, cost between $8 and $12 billion.
[45:47.380 --> 45:54.140] Those were $1964, so that's maybe $60 to $90 billion today.
[45:54.140 --> 45:58.420] It would have been so cool if it had worked.
[45:58.420 --> 46:03.900] It didn't work, and it was kind of obsolete before it was completed.
[46:03.900 --> 46:07.900] There's a number of reasons it didn't work, but a big one is that Sage was the largest
[46:07.900 --> 46:12.220] one built to shoot down airplanes.
[46:12.220 --> 46:14.940] Built to shoot down bombers.
[46:14.940 --> 46:21.300] By the time Sage was operational in 1961, the Soviets and the Americans were both moving
[46:21.300 --> 46:28.100] their nuclear arsenal off of the bombers and onto missiles, onto ICBMs.
[46:28.100 --> 46:33.380] Sage was never going to be fast enough to shoot down missiles.
[46:33.380 --> 46:39.340] All it amounted to in the end was a really expensive and elaborate air traffic control system.
[46:39.340 --> 46:42.820] It actually was the basis of the sort of well until the 1980s.
[46:42.820 --> 46:47.740] It was used as an air traffic control system.
[46:47.740 --> 46:52.100] But the fact that it didn't work, it didn't do what it was supposed to do, does not mean
[46:52.100 --> 46:57.300] that Sage was not hugely influential.
[46:57.300 --> 47:00.620] When I was showing my computer, my first computer, and the old computers, when I asked, what
[47:00.620 --> 47:04.940] makes the computers of the 1970s different from the computers in the 1940s?
[47:04.940 --> 47:11.940] If you were to make a list, virtually everything you would come up with was pioneered by Sage.
[47:11.940 --> 47:15.620] Sage computers had screens, they had video displays and graphics.
[47:15.620 --> 47:20.300] Sage computers had keyboards and handheld controls like those light guns.
[47:20.300 --> 47:22.900] Sage computers had magnetic core memory.
[47:22.900 --> 47:26.860] They had parallel logic and multi processing, as opposed to serial processing.
[47:26.860 --> 47:30.780] It was a technical distinction, but it was another big innovation.
[47:30.780 --> 47:33.580] They could convert from analog to digital.
[47:33.580 --> 47:38.540] They networked multiple computers, sending information over telephone lines.
[47:38.540 --> 47:43.220] They simulated the real world in what was called real time.
[47:43.220 --> 47:45.860] All these things, Sage brought them together.
[47:45.860 --> 47:50.860] The reason that our computers look like they do, that they do the things that they do,
[47:50.860 --> 47:54.540] is because of their ancestry and sage.
[47:54.540 --> 48:00.580] The reason you can play video games on computers is because of Sage.
[48:00.580 --> 48:03.860] Sage had another cultural impact, too.
[48:03.860 --> 48:08.780] When the parts from this giant computer system were retired, they would sometimes sell them
[48:08.780 --> 48:17.740] off, and Hollywood producer Irwin Allen in the mid-60s, bought a whole warehouse full of
[48:17.740 --> 48:23.100] old Sage computers and monitors, and he used them for decades as props and movies and TV
[48:23.100 --> 48:24.380] shows.
[48:24.380 --> 48:30.420] So Sage computers actually turn up in dozens of movies and TV shows from Lost in Space
[48:30.420 --> 48:37.980] in the 1960s, the old Batman series, Battlestar Galactica, even as late as Lost.
[48:37.980 --> 48:46.900] Just a few years ago, these are Sage computers in the background of all these TV shows.
[48:46.900 --> 48:49.940] The war room in Dr. Strange Love, this isn't Sage.
[48:49.940 --> 48:55.620] Blue brick, of course, built the set from scratch, but it was inspired by the real life Sage
[48:55.620 --> 49:01.860] war room over here on the right.
[49:01.860 --> 49:08.100] But deeper than these sort of its appearances on TV and movies, Sage converted the Air
[49:08.100 --> 49:16.860] Force to kind of faith in real time digital computing, to faith in air defense, and centralized
[49:16.860 --> 49:24.780] command and control systems.
[49:24.780 --> 49:30.620] Sage lasted until, as I said, until the 1980s, and by finally taking a part in the 1980s,
[49:30.620 --> 49:36.660] by which point was so old, the vacuum tubes that had these thousands of vacuum tubes,
[49:36.660 --> 49:40.460] the only place the Air Force could get, the only place the vacuum tubes that they needed
[49:40.460 --> 49:43.500] were manufactured was the Soviet Union.
[49:43.500 --> 49:49.780] So the Air Force was buying vacuum tubes for Sage from the Soviets to keep their old,
[49:49.780 --> 49:55.100] 30-year-old computers running.
[49:55.100 --> 49:58.300] Now here's where actual video games enter the story.
[49:58.300 --> 50:04.700] In 1961, just about the time Sage was going online, the computer students at MIT got a
[50:04.700 --> 50:05.700] new toy.
[50:05.700 --> 50:08.580] It was a program data processor, a PDP1.
[50:08.580 --> 50:10.420] They call it a mini computer.
[50:10.420 --> 50:15.660] It's mini because instead of being a four-story building, it's about the size of a deep freezer
[50:15.660 --> 50:16.660] or a closet.
[50:16.660 --> 50:19.780] It's behind this gentleman on the left.
[50:19.780 --> 50:23.780] And as you can see, it had the same kind of round display monitor that Sage did or that
[50:23.780 --> 50:28.140] an old World War II radar screen did.
[50:28.140 --> 50:31.260] And the students at MIT were brainstorming applications.
[50:31.260 --> 50:36.780] Again, the technology comes before the application, the search for the application, the killer app.
[50:36.780 --> 50:40.300] They were looking for ideas for programs they could write that would show off the cool
[50:40.300 --> 50:42.820] visual display in a kind of compelling way.
[50:42.820 --> 50:45.260] They made a drawing program, sort of like an etch-a-scatch.
[50:45.260 --> 50:48.900] They made a kaleidoscope generator.
[50:48.900 --> 50:54.020] But MIT student named Steve Russell looked at the little lights on the screen and thought
[50:54.020 --> 50:55.020] about spaceships.
[50:55.020 --> 50:58.540] And he came up with a game with battling spaceships.
[50:58.540 --> 51:04.340] And he typed out a few thousand lines of code and produced space war, the first computer
[51:04.340 --> 51:06.620] video game.
[51:06.620 --> 51:11.980] These were two person game, these two little rocket ships would float around and try to
[51:11.980 --> 51:12.980] shoot each other.
[51:12.980 --> 51:17.180] And you had switches that you could turn, you could thrust, and you could fire.
[51:17.180 --> 51:18.500] You could play it online now.
[51:18.500 --> 51:21.220] And it's actually extremely hard, or at least I find it extremely hard.
[51:21.220 --> 51:26.340] But anyway, Sage borrowed all of its technology from, I mean, space war, borrowed its technology
[51:26.340 --> 51:27.340] from Sage.
[51:27.340 --> 51:32.780] It also borrowed the deeper logic of what the computer is for, this idea of simulation,
[51:32.780 --> 51:33.780] real time.
[51:33.780 --> 51:38.260] It's an extremely simple game, but I would say that there's not, it's not a coincidence
[51:38.260 --> 51:44.580] that what it was about was war and space, sort of simulation of high-tech war.
[51:44.580 --> 51:51.300] It's kind of the raison d'etre of these machines, whether serious or for fun.
[51:51.300 --> 51:58.340] And space war spread rapidly through the computing centers, universities of the 1960s.
[51:58.340 --> 52:03.100] There are also variations, there's California space war and Michigan space war.
[52:03.100 --> 52:07.420] If you're involved in computers in the early 1960s or in the 1960s, you may well remember
[52:07.420 --> 52:09.500] playing space war.
[52:09.500 --> 52:14.940] As early as 1963, it had become so popular that Stanford's computer science department
[52:14.940 --> 52:18.860] had to have a strict policy of no space war during business hours.
[52:18.860 --> 52:25.300] So this is the start of video games as we know them.
[52:25.300 --> 52:30.940] There's this great article on space war was published in Rolling Stone magazine in 1972.
[52:30.940 --> 52:37.460] It's called space war, fanatic life and symbolic death among the computer bombs.
[52:37.460 --> 52:41.420] And the article is written by Stuart Brand, who's a kind of hippie entrepreneur.
[52:41.420 --> 52:46.020] He also produced the whole Earth catalog, which was a sort of counterculture, how to manual,
[52:46.020 --> 52:48.460] those popular at the time.
[52:48.460 --> 52:51.860] Brand said, ready or not, computers are coming to the people.
[52:51.860 --> 52:55.500] That's the best news since psychedelics.
[52:55.500 --> 52:59.820] What excited Brand and the reason that something like this was in Rolling Stone magazine was
[52:59.820 --> 53:04.580] this idea that this computer, the product of the military industrial complex, was being
[53:04.580 --> 53:06.500] appropriated by the counterculture.
[53:06.500 --> 53:14.540] That these games were being played by guys with long hair, by people who smoked pot.
[53:14.540 --> 53:16.300] And this seemed like a real clash to them.
[53:16.300 --> 53:21.500] The computers, which are the most sort of like, you know, IBM, crew cut, Horn Rim glasses
[53:21.500 --> 53:29.500] kind of a phenomenon at the time, were being adopted by the counterculture.
[53:29.500 --> 53:32.380] Brand said, space war serves Earth peace.
[53:32.380 --> 53:35.460] And he waxed wraps out about hackers.
[53:35.460 --> 53:42.980] And the revolutionary potential of playing around hacking these things.
[53:42.980 --> 53:47.220] He said, space war was a crystal ball of things to come in computer science.
[53:47.220 --> 53:48.980] That it was interactive in real time.
[53:48.980 --> 53:51.060] That it bonded human and machine.
[53:51.060 --> 53:53.260] That it was a game.
[53:53.260 --> 53:55.500] Brand was right about all these things.
[53:55.500 --> 53:59.820] But I would say, you know, if you changed the word game to simulation, you could say all
[53:59.820 --> 54:03.540] those things that were true about sage as well.
[54:03.540 --> 54:09.340] So my point would be that sage and space war were not opposites.
[54:09.340 --> 54:13.940] That sage, which was the very epitome of the military industrial complex and space war,
[54:13.940 --> 54:20.980] which was this countercultural fun hacker phenomenon, are really two sides of the same coin.
[54:20.980 --> 54:24.460] All right.
[54:24.460 --> 54:26.100] And we're just about at a time.
[54:26.100 --> 54:27.100] I do want to say something.
[54:27.100 --> 54:31.060] I've sort of told you this story, but I want to say something about what I think is the
[54:31.060 --> 54:35.540] moral weight of the story, the implication of it for us today.
[54:35.540 --> 54:41.980] And for that, I want to take a short side trip to the war in Vietnam.
[54:41.980 --> 54:48.180] This whole faith in simulation and computers, modeling the real world, probably was at its
[54:48.180 --> 54:51.740] zenith when Robert McNamara was the Secretary of Defense.
[54:51.740 --> 54:55.940] McNamara was a statistician and economist.
[54:55.940 --> 55:01.580] He was one of the Wiz kids in the Army Air Corps in World War II planned and improved
[55:01.580 --> 55:04.900] the cost efficiency of bombing raids in Japan.
[55:04.900 --> 55:09.820] After the war, he went to work for Ford Motor Company, was the president of Ford Motor Company.
[55:09.820 --> 55:13.340] And in 1961, Kennedy brought him in to be Secretary of Defense.
[55:13.340 --> 55:18.340] And McNamara brought in all these Wiz kids, the civilian computer experts, and staffed
[55:18.340 --> 55:20.380] the Pentagon with them.
[55:20.380 --> 55:25.380] His faith in computers and quantitative data was legendary, his famous quote he said to
[55:25.380 --> 55:26.380] it.
[55:26.380 --> 55:28.180] It might have been Ellsberg that he said this to actually someone was saying that we're
[55:28.180 --> 55:31.140] losing the war in Vietnam and he said, where is your data?
[55:31.140 --> 55:35.860] Don't get me poetry, give me something I can put in the computer.
[55:35.860 --> 55:41.060] The Vietnam War became a test case for this kind of war fighting, for computerized planning,
[55:41.060 --> 55:45.300] computerized warfare, for all these dreams about automation and simulation.
[55:45.300 --> 55:48.380] Certainly it was the most computerized war fought to date.
[55:48.380 --> 55:53.220] So the third of the material, the US brought to Vietnam, physical material, was electronic
[55:53.220 --> 55:57.220] communications equipment.
[55:57.220 --> 56:01.900] The building up there, it's called, it was the largest building in Southeast Asia in
[56:01.900 --> 56:03.740] 1968.
[56:03.740 --> 56:05.820] It housed US Air Force igloo white.
[56:05.820 --> 56:07.140] It was in Thailand.
[56:07.140 --> 56:10.300] It's called the Infiltration Surveillance Center.
[56:10.300 --> 56:14.180] It contained banks of video displays that were connected to sensors and these little
[56:14.180 --> 56:19.420] sensors were strewn all over Vietnam along the Ho Chi Minh Trail, the roots by which the
[56:19.420 --> 56:24.900] North Vietnamese were smuggling manpower and material into the South to the Vietnam War
[56:24.900 --> 56:29.580] guerrillas to fight the Americans in South Vietnam.
[56:29.580 --> 56:32.980] And if this looks familiar, that's because of course it's sage.
[56:32.980 --> 56:37.060] These are the sage computers repurposed and the thinking was, well okay, sage isn't fast
[56:37.060 --> 56:43.260] enough to shoot down ICBMs, but maybe we can use it to shoot the Vietnam War trucks.
[56:43.260 --> 56:46.660] And people walking through the jungle.
[56:46.660 --> 56:52.980] So the sensors in the jungle would go off, would trigger blips of light on the screens back
[56:52.980 --> 56:54.980] at the ISC.
[56:54.980 --> 56:56.660] They would radio coordinates to jets.
[56:56.660 --> 57:00.660] The jets would fly out, the fighter jets would fly out to the grid square on the map and
[57:00.660 --> 57:04.820] they would drop bombs on the jungle in that spot.
[57:04.820 --> 57:09.220] The Air Force bragged that the whole process could happen in as little as five minutes and
[57:09.220 --> 57:13.580] then many cases no American would ever see the target at all.
[57:13.580 --> 57:14.580] It was that automated.
[57:14.580 --> 57:18.860] I mean the pilot would fly over but he wouldn't see the thing he was dropping the bomb on.
[57:18.860 --> 57:21.780] And you might ask was that a good thing or a bad thing?
[57:21.780 --> 57:26.660] Is that in software terms, is that a feature or a bug if you can drop the bomb without any
[57:26.660 --> 57:30.140] human being ever laying eyes on the target?
[57:30.140 --> 57:32.940] The Air Force said that igloo white was a massive success.
[57:32.940 --> 57:39.660] They claim to have destroyed over 35,000 North Vietnamese trucks.
[57:39.660 --> 57:43.140] If that's true, I mean the operation cost about a billion dollars a year.
[57:43.140 --> 57:52.260] So even if that's true, the US was spending about $100,000 for every truck that it destroyed.
[57:52.260 --> 57:58.580] But even those numbers are kind of suspect because the Senate pointed out in 1971 that there
[57:58.580 --> 58:05.380] were not 35,000 trucks in all of North Vietnam.
[58:05.380 --> 58:09.660] Also daytime recon flights, they never seemed to see the destroyed trucks afterwards.
[58:09.660 --> 58:13.860] The Air Force said well the Vietnam must have dragged them off into the jungle.
[58:13.860 --> 58:17.220] Which would seem like a strange thing to do.
[58:17.220 --> 58:21.380] Of course the VC had learned to confuse or hack the sensors.
[58:21.380 --> 58:26.940] They could set off the sensors, get out of the way, let the bombs drop, come through,
[58:26.940 --> 58:31.020] make the Air Force busy and traverse at their leisure.
[58:31.020 --> 58:39.220] I'll skip this one and just say there's lots of examples like this, lots of stories like
[58:39.220 --> 58:46.140] this about how the jungles and villages of Vietnam did not fit into the flowcharts and the algorithms
[58:46.140 --> 58:47.780] that they were supposed to.
[58:47.780 --> 58:53.940] And efforts to make them fit again and again caused carnage and destruction.
[58:53.940 --> 59:01.060] The US failure in Vietnam is a parable for the limits of this kind of simulation.
[59:01.060 --> 59:07.500] For there's this allure to push button war, this dream that we can make conflict painless
[59:07.500 --> 59:09.260] and bloodless.
[59:09.260 --> 59:12.780] But the real world bites back in all kinds of ways.
[59:12.780 --> 59:18.700] The world refuses to be perfectly modeled by computers, by algorithms, by equations
[59:18.700 --> 59:21.500] and social science.
[59:21.500 --> 59:26.580] This is where I reassure my students, I'm not anti social science and I'm not anti video
[59:26.580 --> 59:27.580] games.
[59:27.580 --> 59:29.340] I love games, I'm a gamer.
[59:29.340 --> 59:35.340] I don't think simulations are evil because they were used to fight a bad war.
[59:35.340 --> 59:40.500] But the games we play have histories and the tools we use have histories.
[59:40.500 --> 59:45.300] If you have a hammer, every problem looks like a nail and if you have satellites and drones
[59:45.300 --> 59:51.300] and smart bombs, that shapes the way you see the world's problems too.
[59:51.300 --> 59:57.620] And the Cold War was a time when the simulation threatened to become more real than the reality.
[59:57.620 --> 01:00:04.580] Computer models, war games, statistical simulations, these were abstractions that had real costs,
[01:00:04.580 --> 01:00:09.940] real political costs, real costs in blood and lives.
[01:00:09.940 --> 01:00:17.100] We've left the Cold War behind, of course, I hope, but not the computers or the games,
[01:00:17.100 --> 01:00:20.980] and not the dream of push button war.
[01:00:20.980 --> 01:00:24.780] Not the inflated faith in models and simulations.
[01:00:24.780 --> 01:00:30.100] I was pleased when I got this, these are pilots flying remote drones, dropping shooting missiles
[01:00:30.100 --> 01:00:32.500] over Afghanistan.
[01:00:32.500 --> 01:00:39.180] I've used this slide in a couple of talks and then only recently I read that this image
[01:00:39.180 --> 01:00:41.780] of a predator drone is not real.
[01:00:41.780 --> 01:00:47.620] It's a 3D model that somebody cooked up in Photoshop or in a 3D rendering program.
[01:00:47.620 --> 01:00:51.580] But if you search Google image search for predator drone, this is the one that comes up.
[01:00:51.580 --> 01:00:53.500] This is the one you see in all sorts of stories.
[01:00:53.500 --> 01:00:58.860] What I think is just another kind of metaphor for how, I don't know, that I couldn't even
[01:00:58.860 --> 01:01:00.180] find a picture of a drone.
[01:01:00.180 --> 01:01:05.140] Instead I found this, not knowing, I found this 3D computer simulation of a drone.
[01:01:05.140 --> 01:01:13.100] This metaphor for the way in which our tools distance us from real engagement with what
[01:01:13.100 --> 01:01:16.100] is actually happening out there in the world.
[01:01:16.100 --> 01:01:17.740] That's not just the military, it's all of us.
[01:01:17.740 --> 01:01:23.380] We see it in public policy, we see it in economics, even in our own day-to-day personal interactions.
[01:01:23.380 --> 01:01:27.900] The way we see the world, the way we think it works is shaped by our interactions through
[01:01:27.900 --> 01:01:33.980] computers, through simulations, models, and games.
[01:01:33.980 --> 01:01:38.140] Video games replace a real world, not just video games, computers replace a real world
[01:01:38.140 --> 01:01:44.500] that is big and messy and complicated with a little toy world that is knowable and controllable
[01:01:44.500 --> 01:01:46.740] and predictable.
[01:01:46.740 --> 01:01:47.740] That's why we love them.
[01:01:47.740 --> 01:01:52.300] I mean, that's why your children, your grandchildren play video games for that very reason.
[01:01:52.300 --> 01:01:54.140] We can't blame computers for that.
[01:01:54.140 --> 01:01:56.260] That's what we built them to do.
[01:01:56.260 --> 01:02:01.300] Computers will always pull us back from reality towards abstraction, towards simulations.
[01:02:01.300 --> 01:02:04.500] That's their whole purpose.
[01:02:04.500 --> 01:02:09.980] They'll tempt us to doubting or diminishing the role of chaos and contingency in human
[01:02:09.980 --> 01:02:11.780] lives.
[01:02:11.780 --> 01:02:13.740] So I don't blame computers for that.
[01:02:13.740 --> 01:02:19.060] I say what they're for, but we have to recognize that tendency and guard against it.
[01:02:19.060 --> 01:02:25.180] And I think we can only do that by seeing that the toys we play with, the tools we use,
[01:02:25.180 --> 01:02:29.540] have a history, that they are always going to be artifacts of the history that created
[01:02:29.540 --> 01:02:36.460] them, the political moment that created them, just as we are artifacts of that history.
[01:02:36.460 --> 01:02:37.460] Thank you.