THOMAS HAIGH: I mean, Klára von Neumann, she’s like at Los Alamos, as someone with absolutely no training in physics or mathematics, talking one-on-one with Nobel prize winners, the people who invented nuclear weapons and like telling them, “This is what the simulation did. Here's my ideas for it. Here’s what these calculations mean…”
Which is pretty incredible.
KATIE HAFNER: I’m Katie Hafner, and this is Lost Women of Science, where we unearth the stories of scientists who haven’t gotten the recognition they deserve.
This season, we focus on Klára Dán Von Neumann, who wrote some of the earliest lines of computer code, in the 1940s.
In the last episode we took you deep into the workings of the ENIAC, that groundbreaking early electronic computer Klári programmed.
This episode is all about what Klári’s programs actually did. And it raises some uncomfortable questions–particularly now–because the force pushing the burgeoning field of computing forward was the nuclear arms race.
We’re zooming in on a particular moment in time. A time span of about seven years after the second World War…It’s a time when a lot of big questions were cropping up: questions about the ethics of weapons development and what the U.S. had done to Japan.
And it’s a time before Los Alamos had its own computer…when the ENIAC, which lived in Aberdeen, Maryland, was the most powerful machine around. So, this post-war moment…it was a strange one for Los Alamos.
ANNE FITZPATRICK: Roughly for about a year and a half after the war ended, the lab’s future was indeed uncertain.
KATIE HAFNER: That’s Anne Fitzpatrick, a technical director at the Department of Defense. Her PhD thesis on computers and the hydrogen bomb is one of the few declassified sources we have about that bomb’s development.
ANNE FITZPATRICK: There was one school of thought that the lab should be closed permanently. People felt they’d completed the wartime mission–close it up, make it a monument or landmark.
KATIE HAFNER: The destructiveness of the Hiroshima and Nagasaki bombs was at once immediate and lingering.
Several of the scientists who worked on these weapons developed grave misgivings about what they’d done…
Some of them left Los Alamos because of this. It was almost as if they were recanting.
But, scientists were deeply divided on this.
ANNE FITZPATRICK: And then there was another school of thought. They wanted to, to continue work on nuclear weapons development.
We had so many refugee scientists who had fled fascist Europe and the rise of Hitler and they were terribly afraid of the Russians.
KATIE HAFNER: One of these refugee scientists was John von Neumann.
The United States and the Soviet Union had been allies during the war, but soon after, their relationship turned hostile. The U.S. had one piece of insurance, and that was its exclusive possession of nuclear weapons–but that was bound to change at any moment, as the Soviets came closer to perfecting nuclear weapons of their own.
NIC LEWIS: The thought was we can't let the Soviets have a monopoly on either atomic or thermonuclear weapons. We need to get there first, so that we're in a position of strength, uh, against Stalin.
KATIE HAFNER: That’s Nic Lewis, the historian of technology at Los Alamos from earlier episodes.
Despite how destructive and harrowing and tragic the atomic bombs dropped on Japan were, they were primitive and inefficient compared to what was possible.
NIC LEWIS: The bombs that had been built during the war were very crude. They weren't suitable military weapons. They were more laboratory pieces.
KATIE HAFNER: So, those who remained at Los Alamos turned their sights to building better nuclear weapons. But, there were roadblocks.
THOMAS HAIGH: They have limited amounts of uranium and plutonium, and they want to produce weapons that can make larger explosions using smaller amounts of material.
KATIE HAFNER: That’s Thomas Haigh, the history professor you’ve been hearing throughout the season. He co-authored ENIAC in Action, a book that gives us the most thorough account of Klári’s work.
THOMAS HAIGH: You can't just be like the Hiroshima bomb and use a large lump of enriched uranium. Only one percent of that fissioned during the millionth of a second it took to destroy itself.
KATIE HAFNER: Fission is the process that powered the atomic bomb–and that powers nuclear energy today. Fission occurs when a neutron collides with an atomic nucleus, breaking it apart. That nucleus then shoots out a few more neutrons that go on to split more nuclei. Meaning that fission creates a chain reaction–the splitting of one atom causes the splitting of another and another...The result is the release of a huge amount of energy–an explosion.
A perfect nuclear weapon would maximize fission, releasing the most energy using the least amount of material.
THOMAS HAIGH: But just taking like 10 different designs, making 10 different test weapons and exploding them in the desert, would use up the entire stockpile. You need to be more creative.
KATIE HAFNER: When you can’t learn by doing–well, that’s where math is useful.
Those scientists who wanted to stay at Los Alamos and continue advancing nuclear weapons technology–this is what they were trying to do. They wanted to figure out, through some equation, how to up the fission, so that they could build more efficient weapons.
THOMAS HAIGH: And none of them could come up with a brilliant equation like that, that was going to tell them the answer of what is the most efficient way to configure our weapon.
KATIE HAFNER: The problem was, fission has so many tiny moving parts–literally. By definition. And the chain reaction it causes doesn’t happen all at once–it unfolds over time–granted, a tiny fraction of a second. But still, one event begets another, begets another. No single equation could account for everything that is happening as these chaotic uranium atoms bounce around.
This is where someone named Stanislaw Ulam enters the picture.
ANANYO BHATTACHARYA: He's a Polish mathematician and one of von Neumann’s very best friends.
KATIE HAFNER: That’s Ananyo Bhattacharya, who wrote a new biography of John von Neumann titled The Man from the Future.
In 1946, Stan Ulam was diagnosed with viral encephalitis. The order from his doctor was a mathematician’s worst nightmare: stop thinking.
ANANYO BHATTACHARYA: Now he’s told by his doctors to rest his inflamed brain, but Ulam can't help himself. And he begins to play solitaire because he's so bored. And he goes um, “hmm, I wonder if I can work out what the chances of, of winning solitaire are”
And he, he just realizes the, the numbers, they quickly get out of hand. There are too many possible card combinations.
KATIE HAFNER: He can’t figure out a single equation that would account for all the card combinations that could possibly occur.
ANANYO BHATTACHARYA: So then he starts thinking a bit differently about it. He says, maybe the best way of figuring out what the chances are of winning is just to play as many hands as possible and note down the results.
And so this is a statistical method. And this is, uh, an incredible insight because it becomes a way of simulating almost any complex physical process that you can’t do just by doing the maths directly.
KATIE HAFNER: Stan Ulam shared this idea with John von Neumann during his next visit to Los Alamos and they quickly developed something we now call the Monte Carlo method.
We talked about Monte Carlo in the first episode. Remember, the Monte Carlo method involves simulating a course of events, and letting it play out over and over again. In doing this, you get a sense of what could happen, and what is likely to happen. In Tom Haigh’s words:
THOMAS HAIGH: Let’s take these complicated probabilities, simulate them a bunch of times and see what the distribution of outcomes is.
KATIE HAFNER: And Stan Ulam realized that this method of simulating a course of events, over and over again…it’s perfect for figuring out the probabilities of winning a hand of solitaire.
Because, in solitaire, there are just so many different possible outcomes.
And, in this respect, solitaire actually has something in common with fission: in both cases, one event closes off some possibilities and opens others. In solitaire, this event is playing a new card. In fission, it’s the splitting of a nucleus.
ANANYO BHATTACHARYA: It turns out while you can't solve mathematically the problem of neutrons bouncing around inside a bomb definitively with mathematics, because it's just too complicated, what you can do is use the Monte Carlo method.
KATIE HAFNER: When you use the Monte Carlo method to figure out what happens inside a bomb, you follow the path of individual neutrons. In doing so, you learn how likely it is that a neutron will undergo fission, and whether that fission will trigger a chain reaction. If enough chain reactions occur in the simulation, the bomb will explode in real life.
THOMAS HAIGH: If a neutron is traveling at a certain speed through the core sector of the weapon, what are the odds it has hit another nucleus. If it hits another nucleus, what are the odds that it causes a fission to continue the chain reaction or that it's just harmlessly absorbed?
KATIE HAFNER: The scientists at Los Alamos were asking these questions because they wanted to optimize nuclear weapons. And the Monte Carlo Method could provide the answers. The problem now was just: Well how do you do it…
THOMAS HAIGH: So obviously conceptually, there's nothing in the process of Monte Carlo simulation that has to be done with a computer, but because it involves carrying out this laborious chain of simulation over and over again, if humans were doing it, it would take them a long time.
KATIE HAFNER: It was pretty much impossible to do these simulations without a computer…
But, there was no sophisticated computer on site at Los Alamos. The lab needed people off site with access to powerful computers. So Los Alamos was…
THOMAS HAIGH: …much more reliant on people who are not even necessarily there full-time and working as contractors.
KATIE HAFNER: And because this is a brand new field, when the lab calls on these off-site contractors, they define the work of computer programming as they go.
THOMAS HAIGH: It lowers the barrier of entry for getting other people involved.
KATIE HAFNER: In this post-war moment, Los Alamos became…
THOMAS HAIGH: A netherworld that was at once nowhere and everywhere.
KATIE HAFNER: A lab with employees spread out across the country, where secrecy and circumstance meant that typical rules and requirements did not apply.
NIC LEWIS: And that's when Klári von Neumann comes into the picture.
KATIE HAFNER: That's Nic Lewis again.
Klári’s connection to Johnny meant she had access to a computer. So in the summer of 1947, she joined this Los Alamos netherworld from all the way in Princeton…She was a remote worker before working remotely was even a thing.
NIC LEWIS: And the lab hired Klári von Neumann as an outside consultant to write the program for the very first Monte Carlo calculation. The issue was, well, how do we do that? We’ve never coded for something like this before.
KATIE HAFNER: Fortunately, there was now a machine that was up to the challenge: the ENIAC, the first general purpose electronic computer. If you listened to our last episode, you’ve definitely heard of the ENIAC.
Right around the time Stan Ulam and John von Neumann developed the Monte Carlo method, Johnny decided to reconfigure the ENIAC so that it could run bigger, more complex problems. By which I mean…Monte Carlo nuclear weapons simulations.
Klári was brought in to do something that had never been done before: create a series of Monte Carlo simulations that the ENIAC could understand. In essence, her job was to turn physics problems into computer code.
After months of working on this from her desk in Princeton, in April of 1948, Klári arrived at the Army’s Aberdeen Proving Ground in Maryland, where the ENIAC was.
Klári called each group of problems executed on the ENIAC, a “run” – as in, “running a program.”
This first run, which happened in April and May of 1948, consisted of trial problems to test if the simulation worked.
And this first run remains completely historic.
THOMAS HAIGH: There’s reason to believe this code was run before any other modern style code was run on any computer.
KATIE HAFNER: This modern style code Tom refers to is the forebear of the code run on your computer, your phone, your smart watch.
Klári had started out just writing that code, but…
THOMAS HAIGH: What we see is her role rapidly expands from that. She's very hands-on, completely involved all the way through the process of writing, running and interpreting the results of the code.
KATIE HAFNER: Klári even wrote the first ever report on Monte Carlo computer simulations.
THOMAS HAIGH: If you look at the written report, it still stands up as a very nice, concise description of what’s happening.
KATIE HAFNER: But Klári paid a price.
THOMAS HAIGH: She certainly was deprived of sleep. The machine was working 16 hours a day. It had been like very intense, very challenging. And then, she'd lost 15 pounds during that month. She’s, you know, somehow like, so shaken up that she's, you know, in ill health, getting medical checkups, tests, and treatments afterwards.
So there has to be that kind of tension that this work is making her like physically and mentally ill, but at the same time, she must be finding it satisfying that she, she keeps doing it.
KATIE HAFNER: And we see this in the historical record. As soon as she recovered from that first run, Klári got back to work.
Five months later, she returned to Aberdeen for the second run. And this was the real deal…
THOMAS HAIGH: With the second run, that's the one where they're doing work with real weapons configurations for the first time.
So they cared about the specific results of this, not just as a proof of principle, but, you know, for the basis of, um, extremely important decisions involving millions of dollars-worth of enriched materials and national security.
KATIE HAFNER: The stakes were really high. And as the person who had done the coding and overseen the run, Klári traveled alone to Los Alamos to share the results with the scientists there.
THOMAS HAIGH: Something goes wrong with that, in terms of a bug.
KATIE HAFNER: In one of the problems of the second run, there was an error in the code.
THOMAS HAIGH: And because of that mistake, the results for those problems turned out not to be reliable.
KATIE HAFNER: Word of that mistake got around. And this created a general sense of skepticism about the viability of any Monte Carlo computer simulation.
THOMAS HAIGH: So, there's the feeling there that the whole usefulness of this as a serious tool is very much up in the air.
KATIE HAFNER: Not only had Klári made a coding error but now this failed calculation put all of this Monte Carlo work in jeopardy.
THOMAS HAIGH: And the letters that John von Neumann sent her during this trip also give a picture of someone in mental distress.
JOHN VON NEUMANN: Was it wrong to let you go alone to Los Alamos?
THOMAS HAIGH: John von Neumann said he was scared out of his wits, um, that the stress would leave her, as he put it, ruined physically and emotionally.
JOHN VON NEUMANN: But I believed that this was what you wanted most. That this was the proof of your intellectual independence…
THOMAS HAIGH: So, you know, you imagine that situation, right? You're someone who, you don't have a PhD, you don't have a professorship, you don't have any training in mathematics. You're doing this work because you're someone's wife. You've done it. There are problems with it. You're going to Los Alamos. There are Nobel Prize winners there, and you're telling them what you found and they're like asking tough questions. Right? I mean, I, you know, would be pretty terrified in that circumstance.
KATIE HAFNER: And you're a man.
THOMAS HAIGH: Yeah. So add it. You're, you're a woman with no qualifications. Who's plagued with depression and anxiety.
KATIE HAFNER: Imagine it: You’re writing code for the most destructive weapons ever created…It’s a project that the government is funneling millions of dollars into. You have limited time with the computer. And then you have to present your findings to a room full of the smartest, most powerful scientists in the world.
Every number she wrote down, she had to defend. And the stress got to her.
But from everything I’ve learned about Klári at no point in her life did she allow herself to feel defeated for very long. So, in May of 1949, Klári led the work on the third run of Monte Carlo fission simulations.
THOMAS HAIGH: Klára von Neumann took a very prominent role again, in this one.
KATIE HAFNER: During the third run, Klári trained some new recruits. Among them was physicist Harris Mayer and his wife Rosalie. Ten years ago, George Dyson, a science historian featured in earlier episodes, spoke with Harris about his experience…
HARRIS MAYER: And I told him I don't know anything about machines.
KATIE HAFNER: That’s Harris Mayer. He was concerned about his lack of experience, but the higher ups at Los Alamos assured him that this was okay. He’d be trained by the veterans…meaning Klári.
HARRIS MAYER: And Klári was one of the first people that knew how to program. And she felt she was very important in this. So Klári thought that we were taking things away from her by taking her baby, which is programming. And even surpassing her very easily, and this really disturbed her.
KATIE HAFNER: Think about Harris’s position here; this is the 1940s. He has a PhD, and he worked on the Manhattan Project, but when it comes to computers, he’s a newbie. In comes a young woman with a high school diploma showing him the ropes.
Now think about how Klári must have seen this. Here’s a man–a smart, accomplished, credentialed man. It might have crossed her mind that she was training her replacement.
And her fears were valid. Klári, like some of the other women she was working with, was brought into this world by way of a husband. This set-up was actually common…It even has a name:
NIC LEWIS: Sure, the, the couples phenomenon.
KATIE HAFNER: The couples phenomenon…This was a common way women got jobs: teaming up with their spouses.
On the one hand, these husbands really believed in their wives and they vouched for them, giving them access to this groundbreaking work.
But on the other hand, this circumstance made these women dependent on men and their approval. Husbands held the keys to their wives’ careers. And the wives could be locked out at any moment.
It is important to note, though, that at this moment, Klári wasn’t actually being pushed out. In fact, during the third run…the run Harris Mayer was talking about…
THOMAS HAIGH: She has these different kinds of expertise and skills that make people respect her and let her make these connections on her own, organizing the third run with very little involvement as far as we can see with John von Neumann.
KATIE HAFNER: And Klári would remain an authority on ENIAC coding for the next few years.
THOMAS HAIGH: So for that period, she was absolutely essential as one of the main contributors to Los Alamos and its use of ENIAC at a time when ENIAC was the only programmable, electronic computer available anywhere in the world.
KATIE HAFNER: Coming up, Klári’s last big computer run. I’m Katie Hafner and this is Lost Women of Science.
THOMAS HAIGH: And then the final block of calculations that Klára von Neumann was deeply involved with were the calculations in 1950 for the Super. So that was Teller’s design for a hydrogen bomb.
KATIE HAFNER: In the first three Monte Carlo runs, Klári had been working on simulations for weapons that relied on fission, the process we described earlier. Now, in the last leg of her coding work, Klári began work on the hydrogen bomb–also called a thermonuclear weapon, or “Super.” This relied on something called fusion.
The hydrogen bomb was the brainchild of Edward Teller, another Hungarian-born theoretical physicist.
THOMAS HAIGH: From what I read about him, it seems like he was a bit like Steve Jobs with the reality distortion field.
KATIE HAFNER: The folks at Apple would throw around this phrase to describe how Steve could convince himself and others of pretty much anything, almost like he was bending reality. Just like Steve Jobs, Edward Teller got a lot of people to carry out his vision.
THOMAS HAIGH: He would have a strong opinion that something would work and argue for it, charismatically and forcefully. And I think to some extent, just wear people down with perseverance.
KATIE HAFNER: But, while Jobs’s vision was a phone slash computer you could carry around in your pocket, Teller’s vision was the most destructive weapon ever imagined.
And at that time…
THOMAS HAIGH: He believed that he had a design that would produce a hydrogen bomb working by fusion rather than fission and creating an enormously greater release of explosive force than the atomic bombs that were used on Hiroshima and Nagasaki.
KATIE HAFNER: Fusion is basically the opposite of fission. Instead of splitting a nucleus, two nuclei fuse to form a larger, heavier one, releasing energy. And while a fission weapon uses radioactive elements like uranium and plutonium, the particles in a fusion process are hydrogen molecules. Hence: hydrogen bomb.
Fusion is much much harder to achieve than fission. In fact, we are still struggling to create fusion power plants today. The reason for this is that the fusion process requires extreme heat. To give you a sense of just how much heat, fusion is the process that powers the sun. In 1950, the idea of that much power was fantastical…and alarming.
NIC LEWIS: There was a fear of what a thermonuclear device could do.
KATIE HAFNER: That’s Nic Lewis again.
NIC LEWIS: So the thought was, well, we need to see if it's even feasible. Will one of these even work? If not, then there’s no need for us to worry about it.
KATIE HAFNER: Los Alamos had been researching thermonuclear weapons since the early 1940s. But because fusion was harder to accomplish, fission had been its main priority.
ANNE FITZPATRICK: And then of course, 1949, the USSR detonates its first fission device.
KATIE HAFNER: That’s Anne Fitzpatrick again–She’s a nuclear weapons expert. Because the Soviet Union now had nuclear capabilities, the U.S. needed something greater…
ANNE FITZPATRICK: The attitude at the lab was like, okay, all hands on deck. We need every single person working on this problem because we, you know, we've got to develop the H bomb.
There were people who had ethical concerns, but, um, most of the people I've talked to felt pretty strongly in favor of building the H-bomb.
KATIE HAFNER: Plus, those who stood strongly against developing more powerful weapons had simply left Los Alamos at the end of the war.
We don’t know what Klári thought of the news that the Soviets were developing nuclear weapons. And we don’t know what she thought of the work she was doing. We don’t even know if she was given all the details.
What we do know is that her day-to-day work must have felt only distantly related to its implications…
THOMAS HAIGH: For Klára von Neumann whose entry point to this is a flow diagram where everything has already been abstracted to mathematics in boxes. And then her job is to turn that into computer code. I think she’s so far away from even the materiality of an atomic weapon, let alone the effects of dropping it on people or the foreign policy questions of, you know, what should be happening that I think she’s experiencing it completely as an interesting intellectual challenge.
KATIE HAFNER: Tom’s so right…Klári was so removed from the consequences of it all. She was working on a simulated reality…in a world that existed within the computer. It’s another netherworld…something that is at once there and nowhere.
That’s a danger of computer simulations, both then and now. They create this distance between the programmer and the real world, between code and its consequences.
Many of the scientists who surrounded Klári didn’t see this disconnect as an issue. Stan Ulam, the guy who had the idea for Monte Carlo, once said to his wife Francoise, “Even the simplest calculation in pure mathematics can have terrible consequences…What would Archimedes and Newton have done if they had cared about the consequences of their thoughts?”
And so, in this theoretical way, Klári kept working towards a terrifying potential reality: a workable hydrogen bomb.
In the summer of 1950, her work on the Super began in Aberdeen. And unfortunately for Edward Teller…
THOMAS HAIGH: Those calculations put the final nail in the coffin of the idea that his plan for a hydrogen bomb was workable.
KATIE HAFNER: But Teller, a fervid hawk, didn’t give up. He soon teamed up with Stan Ulam, and…
THOMAS HAIGH: He and Ulam get together and come up with the design for an actual workable hydrogen bomb using different principles. And that's the one that the hydrogen bombs that we have today are directly descended from.
KATIE HAFNER: In her memoir, Francoise Ulam, Stan’s wife, writes that the day the two men figured this out is etched in her memory.
She came home to find her husband staring blankly out the window. In a small voice, he said to her: “I found a way to make it work. It will change the course of history.”
A hydrogen bomb has never been used in war. It’s hard to comprehend the level of destruction it would cause. It’s at least a hundred times more deadly than the atomic bombs dropped on Japan. And we’re sitting on a stockpile of these bombs…banking on nuclear deterrence, just hoping they’ll never be used.
After Klári's super calculations in 1950, something changed.
THOMAS HAIGH: By the summer of 1951, the Institute for Advanced Study’s computer was finally becoming operational. And one of the computers that was modeled on it at Los Alamos was also nearing completion there and I believe started to run in 1952.
KATIE HAFNER: The ENIAC, a kind of Frankenstein machine made of older parts, was always on the verge of being made obsolete by newer technologies. And finally, it happened; by 1952, both the Institute for Advanced Study and Los Alamos had their own more powerful computers. And the ENIAC went from being the most powerful machine in use, to a large, clanging historical artifact.
Pieces of it are spread all over—for instance, there are pieces at the Smithsonian, the Institute for Advanced Study, the Computer History Museum…etc. And for Klári…
THOMAS HAIGH: So it’s clear that her expertise was still valued by Los Alamos as they get ready to shift to their own computer. But at the same time, I think it was in many ways, a natural end point for an involvement because now they have their own computer on site. They're not going to be relying on outside contractors the same way they were when they were scavenging time on computers around the country.
KATIE HAFNER: And this is the moment when Los Alamos began training on-site programmers. So, this netherworld, as Los Alamos had been during the postwar period, it became something much more conventional.
THOMAS HAIGH: What’s unusual in Klára von Neumann's case is she’s not coming from a scientific discipline. And I think it's that kind of transition into scientific computing that goes away as the field becomes more professional.
KATIE HAFNER: As the field started to professionalize, Klári’s coding work started to taper off.
At Los Alamos, computing, coding…This was initially open to someone like Klári -- a consultant who happened to be in the right place at the right time. But once the expectations and requirements for the work became more standardized, this started to change.
The netherworld and its shadowy opportunities were replaced with the real world, complete with its credentials, standard procedures, limitations and biases. But Klári did have another big contribution to make before she left the field for good. She helped convert her original Monte Carlo codes to run on Los Alamos’s new computer, the MANIAC …
NIC LEWIS: So in a way, her work in developing the Monte Carlo for the ENIAC survived into the work done on the MANIAC, which has survived since then in the continued work at the lab today in computing.
KATIE HAFNER: Monte Carlo isn’t some obscure algorithm used only at Los Alamos–far from it. In the past 74 years since those first ENIAC runs, the Monte Carlo method has been used in just about every professional field you can think of. Election predictions – that’s Monte Carlo. Weather forecasts – that’s it too.
Monte Carlo is used to determine risk in financial investments, to estimate the chances of a nasty, new COVID variant–and so much more.
And Klári was there right from the start. She helped put all of this in motion.
After 1952, Klári did continue to share her expertise with scientists and programmers at Los Alamos, but she never had the same leadership role that she took on in those early Monte Carlo runs.
Her gigs got fewer and fewer. And then, just a few years later, in 1955, something took her out of the work entirely.
KLARA VON NEUMANN: On the 9th of July of that exceptionally hot summer, Johnny collapsed.
KATIE HAFNER: At first, Johnny was told he was suffering from nervous exhaustion. But, after returning to the doctor less than a month later, he got a new diagnosis.
KLARA VON NEUMANN: This was cancer in an already very advanced and metastasized form.
THOMAS HAIGH: And that’s probably a pretty hard stop on Klára von Neumann having, you know, the ability to engage in this kind of correspondence. You know, it’s going to, like, just shake everything up, fundamentally.
KATIE HAFNER: Johnny’s diagnosis was bone cancer, and the cancer had already spread to many parts of his body. He soon needed a wheelchair to get around.
In March of 1956, he was admitted to Walter Reed hospital in Washington D.C.
Johnny deteriorated rapidly.
MARINA VON NEUMANN WHITMAN: I don't think anything comforted him. He was simply inconsolable about the fact that he was dying,
KATIE HAFNER: That’s Marina von Neumann Whitman, Johnny’s daughter. By all accounts, Johnny was frightened. In a surprise to everyone close to him, he even turned to Catholicism, asking to see the hospital’s priest.
When the cancer affected his brain, he started sleep-talking in Hungarian.
MARINA VON NEUMANN WHITMAN: He tried to do simple arithmetic and couldn’t, and it was more than I could handle emotionally.
KATIE HAFNER: Over the year and a half between Johnny’s diagnosis and his death, Klári was his unflagging caregiver. As part of that job, she facilitated final visits from everyone in his life. But…
KLARA VON NEUMANN: There was really only one man, one person whom he wanted to see.
KATIE HAFNER: And that was the mathematician Oswald Veblen, who’d been a father figure to Johnny, helping him get his jobs at Princeton and at the Institute for Advanced Study. Klári wrote letters begging him to come. But in the end…
KLARA VON NEUMANN: He did not come.
KATIE HAFNER: Johnny’s hawkish tendencies, his willingness to push forward with a weapon still deadlier than those dropped on Japan–that was just too much for some of his peers to bear…even those who had been closest to him.
KLARA VON NEUMANN: Veblen never forgave him.
KATIE HAFNER: That’s a harsh rejection for a man on his deathbed. And I can’t imagine how dark that entire episode must have been for Klári, as well. She was unable to give her husband his last wish.
John von Neumann died on February 8, 1957. He was fifty-three.
Johnny opened doors for Klári, into a brief career where her talents could shine. He showered her with praise and bolstered her confidence. But he did not do the work for her….
But for reasons of time, place and who knows what else, she didn’t continue to pursue a career in this thing she’d grown expert in.
As for the legacy of Klári’s work: I think of this less in terms of what it did, and more in terms of what it meant.
Klári wrote what is probably the first modern-style computer code ever run on a computer. In doing so, she wrote history. The origins of coding we do today can be traced to the programs that came from Klári’s own hand: the Monte Carlo runs she coded in the late 1940s.
Next time on Lost Women of Science, Klári’s brush with paradise.
This has been Lost Women of Science. Thanks to everyone who made this initiative happen, including my co-executive producer Amy Scharf, producer Sophie McNulty, associate producer Ashraya Gupta, senior editor Nora Mathison, composer Elizabeth Younan, and the engineers at Studio D Podcast Production.
Thanks also to our voice actors Eva Szabo and Nandor Tary, as well as our many Hungarian translators: Agi Antal, Rick Esbenshade, Charles Hebbert, Laszlo Marcus, Alina Bessenyey Williams, and Lehel Molnar.
We’re grateful to Mike Fung, Cathie Bennett Warner, Dominique Guilford, Jeff DelViscio, Meredith White, Bob Wachter, Maria Klawe, Susan Kare, Jeannie Stivers, Linda Grais, Rabbi Michael Paley, Marina von Neumann Whitman, George Dyson, Thomas Haigh, and our interns, Hilda Gitchell, Kylie Tangonan, Leeza Kopaeva, and Giuliana Russo. Thanks also to the Computer History Museum, to Paula Goodwin, Nicole Searing and the rest of the legal team at Perkins Coie, and to the Institute for Advanced Study, the Library of Congress, and the UCSD Special Collections for helping us with our search. Many thanks to Barnard College, a leader in empowering young women to pursue their passion in STEM for support during the Barnard Year of Science.
A special shout out to Celia Bolgatz at the Women’s Audio Mission in San Francisco, where this podcast was recorded.
Lost Women of Science is funded in part by the Gordon and Betty Moore Foundation, Schmidt Futures and the John Templeton Foundation, which catalyzes conversations about living purposeful and meaningful lives.
This podcast is distributed by PRX and published in partnership with Scientific American.
You can learn more about our initiative at lostwomenofscience.org or follow us on Twitter and Instagram. Find us @lostwomenofsci.
Thank you so much for listening. I’m Katie Hafner.
Katie Hafner was a longtime reporter for The New York Times, where she continues to be a frequent contributor. Katie is uniquely positioned to tell the stories of lost women of science. Not only does she bring a skilled hand to complex narratives, but she has been writing about women in STEM for nearly 30 years. She is the author of six books of non-fiction, and her first novel, The Boys, was published in July 2022 by Spiegel & Grau. Katie is also the host and executive producer of Our Mothers Ourselves, an interview podcast that celebrates extraordinary mothers.
This episode, Klára is not the computer scientist responsible for your phone's weather app.