Feynman Exposes Los Alamos Security Lapses, Criticality Risks
The Quagmire
The technical thrill of solving the atomic bomb caused Los Alamos scientists to stop questioning the moral consequences of their work until after the Trinity blast.
Richard Feynman recalls how excessive military compartmentalization nearly caused a nuclear criticality catastrophe at the Oak Ridge enrichment plant, which was averted only after Los Alamos demanded plant personnel be taught the fundamental physics of neutron moderation. (35:08)
By reorganizing enlisted personnel from the Special Engineer Detachment into an assembly line and explaining the physical purpose behind the calculations, Feynman boosted IBM punch-card computation output threefold in preparation for the Trinity test. (51:35)
Feynman exposed critical vulnerabilities in Los Alamos security protocols by exploiting mechanical tolerances and psychological biases to crack Mosler filing cabinets, ultimately unlocking all nine duplicate archives of atomic bomb secrets using Euler's number (27-18-28). (75:35)
Reflecting on the Trinity test and postwar celebrations, Feynman noted that scientific momentum and problem-solving euphoria blinded researchers to the moral consequences of the weapon, inducing deep personal anxiety about nuclear annihilation after the war. (67:15)
Detailed notes
## Background
Published on YouTube by the channel The Quagmire on July 12, 2016, this video presents a complete recording of Richard Feynman's classic lecture "Los Alamos From Below." Unlike most circulating copies that cut off early, this version includes an audience question-and-answer session where Feynman spends ten minutes recounting three distinct stories about cracking safes at Los Alamos National Laboratories.
## TL;DR
Feynman reflects on his time as a young, irreverent physicist recruited to the Manhattan Project, detailing both the chaotic early organization of the Los Alamos mesa and his clashes with military censors. He recounts solving dangerous criticality hazards at Oak Ridge, revolutionizing early computing workflows with Marchant calculators and IBM punch-card machines, and working alongside luminaries such as Hans Bethe, John von Neumann, and Niels Bohr. After describing his unshielded view of the Trinity detonation and the lingering postwar anxiety that followed the weapon's creation, Feynman closes with an entertaining breakdown of how he exploited human psychology and lock tolerances to crack high-security Mosler safes across the facility.
## Atmosphere & style
The lecture carries an intimate, conversational, and mischievous charm, filled with Feynman's trademark Brooklyn cadence, playful self-deprecation, and sharp comic timing. The audience is frequently in stitches as Feynman details bureaucratic absurdities, dorm pranks, and lock-picking antics, giving the feeling of an affectionate campfire yarn rather than a stiff academic address. Beneath the rollicking storytelling, however, runs a poignant emotional undercurrent—marked by the personal tragedy of his wife Arline's illness and a sobering, contemplative stillness when confronting the moral dread that settled over the scientists after Trinity.
## Bulletin
- [35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) Richard Feynman recalls how excessive military compartmentalization nearly caused a nuclear criticality catastrophe at the Oak Ridge enrichment plant, which was averted only after Los Alamos demanded plant personnel be taught the fundamental physics of neutron moderation.
- [51:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3095s) By reorganizing enlisted personnel from the Special Engineer Detachment into an assembly line and explaining the physical purpose behind the calculations, Feynman boosted IBM punch-card computation output threefold in preparation for the Trinity test.
- [01:15:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4535s) Feynman exposed critical vulnerabilities in Los Alamos security protocols by exploiting mechanical tolerances and psychological biases to crack Mosler filing cabinets, ultimately unlocking all nine duplicate archives of atomic bomb secrets using Euler's number (27-18-28).
- [01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) Reflecting on the Trinity test and postwar celebrations, Feynman noted that scientific momentum and problem-solving euphoria blinded researchers to the moral consequences of the weapon, inducing deep personal anxiety about nuclear annihilation after the war.
## Key takeaways
- [02:01](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=121s) Feynman initially declined recruiting efforts by Bob Wilson but joined the Manhattan Project out of concern that Nazi Germany might develop an atomic weapon first.
- [13:54](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=834s) Hans Bethe selected Feynman as his primary sparring partner because Feynman was completely unintimidated by authority and challenged physical assumptions directly.
- [13:54](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=834s) Feynman routinely engaged in mental calculation duels with Bethe, contrasting mechanical calculator operation with Bethe's rapid mental arithmetic and logarithmic interpolation.
- [22:24](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1344s) Strict and often irrational military censorship led Feynman into persistent battles with base authorities, demonstrating that absurd security regulations invited mockery rather than compliance.
- [32:22](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1942s) Physical security at Los Alamos was deeply flawed, with classified data originally stored in wooden desks and padlocked cabinets that could be easily bypassed.
- [35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) Military compartmentalization prevented Oak Ridge workers from understanding that water-moderated uranyl nitrate drastically lowers the mass needed for criticality, creating severe radiation hazards.
- [40:51](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2451s) Feynman forced the military to drop secrecy barriers at Oak Ridge by threatening that Los Alamos would renounce safety responsibility unless workers understood how neutron physics worked.
- [43:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2593s) While inspecting complex chemical schematics for Oak Ridge, Feynman accidentally identified an unvented valve symbol that averted dangerous accumulations of fissile material.
- [46:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2769s) To bypass repair backlogs, Feynman routinely disassembled and repaired broken Marchant mechanical calculators himself in violation of factory warranties.
- [51:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3095s) Informing Special Engineer Detachment soldiers about the real physical purpose of their IBM tabulator work transformed team morale and accelerated calculation completion speeds.
- [55:43](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3343s) To deliver the Trinity test bomb yield calculation in weeks, Feynman devised a color-coded sub-deck pipeline that isolated numerical errors without halting the calculation flow.
- [01:00:12](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3612s) Niels Bohr sought Feynman out for private consultations because Feynman cared only about the validity of the physics and was unafraid to tell Bohr when an idea looked crazy.
- [01:03:30](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3810s) Feynman observed the Trinity detonation directly through a truck windshield without dark glasses, knowing ordinary automotive glass would block harmful ultraviolet radiation.
- [01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) The euphoric scientific momentum of completing the bomb caused researchers to disregard moral implications until Robert R. Wilson voiced despair immediately following the test.
- [01:09:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4149s) Feynman discovered that Mosler combination locks had a ±2 digit margin of error per wheel, reducing the search space from 1,000,000 combinations down to 8,000.
- [01:11:37](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4297s) By casually reading dials on open safes and recording the final two numbers, Feynman built a catalog that allowed him to crack colleagues' safes in under three minutes.
- [01:15:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4535s) Feynman unlocked all nine filing cabinets containing duplicate Manhattan Project secrets by deducing that physicist Frederic de Hoffmann would use the mathematical constant e (27-18-28).
## Detailed walkthrough
### [00:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=13s) Introduction by Joseph Hirschfelder
The lecture begins with Joseph Hirschfelder opening the session, quipping that the venue might get raided by the police department before introducing the featured speaker: Richard Phillips Feynman, Nobel laureate in physics and "scientist extraordinary." Hirschfelder highlights the unique nature of Feynman's intellect, emphasizing that the difference between him and other people lies in his rare pairing of rigorous logic and expansive imagination. While most logical thinkers lack creative vision, Hirschfelder quotes the renowned Soviet physicist Yakov Frenkel, who observed that "knowledge progresses by illogical steps." He compares Feynman’s knack for probing scientific frontiers beyond the bounds of pure logic to the way mathematicians utilize complex variables to achieve analytic continuations.
Hirschfelder further recounts that during the Manhattan Project at Los Alamos, J. Robert Oppenheimer regarded Feynman as his "most valuable player." Beyond Feynman's extraordinary gifts as a theoretical physicist, Hirschfelder describes him as a "super duper troubleshooter" who could walk into an experimental laboratory and pinpoint exactly why an apparatus was malfunctioning simply by asking the right probing questions.
[02:01](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=121s) "Los Alamos from Below": Wilson Recruits Feynman
Taking the stage to applause, Feynman opens by dispelling any illusions about his standing during the war, wryly noting that Hirschfelder's flattering introduction is completely inappropriate for a talk titled "Los Alamos From Below." Feynman emphasizes that while he may be somewhat famous in the present day, he was a total nobody back then—a young graduate student at Princeton who had not even finished his Ph.D. thesis. Unlike higher-echelon leaders who wrestled with grand policy decisions, Feynman was merely "flittering about underneath somewhere."
```mermaid
flowchart TD
accTitle: Feynman's Path to the Manhattan Project
accDescr: Diagram illustrating Feynman's transition from graduate student to Los Alamos researcher.
subgraph Princeton["Princeton University"]
A["Working on Ph.D. thesis"] -->|"Bob Wilson pitches secret project"| B{"Initial refusal"}
B -->|"Contemplates Hitler & Nazi bomb threat"| C["Agrees to 3:00 PM meeting"]
C -->|"By 4:00 PM"| D["Calculates ion beam limits"]
D -->|"6-week writing leave"| E["Completes Ph.D. dissertation"]
end
subgraph Transition["Project Shift"]
E --> F["Princeton isotope method abandoned"]
F --> G["Mission to Chicago Met Lab"]
end
subgraph LosAlamos["Los Alamos"]
G --> H["Wilson & Manley trigger early move"]
end
```
He recalls sitting in his Princeton office when experimental physicist Bob Wilson walked in. Wilson revealed that he had received funding for a classified project; although sworn to secrecy, Wilson confided in Feynman because he knew Feynman would find the mission irresistible. Wilson explained the objective of separating uranium isotopes to ultimately build an atomic bomb, detailing an innovative electromagnetic separation technique he wished to develop. Feynman's initial reaction was resistance:
> “I said, ‘It's all right that you told me the secret because I'm not going to tell anybody, but I'm not going to do it.’”
Wilson nonchalantly countered that a meeting was scheduled for 3:00 PM anyway. Returning to his desk, Feynman managed to focus on his thesis for only about three minutes before anxiety set in. Pacing his floor, he realized the terrifying reality of Nazi Germany under Adolf Hitler and the genuine threat that the Germans might develop an atomic weapon first. Compelled by this fright, Feynman abandoned his thesis work and showed up at the 3:00 PM meeting.
[04:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=275s) Princeton Research and the Evaluation Committee
By 4:00 PM that same afternoon, Feynman had been assigned a desk and paper, rapidly calculating theoretical limits on ion beam currents to determine whether Wilson’s separation method was physically viable. Meanwhile, the experimentalists were cannibalizing equipment from their existing research projects to construct the test apparatus. Feynman watched the machinery mushroom across the room like a time-lapse film. Pure scientific research across the country ground to a halt as physicists repurposed their lives for wartime engineering. Feynman halted his own research as well, though he took a six-week leave of absence down the line to finish his dissertation and obtain his doctorate just before moving west.
Working on the Princeton project also gave Feynman his first taste of working among scientific titans. An evaluation committee convened to determine which uranium isotope separation method the government should pursue, bringing Feynman into discussions with figures like Richard C. Tolman, Henry DeWolf Smyth, Harold Urey, I. I. Rabi, Robert Oppenheimer, and Arthur Compton.
Feynman was initially startled by the committee dynamics. During one heated debate where Arthur Compton argued a valid point, several other committee members raised counter-objections one by one. Feynman expected Compton to aggressively defend himself and repeat his stance to the detractors. Instead, Tolman, serving as chairman, neatly summarized the discussion by incorporating every facet and viewpoint that had been raised, allowing Compton's insight to prevail naturally without repetitive argument. Feynman came away deeply impressed by how these senior scientists listened to each other and synthesized complex arguments without wasting words.
- The evaluation committee included leading scientific administrators: Richard Tolman (chairman), Henry D. Smyth, Harold Urey, I. I. Rabi, J. Robert Oppenheimer, and Arthur Compton.
- Wilson's isotope separation method was ultimately dropped in favor of other approaches (such as gaseous diffusion and Lawrence's calutron electromagnetic method), shifting the team's focus entirely to weapon design at Los Alamos.
[07:23](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=443s) Reconnaissance in Chicago and the Rush to Los Alamos
When the Princeton separation method was abandoned, the team learned that a new laboratory was being founded in Los Alamos, New Mexico, to design and assemble the atomic bomb. Because the site was not yet constructed, Bob Wilson sought to avoid wasting time by dispatching Feynman on an intelligence-gathering mission to the Metallurgical Laboratory at the University of Chicago. Feynman's assignment was to embed himself in each research group under the pretense of collaborating, learn their work in complete technical detail, and then move on to the next group once he had mastered the problem.
Feynman felt slight pangs of guilt about the ruse, but he redeemed himself almost immediately. While listening to a Chicago researcher outline a problem that had stalled the group for three months, Feynman asked, "Why don't you do it that way?" Within thirty minutes, the issue was solved.
Upon his return to Princeton, Feynman briefed his colleagues on the physics of the weapon, the expected energy yields, and the challenges ahead. His friend, the mathematician Paul Olum, noted the amusing contrast between Hollywood cinema and real wartime science:
> “He says, ‘When they make a moving picture about this, they'll have the guy coming back from Chicago telling the Princeton men all about the bomb and he'll be dressed in a suit and carry a briefcase inside and you're in dirty shirt sleeves and just telling us all about it.’”
With Los Alamos still lagging behind schedule, Bob Wilson traveled out to the mesa to investigate the delay. He discovered that the contractors, lacking clear instructions on laboratory requirements like water and gas lines, had simply built the movie theater and other familiar buildings instead. Wilson took charge on the spot, arbitrarily dictating the plumbing and gas layouts so laboratory construction could finally begin. Back east, Oppenheimer was entangled in administrative disputes with General Leslie Groves, making the young scientists increasingly impatient. Wilson called John Manley in Chicago, and together they made the executive decision to pack up their crews and head out to Los Alamos before the site was even finished. Oppenheimer actively recruited them, demonstrating immense patience with everyone throughout the chaotic transition.
[10:18](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=618s) Security Precautions and the Princeton Train Station
Richard Feynman recalls his early impressions of J. Robert Oppenheimer, noting how attentive the director was to personal matters: “He worried about my wife who had TB and whether there would be a hospital out there and everything.” It was Feynman's first time encountering Oppenheimer on such an intimate level, and he found him to be a wonderful man.
Security guidelines surrounding the departure were strict. The scientists were specifically warned never to purchase their train tickets at the Princeton station. The rationale was that Princeton was a tiny station, and a sudden surge of people buying tickets to Albuquerque, New Mexico, would immediately arouse suspicion. Feynman, thinking contrarianly, assumed that since everyone else would obey orders and buy their tickets elsewhere, buying his in Princeton would be harmless. However, when he arrived at the counter and asked for a ticket to Albuquerque, the ticket agent immediately exclaimed: “Oh, so all the stuff is for you. We've been shipping out crates full of counters for weeks and expecting they didn't notice that the address was Albuquerque.” Rather than blowing cover, Feynman realized his purchase at least provided the local agent an explanation for the mountain of freight heading southwest.
- Oppenheimer actively assisted Feynman in checking whether a local sanatorium or hospital in New Mexico could care for his wife, Arline.
- Project staff were ordered to buy train tickets in larger surrounding cities to preserve operational secrecy.
- The Princeton ticket agent had already observed weeks of heavy freight and crate shipments labeled for Albuquerque.
---
[11:18](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=678s) Arrival at the Mesa and Taking Over Construction
When Feynman and his colleagues arrived in New Mexico, they were well ahead of schedule. The dormitories, housing, and laboratories were far from ready. Because the scientists were pushing the schedule and arriving early, the project management scrambled, renting out ranch houses across the surrounding area. The scientists lived in these ranches initially and commuted up each morning. For an East Coast native unaccustomed to traveling, the scenery was staggering. Driving up toward the high mesa, Feynman marveled at the great cliffs and wondered out loud whether Native Americans had ever lived there. The driver promptly stopped the car, led him around a corner, and pointed directly to preserved Indian cliff dwellings carved into the rock.
Up on the mesa, the technical area was planned to be enclosed within an inner fence, surrounded by a town, which in turn would be enclosed within a larger outer perimeter fence. Because construction was actively underway, the boundaries were still wide open. Physicists found themselves stepping directly into logistics and civil engineering:
```mermaid
flowchart TD
accTitle: Physicist Roles During Los Alamos Site Setup
accDescr: How experimentalists and theorists split work when arriving at the unfinished Los Alamos site.
subgraph SiteSetup["Los Alamos Site Arrival"]
Arrival["Scientists Arrive Early<br/>(Labs & Dorms Unfinished)"]
end
subgraph Experimentalists["Experimental Physicists"]
Arrival --> Exp["No Active Apparatus"]
Exp --> Logistics["Directing Supply Trucks<br/>(Paul, John Williams)"]
Exp --> Build["Finishing Construction"]
end
subgraph Theorists["Theoretical Physicists"]
Arrival --> Theo["No Apparatus Required"]
Theo --> Housing["Live on the Mesa<br/>(Old School Lodges)"]
Theo --> Lectures["Robert Serber Lectures<br/>(Rolling Blackboard)"]
end
```
Feynman's assistant, Paul, stood at the entrance with a clipboard, directing delivery trucks across the site. Inside the unfinished laboratory, Feynman encountered renowned experimentalists whose work he had previously only read about in the *Physical Review*. He met John Williams standing over a desk covered in blueprints, sleeves rolled up, barking orders out the window to direct trucks and materials. Because their experimental apparatus had not yet been delivered or set up, the experimentalists simply took over the construction management to finish the job.
- The theoretical physicists chose to live directly on the mesa rather than in the ranch houses so they could start work immediately.
- The theorists shared cramped rooms with only a single rolling blackboard on wheels, moving it from room to room.
- Robert Serber delivered the introductory lectures detailing the theoretical bomb physics developed during the earlier Berkeley conferences.
- Feynman had previously worked on separate problems and spent late nights intensely reading and studying to catch up on nuclear physics.
---
[13:54](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=834s) Sparring with Hans Bethe and Mental Arithmetic Duels
Feynman gained early prominence within the theoretical division largely by chance. Senior luminaries such as Victor Weisskopf and Edward Teller were temporarily away from the site on other business, leaving Hans Bethe as the sole senior theorist present. Bethe needed someone to bounce ideas against and walked into the office of the young Feynman. As Bethe laid out his concepts, Feynman reacted purely to the physics without any deference to authority: “So he came into this little squirt in an office and he starts to explain this idea and I says, 'No, no, you're crazy, it'll go like this,' and he says, 'Just a moment,' and he explains that he's not crazy, that I'm crazy, and we keep on going like this.” Feynman noted that Bethe desperately needed a sparring partner who would bluntly point out mistakes, and this rapport led to Feynman being appointed a group leader with four men working under him directly beneath Bethe.
Their collaboration quickly evolved into rapid calculation contests between Bethe’s remarkable mental math skills and Feynman operating a hand-cranked Marchant calculator:
| Calculation | Bethe's Mental Method / Result | Feynman's Machine / Method |
|:--- |:--- |:--- |
| **$48^2$** | **2304**<br>Rule for numbers near 50: $(50 - d)^2 = (25 - d) \times 100 + d^2$. Since $48$ is 2 below 50: $25 - 2 = 23$, and $2^2 = 04 \rightarrow 2304$. | Began cranking the Marchant calculator before Bethe called out ~2300, followed immediately by the exact figure 2304. |
| **$\sqrt[3]{2.5}$** | **1.35**<br>Mental logarithm interpolation: divides $\log_{10}(2.5)$ by 3, notes values of $\log_{10}(1.3)$ and $\log_{10}(1.4)$, and interpolates between them. | Pulled out the Marchant reference chart from a drawer to look up trial divisor roots, but was beaten before starting. |
| **$174 \times 140$** | Mentally races using standard or alternate approximations. | **~24,500**<br>Feynman approximated: $174 \times 140 \approx 100 \times \sqrt{3} \times 100 \times \sqrt{2} = 10,000 \times \sqrt{6} \approx 24,500$ ($245 \times 10^2$). |
Bethe's computational speed challenged Feynman to practice relentlessly. Over time, Feynman managed to beat Bethe roughly one out of four times by spotting idiosyncratic numerical shortcuts and factorization tricks before Bethe completed his formal steps.
---
[17:00](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1020s) Rough Lodging and the Dormitory Roommate Ruse
Before proper housing was erected, the theoretical group was lodged in the rustic cabins of the former Los Alamos Ranch School for boys. Feynman was initially assigned to the Mechanics Lodge, packed tightly into bunk beds. The layout was poorly planned: theorist Bob Christy and his wife had to walk straight through Feynman's bedroom every morning simply to reach the bathroom. When the occupants were relocated to the "Big House," conditions were hardly better. Cots were lined up side-by-side along the second-floor veranda patio. Downstairs, a master chart listed bed assignments and dressing quarters; under Feynman's name, the chart read only "Bathroom C, no bed number."
Thoroughly irritated by the living arrangements, Feynman jumped at the chance to select a room as soon as the first new dormitory opened:
- Feynman inspected the camp map and deliberately selected a room overlooking the women’s dormitory, only to discover later that a large tree completely blocked the view.
- Management announced that rooms would temporarily house two men in double-decker bunks, with two rooms sharing an adjoining bathroom.
- Determined to maintain privacy and avoid a roommate, Feynman took advantage of belongings he was storing for his wife Arline, who was hospitalized with tuberculosis in Albuquerque.
- Feynman unpacked a women's nightgown and tossed it casually onto the upper bunk, placed women's slippers by the bed, and sprinkled cosmetic powder across the shared bathroom floor to create the unmistakable appearance of a cohabiting woman.
- The scheme took an unexpected turn when he returned to his room that evening to find that the cleaning staff had treated the clothes with complete, unfazed decorum: his pajamas were folded beneath the bottom pillow, while the woman's nightgown was neatly folded and tucked beneath the pillow of the upper bunk, with her slippers straightened below.
[19:18](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1158s) The Bunk Bed Prank and the Dormitory Politics
Richard Feynman concludes the tale of how he defended his dormitory room from being assigned a roommate. For four straight nights, he continued the ruse: every morning when he awoke, he ruffled up the top bunk, scattered his wife's nightgown, and left her powder around the bathroom before slipping out. Each evening, he would return to find the bed made up neatly and the bathroom cleaned, confirming that nobody had been moved in. By the end of four days, housing assignments across Los Alamos had settled down, and the threat of getting a roommate vanished.
```mermaid
flowchart TD
accTitle: The Dormitory Prank Chain of Command
accDescr: Diagram tracing the panic caused by Feynman's prank from the dorm charwoman up to the governing board and back down to create a new rule.
subgraph Escalation ["The Report Escalates"]
A["Charwoman finds women's items in men's dorm"] --> B["Chief Charwoman"]
B --> C["Lieutenant"]
C --> D["Major"]
D --> E["Governing Board & General"]
end
subgraph Reaction ["The Fallout"]
E -->|"Instructions down chain of command"| F["Clean up and observe for 4 days"]
F --> G["New Rule: No women in men's dormitories"]
G --> H["Dormitory faction forms protest"]
H --> I["Feynman elected to Town Council"]
end
```
The prank, however, sparked an unexpected political crisis. Los Alamos had a town council where civilian factions—such as the housewives, mechanics, and technical staff—voted on administrative issues alongside army leadership. When a sweeping new edict was handed down declaring, "No women in the men's dorms," the bachelors and single women living in the dormitories were outraged by the infantilizing restriction on grown adults. They formed their own political faction, debated the issue, and elected Feynman as their representative on the town council.
Over a year later, Feynman recounted the bunk bed prank to Hans Bethe, who sat on the project's governing council. Bethe burst into laughter, revealing the prank's full ripple effect:
- A terrified charwoman had discovered the women's clothing and powder in the men's dorm and reported that someone was illicitly sleeping with a resident.
- The panic traveled up the military chain of command: charwoman to chief charwoman, to lieutenant, to major, and ultimately up to the general and the governing board.
- Unsure of what action to take, commanders issued orders back down the chain to simply clean up the room each day and monitor the situation.
- After four days of hand-wringing at the highest levels, the administration enacted the sweeping ban on women in the dorms, which directly led the dormitory residents to elect Feynman to fight it.
- As Bethe explained to Feynman: “Ah, that's how you got on the town council.”
---
[22:24](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1344s) The Voluntary Censorship Scheme and Encrypted Mail
Feynman turns to the project's censorship apparatus, calling it an entirely illegal operation because authorities had no constitutional right to censor domestic civilian mail within the continental United States. To circumvent this, the administration instituted a delicate "voluntary" system: scientific personnel agreed to leave their outgoing mail unsealed and consented to having their incoming letters inspected. If censors approved an outgoing letter, they sealed it; if it violated security rules, they returned it to the sender citing the specific infractions. The arrangement also included express permissions for personnel to critique the administration or write to their senators.
On the very first day censorship took effect, Feynman was summoned down to the censor's office over incoming mail:
> “Precisely.” I said, “I have a game. I challenge them to send me a code that I can't decipher. See?”
```mermaid
flowchart LR
accTitle: The Coded Mail Compromise
accDescr: Flowchart illustrating the agreement between Feynman and the military censors regarding encrypted family letters.
A["Father / Wife send coded letters"] --> B["Censors intercept mail"]
B --> C{"Censors require key"}
C -->|"Feynman wants puzzle unsolved"| D["Censors extract key from letter"]
D --> E["Feynman receives encrypted text only"]
```
The censors confronted him with a letter from his father written on lined paper, covered in systematic patterns of dots (such as four dots below a line and one dot above), followed by another letter from his wife, Arline, containing random letter sequences like `TJXY WZ TWW1X3`. Feynman admitted both were ciphers, but freely confessed he had no idea what they said or what the keys were. He explained that he and his family played an ongoing intellectual game where they challenged him to break their codes without a key. Because censorship regulations promised not to disrupt normal correspondence habits, Feynman and the censors struck a compromise: his family would mail the decryption key along with the encrypted letters, and the censors would confiscate the key before passing the coded text to Feynman.
---
[24:56](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1496s) Bureaucratic Clashes and Censor Scissors
The censorship system proved full of operational friction. Soon after, Feynman received a letter from his wife stating it felt awkward to write knowing that someone was looking over her shoulder—with the word identifying the censor completely erased using chemical ink eradicator. Feynman marched down to the office to protest, reminding the staff that incoming mail was not supposed to be altered. The censors denied responsibility, scoffing that legitimate censors never use ink eradicator and always snip forbidden passages out with scissors. When Feynman wrote back asking Arline whether she had used eradicator, she replied that she had not and that it must have been the censors—except this time, the word identifying them was neatly snipped out of the paper with scissors.
Feynman escalated the issue to the major overseeing censorship, arguing that the censors were poorly trained in the delicate agreements made with the scientists:
- Feynman spent several days challenging the major, serving as an informal ombudsman since he corresponded with his wife daily.
- When the major asked, “What's the matter? Don't you think I have goodwill?”, Feynman countered, “Yes, you have perfectly goodwill, but I don't think you have power.”
- Stung by the remark, the major immediately picked up the telephone, issued orders, and permanently halted the physical mutilation of Feynman's letters.
Absurd mix-ups continued regardless. While visiting Arline at the Presbyterian Sanatorium in Albuquerque, she asked him why he had not brought the items she requested. When Feynman asked what items, she recited, "Lith, glycerin, hot dogs, laundry." It turned out the censors had confiscated her innocent shopping list because they mistook words like "lithog" and "glycerin" for a dangerous sabotage cipher.
---
[26:44](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1604s) Mathematical Curiosities and Arabic Numerals
During the early weeks of settling the mail rules, Feynman spent time playing with an adding machine and stumbled onto an interesting arithmetic pattern: dividing 1 by 273 yielded `0.00366...` with repeating sequences like `04115`, `226`, `337`, where carry digits briefly disturbed the pattern before realigning with `10 1013` (equivalent to `114` or `115`). Finding the periodic cycle amusing, he wrote an explanation of the calculation in a routine letter.
The letter was promptly rejected and returned with a note directing him to Paragraph 17B:
- Paragraph 17B mandated that letters could only be written in recognized languages—such as English, Russian, Spanish, Portuguese, Latin, and German.
- Any other language required explicit written authorization.
- The rule strictly reiterated: no codes.
Feynman fired back with an explanatory note showing the explicit arithmetic division of 1 by 273. He proved mathematically that the long repeating decimal contained no more intrinsic information than the number 273 itself, meaning it could not conceal secret data. He concluded his appeal by formally requesting official administrative permission to write his personal letters in "Arabic numerals"—a request the censors were forced to grant.
[28:34](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1714s) Censorship Shenanigans and Coded Letters
The ongoing feud with the camp censors over Feynman’s correspondence with his ailing wife, Arline, reached absurd heights when the authorities took issue with her mentioning the censorship itself. The censors repeatedly sent Feynman notes complaining that his wife had brought up the subject, culminating in a formal demand: “Please inform your wife not to mention censorship in her letter.” Taking their instructions literally, Feynman began his reply, “I have been instructed to inform you not to mention censorship in your letter,” only for the letter to come bouncing right back rejected. When he challenged the absurdity—demanding to know how else he was supposed to tell her without mentioning it, and pointing out the irony that the censor had to order him to tell his wife not to tell him what was being censored—the censors revealed their genuine concern. They worried that mail could be intercepted between Los Alamos and Albuquerque, alerting outsiders to the existence of censorship, and begged that Arline simply act normal.
Feynman visited Arline in Albuquerque and agreed to tone down the direct references, but the couple had already devised an illicit signaling system. Whenever Feynman placed a dot at the end of his signature, it signaled that the censors had intervened again, prompting Arline—who had endless free time while confined to her hospital bed—to unleash the next prank she had cooked up. Her subsequent volley took inspiration from an advertisement offering blank jigsaw puzzles for writing letters to boyfriends. She wrote her message across the blank puzzle, disassembled the pieces into a little sack, and mailed it. The censors were unamused, returning it with an exasperated reprimand: “We do not have time to play games. Please instruct your wife to confine ourselves to ordinary letters.” The couple already had their next escalating tactic prepared: an opening line reading, “I hope you remembered to open this letter carefully because I have included the Pepto-Bismol for your stomach as we arranged,” accompanied by loose powder that would burst all over the inspection office when opened roughly. However, they never had to deploy that particular trick.
```mermaid
flowchart TD
accTitle: Arline and Richard Feynman Mail Escalation Loop
accDescr: Diagram illustrating the escalating cycle of pranks and censorship between Richard Feynman, the army censors, and Arline Feynman.
subgraph Letters ["Communication Loop"]
A["Arline mails unconventional letter / puzzle"] --> B["Army censors inspect mail"]
B -->|"Rejection & warning note"| C["Feynman receives reprimand"]
C -->|"Signature with secret dot"| A
end
subgraph Planned ["Held in Reserve"]
C -.-> D["Pepto-Bismol powder envelope trap"]
end
```
[30:48](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1848s) Exploiting Blind Spots and Testing Fence Security
Through his relentless back-and-forth with the censors, Feynman learned the practical boundaries of what could and could not slip through the mail far better than anyone else on the mesa, which allowed him to win a series of bets. Outside the office, he turned his critical eye toward physical site security. He observed that civilian workmen living outside the perimeter disliked walking all the way around to the official checkpoint and had cut a hole in the perimeter fence. To illustrate the security lapse, Feynman repeatedly exited through the guard gate with his pass, walked down along the perimeter, slipped back in through the fence breach, and walked right back out through the gate again. The gate sergeant grew increasingly baffled by a man who was constantly exiting the facility without ever coming back in, and his instinct was to call a lieutenant and have Feynman arrested.
After pointing out the hole directly, Feynman bet a colleague that he could mail a description of the breach out of Los Alamos without being stopped by censorship. He outsmarted the screeners by framing the secret information as an administrative complaint:
> “You should see the way they administer this place.”
He followed this permitted critique with the precise coordinates: a hole of specific dimensions situated 71 feet away from a designated landmark, big enough for a man to walk through. Because it exposed their own failure, the censors could not claim the hole did not exist, and the letter passed through untouched. On another occasion, Feynman successfully mailed an account detailing how a young man in one of his groups—named Key, who later became well known—was dragged out of bed in the middle of the night and interrogated under bright lights by army officers suspicious that his father was a communist.
- Workmen cut an illicit shortcut through the perimeter fence to avoid walking the long detour to the official gate.
- Gate guards attempted to jail Feynman because his log showed him repeatedly leaving the facility without entering.
- Feynman slipped physical vulnerability reports past censors by embedding exact measurements into letters framed as grievances about site administration.
[32:22](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1942s) Wooden Cabinets, Padlocks, and Edward Teller's Desk Drawer
Security around the laboratory's core scientific data was remarkably primitive in the early days of the project. Highly sensitive documents regarding uranium and weapon physics were stored inside wooden filing cabinets fastened with standard, commercial padlocks and shop-fabricated drop rods. These protections were essentially decorative; an intruder did not even need to pick the padlock, because tipping the wooden cabinet backward allowed someone to slide documents straight through the bottom rod opening. Feynman constantly picked the locks to demonstrate their inadequacy and used general meetings to urge the installation of secure storage. During one meeting, Edward Teller pushed back, suggesting that filing cabinets were unnecessary because he kept his most vital secrets inside his personal desk drawer.
Feynman, seated toward the back, quietly slipped out while the meeting was still underway to inspect Teller’s desk. Finding that the drawer functioned like a commercial paper dispenser—allowing someone reaching underneath the back to pull out one sheet after another—Feynman emptied the entire contents of the drawer and stashed the papers away. He slipped back into the hallway upstairs just as the meeting adjourned and casually fell into step alongside Teller on the walk back to the office. When Teller proudly offered to demonstrate how secure his locked desk was, Feynman looked on innocently until Teller turned the key, pulled the drawer open, and stared into empty space:
> “The trouble with playing a trick on a highly intelligent man like Mr. Teller is the time that it takes him to figure out from the moment that he sees that there's something wrong till he understands exactly what happened is too damn small to give you any pleasure.”
[35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) The Criticality Peril at Oak Ridge
Feynman turned to technical problems outside Los Alamos, most notably the severe safety oversights at the Oak Ridge enrichment facility in Tennessee. While Los Alamos was tasked with weapon design, Oak Ridge was charged with separating uranium isotopes—specifically isolating the fissile uranium-235 from the predominant uranium-238. At the time, the facility was handling minute experimental quantities of U-235 while simultaneously running large-scale chemical operations designed to repeatedly purify the output across multiple stages. However, the Oak Ridge chemists could never obtain consistent assay measurements to establish the enrichment level of their batches, regardless of the instructions sent from Los Alamos.
Exasperated by the flawed assays, Emilio Segrè insisted on traveling to Oak Ridge in person to investigate the procedure on the ground. The army fiercely resisted his visit, insisting on strict compartmentalization:
> “The army people said no it's our policy to keep all the information loss in one place and that the people in Oak Ridge would not know anything about what it was used for.”
Under military secrecy rules, plant laborers were kept totally ignorant of the bomb project, while upper-level staff were informed only that they were separating uranium, without understanding critical mass, bomb dynamics, or yields. Segrè finally forced the issue, arguing that the entire enterprise risked going up in smoke if the assays remained unreliable. Walking through the Oak Ridge plant, Segrè was stunned to see workers wheeling around large carboys of green water—uranyl nitrate solution. When he asked whether they planned to handle the enriched material in the exact same manner, the personnel saw no reason not to. Segrè asked whether it would explode, to which the uncomprehending workers replied, "Huh? Explode? How?" The army viewed Segrè’s alarming questions as proof that outside scientists should never have been admitted. In truth, the army’s safety calculations assumed that since the plant would never hold the 20-odd kilograms of pure metal required for a bomb, no critical explosion was possible—completely overlooking the fact that moderating neutrons with water vastly reduces the mass needed to trigger a critical reaction.
[37:27](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2247s) The Oak Ridge Criticality Threat and Christy's Pneumonia
Feynman explains the critical physics hazard discovered at the Oak Ridge plant. Neutrons are slowed down significantly in water, meaning it requires less than one-twentieth or one-hundredth as much fissile material to spark an unintended chain reaction. While such an accident would not produce a full-scale nuclear explosion, it would unleash lethal bursts of radioactivity, killing anyone nearby. Because the plant's planners had paid zero attention to safety, Emilio Segrè was dispatched to survey the whole facility to identify where concentrations of enriched material would occur. Simultaneously, Los Alamos calculated safety limits. Robert Christy's group tackled water solutions, while Feynman's group calculated thresholds for dry powder stored in boxes.
Christy was originally scheduled to travel to Tennessee to report their findings. However, when Christy suddenly came down with pneumonia, the responsibility fell unexpectedly on Feynman. It was Feynman's first time ever flying on an airplane. Classified documents were strapped to his back with a harness belt. Air travel at the time resembled a bus route with frequent layovers. During one stop, a fellow passenger casually complained about travel difficulties, but Feynman held a high-tier travel clearance:
> “I couldn't resist. I said, 'Well, I don't know.' Oh, I said I have a priority a little bit later. It looks like some generals are coming. They're going to put out some of us number threes. It's all right. I'm a number two.”
Feynman joked that the man probably wrote to his congressman—if he was not one himself—to complain about kids being given number-two priorities during wartime.
```mermaid
flowchart TD
accTitle: Oak Ridge Criticality Safety Task Division
accDescr: Diagram illustrating the parallel calculations at Los Alamos and how Christy's illness forced Feynman to travel.
subgraph LosAlamos["Los Alamos Calculations"]
A["Emilio Segrè Surveys Plant"] --> B["Calculate Explosion Limits"]
B --> C["Christy's Group: Water Solutions"]
B --> D["Feynman's Group: Dry Powder in Boxes"]
end
subgraph Deployment["Delivery to Oak Ridge"]
C --> E{"Christy Gets Pneumonia"}
D --> E
E -->|"Assignment transfers"| F["Feynman Flies to Oak Ridge"]
end
```
[39:14](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2354s) Inspecting the Plant and Oppenheimer's Instructions
Arriving at the Oak Ridge facility, Feynman conducted a walk-through without speaking, observing the entire layout. The plant's danger was far worse than Segrè had initially realized; Segrè had become confused and failed to recognize that large stacks of boxes in different rooms were actually positioned against the same shared wall, dramatically raising the risk of critical mass. Relying on an intense short-term working memory, Feynman memorized specific plant details, including vat numbers and locations in Building 9207.
That night, he analyzed the safety hazards and formulated straightforward remedies: introducing cadmium into liquid solutions to absorb thermal neutrons, and spacing dry powder containers far enough apart to prevent critical interactions. Before Feynman had departed Los Alamos, J. Robert Oppenheimer gave him explicit, strict operational instructions regarding the plant's leadership:
> “He said, 'Then you should say Los Alamos cannot accept the responsibility for the safety of the Oak Ridge plant unless these guys'—I said, 'You mean me, little Richard's going to go in there and say?' He says, 'Yes, little Richard, you go and do that.'”
[40:51](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2451s) Confronting Military Secrecy and Educating the Plant Leaders
The next day, Feynman attended a massive meeting filled with corporate executives, engineers, technical experts including Julian Webb, and military generals. A lieutenant escort informed Feynman that a colonel had ruled against explaining how neutrons functioned; Feynman was supposed to give safety rules without revealing the underlying physics. Feynman refused to comply with this restriction:
> “I said in my opinion it's impossible for them to understand or to obey a bunch of rules that they don't understand unless they understand how it works. And so it's my opinion that it's only going to work if I tell them, and so [Los Alamos] cannot accept the responsibility for the safety of Oak Ridge plant unless they are fully informed as to how it works.”
The lieutenant relayed the ultimatum to the colonel, who stood gazing out the window for five minutes before deciding to authorize the full technical briefing. Feynman praised the military's capacity to make crucial high-stakes decisions so rapidly. When he addressed the assembly, he laid out nuclear fundamentals: separating fissile masses, utilizing cadmium absorbers, and accounting for the higher efficiency of slow neutrons over fast neutrons. Though these were basic, elementary concepts at Los Alamos, the Oak Ridge personnel had never encountered them, viewing Feynman as an astonishing genius who had dropped down from the sky. Consequently, the site formed specialized calculation groups and resolved to completely redesign the processing plants.
- **Cadmium**: Introduced into water solutions because of its effectiveness at absorbing thermal neutrons.
- **Elementary Primer vs. Genius**: Basic Los Alamos concepts struck the compartmentalized engineers as revolutionary physics insights.
- **Redesign Mandate**: The design teams, chemical engineers, and construction crews agreed to overhaul plant blueprints before processing enriched material.
[43:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2593s) Reviewing Blueprints and the Accidental Valve Discovery
A month or two later, Feynman returned to evaluate the newly drafted plant blueprints prepared by Stone & Webster Company engineers. Accompanied by his regular military escort, Lieutenant Zumwalt, he was led into a conference room with two engineers and an enormous table covered in thick stacks of oversized blueprints. Feynman had taken mechanical drawing in school, but he struggled to interpret complex engineering schematics.
The engineers rapidly walked him through the complicated chemical flow of uranium nitrate and carbon tetrachloride cycling across floors and pipes, reassuring him that the system was fail-safe and required at least two independent valve failures for material to dangerously accumulate. Feynman was completely overwhelmed and could not decipher the blueprint symbols, particularly a small square with a cross inside that appeared repeatedly along the lines:
> “You didn't ask them right away. Right away it would have been okay, but they've been talking a little bit too long. You hesitated too long. If you ask them, they'll say, 'What are you wasting my time all this time for?'”
Afraid to reveal his ignorance, Feynman took a calculated gamble to find out whether the recurring symbol represented a valve or an architectural window. Putting his finger squarely on page three of the blueprints, he asked the engineers what would happen if that particular valve got stuck, expecting them to correct him and reveal the symbol's true identity. Instead, the two engineers spent minutes tracing the piping up, down, and across the drawings before turning around in shock to confirm that he was entirely correct: the plant would indeed accumulate dangerous material. Feynman's blind guess appeared as a stroke of sheer technical brilliance, prompting the engineers to roll up their plans to fix the oversight.
[46:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2769s) The Genius of Finding the Valve and Fixing Marchant Calculators
Following the plant inspection, Mr. Zumwalt—who had been shadowing Feynman every step of the way—expressed sheer awe. Zumwalt declared Feynman a genius, recalling how on a previous walk-through Feynman had memorized details like evaporator C-21 in building 9207 by the next morning and had demonstrated complete command over neutron safety. But what Feynman had just done on the blueprints amazed Zumwalt so thoroughly that he demanded to know the secret. Feynman pragmatically punctured the mystique: “I told him you try to find out whether it's a valve or not.”
Feynman then transitions to another major challenge at Los Alamos: handling immense volumes of numerical calculations. At the time, calculations were performed using mechanical Marchant computing machines. These hand-operated, motorized desktop gadgets multiplied, divided, and added mechanically through gears and levers—far removed from modern push-button electronics. Because of strict policy, whenever a Marchant broke down, it had to be shipped all the way back to the factory for repair since there was no resident mechanic. With machines constantly in transit, the laboratory faced a severe shortage.
Defying the warning labels stating the factory would void responsibility if opened, Feynman and a few colleagues began taking the covers off. They received what Feynman called a "series of lessons by God" in machine repair: the first machine they opened simply had a loose spring hanging beside a shaft with a hole, meaning the fix was as straightforward as hooking the spring back in place. They gradually tackled increasingly elaborate repairs, sending only genuinely impossible cases back to the factory while keeping the laboratory running. While a machinist in the shop eventually took over typewriter maintenance, Feynman ended up repairing all the Marchants himself after the other physicists lost interest.
- Marchant calculators required tedious mechanical cycling for arithmetic, making them prone to jams and wear under constant wartime use.
- The factory warranty strictly forbade opening the cases, but lab personnel routinely bypassed the rule out of sheer operational necessity.
- Feynman took over typewriter servicing briefly before deferring to a superior machinist, ultimately focusing entirely on mechanical calculator repair.
[47:57](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2877s) Stanley Frankel and Mass-Production Computing on Punch-Card Machines
As the project advanced, researchers confronted their largest numerical bottleneck: modeling the hydrodynamics of the bomb's implosion. Accurately determining the energy release required simulating the material being driven inward by high explosives and then rebounding outward, a task that vastly exceeded the capacity of manual Marchant operators. A clever physicist named Stanley Frankel realized the calculation could be mapped onto business punch-card equipment manufactured by IBM.
IBM's inventory included tabulators for listing sums, sorters, collators, and large multiplying punches that could read two numbers off a punch card, compute their product, and punch the answer onto the same card. Frankel designed a cyclic pipeline where batches of punch cards rotated through a room from one specialized machine to the next. To prepare for the equipment and debug their numerical cycles while waiting for delivery, Feynman organized a human assembly line using Marchant calculators:
```mermaid
flowchart LR
accTitle: Human Assembly-Line Simulation of IBM Cycle
accDescr: Flowchart illustrating the pipeline of card movement between human operators simulating IBM punch-card machines.
subgraph HumanPipe["Human Calculation Pipeline"]
A["Adder Station (Marchant)"] -->|"passes index card"| B["Multiplier Station (Marchant)"]
B -->|"passes index card"| C["Cubing Station (Marchant)"]
C -->|"returns to cycle"| A
end
```
By assigning each woman at a Marchant to perform only a single operation (such as adding, multiplying, or cubing) before handing the index card to her neighbor, the group achieved Henry Ford–style mass production. “The damn thing works a hell of a lot faster the other way,” Feynman observed, noting that their human room attained the exact calculation speed predicted for the IBM hardware. The only difference was that punch-card machines could work three continuous shifts without fatigue.
When the IBM hardware finally arrived in crates, the assigned military repairman was nowhere to be found. The delicate, electromechanical tabulators and multipliers arrived partially disassembled, accompanied by piles of loose wiring and blueprints. Frankel and Feynman took it upon themselves to assemble the machines, enduring constant scolding from laboratory brass who worried they would permanently ruin the apparatus. Feynman recalled hesitating to straighten a visibly bent part inside one multiplier out of fear it might snap. When the IBM field technician finally arrived right on his original schedule, he readily repaired the rest of the machines and spent three full days troubleshooting that same stuck multiplier until Feynman pointed out the bent piece—which the technician quickly straightened, resolving the problem instantly.
[51:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3095s) "The Computer Disease" and Reorganizing the IBM Group
Once the IBM installation became operational, Frankel succumbed to what Feynman diagnosed as the very first recorded case of "the computer disease." Captivated by the logic of the machine—such as utilizing internal X-relays to trigger conditional branches depending on whether an intermediate number was even or odd—Frankel spent his time writing clever programs instead of directing production. He withdrew into tinkering with automated routines to make a single tabulator tabulate and print tables of the arc tangent function by on-the-fly numerical integration, completely neglecting oversight while the critical implosion calculations ground to a halt. As Feynman pointed out: “The disease with computers is you play with them.”
With implosion calculations stalled, Feynman was instructed to leave his theoretical physics group and take charge of the IBM computation unit. Upon taking over, he recognized the core cause of the team's sluggish output: the workforce consisted of enlisted soldiers from the Special Engineer Detachment (SED)—bright young men with technical and high-school engineering backgrounds drafted into the Army and sent to Los Alamos. Housed in barracks under rigid secrecy, they were assigned to punch numbers onto cards eight hours a day without any knowledge of what the numbers represented.
Feynman went directly to security, secured authorization, and gave the enlisted men a comprehensive lecture on the physics of the atomic bomb and the stakes of the war. Understanding that a higher pressure output signified greater explosive energy transformed the detachment's morale overnight:
> “They knew what they were doing. Complete transformation. They began to invent ways of doing it better. They improved the scheme. They worked at night.”
The SED crew voluntarily worked overnight shifts without needing any supervision and developed their own process improvements. Under Frankel's supervision, the group had taken nine months to complete three calculations; under Feynman and the revitalized enlisted staff, they completed nine calculations in three months—nearly a tenfold increase in speed. To maximize throughput, Feynman devised a method of running distinct problems simultaneously out of phase by color-coding decks of punch cards, keeping the adders, multipliers, and sorters continuously saturated.
[54:30](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3270s) The High-Speed Computing Crunch for the Test
Near the end of the war, just before the Trinity test in the desert, Los Alamos faced an urgent question: how much energy would be released by the bomb? While calculations had been run for various experimental setups, the specific design ultimately chosen for the test had never been computed. Bob Christy came to the computing group and demanded an answer within three weeks to a month. Feynman initially told him it was impossible: although the group produced several completed problems a month, they did so via pipelining and parallel processing across the IBM punched-card machines. A single problem took months to make its way through the entire cycle, and the process could not simply be sped up. Christy pointed out their aggregate output rate, but Feynman insisted, “I know it takes much longer to do the problem but we're doing them in parallel.”
After Christy left, Feynman wondered if they could clear the deck and run a single problem completely unimpeded. He wrote a challenge on the blackboard asking the Special Engineer Detachment (the “boys”) if they could pull it off. Enthusiastic about the challenge, the crew agreed to work overtime and double shifts. The directive was established: suspend all other work and push this single critical problem through the cycle with maximum concentration.
```mermaid
flowchart TD
accTitle: Card Error Propagation and Fast-Track Correction
accDescr: Diagram illustrating how numerical errors spread like a disease through punch cards and how the team used smaller decks to patch them.
subgraph MultiDeckFix["Parallel Patching Solution"]
WD["Main 50-card Deck (White)<br/>Runs slowly through machines"]
BD["10-card Patch Deck (Blue)<br/>Cycles rapidly to correct region"]
YD["Emergency Deck (Yellow)<br/>Patches nested error in blue deck"]
WD -.->|"Error detected"| BD
BD -.->|"Second error found"| YD
YD ==>|"Corrected data merged"| BD
BD ==>|"Spliced back into place"| WD
end
```
[55:43](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3343s) Fuchs's Car and the Three Colored Decks
Just as the intensive computing run began, Feynman's wife, Arline, died in the hospital in Albuquerque, forcing him to leave Los Alamos for three days. To make the trip, he borrowed a car from Klaus Fuchs, a dorm friend who had an automobile. At the time, Feynman had no idea that Fuchs was using the vehicle to transport atomic secrets to a Soviet contact in Santa Fe. The trip proved harrowing in its own right, suffering three flat tires along the road to Albuquerque.
When Feynman returned to the computing room, he was greeted by sheer chaos. Despite the strict order to work on only one calculation, the room was filled with three different colored decks of cards: white, blue, and yellow. When he demanded to know why they were running multiple problems, the exhausted crew shouted him down:
> “Get out. Get out. Get out. Wait. We'll explain everything.”
Feynman waited, and the team revealed an ingenious improvisational workaround. Each deck of fifty cards represented spatial positions or depths in the imploding core. When a machine miscalculated or a wrong number was entered, the numerical error did not instantly ruin the whole calculation; it spread out like a disease from the source to adjacent cards with each successive computational cycle. Rather than halting the machines and discarding days of progress, the operators pulled a small sub-deck of ten cards around the error. Because a ten-card deck could run through the card-punching and tabulating machinery five times faster than the full fifty-card deck, they raced the smaller blue deck through to catch up with and heal the spread of the error while the main white deck continued cycling. When they made another error inside the blue deck, they introduced a tiny yellow deck that cycled even faster. When Feynman walked in, they were frantically trying to splice the corrected numbers back into the master deck before the error outpaced them. The boss had only walked in to disrupt the delicate operation, but the gamble worked, and they delivered the final yield calculation on time.
- Klaus Fuchs lent his car to Feynman, completely hiding his secret life as a Soviet espionage courier.
- The 10-card sub-decks cycled substantially faster than the 50-card master deck, exploiting the localized propagation speed of numerical errors across spatial coordinates.
- The multi-colored cards allowed visual tracking to prevent splicing incorrect stages of computation together.
[58:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3490s) Lessons from Fermi and von Neumann
Looking back on his transition from an underling to a group leader at Los Alamos, Feynman reflected on the privilege of interacting with the world's foremost scientific minds. One of his early humbling encounters was with Enrico Fermi, who came down from Chicago to consult. Feynman had spent weeks wrestling with a calculation so intricate and difficult that he could not understand or explain the physical intuition behind the numbers, even though physical intuition was normally his strongest asset. When he began laying out the problem, Fermi cut him off:
> “Before you tell me the result let me think: it's going to come out like this... and it's going to come out like this because of so-and-so.”
Fermi instantly deduced the exact outcome and provided an obvious physical explanation that Feynman had missed after weeks of labor, doing what Feynman was supposed to be best at ten times better.
Feynman also developed a close rapport with the mathematician John von Neumann, who provided vital technical advice regarding numerical instabilities in the machine calculations. On Sundays, Feynman often joined von Neumann and Hans Bethe for walks through the local canyons. During these conversations, von Neumann gave Feynman a piece of advice that profoundly shaped his temperament:
> “The one thing that von Neumann gave me was an idea that he had, which is interesting: that you don't have to be responsible for the world that you're in.”
Feynman credited this single remark with instilling in him a powerful, lifelong sense of “social irresponsibility” that brought him enduring personal happiness.
[01:00:12](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3612s) Niels Bohr and the Man Who Wasn't Afraid
Niels Bohr arrived at Los Alamos under the government pseudonym “Nicholas Baker,” accompanied by his physicist son Aage Bohr (known as “Jim Baker”). To the young physicists and senior project leaders alike, Bohr was regarded as an absolute deity. During his first visit, Feynman could only glimpse Bohr's head from the back corner of a packed meeting where project leaders stood in awe, nodding along with whatever Bohr said while the great man mumbled around a pipe that perpetually went out.
Before Bohr's next visit, Feynman received an unexpected telephone call at eight o'clock in the morning from Aage Bohr, summoning him to an office in the technical area before anyone else had arrived. Once behind closed doors, Bohr began explaining ideas for increasing the bomb's efficiency. True to his nature, Feynman disregarded authority entirely and bluntly attacked flawed reasoning:
> “I was always dumb about one thing: I never knew who I was talking to; I would only worry about the physics: if the idea looked lousy, I said it looked lousy; if it looked good, I said it looked good.”
For two hours, Feynman and Bohr passionately argued ideas back and forth, with Bohr perpetually relighting his pipe and speaking in his characteristically muffled, nearly unintelligible cadence while his son translated when necessary. Once satisfied that they had hammered out the flaws, Bohr lit his pipe and said, “Well, I guess we can call in the big shots now.” Aage later explained to Feynman why he had been summoned: after the previous visit, Bohr had instructed his son to track down the name of the junior fellow sitting in the back corner because “he's the only guy that's not afraid of me, and will say when I've got a crazy idea.” Bohr knew that senior scientists who simply answered, “Yes, yes, Dr. Bohr,” were useless for pressure-testing new thoughts.
- Niels Bohr operated under the cover name Nicholas Baker; his son Aage Bohr used Jim Baker.
- Bohr explicitly sought out Feynman because the established lab leaders were too awestruck to challenge his physics.
- Shortly after these computational and theoretical breakthroughs, Feynman took a brief leave of absence following his wife's passing, only to receive a coded message stating that “the baby was expected on such-and-such a day,” signaling his immediate flight back to the desert for the Trinity detonation.
[01:03:07](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3787s) Watching the Trinity Test Through a Truck Windshield
Feynman recalls waiting at the test site, twenty miles away from the detonation point, while others were stationed much closer at six miles. The observers had a radio meant to announce the countdown, but it failed to work for most of the wait, leaving them completely in the dark. Only a few minutes before the detonation did the radio finally come to life, announcing that there were about twenty seconds left. The observers had been issued dark glasses, but Feynman refused to wear them, reasoning that from twenty miles away, dark glasses would prevent anyone from seeing a thing. Concluding that bright visible light does not cause optical damage, but ultraviolet light does, he positioned himself behind a glass truck windshield to block the UV rays. Feynman reflects, “I'm about the only guy in the world who actually looked at the damn thing.” The observers six miles away had been instructed to lie face down on the ground, and those around Feynman were wearing dark protective goggles, leaving him as the sole person to witness the test directly with the naked eye.
When the bomb went off, the initial flash was overwhelmingly bright, forcing Feynman to jerk his head down. Catching sight of a purple splotch on the truck's floor, he instantly told himself, “That ain't it. That's an after image,” and immediately looked back up. Over the course of approximately one minute, he watched the intense white light shift to yellow, then to orange, while clouds formed and dissolved rapidly in response to the blast's compression and expansion waves. The blazing core transformed into an immense orange ball that billowed upward, darkened along its outer edges, and turned into a towering sphere of smoke pierced by inner flashes of emerging fire. For that entire minute, everyone watched in dead silence. Suddenly, a minute and a half after the flash, a tremendous bang struck, followed by rumblings like thunder. That solid wall of sound broke the silence and released everyone's tension, proving to Feynman beyond doubt that the bomb had really worked.
- Detonation observers: stationed 20 miles away (Feynman's group) and 6 miles away (lying on the floor).
- Protection rationale: ordinary glass stops ultraviolet radiation, which Feynman determined was the only component capable of burning the eyes.
- Visual progression of the fireball: blinding white flash $\rightarrow$ purple retinal afterimage $\rightarrow$ yellow $\rightarrow$ orange $\rightarrow$ rising billow of black smoke with interior fire flashes.
- Sound delay: approximately 90 seconds between the light flash and the arrival of the shockwave thunder.
---
[01:05:32](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3932s) William Laurence, the Smyth Report, and the Gold Doorstop
Right after the shockwave tore through the site, the man standing beside Feynman asked, “What's that?” Feynman answered: “That was the bomb.” The man was William L. Laurence, the science reporter brought in to write the official news dispatches on the project. Feynman had initially been tasked with showing Laurence around Los Alamos, but Laurence found the physics too technical, so Henry DeWolf Smyth arrived later to take over the public documentation, producing the famous Smyth Report.
While giving Smyth a tour, Feynman brought him into a room where a silver-plated ball of plutonium—slightly smaller than a pedestal head, pleasantly warm to the touch from radioactive decay—sat on display. As they lingered by the doorway discussing the philosophical marvel of handling a synthetic element never before seen on Earth since the dawn of creation, Smyth casually began kicking the doorstop with his foot. Feynman pointed down and remarked, “And the doorstop is appropriate to the door.” The doorstop was a large, solid hemisphere of yellow gold. To minimize the amount of precious fissile material required for critical mass, Los Alamos scientists had performed neutron-reflection experiments on an array of candidate tamper materials—including platinum, zinc, brass, and gold. Once the reflective tests were finished, someone decided that the heavy lump of gold made the ideal doorstop for the vault holding the world's most valuable substance.
---
[01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) Post-Trinity Euphoria, Bob Wilson's Regret, and the Illusion of Progress
The successful test triggered wild celebration across Los Alamos. Staff members held impromptu parties, and Feynman joined the revelry, sitting on the back of a jeep beating bongo drums. Amid the ecstasy, he noticed Robert R. Wilson—the very man who had originally recruited him to the Manhattan Project at Princeton—sitting alone, completely despondent. When Feynman asked him what was wrong, Wilson replied, “This terrible thing we made.” Feynman countered, “But you started. You got us into it.” Reflecting on the shift in mindset, Feynman explains: “You see what happened to me or what happened to the rest of us is we started it for a good reason but then we're working very hard to do something and to accomplish it is a pleasure is excitement and you don't stop to think any you know you just stop after you thought at the beginning you stop thinking.” Wilson was the rare individual who had resumed thinking about the moral consequences at that exact moment.
Shortly after the war ended, Feynman moved to Ithaca, New York, to teach at Cornell University, but he remained deeply haunted by nuclear devastation. Sitting in a New York City restaurant, he would look out over the skyline, calculating the radius of Hiroshima-level destruction and visualizing Manhattan flattened all the way down to 34th Street. Seeing construction crews laying down new roads or raising bridges struck him as sheer lunacy: “I would go along and I'd see people building a bridge or they're making a new road. And I thought, they're crazy. They just don't understand. They don't understand. Why are they making new things? It's useless.” He closes his main lecture by noting with relief that he has been proven wrong for nearly thirty years, glad that humanity continued building even when his immediate postwar reaction had been that creating anything new was entirely futile.
---
[01:09:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4149s) Cracking the Mosler Combination Safes
Following applause for the main address, an audience member asks Feynman if he would share his stories about picking safes. Agreeing to take ten minutes to recount three separate safe-cracking exploits, Feynman explains that his obsession began once he learned lock-picking and encountered the newly installed Mosler filing cabinets equipped with combination locks. Driven by what he terms a personal disease—the irresistible urge to unlock anything kept secret—he resolved to master the new safes. Traditional safecracking lore, which claimed an operator could listen for tumblers or feel dial resistance, proved useless because Mosler's modern locks were specifically engineered so that nothing pressed against the internal wheel pack while the dial turned. Popular locksmith manuals were equally unhelpful, opening with absurd melodramas about drowning women in underwater safes and concluding with vague, unscientific advice about psychologizing the owner's habits.
Treating the cabinets as an experimental problem during two years of idle downtime at the laboratory, Feynman systematically mapped out the lock tolerances. He discovered that a combination allowed a margin of error of two digits on each wheel, which meant testing only every fifth number out of a hundred. This cut the total search space down from one million theoretical combinations to $20 \times 20 \times 20$, or 8,000 sequences. By designing a movement routine that adjusted subsequent wheels without disturbing previously indexed numbers, he managed to reduce the physical time required to test every single possibility down to eight hours. Ultimately, he discovered an even faster shortcut: whenever an authorized user left an office safe drawer open, a little manual manipulation of the visible bolt mechanism allowed him to harvest the final two numbers of the combination in moments.
```mermaid
flowchart TD
accTitle: Combination Space Reduction Method
accDescr: Flowchart illustrating how Feynman systematically reduced the search space of the Mosler safes from one million to eight thousand combinations.
subgraph MechanicalLimits["Lock Tolerance Analysis"]
A["100 Numbers per Wheel<br/>(1,000,000 combinations)"] --> B["±2 Digit Margin of Error<br/>per Wheel"]
B --> C["Sample Every 5th Number<br/>(20 steps per wheel)"]
end
subgraph Optimization["Execution Routine"]
C --> D["Search Space:<br/>20 × 20 × 20 = 8,000 Combinations"]
D --> E["Optimized Wheel-Spin Sequence<br/>Preserving Prior Settings"]
E --> F["Full Dial Exhaustion<br/>in 8 Hours"]
end
```
[01:11:37](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4297s) The Art of Practicing and Building a Reputation
Richard Feynman explains that in order to pick up safe combinations unobtrusively, he practiced constantly, just like a card sharp practices card tricks over and over again to become quicker and completely inconspicuous. Whenever he visited a colleague's office to chat, he would lean casually against their open filing cabinet, subtly fiddling with the dial just as though he were fidgeting with a wristwatch. To an outside observer, he was doing nothing at all, but he was actually feeling out and reading the last two numbers of the three-number combination. He would then walk back to his own office and note them down. Because safe locks had loose tolerances of roughly two units, having the final two numbers meant there were only twenty possible values for the first number, allowing the safe to be opened in under three minutes.
```mermaid
flowchart LR
accTitle: Feynman Safe-Cracking Technique
accDescr: Diagram illustrating Feynman's method of harvesting two numbers and testing the remainder.
subgraph S1["Information Gathering"]
A["Chat in colleague's office"] --> B["Lean on open cabinet & fiddle dial"]
B --> C["Harvest final 2 numbers"]
end
subgraph S2["Office Record & Cracking"]
C --> D["Record in office list"]
D --> E["Test 20 possibilities for 1st number"]
E --> F["Safe opens in ~3 minutes"]
end
```
Eventually, Feynman maintained an office catalog containing the final two combination numbers for nearly everyone's safe, earning himself a formidable reputation across Los Alamos. When someone like "Mr. Schmaltz" was away and staff urgently needed an enclosed document, they would ask Feynman if he could crack it. He would feign the need for specialized equipment, pop a standard screwdriver into his back pocket purely for theatrical effect, and head into the target office. He made a point of shutting the door behind him under the pretense that safe-cracking methods were too dangerous for general knowledge: “I close the door and I sit down and read a magazine or do something. I'd average about 20 minutes of doing nothing and then I'd open it, you see, or I open it right away, make sure everything was all right and I'd sit there for 20 minutes to give myself a good reputation that it wasn't too easy.” He would then emerge, slightly sweaty, announcing that the safe was open.
- To maintain the illusion of difficulty, Feynman deliberately wasted about 20 minutes reading or sitting idle inside the locked office.
- A single accidental safe opening during the project further cemented his sensationalized reputation as a master cracker.
[01:13:27](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4407s) Hunting the Los Alamos Secrets in the Declassification Office
After the war, Feynman returned to Los Alamos to finish writing reports and needed a specific classified document on a Saturday. He assumed the laboratory still operated on its wartime schedule, but found the central document library closed. The library's main vault possessed an unfamiliar mechanism and was guarded by armed sentries pacing back and forth, putting it out of reach. Feynman then remembered Frederic de Hoffmann in the declassification section. Because de Hoffmann grew weary of constantly running back and forth to the library to review materials for declassification, he had made duplicate copies of every single document in the Los Alamos library and filed them sequentially into nine filing cabinets spanning two rooms.
Feynman walked into de Hoffmann's office; the lights were on and the door stood open, giving the impression that de Hoffmann would return at any moment. While waiting with nothing to do, Feynman began turning the dials on the cabinets, testing standard combinations such as 20-40-60 to no avail. He began reflecting on locksmith books he had studied, recalling their advice about safe-cracking through psychology rather than mechanical manipulation: “Then I begin to think you know those locksmith people. I ain't never been able to figure out how to open them cleverly. Maybe they don't either. Maybe all the stuff they're telling me about psychology is right.”
[01:15:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4535s) Cracking de Hoffmann's Safes via Mathematical Constants
The locksmith literature noted that secretaries, terrified of forgetting combinations entrusted to them, routinely wrote them down in semi-hidden places. Feynman recalled that one classic hiding spot was along the unpainted exterior wooden side of an upper desk drawer. Finding the desk locked, he picked the desk lock immediately and inspected the wood, finding nothing. Digging through the papers inside, he found a reference sheet for the Greek alphabet (alpha, beta, gamma, delta) kept by the secretaries so they could pronounce and transcribe physics formulas properly. Carelessly jotted across the top was the value of pi: $\pi = 3.14159$.
Recognizing that a secretary had no reason to do numerical calculations, Feynman took this as a potential clue and tested combinations based on pi (31-41-59, 13-41-95, and other permutations) for twenty minutes, but none succeeded. As he started to leave, he realized his psychological intuition had targeted the wrong person: “Freddy de Hoffman is just the kind of a guy to use a mathematical constant for his safe combination.” If pi was not the answer, the other paramount mathematical constant was Euler's number ($e = 2.71828$). Feynman walked up to the cabinets and dialed 27-18-28. The mechanism clicked open instantly. Feynman checked the remaining cabinets and discovered that all nine filing cabinets containing the entire collection of atomic bomb secrets—the theoretical calculations, neutron liberation rates, critical masses, and formulas—were set to the exact same combination. Concluding his talk, Feynman noted that although he had many more stories, the hour was late, ending the lecture to audience applause.
## Notable quotes
> [01:00:12](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3612s) I was always dumb about one thing: I never knew who I was talking to; I would only worry about the physics: if the idea looked lousy, I said it looked lousy; if it looked good, I said it looked good. — Richard Feynman
> [58:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3490s) The one thing that von Neumann gave me was an idea that he had, which is interesting: that you don't have to be responsible for the world that you're in. — Richard Feynman
> [51:35](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3095s) The disease with computers is you play with them. — Richard Feynman
## Mentioned
Aage Bohr, Adolf Hitler, Arline Feynman, Arthur Compton, Bob Christy, Bob Wilson, Cornell University, Edward Teller, Emilio Segrè, Enrico Fermi, Frederic de Hoffmann, Hans Bethe, Harold Urey, Henry DeWolf Smyth, I. I. Rabi, IBM, J. Robert Oppenheimer, John Manley, John von Neumann, John Williams, Joseph Hirschfelder, Julian Webb, Klaus Fuchs, Leslie Groves, Lieutenant Zumwalt, Los Alamos Ranch School, Marchant calculator, Metallurgical Laboratory, Mosler, Niels Bohr, Paul Olum, Presbyterian Sanatorium, Princeton University, Richard C. Tolman, Richard Feynman, Robert R. Wilson, Robert Serber, Smyth Report, Special Engineer Detachment (SED), Stanley Frankel, Stone & Webster Company, University of Chicago, Victor Weisskopf, William L. Laurence, Yakov Frenkel
## Video description
There are quite a few copies of this Feynman lecture floating around out there, but most end prior to the question from the audience.
After the lecture, a guy in the audience asks Feynman about his safe-cracking stories and Feynman goes on for about another ten minutes relating three different stories on his safe-cracking while at Los Alamos National Laboratories. Enjoy!
In his lecture 'Los Alamos From Below,' Richard Feynman reflects on his wartime work on the Manhattan Project as a junior physicist, recounting how institutional improvisation, irreverent problem-solving, and flawed security practices underpinned major scientific and computational milestones.
Scientific Pragmatism Over Authority (13:54)
Feynman joined the project as an unfinished Princeton graduate student, yet his disregard for scientific hierarchy quickly made him an essential collaborator for the project's leadership. When senior theorists were away from Los Alamos, Hans Bethe sought out Feynman as an uninhibited sparring partner willing to bluntly dispute his calculations, leading Bethe to assign Feynman as a group leader. Similarly, Niels Bohr sought private two-hour debates with Feynman during visits because other senior researchers were too deferential to point out flaws in Bohr's ideas. Feynman also learned practical physical intuition from Enrico Fermi, while John von Neumann advised him that an individual does not have to feel responsible for the entire world—a philosophical remark Feynman credited with shaping his personal outlook.
Breaching Secrecy to Prevent Disasters (35:08)
Military compartmentalization frequently created acute operational hazards that technical transparency had to resolve. When Emilio Segrè discovered that the Oak Ridge enrichment plant was dangerously mishandling uranyl nitrate solutions without criticality awareness, Feynman was dispatched to intervene after Robert Christy fell ill with pneumonia. Facing military orders that barred him from explaining neutron behavior to plant personnel, Feynman gave an ultimatum on behalf of Los Alamos: the laboratory would not accept responsibility for the safety of the plant unless workers were fully taught the physics behind critical mass. The military relented, enabling Feynman to teach principles such as cadmium absorbers, vessel spacing, and slow neutron absorption. Later, Feynman took a chance while reviewing blueprints by pointing at an unfamiliar symbol and questioning what would happen if that valve jammed, which accidentally revealed an overlooked chemical loop that would have triggered a fatal critical accumulation.
Mass Computation and Error Management (47:57)
Modeling the hydrodynamics of the implosion weapon required computational throughput that manual Marchant desk calculators could not achieve. After Stanley Frankel developed 'the computer disease'—becoming consumed with programming elegant subroutines rather than directing production—Feynman took charge of the unit of IBM punch-card machines. The unit was staffed by enlisted soldiers from the Special Engineer Detachment who were initially given numbers to process without context. Output soared only after Feynman obtained security clearance to brief the soldiers on the physics of the weapon and the high stakes of their calculations. In the rush to compute energy yields ahead of the Trinity test, the team devised an improvisational error-correction system: rather than discarding runs when an arithmetic error spread across a 50-card deck, operators ran a smaller 10-card sub-deck through the pipeline at five times the speed to catch up and repair the numerical error in stride.
Absurdities of Military Security and Censorship (22:24)
Feynman constantly exposed vulnerabilities and contradictions in the project's security systems. The base mail censorship—framed as a voluntary agreement to circumvent legal bans on domestic civilian censorship—led to ongoing friction as Feynman challenged arbitrary excisions and sent encoded communications to his hospitalized wife Arline. Physical security was similarly lax: perimeter fencing was breached by civilian workers taking shortcuts, while vital technical papers were kept in wooden cabinets that could be accessed by sliding sheets out from underneath. When Mosler combination safes were introduced, Feynman treated them as an experimental challenge, noting a two-unit dial tolerance that reduced the search space from one million to 8,000 combinations. By subtly manipulating dials while leaning on colleagues' open cabinets, he cataloged their final two numbers, allowing him to open their safes in minutes. He later opened all nine cabinets in the declassification office holding complete bomb secrets after deducing that Frederic de Hoffmann had used Euler's constant, 27-18-28, as the combination.
The Moral Aftermath of the Trinity Test (63:07)
Stationed twenty miles away during the Trinity test, Feynman chose not to wear the issued protective goggles, reasoning that visible light was harmless and that a glass truck windshield would adequately absorb dangerous ultraviolet radiation. He directly watched the blinding white flash shift to orange as the expanding fireball climbed. While the test provoked wild celebration among the staff, Feynman noticed project recruiter Bob Wilson sitting in deep despondency over the weapon they had created. Feynman reflected that while the team had started the project for a sound moral purpose—countering Nazi Germany—the relentless technical excitement of the challenge had blinded them to the ethical implications of their success, a realization that left Feynman deeply alienated and unsettled by normal civilian reconstruction projects after the war.
Feynman depicts the Manhattan Project not as a seamless military triumph, but as a precarious technical enterprise salvaged by young researchers willing to dismantle bureaucratic compartmentalization. While Feynman embraced the technical and personal thrill of problem-solving, his post-war reflection aligns with Wilson's remorse, underscoring how focused scientific momentum can easily outpace moral reflection until devastating consequences are realized.
## Background
该视频由 YouTube 频道 The Quagmire 于 2016 年 7 月 12 日发布。视频内容为诺贝尔物理学奖得主理查德·费曼关于战时曼哈顿计划的经典演讲《底层视角的洛斯阿拉莫斯》，并在正篇演讲结束后完整收录了费曼回答观众提问、大谈洛斯阿拉莫斯开锁秘闻的精彩后半段内容。
## TL;DR
费曼以当年一名未获博士学位的普通科研人员视角，生动还原了曼哈顿计划在洛斯阿拉莫斯的日常攻关与生活琐事。他不仅讲述了与汉斯·贝特、奥本海默、玻尔等学术巨擘的激烈争鸣，还细致回顾了只身前往橡树岭排除化工厂临界爆炸隐患、利用女算手与 IBM 打孔卡构建高效计算流水线，以及肉眼目睹三位一体核试验的震撼经历。讲座尾声，他反思了战后核阴云带来的虚无感，并在观众问答环节幽默复盘了自己通过机械容差与心理学破译莫斯勒绝密保险柜的传奇往事。
## Atmosphere & style
整场讲座充满着费曼标志性的纽约口音、机智幽默与桀骜不驯的极客色彩，现场笑声与惊叹声此起彼伏。费曼以第一人称的自嘲剥离了宏大叙事的严肃外衣，在荒诞的恶作剧日常与分秒必争的战时核危机排查之间自如切换。从最初反抗荒唐官僚体制的戏谑欢快，到亲历三位一体核爆强光时的屏息肃穆与狂欢后的战后虚无，整场演讲呈现出一段层次分明、既令人捧腹又发人深省的情感弧线。
## Bulletin
- [02:02](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=122s) 理查德·费曼在洛斯阿拉莫斯以底层年轻研究员的独特视角亲历曼哈顿计划，凭借敏锐的物理直觉与工程排障能力成为团队不可或缺的核心骨干。
- [35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) 军方在橡树岭工厂推行的极端保密壁垒险些酿成严重核临界事故，费曼力排众议打破信息封锁并普及中子物理机制，成功化解危机。
- [52:40](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3160s) 费曼革新了战时科学计算流程，通过打破信息孤岛激发年轻士兵主动性，并引入错相位多任务流水线，将 IBM 打孔卡计算效率提升近十倍。
- [01:09:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4149s) 通过对莫斯勒保险柜机械容差与物理学家思维心理的深入剖析，费曼破解了多台机密保险柜，揭示了看似严密的保密系统背后巨大的安全漏洞。
## Key takeaways
- [03:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=190s) 费曼受纳粹德国可能率先研制出原子弹的威胁驱使，毅然中断普林斯顿博士论文工作投身战时工程。
- [05:50](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=350s) 在普林斯顿铀分离技术评估会议上，泰斗科学家阿瑟·康普顿倾听质疑、托尔曼融会贯通的高效决策格局给费曼留下深刻启发。
- [12:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=733s) 初抵洛斯阿拉莫斯时由于设施严重滞后，理论组与实验组物理学家甚至直接接管现场施工调度以推动工程进度。
- [13:16](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=796s) 费曼与理论部主管汉斯·贝特因直率而毫无顾忌的学术争论建立深厚信任，并在频繁的心算较量中锤炼出独特的数字速算技巧。
- [22:24](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1344s) 面对基地严苛甚至违法的邮件审查制度，费曼夫妇利用暗号、拼图信件及“阿拉伯数字”申请书等幽默手段频频抗争并揭露官僚荒谬。
- [35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) 军方在橡树岭推行极端的信息隔离政策，忽视了水中中子慢化引发的剧烈临界反应，险些造成大批量浓缩铀溶液致命核事故。
- [41:03](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2463s) 奥本海默授予费曼关键通牒策略，促使军方高层在五分钟内妥协，允许其向橡树岭技术骨干公开中子慢化与临界原理以彻底重构防爆设计。
- [43:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2593s) 在审核斯通-韦伯斯特公司繁复管网蓝图时，费曼凭借超常运气随手指出的符号碰巧正是全系统中唯一未设双重冗余的致命故障阀门。
- [47:57](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2877s) 在等待 IBM 硬件到货期间，费曼召集女算手分别模拟加法、乘法等运算单元，开创出速率匹敌机器理论上限的“真人计算机流水线”。
- [52:40](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3160s) 接管 IBM 计算组后，费曼向特种工程分队（SED）士兵公开项目全貌并引入多色卡片错相位并发调度，使计算效率实现了十倍爆发。
- [56:01](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3361s) 在核爆能量释放计算面临死线且出现迭代误差时，团队独创性地采用小规模蓝黄卡片快速追赶缝合方案，在不中断主流水线的前提下成功交差。
- [58:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3490s) 冯·诺伊曼“不必为所处世界负全责”的人生建议，深刻塑造了费曼日后摆脱宏大社会压力、专注于纯粹科学探索的豁达心态。
- [01:00:11](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3611s) 尼尔斯·玻尔专程单独召见费曼探讨物理设想，正是看中他不畏惧权威、敢于直言不讳批驳糟糕想法的求真特质。
- [01:03:07](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3787s) 在三位一体核试验中，费曼凭借普通玻璃能完全阻隔紫外线的物理常识躲在卡车挡风玻璃后，成为唯一用肉眼完整直视核爆全过程的人。
- [01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) 试验成功的狂欢让费曼目睹威尔逊对“创造出可怕怪物”的深沉反思，战后返回城市后更一度陷入对现代文明脆弱性的强烈虚无感。
- [01:09:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4149s) 通过测算莫斯勒文件柜正负2个刻度的机械容差，费曼将原本上百万种的三位密码组合骤降至8000组，开创出地毯式高效穷举算法。
- [01:15:33](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4533s) 费曼利用同事倾心科学的思维惯性，结合自然常数 $e$（27-18-28）一举打开存放基地全部原子弹绝密核心机密文件的九台保险柜。
## Detailed walkthrough
### [00:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=13s) 主持人致辞：超越逻辑的物理学家
讲座在主持人的开场白中拉开序幕。主持人开玩笑说警察局随时可能来查抄这个地方，但无论如何活动必须正式开始。他向听众隆重介绍了当天的演讲嘉宾——诺贝尔物理学奖得主、传奇科学家理查德·菲利普斯·费曼（Richard Phillips Feynman）。主持人评价道，费曼与其他人的最大不同在于，他不仅拥有极其严密的逻辑，还拥有异常宏大的想象力。这是一种极为罕见的组合，因为大多数讲究逻辑的人往往缺乏想象力；正如伟大的苏联物理学家雅科夫·弗伦克尔（Yakov Frenkel）所言：“知识是通过非逻辑的跳跃向前推进的。”主持人指出，费曼正是凭借其丰富的想象力，越过逻辑的边界去探索前沿科学，这种方式就像数学家利用复变函数进行解析延拓一样。
主持人还特别提到了费曼当年在战时曼哈顿计划中的特殊地位。在洛斯阿拉莫斯国家实验室，实验室主任罗伯特·奥本海默（J. Robert Oppenheimer）将费曼评为团队中“最有价值的选手（most valuable player）”。在理论物理学家的本职之外，费曼更是一位超级故障排除专家（super-duper troubleshooter）。他常常直接走进实验实验室，仅仅通过提出几个切中要害的问题，就能立刻找出实验装置无法正常运转的根本原因。
```mermaid
flowchart LR
accTitle: 主持人对费曼学术特质的评价
accDescr: 展示主持人如何将逻辑与想象力的结合、解析延拓比喻以及洛斯阿拉莫斯的排障能力联系在一起
subgraph Mindset["思维特质"]
A["严密逻辑"]
B["非凡想象力"]
end
subgraph Method["科研方法"]
C["类比复变函数<br/>解析延拓"]
end
subgraph Practice["洛斯阿拉莫斯实践"]
D["奥本海默评价：<br/>最有价值选手"]
E["实验室超级排障专家<br/>精准提问定位问题"]
end
A --> C
B --> C
C --> D
D --> E
```
### [02:02](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=122s) “底层视角”：一个未拿到学位的研究生
费曼走上讲台，热情地向台下观众问好，但随即指出刚才的开场介绍对于他当天的演讲主题并不合适。他当天的讲题是《底层视角的洛斯阿拉莫斯》（Los Alamos from Below）。费曼强调，这里的“底层”指的是字面意义上的底层：尽管在演讲当下他已经小有名气，但在曼哈顿计划初期，他根本不是什么名人，甚至连博士学位都还没有拿到。
费曼坦言，市面上关于洛斯阿拉莫斯的回忆大多出自政府高层或组织管理者之口，那些人整天操心的是宏观的大决策；而他当年什么大决策都不用管，只是在实验室底层到处乱窜。尽管随着项目推进他稍微往上挪了几个台阶，但绝非决策层成员。费曼希望听众忘掉主持人那些耀眼的夸赞，将他想象成当年那个还没拿到学位、正为博士论文发愁的年轻研究生，本场讲座仅仅是他个人在这段历史中的经历记录。
- 费曼加入曼哈顿计划相关工作时，尚未取得普林斯顿大学的博士学位。
- 演讲的核心视角聚焦于普通科研人员的亲历细节，而非战略决策层的宏大叙事。
### [03:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=190s) 鲍勃·威尔逊的秘密来访与普林斯顿的转向
费曼回忆起自己卷入这项计划的开端：当时他正待在普林斯顿大学的宿舍里写博士论文，同校的实验物理学家鲍勃·威尔逊（Robert R. Wilson）突然推门走了进来。威尔逊神情神秘地告诉费曼，自己拿到了一笔绝密项目的科研经费，按保密规定他不能向任何人透露，但他打算破例告诉费曼，因为威尔逊确信费曼一旦知晓内容，就一定会忍不住加入。威尔逊向费曼解释了通过同位素分离来制造原子弹的设想，并详细说明了一种由他自主构思的铀同位素分离新方法。
听完这番话后，费曼当场表示拒绝，称自己打算回去继续写论文。威尔逊并不纠缠，只说下午3点实验室有一场讨论会，到时见。费曼回到桌前继续写论文，但仅仅坚持了大约三分钟就再也坐不住了。他开始在房间里来回踱步，满脑子都是希特勒统治下的纳粹德国以及原子弹的现实威胁。费曼事后回忆起那一刻的恐惧与警醒，毅然做出了改变人生轨迹的决定：
> “The Germans had Hitler and the possibility of developing an atomic bomb was obvious and the possibility that they would develop before we did was very much of a fright.”（“德国人有希特勒，制造原子弹的可能性显而易见，而他们赶在我们之前造出原子弹的可能性令人极度恐惧。”）
下午3点，费曼准时走进了会议室。到了下午4点，他已经在实验室分配到了自己的办公桌和演算纸，拼命计算该方法是否会受到离子束总电流的物理限制。与此同步，其他年轻物理学家也纷纷停下手头的纯科学研究，直接从各自原先的科研装置上强拆零件，拼凑成用于同位素分离实验的新设备。整个实验室就像快进播放的电影一样，每次抬头看，实验装置都在飞速变大。战争期间，全美的基础科学研究几乎停滞，所有精力都转向了这类战时工程。后来费曼专门请了六周的假，突击写完论文并通过了答辩，在奔赴洛斯阿拉莫斯前夕正式拿到了博士学位。
### [05:50](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=350s) 普林斯顿评审委员会：初见学界泰斗
在普林斯顿参与该项目期间，费曼经历了一段极富启发的体验——他平生第一次面对面接触到了如此多的物理学巨擘。当时军方设立了一个高级评估委员会，负责审查并决定究竟采用哪种技术路线来分离铀同位素。委员会成员包括理查德·托尔曼（Richard C. Tolman）、亨利·德沃尔夫·史迈斯（Henry DeWolf Smyth）、哈罗德·尤里（Harold C. Urey）、伊西多·拉比（I. I. Rabi）、罗伯特·奥本海默以及阿瑟·康普顿（Arthur Compton）等顶级科学家。
因为费曼彻底掌握了威尔逊方案背后的理论机制，他受邀列席会议回答技术质询。令初出茅庐的费曼大为震撼的，是这些顶尖科学家开会讨论的方式。会议中，康普顿提出了一个极具洞察力且完全正确的观点，但在座的其他委员却接连抛出各种反对意见和潜在风险。费曼在心里暗自替康普顿着急，认为康普顿应该立刻站起来反驳并重复自己的正确主张。然而出乎费曼意料的是，康普顿自始至终没有抢着自我辩解，而是认真倾听并吸收每个人的质疑点。直到会议最后，作为主持人的托尔曼在总结发言时，不仅全面采纳了康普顿的核心观点，还将其他委员提出的反思有机地融为一体，整场讨论没有一句多余的重复论述。这让年轻的费曼深刻见识到了真正伟大科学家的胸襟与高效决策智慧。
- 普林斯顿铀分离评估委员会主要成员：
- 理查德·托尔曼（Richard C. Tolman，委员会主席）
- 罗伯特·奥本海默（J. Robert Oppenheimer）
- 阿瑟·康普顿（Arthur Compton）
- 伊西多·拉比（I. I. Rabi）
- 哈罗德·尤里（Harold C. Urey）
- 亨利·德沃尔夫·史迈斯（Henry DeWolf Smyth）
### [07:23](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=443s) 芝加哥“卧底”刺探与现实与电影的差距
最终，评估委员会决定放弃威尔逊在普林斯顿研发的同位素分离路线，所有人被告知该项目终止。随后军方决定在新墨西哥州的洛斯阿拉莫斯建立专门研制原子弹的基地，普林斯顿的整个团队将全员迁往新基地。在普林斯顿项目停摆而洛斯阿拉莫斯尚未完工的空档期，鲍勃·威尔逊为了不浪费时间，派费曼前往芝加哥大学冶金实验室（Metallurgical Laboratory）搜集有关原子弹的所有最新理论与实验情报，以便普林斯顿团队提前制作好计数器等探测设备。
费曼接到的指令相当巧妙：去芝加哥逐一拜访各个研究小组，声称自己要加入他们并开始上手计算；一旦弄懂问题的核心细节，便立刻换去下一个小组继续提问。虽然这种做法让费曼有些良心不安，但期间发生了一个戏剧性的意外插曲：一个芝加哥团队花了三个月都没能解决一道难题，费曼在听他们讲解时顺口问了一句“你们为什么不试试那样做”，结果对方在半小时内彻底解开了这个困扰数月的死结。
费曼搜集完爆炸威力与弹体设计的参数回到普林斯顿向同事汇报时，卷着衬衫袖子、满身灰尘。他的数学家好友保罗·奥卢姆（Paul Olum）打趣道：
> “When they make a moving picture about this, they'll have the guy coming back from Chicago telling the Princeton men all about the bomb and he'll be dressed in a suit and carry a briefcase inside and you're in dirty shirt sleeves and just telling us all about it.”（“将来如果有人把这段拍成电影，那个从芝加哥带回原子弹秘密告诉普林斯顿团队的家伙肯定西装革履、手提公文包，而现实里的你却只穿着脏兮兮的衬衫卷着袖子跟我们高谈阔论。”）
这也让在场众人深刻体会到现实生活与影视幻想之间的巨大反差。
### [09:19](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=559s) 失去耐心的科学家：进军未完工的秘密基地
芝加哥之行结束后，前方的基地建设依然延宕不前。鲍勃·威尔逊专程赶往洛斯阿拉莫斯现场查看究竟是什么拖延了进度。到达基地后，威尔逊哭笑不得地发现施工队伍确实干得热火朝天，但他们已经建好的竟然是一座电影院和几栋普通附属建筑——因为承包商只知道怎么造常规建筑，却完全没有收到如何建设专业实验室的具体施工图纸，比如到底需要铺设多少条燃气管和水管。威尔逊毫不犹豫地站在工地上当场拍板，凭直觉直接给出了水管和燃气管道的数量指标，命令施工队立即动手建造实验室。
与此同时，奥本海默正与负责整个工程的莱斯利·格罗夫斯将军（General Leslie Groves）就诸多管理和权限问题争执不下，而普林斯顿这群热血沸腾的年轻科学家早已等得极度不耐烦。在威尔逊与芝加哥大学的约翰·曼利（John H. Manley）通电话商议后，两地团队决定不再等待官方通知，直接强行出发前往新墨西哥州。于是，这批由奥本海默耐心招募的科学家，在洛斯阿拉莫斯基地的物理设施尚未完全就绪时，便浩浩荡荡地直接开进了高山峡谷之中。
### [10:18](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=618s) 秘密买票与初抵洛斯阿拉莫斯
理查德·费曼（Richard Feynman）谈到罗伯特·奥本海默（J. Robert Oppenheimer）对每个人的困难都关照备至。费曼的妻子阿琳患有肺结核，奥本海默十分体贴地关心她的病情，并细心打听当地是否会有合适的疗养院与医院设施。费曼动情地感慨：“这是我第一次以如此私人的方式认识他，他真是一位了不起的人。”（“It was the first time I met him in such a personal way. And he was such a wonderful man.”）前往洛斯阿拉莫斯前，大家接到了一项特别保密指令：千万不要在普林斯顿买火车票。因为普林斯顿只是个小车站，如果人人都去买直达新墨西哥州阿尔伯克基的客票，势必会引人生疑。于是所有人都跑去外地买票。但费曼转念一想，既然大家都避开普林斯顿，那自己反其道而在当地买票应该毫无破绽。谁知当他走到普林斯顿售票窗口说“我要一张去新墨西哥州阿尔伯克基的票”时，售票员脱口而出：“哦！原来这些东西全是给你运的啊！几个星期以来我们一直在往外发运装满计数器的板条箱呢。”原来军方虽然保密，箱子上却清清楚楚写着送往阿尔伯克基的地址。
物理学家们动身太早，提前到达了工地。彼时宿舍楼和实验室全都没建好，他们的突然涌入让基地的后勤管理人员措手不及。军方手忙脚乱地租下周边各处的农场民房给他们暂住，科学家们每天清晨乘车上山。对于一个很少远游的东部年轻人来说，沿途的壮丽景致令人叹为观止——高耸入云的台地、雄伟绝伦的悬崖峭壁。上山路上费曼随口问身旁的司机：“这里过去会不会住过印第安人？”司机二话没说踩下刹车，领着他转过岩角，眼前赫然出现了几处古老的印第安崖洞，让人大开眼界。
### [12:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=733s) 物理学家“接管”建筑工程
第一天抵达施工现场时，眼前的景象一片杂乱。按照规划，基地核心是工作用的“技术区”（Technical Area），周围会围上一圈铁丝网；技术区外是一座配套城镇，城镇最外围还会建起一道更大的防护铁丝网。但当时一切都在赶工，四周连围栏都没立起来。费曼看见自己的助手保罗正拿着写字夹板站在路口，充当交通调度员，指挥进出的卡车该往哪个区域卸料。
走进简陋的实验室后，费曼陆续见到了许多只在《物理评论》（*Physical Review*）学术论文上拜读过大名、却从未谋面的顶级学者。有人指着一位卷着袖子、正趴在满桌蓝图上大呼小叫的男子介绍：“这位是约翰·威廉姆斯（John Williams）。”只见威廉姆斯正趴在窗前，扯着嗓子指挥各路卡车往不同方向运建材。当时实验物理学家们的实验大楼和仪器都还没影子，无事可做，索性就“接管”了施工公司，直接下场监督和协助把楼盖好。
```mermaid
flowchart TD
accTitle: 洛斯阿拉莫斯初期理论组与实验组的分工
accDescr: 展示初期实验物理学家与理论物理学家的不同工作状态
subgraph Construction ["施工与准备阶段"]
A["实验物理学家"] -->|"设备未到/无实验可做"| B["协助指挥卡车与盖楼"]
B --> C["完成技术区与实验室建设"]
end
subgraph Theoretical ["理论组攻关阶段"]
D["理论物理学家"] -->|"入住工地现场"| E["推着流动黑板听课讨论"]
E --> F["罗伯特·瑟伯尔讲解伯克利方案"]
end
```
### [13:16](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=796s) “两个疯子”的碰撞与速算较量
理论物理学家们没有住进山下的农场民房，而是坚持直接住在工地上，以便立即展开演算。他们占了几间空屋子，全组只有一块装有滑轮的移动黑板。大家推着黑板走动，由罗伯特·瑟伯尔（Robert Serber）向所有人系统传授伯克利团队构思的原爆理论和核物理知识。费曼此前并未深入接触过这个领域，只能拼命研读文献、日夜恶补。
碰巧的是，理论组的核心大腕们因各种突发状况同时离开了基地——维克托·魏斯科普夫（Victor Weisskopf）回麻省理工学院处理杂务，爱德华·泰勒（Edward Teller）也恰好外出，基地里只剩下了汉斯·贝特（Hans Bethe）。贝特亟需找人争论来碰撞灵感，便推门迈进了名不见经传的小年轻费曼的办公室。贝特刚开始阐述自己的构想，费曼便脱口而出：“不对不对，你疯了吧，这应该这样……”贝特立刻回击反驳：“等一下！”并据理力争费曼才是疯了的那一个。费曼事后回忆说，自己在物理面前从来不管对方头衔有多大，张嘴就说“这不对”或者“你疯了”，但这恰恰是贝特最渴望的毫无保留的学术交锋。这次火花四射的交流让贝特对他青睐有加，不久费曼便被提拔为贝特手下统领四名组员的研究小组长。
贝特极度擅长心算，两人在此后展开了频繁的“人脑对抗手摇计算机”比赛：
- **平方运算：** 贝特提到某个公式涉及压力平方，数值是 $48$。费曼伸手去够马钱特（Marchant）手摇计算机，贝特已脱口而出“大概 2300 左右”。费曼摇动把手算完，贝特补充道：“想要精确值吗？2304。”贝特随即倾囊相授接近 $50$ 的数字速算法：利用 $(50-x)^2 \approx 2500 - 100x + x^2$，例如 $47^2$ 就比 $25$ 少 $3$ 等于 $22$，再加上余数 $3$ 的平方 $9$，即为 $2209$。
- **开立方运算：** 遇到计算 $2.5$ 的立方根时，费曼从抽屉翻出马钱特公司附带的查表卡片准备试算，贝特沉吟片刻给出了答案：“1.35。”原来贝特记住了一连串常用对数表，他在脑中调出 $\log(2.5)$ 除以 $3$，再利用 $\log(1.3)$ 和 $\log(1.4)$ 进行线性内插推算。
- **日常速算竞争：** 费曼受到巨大刺激并勤学苦练，几年后也总结出了寻找数字巧合的技巧（如计算 $174 \times 140$ 时，将其近似转换为 $\sqrt{3} \times \sqrt{2} = \sqrt{6}$ 的比例），能在四次较量中险胜一次。
### [17:00](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1020s) 住宿风波与假扮有女伴的恶作剧
初期的住宿条件极为局促与混乱。最早理论物理学家们被挤在原农场男校的“机工宿舍”（Mechanics Lodge）里睡双层架子床。房间布局极为尴尬，物理学家罗伯特·克里斯蒂（Bob Christy）和妻子每天清晨去洗手间，都必须径直穿过费曼等人的卧室。随后他们搬到了“大屋”（Big House），二楼的整个大回廊密密麻麻排满了行军床，一楼则挂着一张统筹大表，列出每个人的床号和更衣用的浴室代号。当费曼找到自己的名字时，上面只写着“C浴室”，床号一栏竟是空白，让他哭笑不得。
好不容易盼到单身汉宿舍楼竣工，费曼兴冲冲去挑选房间。他特意挑选了一间能透过窗户遥望对面女生宿舍的朝向，尽管搬进去后才郁闷地发现窗前正好挡着一棵参天大树。管理人员通知他，眼下宿舍紧张，每个房间得临时塞两个人，中间共用一个卫生间。一心想要独享空间的费曼动起了歪脑筋。由于妻子阿琳在阿尔伯克基的医院养病，费曼身边留存着几箱她的衣物行李：
- 趁第一晚同屋室友还没分配进来，费曼翻箱倒柜找出了妻子的一件睡袍。
- 他将被子扯开，极其随意地将女式睡袍扔在靠上的双层床铺上，并将一双女式拖鞋摆在床边。
- 随后，他又在卫生间地板上故意撒落了一些香粉，制造出已有女性在此留宿的逼真假象，企图以此吓退分房管理员。
然而第二天晚上费曼回到房间，看到的场面却让他瞠目结舌：宿舍保洁员不仅没有被吓到或举报，反而尽职尽责地将费曼下铺的男式睡衣折叠整齐塞在枕头下、拖鞋摆齐；更绝的是，上铺那件用来恶作剧的女式睡袍也被抚得平平整整，折好温柔地塞进了上铺的枕头底下，女式拖鞋同样在床底整齐排放。
### [19:18](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1158s) 假装有同居女郎与竞选镇议会代表
理查德·费曼继续讲述他在洛斯阿拉莫斯单身汉宿舍保住单人间的故事。第一天得手后，第二天清晨醒来，他如法炮制：将上铺的被褥弄得乱七八糟，在浴室里撒上爽身粉、扔下睡袍。这套伪装足足持续了四个早晨，直到宿舍分配彻底定局，不再有被塞进室友的危险，他才停止。整个过程中，虽然这是一栋男士宿舍，但他每晚制造假象、每天早晨房间被打扫得整整齐齐，独占单间的计划大获成功。
不久后，当地成立了一个管理社区日常的“镇议会”（Town Council）。军方与平民组成的管委会共同决定城镇运营，小镇上随之涌现了各种政治派系，比如家庭主妇派、机械师派、技术人员派等。这时，军方颁布了一条新规定：“严禁女性进入男士宿舍。”单身汉和单身女郎们对此极为愤慨，费曼回忆道：
> “Well, this is absolutely ridiculous. All grown people, of course.”
> （“这简直荒谬透顶。大家可全都是成年人啊。”）
宿舍住户们认为必须成立自己的阵营进行抗争。经过激烈的辩论和动员，费曼被选为宿舍代表，正式进入了镇议会。
```mermaid
flowchart TD
accTitle: 假伪装引发镇议会选举流程
accDescr: 展示清洁女工发现睡袍与爽身粉后逐级上报至最高层，导致出台禁令并最终让费曼当选代表的全过程。
subgraph Report["层层上报"]
A["清洁女工发现上铺有女性睡袍与爽身粉"] --> B["清洁领班与值班中尉"]
B --> C["少校与将军"]
C --> D["洛斯阿拉莫斯管理委员会"]
end
subgraph Reaction["政策与反弹"]
D -->|"恐慌商议四天后颁布"| E["颁布禁规：严禁女性进入男士宿舍"]
E -->|"引发宿舍单身汉强烈抗议"| F["宿舍住户成立政治派系进行抗争"]
F --> G["费曼当选镇议会代表"]
end
```
在镇议会任职约一年半后，费曼与理论部主管汉斯·贝特（Hans Bethe）聊天，随口提起了自己当年在上铺扔睡袍、撒爽身粉来霸占房间的恶作剧。贝特听完哈哈大笑，当场揭晓了谜底：原来贝特当时就坐在高级管理委员会里。当年清洁女工撞见上铺像是睡了女人，吓得浑身发抖，逐级向清洁领班、中尉、少校上报，消息一路惊动了将军和最高管理委员会。高层们急得团团转，连续四天下令“先把东西原样放回去，打扫干净，看看第二天会怎样”。整整恐慌商议了四天后，委员会最终憋出了一条正式禁令——“严禁女性进入男士宿舍”。这道禁令直接引发了全宿舍的抗议狂潮，并最终把始作俑者费曼推上了镇议会代表的席位。
---
### [22:24](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1344s) 战时邮件审查制度与双向密码暗号
随后，费曼谈到了基地内部推行的邮件审查制度。在美国本土对本国公民的信件进行审查实际上是完全违法的，为了绕过法律限制，军方只能将其包装成一种“自愿协议”：
> “They decided to do something utterly illegal, which is to censor mail of people inside the United States in the continental United States, which they have no right to do.”
> （“他们决定干一件完全违法的事，那就是在美国本土内部审查本国公民的信件，他们根本没有权力这么做。”）
军方与这群崇尚自由的科学家达成微妙的君子协定：大家寄信时不封口，供审查人员检查无误后替其封口；收到的来信也允许被开封检查。规则还特意规定科学家有权在信中批评基地管理层甚至给参议员写信抱怨，并且承诺绝不干涉大家正常的通信习惯。
然而审查开始的第一天，费曼就被叫进了审查办公室。长官指着他父亲寄来的一张画满线条和黑点的信纸（线上四个点、线下一个点、两个点等）质问这是什么。费曼爽快地回答：“这是密码。”对方追问密码写了什么、密钥是什么，费曼如实回答自己一概不知。紧接着对方又拿出一封费曼妻子阿琳（Arline）寄来的信，上面写着“TJXY WZ TWW1X3”之类的乱码。审查官震惊地质问他怎么可能收到了密码却不知道密钥，费曼解释说这是他和家人的智力挑战游戏：他向家人挑战，让他们编造自己解不开的密码寄过来，所以绝不能提前寄密钥。审查官提出折中方案：必须让家属寄信时附上密钥，由审查办公室把密钥扣下，只把密文发给费曼。
- **自愿审查机制规则**：信封不得封口，审查人员若发现违规，须退回信件并附上违反条款的说明便条。
- **父亲的暗号形式**：横线上下一组组排列的点阵。
- **妻子的密文示例**：由无序字母与数字组成的字串（如 TJXY WZ TWW1X3）。
---
### [24:56](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1496s) 消字灵、剪刀剪洞与长官的权威
第二天，审查引发的摩擦进一步升级。费曼收到妻子阿琳的来信，信中写道：“写信变得非常困难，因为我总觉得……在看着我的肩膀。”而在省略的地方，字迹被消字灵（ink eradicator）擦拭得干干净净。费曼立刻跑去审查局抗议，指出根据协议，审查员无权涂改或抽出来信内容，若有异议只能通知收信人。审查人员却嘲弄他不懂行，称审查官审查违规内容历来都是用剪刀剪洞，绝不会用消字灵。
费曼随即写信询问妻子是否自己用了消字灵，阿琳回信写道：“不，我没在信里用消字灵，那肯定是……”结果信纸上的那个词被剪刀整整齐齐剪掉了一个大洞。
费曼拿着带洞的信直接找到负责审查的少校交涉。少校辩解称底层审查员受过传统军管训练，一时无法适应这种对科学家必须“小心翼翼”的新模式。费曼一连抗议了几天，少校有些挂不住面子：
> “He said, 'What's the matter? Don't you think I have goodwill?' I says, 'Yes, you have perfectly goodwill, but I don't think you have power.'”
> （“他说：‘怎么回事？难道你认为我没有诚意吗？’我说：‘不，你充满诚意，但我认为你没有实权。’”）
这句讽刺激怒了少校，他当场拿起电话把下属训斥了一通，此后剪信和涂改的现象确实被制止了。然而误会依然层出不穷：某天审查处扣押了妻子寄来的一份“没有密钥的密码”，直到费曼去阿尔伯克基的疗养院探视妻子时，妻子询问为何没买“洗衣粉、甘油、热狗、脏衣服”等生活用品，费曼才恍然大悟——那只是一张普通的购物清单（list），却被审查员误认成了高深莫测的密码。
---
### [26:44](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1604s) “阿拉伯数字”语言申请书
在审查制度推行的最初几周里，费曼在办公室摆弄机械计算机时发现了一个奇妙的数学规律：计算 $1 \div 273$ 时，得出的循环小数展开呈现为极为工整的递推数列（0.00366...中呈现类似 04115、226、337 的有趣步进结构，且在进位时规律同样优美）。由于他每天都给妻子写信分享日常琐事，便顺手把这个长串数字和自己的数学分析写进了家书寄出。
信件随即被审查员拦截退回，并附上一张警告便签：“请参阅第 17B 条款”。费曼翻开条款一看，上面赫然写着：通信仅限使用英语、俄语、西班牙语、葡萄牙语、拉丁语、德语等指定语种，使用其他任何语言必须事先获得书面许可，且严禁使用代码暗号。
面对这种官僚教条，费曼决定用物理学家的幽默进行反击：
> “I therefore ask for permission to write my letters in Arabic numerals. I like to use Arabic numerals in my letters.”
> （“我因此特此申请：准许我在信件中使用阿拉伯数字。我喜欢在信里使用阿拉伯数字。”）
费曼在信中给审查员列出了详尽的数学除法推导，严谨地证明这个除法计算结果所包含的信息量，根本不比除数“273”本身所具有的信息量更多，因此在信息论上绝不可能构成密码，并正式递交了在信中使用“阿拉伯数字”的书面申请。最终，这封信顺利通关，但在日后的通信中，费曼夫妇与审查制度之间围绕“隐私被窥视”的智斗依然接连不断。
### [28:34](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1714s) 审查员的死循环与阿琳的拼图恶作剧
费曼讲述了他与审查官之间陷入的一场荒诞逻辑怪圈。基地规定信件中绝对不能提及“审查”二字。然而，审查员却接连给费曼递便条，抱怨道：“你妻子在信里提到了审查制度。”最终审查官要求费曼写信告诫妻子，千万别在信中提这回事。费曼于是照办，在信的开头写道：“我接到指示，要通知你不要在信里提到审查。”结果信件立刻被审查处打了回来。费曼反驳道，审查员既然不准妻子提审查，却非要通过他去通知妻子不要提，这本身就自相矛盾。审查部门解释说，他们担心从阿尔伯克基寄出的邮件在半途被截获，从而泄露基地存在信件审查的秘密，因此恳请费曼的妻子在信里表现得“更正常一些”。
费曼随后亲自前往阿尔伯克基探望妻子阿琳，劝她别再直接提审查。但两人此前被百般刁难，早就不甘示弱地私下约定了一套暗号：只要费曼在署名末尾点上一个句点，就意味着他们又被审查处找了麻烦，阿琳便会立刻启动下一套恶作剧计划。阿琳因病终日卧床，有大把时间构思点子。她在一则广告上看到一种空白拼图商品，上面写着“把信写在拼图上寄给男友：附赠空白拼图，写好后拆散装袋寄出”。阿琳果真买来拼图，写下信拆散成一堆碎片寄了过来。审查处不得不硬着头皮拼完，并在附信中抓狂地警告费曼：“我们没闲心玩游戏，请指示你的妻子只写寻常信件。”
阿琳甚至还准备了更狠的后招：信件开头写着“希望你拆信时格外小心，因为我按照约定在信里给你放了缓解胃痛的 Pepto-Bismol 药粉”。按照设想，审查员一旦匆忙拆开，药粉就会洒满整间办公室的地面，而审查纪律严格规定不得破坏信件内任何物品，审查员势必得狼狈不堪地把粉末一点点收集起来。不过由于审查部门及时服软，这封药粉信最终没有寄出。
- 费曼与妻子通过在署名后加一个句点作为秘密代码，指示对方触发下一个恶搞审查制度的计划。
- 审查部门在拼完拼图信件后递来严厉警告，要求费曼督促妻子遵守寻常通信规则。
---
### [30:48](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1848s) 铁丝网的漏洞与安全保密的讽刺赌局
长期与审查官斗智斗勇，让费曼彻底摸清了审查系统的边界，甚至以此打赌赢钱。某天在基地外围巡视时，费曼发现那些住在基地更外围的工人嫌绕大门太远，擅自把铁丝网剪开了一个大洞穿行。为了证明安保漏洞有多可笑，费曼先正大光明从正门刷卡出去，绕到铁丝网破洞钻进来，接着再从正门出去、破洞进来，如此反复多次。守门军士彻底懵了，只见费曼不断走出大门，却从不见他走进来。军士的第一反应不是排查漏洞，而是立刻叫来中尉要把费曼关进监狱。
费曼坦白了破洞的存在，并决定以此与人打赌：他能把基地围栏有洞的确切信息通过审查部门合法寄出去。费曼抓住了审查规则中允许抱怨行政管理的漏洞，在信里写道：“你真应该看看他们是怎么管理这个地方的——距离某某地点 71 英尺处的铁丝网上就有一个洞，尺寸有多大多大，人完全可以径直走过去。”审查员对此无计可施，因为事实确凿，他们总不能硬说没有这个洞，只能放行信件并赶紧去补网。此外，费曼还成功寄出过一封信，控诉军方某些蠢材仅仅因为某位组员的父亲疑似是共产党人，就在半夜把那名年轻人叫醒、用强光照着眼睛进行严苛审讯。那位遭审讯的人正是日后名声显赫的物理学家戴维·鲍姆（David Bohm）。
---
### [32:22](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=1942s) 抽屉盗宝：爱德华·泰勒的极速顿悟
洛斯阿拉莫斯早期的保密硬件极其简陋。关于铀裂变的大量极其关键的核心机密文件，居然只是放在木质文件柜里，外面挂着普通杂货店买来的廉价挂锁，或是由机械工场随手焊一根铁杆穿过把手锁上。不仅锁极容易被撬开，甚至根本无需撬锁：木柜下方为了固定抽屉滑轨留有空隙，只要把整个柜子向后倾斜，就能直接把手伸进柜底将绝密文件一张张抽出来。费曼频繁在全体大会上当众撬锁，反复疾呼：“我们有重要的机密，我们必须换更好的锁。”
在一次大会上，理论物理学家爱德华·泰勒站起来向费曼挑衅：“好吧，我最核心的机密从不放在文件柜里，而是放在办公桌抽屉里。那样不是安全得多吗？”费曼回答：“我不知道，我还没见识过你的抽屉。”泰勒坐在会场靠前排，费曼坐在后排。会议还在进行，费曼便悄悄溜出会场直奔泰勒的办公室。他甚至用不着开锁，将手伸进抽屉底下的缝隙，就像抽卫生纸一样，抽出一张就会带出下一张。费曼把泰勒抽屉里的核心文件掏得一干二净，藏在一边，然后若无其事地回到楼上。散会人群涌出时，费曼假装刚散会迎面碰上泰勒，提出想看看他的抽屉。泰勒欣然带他走进办公室，费曼装模作样地夸赞抽屉看上去确实很牢固，并提议打开看看里面放了什么。泰勒用钥匙打开抽屉低头一看，立刻反应过来，转头对费曼说：“要是你之前没看过的话，确实挺安全的。”费曼对此感叹道：
> “捉弄泰勒先生这样绝顶聪明的人，最大的弊端就在于，从他发现事情不对劲到彻底想明白前因后果，所花的时间实在太他妈短了，短到你根本来不及体会任何恶作剧得逞的快感。”
---
### [35:08](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2108s) 橡树岭的临界隐患与军方保密的致命盲区
离开那些撬锁逸事，费曼转而讲述了一段在橡树岭（Oak Ridge）工厂发生的惊险技术问题。当时洛斯阿拉莫斯负责原子弹的组装与设计，而橡树岭工厂则负责分离天然铀中的同位素——铀-238与具有裂变爆炸能力的铀-235。彼时橡树岭刚从实验装置中分离出极微量的铀-235，并正筹建大型化工厂，准备在多个巨型大桶中分阶段对提纯物进行化学提纯与浓缩再加工。
橡树岭的技术人员同时还在摸索如何对铀-235的丰度进行化学测定与分析，洛斯阿拉莫斯不断寄去测定指导说明，但橡树岭的化验数据却总是出错。物理学家埃米利奥·塞格雷（Emilio Segrè）最终坚持认为，必须亲自前往橡树岭实地查看，才能找出测定失误的症结。然而军方对此极为抵触，严格奉行将信息隔绝在各孤立地点的壁垒政策。军方认为橡树岭的人绝不需要知道浓缩铀的最终用途：底层工人完全不知道自己在干什么，高级官员也仅仅知道是在提纯铀，根本不知道原子弹的威力及起爆原理，军方严禁两地互通技术情报。
但塞格雷力排众议坚持成行，否则整个化验体系一塌糊涂，项目将陷入瘫痪。当塞格雷走进橡树岭的车间时，眼前的一幕让他大惊失色：工人们正推着装满绿色水溶液（硝酸铀酰）的巨型大罐车大摇大摆地走过。塞格雷连忙询问：“高浓缩铀提纯之后，你们也打算这么大批量地放在一起处理吗？”工人们毫不在意地反问：“对啊，有什么不行？”塞格雷惊骇地问道：“这难道不会爆炸吗？”对方一头雾水：“哈？爆炸？怎么会爆炸？”军方见状反而埋怨，认为塞格雷的话泄露了保密信息。事实的真相极其荒唐：军方只知其一不知其二，他们听说制造一颗原子弹需要几十公斤（如 20 公斤）的高浓铀，便自作聪明地以为只要整座工厂任何一处存放的提纯物达不到这个总质量，就绝不可能发生危险。军方高层完全不知道物理学中的慢化剂效应——当中子处于水等氢介质中时，其引发裂变链式反应的效率会成百上千倍地暴增，即便远低于金属临界质量的浓缩铀溶液，也足以瞬间达到临界甚至引发剧烈爆炸。
```mermaid
flowchart TD
accTitle: 军方信息壁垒导致的橡树岭临界安全隐患
accDescr: 展示军方出于保密政策的信息隔绝如何险些酿成橡树岭浓缩铀溶液临界事故的逻辑链条
subgraph Policy ["军方的极端保密与信息壁垒"]
A["严格单向隔离政策<br>橡树岭不得了解核弹原理与临界参数"] --> B["军方粗暴估算安全线<br>单批次只要低于整颗核弹质量(如20kg)即认为无害"]
end
subgraph Hazard ["橡树岭现场的危险操作"]
C["化验测定频频出错"] --> D["大罐车盛装硝酸铀酰绿色溶液运转"]
B -.->|"忽视水中中子慢化效应<br>临界质量急剧下降"| D
end
subgraph Intervention ["洛斯阿拉莫斯的介入"]
E["塞格雷坚持打破壁垒现场排查"] --> F["发现致命隐患：高浓度溶液聚积极易引发瞬发临界爆炸"]
D --> F
end
```
### [37:27](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2247s) 水中的中子减速与临界危机：克里斯蒂病倒，费曼临危受命
在水中，中子会被减速。一旦中子在水中减速，引发链式反应所需的裂变材料用量将大幅骤降，甚至不到原先的百分之一。这种反应虽不至于引发巨大的核爆炸，但会产生极强的放射性，瞬间杀死周围的人员。然而，田纳西州橡树岭（Oak Ridge）工厂在设计工艺时完全没有考虑安全问题。洛斯阿拉莫斯随即发去电报，要求埃米利奥·塞格雷（Emilio Segrè）彻底排查整个工厂，摸清所有设计流程中浓缩铀的积聚点。与此同时，洛斯阿拉莫斯内部兵分两路展开紧急计算：罗伯特·克里斯蒂（Robert Christy）的小组负责计算水溶液中的临界状况，费曼的小组则负责计算盒装干粉的临界安全量。
按照原定计划，应由克里斯蒂带着两组的全部计算结果前往橡树岭，向工厂管理层全面通报这一险情。费曼高高兴兴地把自己计算的所有数据交给了克里斯蒂，以为自己可以置身事外。不料克里斯蒂突然患上了严重的肺炎，无法成行，这项重任毫无预警地落到了费曼头上。
此前从未坐过飞机的费曼不得不踏上行程。军方用一条小皮带把装有机密文件的包紧紧捆在费曼的后背上。当时的客机飞行体验就像搭乘长途巴士，隔一段时间就要在沿途经停站降落等待，只是各站之间的距离更加遥远。在某次经停等待时，旁边站着一个甩着表链的旅客，用居高临下的口吻对费曼说：“如今坐飞机要是没有优先权级别，一定难得要命吧。”费曼忍不住反击，随口应道自己有优先权。没过多久，广播通知有几位将军登机，工作人员需要把一部分持有三级优先权（Priority 3）的乘客请下飞机。那位乘客以为小年轻费曼肯定要被赶下去，费曼却神气地亮出身份——他是极为罕见且等级更高的二级优先权（Priority 2）。费曼打趣说，那人若不是国会议员，事后也一定会愤愤不平地写信给议员投诉：“战争正打得火热，他们凭什么让这种毛头小子揣着二级优先权到处飞？”
```mermaid
flowchart TD
accTitle: 橡树岭安全危机应对流程
accDescr: 展示洛斯阿拉莫斯发现橡树岭安全隐患到费曼赴实地排查的整个流程
subgraph Problem["安全隐患发现"]
A["水中中子减速"] --> B["临界质量大幅下降"]
B --> C["高致命放射性风险，工厂无防备"]
end
subgraph LA_Task["洛斯阿拉莫斯应急分工"]
D["克里斯蒂组：计算水溶液临界"]
E["费曼组：计算干粉临界"]
end
subgraph Dispatch["人员指派与行程"]
F["克里斯蒂突患肺炎"]
G["费曼顶替赴橡树岭"]
H["携带机密背带乘机 (二级优先权)"]
end
C --> LA_Task
D & E --> F
F --> G --> H
```
### [39:14](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2354s) 穿行于9207号厂房：超强短期记忆与奥本海默的“通牒”
抵达橡树岭后，费曼要求做的第一件事就是让人带他走遍整个工厂。一路上他一言不发，只是默默注视着流水线和容器。费曼很快发现，实际情况比塞格雷报告中描述的还要危急得多。塞格雷初次探访时被复杂的厂房结构搞糊涂了，他只注意到某处堆着大批物料箱子，却没有留意到隔壁房间同样堆积着巨量原料，而那堵墙的另一侧其实属于同一个相连的大开间。核材料堆叠得过近、过多，随时都有可能引发致命失控。
平日里费曼自认记性很差，但在处于高度集中工作时，他的短期记忆却格外惊人。他能清晰记下所有极其繁复琐碎的信息，比如“9207号建筑某某号槽”（Building 9207 VAT number so-and-so）。当天晚上回到住处后，费曼连夜梳理了所有安全漏洞并制定补救措施。从物理原理上看，解决方案相当直接明确：在溶液中添加镉（cadmium）以吸收水中的慢中子；根据严格的间距规则分散存放干料箱，避免铀积聚过密。
早在费曼从洛斯阿拉莫斯出发前，罗伯特·奥本海默（J. Robert Oppenheimer）就向他传授了应对大场面的威慑战术。奥本海默点出橡树岭现场在技术上真正顶用的核心专家名字，比如朱利安·韦伯（Julian Webb），并千叮咛万嘱咐费曼：必须确保这些业务骨干出席次日的安全大会，让他们彻底明白工厂的安危全由他们负责。费曼当时心里犯怵，追问奥本海默：“要是他们不到场，我该怎么办？”奥本海默授意他直接抛出最高通牒：“洛斯阿拉莫斯无法为橡树岭工厂的安全承担责任，除非……”费曼听后目瞪口呆，不敢置信地反问：“你是说让我——小理查德（little Richard）——走进去跟他们这么说？”
> 奥本海默斩钉截铁地回答：“没错，小理查德，你去，就这么办。”（“Yes, little Richard, you go and do that.”）
费曼感叹，这番历练真让自己一夜之间迅速成熟起来。
- 关键防范措施：在水溶液中引入吸中子能力极强的镉元素。
- 物料隔离要求：干粉箱严格按照安全距离与体积规则摆放，打破致密堆积。
- 奥本海默钦点人物：包括朱利安·韦伯等能真正理解裂变原理的橡树岭核心技术骨干。
### [41:03](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2463s) 决策的五分钟：军方妥协与“从天而降的神明”
第二天的大型协调会议现场高朋满座，公司高层大亨、被费曼指名要求的技术骨干，以及掌控整个工程调度的军方将军们齐聚一堂。所有人都意识到橡树岭面临着生死攸关的致命瓶颈——正如费曼所言，如果当时没人重视，这座工厂绝对会引发恶性临界事故。然而，在会议正式开始前，负责陪同费曼的一名陆军中尉转达了一位陆战上校的死命令：费曼只能向下传达具体的安全操作规程，绝对不准向工厂人员透露中子物理的具体机制，因为军方坚持要执行严密的情报物理隔离。
面对军方的层层设限，费曼坚决不肯退让。他指出，如果不搞清楚原理，工人们根本无法理解也不可能遵守一堆莫名其妙的条规；只要盲目操作，工厂早晚会出大祸。费曼直接搬出了奥本海默教他的杀手锏：洛斯阿拉莫斯概不负责。
> 费曼态度坚决地表示：“我认为，他们根本不可能理解或者去遵守一大堆自己搞不懂的规则，除非他们明白这背后的运行原理。因此依我之见，只有让我彻底给他们讲透，事情才能办成；否则，洛斯阿拉莫斯无法为橡树岭工厂的安全承担责任，除非他们完全知晓其工作机制。”（“In my opinion it's impossible for them to understand or to obey a bunch of rules that they don't understand unless they understand how it works. And so it's my opinion that it's only going to work if I tell them and so cannot accept the responsibility for the safety of the Oak Ridge plant unless they are fully informed as to how it works.”）
中尉随即跑去向负责的上校汇报。那位上校走到窗前，背对着众人足足凝神思考了五分钟。费曼由衷佩服这些职业军官雷厉风行的决断力：要在短短五分钟之内，打破绝密壁垒，决定是否允许原子弹核心机密在橡树岭公开，换作自己绝对犹豫不决。五分钟后，上校转过身来拍板：“好吧，费曼先生，讲吧。”
限制一解除，费曼随即坐下倾囊相授。他深入浅出地讲解了中子的扩散、为何物料集中会酿成大祸、镉如何吸收中子，以及慢中子相较于快中子为何具有更高的裂变诱发效率。这些在洛斯阿拉莫斯甚至称得上是入门级的基础知识，对于橡树岭的工程师和高管来说却是闻所未闻的天方夜谭。在那一刻，费曼被在场所有人奉若神明：
> 费曼调侃道：“我就像一个从天而降的神仙，带着所有这些他们从未理解、闻所未闻的奇异现象降临；我不仅深知其中的奥秘，还能随口给出确凿的事实、精确的数据以及所有的一切。于是，我在洛斯阿拉莫斯不过是个底层的后生晚辈，到了这头却摇身一变成了无所不知的超级天才。”（“I was a god coming down from the sky with all all these phenomena that were not understood, never heard of before. I knew all about it. I could give them facts and numbers and everything else. So by being rather primitive back there at Los Alamos, I was a super genius at the other end.”）
听完讲授后，橡树岭各部门立刻行动起来。他们成立了专门的计算小组，开始依据费曼的指导重新核算方案，并召集建筑设计团队、施工工程人员以及负责分离后物料处理的化工专家，全面着手重新设计工厂。
### [43:13](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2593s) 斯通-韦伯斯特图纸之谜：蒙对阀门的奇迹
大约一个月后，费曼依约重返橡树岭，负责审核由知名工程承包商斯通-韦伯斯特公司（Stone & Webster）刚刚赶工出炉的全新工艺图纸。对于如何去审查一座“连地基都还没影的化工厂”，费曼自己心里也没底。在形影不离的护送军官朱姆沃尔特中尉（Lieutenant Zumwalt）的陪同下，费曼走入了一间会议室。两位工程设计师正守在一张巨大无比的长桌旁，桌上铺着一张与桌面等大的巨幅工程蓝图，下方还压着厚厚的一大摞图纸。
费曼虽然在求学时期学过机械制图，但他对阅读极其复杂的化工厂管网施工蓝图并不在行。对面的两位工程师把费曼看作神童专家，语速飞快且毕恭毕敬地展示防爆设计：“费曼先生，我们希望向您说明工厂是这样设计的。您看，为了防止物料积聚——比如浓缩蒸发器在浓缩物料时，一旦阀门卡死导致过度积聚就会爆炸——我们特意设计成了没有任何单一阀门故障会导致事故，系统中任何一个环节都需要至少两道阀门同时卡死才会出问题。”
接着，工程师的手指在层层叠叠的蓝图上飞速比划，滔滔不绝地介绍工艺走向：四氯化碳从这里注入，硝酸双氧铀从那里通入，管道穿过地板伸向二楼，管路上下翻飞、循环交错。站在一旁的费曼看得头晕目眩、大脑一片空白。更糟糕的是，图纸上铺满了一种费曼根本看不懂的符号——一个正方形，中间画着一个小十字。费曼起初以为那代表窗户，但很快否定了自己的猜测，因为这个记号经常出现在管道密集处和建筑内部，根本不可能全都是窗户。
费曼很想开口问那到底是什么，但他错过了最初的最佳提问时机。工程师已经兴致勃勃地讲解了许久，如果这时候突然问一个入门级的符号含义，对方肯定会认定他纯属滥竽充数、白白浪费大家时间。情急之下，费曼迎来了一生中堪称神迹的绝妙运气。他暗自琢磨：这东西八成是个阀门。为了验证这一点，费曼伸出手指，在第三页蓝图正中央的一个十字方块上重重一点，装模作样地质问道：“要是这个阀门卡死了，会发生什么？”
费曼原以为对方会当场纠正说“先生，那不是阀门，那是扇窗户”。然而两位工程师一听，立刻严肃地面面相觑：“嗯，如果这个阀门卡死的话……”随即，两人趴在图纸上疯狂顺着管路追踪，手指在庞杂的图线间上下飞划、反复核对。经过漫长的追踪复核，两人抬起头，嘴巴张得老大，满脸惊骇与敬佩地看着费曼惊呼：
> “先生，您完全正确！”（“You're absolutely right sir.”）
费曼随手蒙中的那个小方块不仅确实是个阀门，而且恰恰是整套冗余设计中唯一漏网的致命单点故障：一旦该阀门卡死，系统将绕过所有安全防线导致物料过量积聚，进而引发致命的超临界爆炸。在全场雷鸣般的掌声中，两位心服口服的工程师默默收起蓝图，彻底折服于费曼深不可测的“神眼”。
### [46:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2769s) 揭开橡树岭阀门的真相与徒手修理 Marchant 计算器
在橡树岭（Oak Ridge）工厂化解危机后，一直紧跟着理查德·费曼（Richard Feynman）的祖姆沃尔特先生（Mr. Zumwalt）彻底被折服了。祖姆沃尔特由衷地赞叹费曼是个不可思议的天才：“我第一次觉得你是个天才，是你仅在工厂走了一圈，第二天早上就能随口指出 9207 号楼的 C-21 蒸发器存在问题；后来你又对中子了如指掌；但你刚才做的事情简直太神了，我很想知道，你究竟是怎么做到的？”面对对方近乎崇拜的追问，费曼却非常接地气地给出了极具个人风格的幽默答复：“你试着去查清楚那到底是不是个阀门就行了。”
回到洛斯阿拉莫斯（Los Alamos）的日常工作，费曼谈到了另一项沉重且繁琐的硬核任务——进行海量的数据计算。当时实验室完全依赖机械式的 Marchant 手摇/电动台式计算器，算起乘除加减非常费劲，极容易磨损损坏。按照正规规定，仪器一旦损坏必须寄回原厂维修，但频繁返厂导致实验室机器严重短缺，眼看科研就要停摆。费曼和几个同事坐不住了，索性违抗“擅自拆盖概不负责”的保修禁令，直接把计算器的金属外壳撬开自己修。
他们把修计算器当成了老天赐予的一连串进阶课。最开始拆开一台时，只看到一根轴上有个孔，旁边耷拉着一根弹簧，显然弹簧就该挂在孔里，挂上机器立马就转了。借由这种摸索，他们的修理技术突飞猛进，越修越熟练，常规故障全能自行解决，只有遇到真正棘手的大故障才会寄回工厂。后来其他几位同事相继放弃，费曼便一个人包揽了全实验室所有的 Marchant 计算器维修；机房虽然还有打字机需要维护，但好在机加工车间有位手艺更精湛的师傅负责，费曼便把全副精力放在了这批机械计算机器上。
### [47:57](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=2877s) 转向 IBM 打孔卡与流水线“真人计算机”实验
随着研究深入，洛斯阿拉莫斯面临一个极为宏大的核心物理课题：精确计算原子弹爆炸过程中的全链条动力学——外层高能炸药向内挤压核心（内爆），随后能量释放再次向外急剧膨胀，这期间到底发生了什么，释放了多少能量。这项工作所需的计算量远超人力摇 Marchant 计算器的极限。此时，一位名叫斯坦利·弗兰克尔（Stanley Frankel）的聪明学者提出，完全可以借助 IBM 的商用制表机来处理。当时的 IBM 设备主要是商业用途的加法制表机（tabulators）和乘法机（multipliers），通过读取打孔卡上的两组数字完成相乘并打印结果，配合整理机（collators）和排序机（sorters）一起使用。弗兰克尔设计了一套循环计算程序，让穿孔卡片在不同机器之间流转，开创了如今数值计算中习以为常的“流水线批量化作业”。
洛斯阿拉莫斯立即向 IBM 下单采购机器，并得知军方有名受过 IBM 专业培训的技术员可以调来负责维护，但人员调配却一拖再拖。在争分夺秒的战时节奏下，费曼他们一刻也不愿空等，虽然编写好了包含加、减、乘各项数值步骤的全部程序，手头却没有一台 IBM 机器可以测试纠错。为了验证算法，费曼想出了一个奇招：他把一间屋子腾出来，召集了一批女算手，每人发一台 Marchant 计算器。每个女孩在流水线上扮演一台特定的机器——有人专门负责做乘法，有人专司加法，还有人专门做三次方计算。
计算卡片在女孩们之间依次传递，上一棒算完三次方把索引卡递给下一棒，下一棒立刻模拟乘法机或加法机继续运算。这种模拟测试不仅把程序中潜藏的缺陷（bugs）全部排查干净，还诞生了惊人的运算速度。此前世上所有的手算工作都是由单个人从头算到尾，而引入福特流水线生产概念后，效率呈几何级提升。这套“真人流水线”的计算速率竟然完全达到了 IBM 机器的设计理论速度；两者唯一的区别在于，IBM 机器不知疲倦、可以一天三班倒二十四小时连轴转，而姑娘们连轴算一阵子就精疲力竭了。
```mermaid
flowchart LR
accTitle: 人力流水线模拟 IBM 打孔卡计算流程
accDescr: 展现费曼在没有 IBM 机器的情况下，通过女算手分别扮演各类运算单元，跑通数值程序的模拟机制。
subgraph Pipeline["女算手流水线（模拟 IBM 打孔卡系统）"]
direction LR
A["索引卡输入"] --> B["算手 1：三次方计算"]
B --> C["算手 2：模拟乘法机"]
C --> D["算手 3：模拟加法制表机"]
D --> E["排查完 Bug 的数值结果"]
end
```
- **拆机违规标语**：厂家明确规定只要用户自行拆下 Marchant 计算器外壳，工厂一概不负责保修。
- **设备配置**：IBM 商业穿孔卡系统核心组件包括加法制表机、乘法机、整理机与分类机。
- **人力流水线产出**：通过分工专精打破传统全流程单人手算，实测速度直接追平了尚未到货的 IBM 硬件理论吞吐量。
### [50:21](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3021s) 硬件组装排障与弗兰克尔的“计算机痴迷症”
IBM 设备运抵基地时，那位军方维修人员依然没到。当时的计算机设备极其复杂，拆成半散件装箱送来，附带了漫天飞舞的电线与接线图纸。弗兰克尔、费曼和另一位同事硬着头皮挽起袖子自己拼装。组装过程中最大的阻碍反倒来自基地的各路大人物，领导们频繁走进来在旁边神经兮兮地警告：“你们要弄坏东西了！你们马上就要把它弄坏了！”
他们跌跌撞撞地组装调试，有些机器顺利开机，有些接错线无法运转。费曼在一台乘法机内部发现了一个弯曲受损的零部件，但由于身边的大官们整天恐吓说“你们会造成不可逆的永久破坏”，他没敢强行将其扳直。不久后，IBM 公司的专业技术员终于如期赶到，顺利调试好了其余机器，却唯独卡在费曼发现问题的那台乘法机上整整折腾了三天。费曼走过去提醒他：“哎，我之前注意到里面那个零件弯了。”对方低头一看恍然大悟：“原来是这儿，这就全说得通了！”伸手一调，机器立刻正常运作。
然而系统跑通后，弗兰克尔却染上了一种现代计算机从业者极为熟悉的顽疾——“计算机痴迷症”（the computer disease）。这种毛病具有极强的诱惑性和破坏性：操作者沉醉于机器的奇妙逻辑，开始痴迷于利用条件分支拨动开关（如 X 开关，判断偶数执行一套指令，奇数执行另一套），变着法子压榨单台机器的极限性能。弗兰克尔完全陷了进去，不仅无心管理团队，甚至把自己关在屋里，专心研究如何让一台制表机自动计算并打印反正切函数 $\arctan(x)$ 的整张数值表——通过边走边积分自动生成数据列。费曼直言这完全是徒劳的无用功，因为实验室书架上早就摆满了现成的反正切函数数值表。
> “The disease with computers is you play with them. They're so wonderful.”
> ——理查德·费曼（关于沉迷于把玩计算机逻辑而荒废主线任务的现象）
### [52:40](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3160s) 接管 IBM 计算组：信息公开与多任务交错处理
由于弗兰克尔沉迷“玩电脑”彻底失职，整个部门的推进速度慢如蜗牛，九个月内总共才完成了三个物理问题的计算。实验室高层当即决定让费曼停下手头原先的小组工作，全权接管 IBM 计算组。费曼上任后深知必须杜绝“计算机痴迷症”，同时他敏锐地察觉到了导致整个小组效率低下的核心病根：体制性的信息黑幕。
为了配合项目，军方从全美各地高中选拔了一批具有工程天赋的聪明小伙子，编成“特种工程分队”（Special Engineer Detachment, 简称 SED）送上洛斯阿拉莫斯山头。军方把他们锁在营房和工位上，却实行严苛的保密隔离，完全不告诉他们这些工作究竟是干什么的。这些年轻士兵日复一日麻木地在穿孔机上给一张张印着枯燥数字的卡片打孔，根本毫无工作热情，进度自然奇慢无比。
费曼立刻去找保密安全部门交涉并据理力争，终于拿到了特殊许可。他召集这批工程士兵开了一次坦诚的大会，把项目的底细全部讲了个通透：他们是在参与一场关乎生死存亡的战争，正在研制原子弹；卡片上的数字代表的是真实的物理状态，最终压强数值算得越高，意味着爆炸释放出的能量就越庞大。这番讲解带来了翻天覆地的转变：
> “Complete transformation. They began to invent ways of doing it better. They improved the scheme. They worked at night. They didn't need supervisors in the night. They didn't need anything. They understood everything.”
> ——理查德·费曼（关于向基层士兵公开项目全貌后引发的巨大转变）
士兵们的主动性被彻底激活，他们不仅夜以继日地主动加班，不需要任何监工，甚至独立发明出了好几套更高效的优化程序方案。仅仅因为把真相告诉了大家，团队战斗力瞬间爆发。此前弗兰克尔带队耗费 9 个月才算完 3 个课题，费曼带队只用 3 个月就完成了 9 个复杂课题，运算速度飙升了近 10 倍。
为了实现如此惊人的吞吐量，费曼还引入了一项绝妙的并行调度技巧。原先一套题目的打孔卡片必须按部就班地在一间机房的各个机器间循环流转，机器经常处于闲置等待状态。费曼采用不同颜色的卡片代表不同的计算任务，并利用时间差进行“错相位”（out of phase）流水线作业——当第一套黄色卡片在乘法机上处理时，第二套不同颜色的卡片已经在前一级的加法制表机上开跑，从而实现两到三个复杂物理问题在同一批硬件设备中同时并行运算。
- **特种工程分队（SED）**：军方在全国范围内网罗的具有理科和工程天赋的年轻士兵，但在保密规定下最初被当成无意识的“打孔人”。
- **产出对比**：
- 接管前（弗兰克尔时期）：耗时 9 个月，仅完成 3 个课题。
- 接管后（费曼时期）：耗时 3 个月，高质量完成 9 个课题（综合效能提升近 10 倍）。
- **调度创新**：利用不同颜色的卡片标记独立计算题目，在空间流水线中错开相位，实现商用机器系统上的多任务并发运算。
### [54:30](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3270s) 紧迫的计算任务与借车奔丧
在临近战争结束、即将在阿尔伯克基附近进行核试验之前，团队面临一个紧迫问题：核爆究竟会释放多少能量。尽管计算团队之前计算过多种设计方案的能量释放，但最终决定采用的那个具体构型却一直未曾计算。物理学家罗伯特·克里斯蒂（Bob Christy）亲自前来，要求在短短三周至一个月内给出该构型的计算结果。费曼最初回绝说这绝不可能，但克里斯蒂指出团队每月的产出速率完全能支撑这个周期。费曼向他解释，之前的流水线模式是并行处理多个问题，单个问题的实际流转周期其实非常长。
克里斯蒂离开后，费曼开始琢磨打破常规流水线的可行性。如果停下机器上的其他所有任务，排除一切干扰全力攻坚，是否能大幅压缩周期？他在黑板上给年轻的计算人员出了一道挑战题，询问大家能否拿下这项任务。大家士气高涨，纷纷表示愿意加班加点、倒双班全力以赴。于是团队定下死规矩：撤下其他所有题目，全机组专攻这一个问题。就在攻坚刚刚展开时，费曼身患肺结核的妻子阿琳（Arline）在阿尔伯克基去世。费曼不得不紧急告假赶往当地。他向住在宿舍的朋友克劳斯·富赫斯（Klaus Fuchs）借了一辆轿车，当时费曼完全不知道富赫斯其实是一名苏联间谍，正借着这辆车往返圣菲传递原子弹机密。这辆车在路上极其不顺，中途接连爆了三次胎。
- 克里斯蒂要求在三周（最多一个月）内拿到最终测试构型的能量释放计算数据。
- 费曼团队确立了“单任务集中攻坚”原则，全员采用双班倒和加班模式推进计算。
- 借给费曼轿车的前往阿尔伯克基的密友克劳斯·富赫斯，实为日后震惊世界的苏联特工。
```mermaid
flowchart TD
accTitle: 三色穿孔卡片错误修正流程
accDescr: 展示主卡片出现数值误差后，团队通过小规模子卡片快速追赶并替换修复的流转机制。
subgraph MainFlow["主计算队列（50张卡片）"]
W1["白色主卡片推进计算"] --> W2["发现单张卡片数值错误"]
W2 --> W3["误差如疾病般向邻近卡片扩散"]
end
subgraph Patch1["一级快速修补（10张卡片）"]
B1["抽取邻近10张卡组成蓝色副卡"]
B1 --> B2["小卡组高速过机追赶计算"]
B2 --> B3["发现蓝色副卡内亦存在计算错误"]
end
subgraph Patch2["二级极速修补"]
Y1["抽取更少卡片组成黄色副卡"]
Y1 --> Y2["极速运算并修正数值"]
end
W2 --> B1
B3 --> Y1
Y2 -->|"回填数据"| B2
B2 -->|"缝合替换"| W3
W3 --> Done["按时完成最终测试计算"]
```
### [56:01](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3361s) 三色穿孔卡片与团队的修错奇策
由于妻子离世，身为团队主管的费曼耽搁了三天。当他心急如焚地赶回洛斯阿拉莫斯实验室的机房时，眼前的一幕却让他大吃一惊：房间里的穿孔卡片居然分成了白、蓝、黄三种不同的颜色在同时运转。费曼当即质疑他们违反了只算一个题目的禁令，但手下的年轻人不仅没有退缩，反而直接冲他喊道：“出去！出去！在外面等会儿，我们待会儿向你解释清楚！”
费曼在门外等候，随后才弄明白了这帮年轻人的惊人巧思。这套物理计算中，一叠卡片对应空间中由浅入深的各个切面。在过机运算时，只要某张卡片出现数值偏差，这种误差在下一轮迭代中就会像传染病一样蔓延到周围的卡片上——比如第39张卡片出错，下一轮就会波及第37至第40张。以往的做法是推倒重来，但面对死线，停机重算无异于前功尽弃。组员们发现，若只抽取出错误中心及其受波及的10张卡片单独处理，小卡组通过机器的速度远快于50张的大卡组。于是他们用蓝色卡片快速追赶重算，与此同时让原本的白色大卡组带着扩散中的误差继续推进，打算等小卡组算完再“缝合”替换进去。令人啼笑皆非的是，蓝色卡片在追赶过程中自身也出了错，他们便依样画葫芦制作了更少张数的黄色卡片进行二次套叠追赶。就在三种颜色卡片交织运转、全员神经紧绷的关头，费曼正好推门而入。对于组员们的自主创新，费曼极为赞赏，并在回忆他们要求自己别插手时感叹道：“‘别管我们，’他们说。于是我就没去管他们，最终一切都顺利解决了。我们按时搞定了那个问题，事情就是这么办成的。”
- 误差扩散机制：50张卡片中的单点数值错误，会在每个迭代周期向两侧相邻卡片扩散（第39张卡片出错会逐步演变为影响36至40张）。
- 快速追赶缝合方案：通过10张蓝卡的小规模子集快速过机抢跑，在不终止主卡组运算的前提下完成局部数据修正并回填。
- 最终成效：这套极具创造性的容错流水线保住了计算进度，赶在沙漠核试验前拿出了准确的能量释放数据。
### [58:10](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3490s) 大师云集：费米的直觉与冯·诺伊曼的“处世哲学”
费曼谈到自己在洛斯阿拉莫斯从一名普通下属成长为小组长的过程中，接触到了极多在物理学史册上赫赫有名的大师，这是他人生中最宝贵的经历之一。其中，恩里科·费米（Enrico Fermi）从芝加哥前来担任顾问，费曼当时正为一项极端繁琐复杂的计算结果苦恼不已。平日里费曼向来以物理直觉敏锐著称，总能预先猜到计算的大致趋势并给出定性解释，但这道题目的复杂程度却让他无法看透背后的物理机理。
当费曼准备向费米陈述这道难题时，费米却示意他先别透露结果。费米在脑海中略加思索，随即准确预言了答案的走向，并给出了一套极其自然、显而易见的物理机制解释。在回忆这场智力冲击时，费曼坦承：“所以，他在我原本擅长的领域，做得比我出色十倍；这对我来说真是一次深刻的教训。”
另一位给费曼留下深刻印记的学者是数学大师约翰·冯·诺伊曼（John von Neumann）。针对计算过程中出现的数值不稳定性，冯·诺伊曼提供了极为透彻的技术指导。在周日休息时，费曼常与汉斯·贝特（Hans Bethe）、冯·诺伊曼等人一同漫步在峡谷之中。在一次散步中，冯·诺伊曼向费曼阐述了一个理念：一个人不必为自己所身处的世界负全责。这番话彻底解开了费曼的心结，促使他形成了一种强大的“社会无责任感”，从而能够摆脱宏大压力的束缚、专注于科学探索，并成为一个快活纯粹的人。
- 费米展现出惊人的定性分析直觉，在完全未看计算过程的情况下准确推导出复杂结果及其物理本质。
- 冯·诺伊曼在技术上解决了计算机数值模拟中的不稳定性难题。
- 冯·诺伊曼的处世哲学深刻影响了费曼：“你大可不必为自己所处的世界负责”，这一观点为费曼注入了伴随一生的豁达心态。
### [01:00:11](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3611s) 尼尔斯·玻尔与不愿唯唯诺诺的“小家伙”
量子力学奠基人尼尔斯·玻尔（Niels Bohr）当时化名“尼古拉斯·贝克”（Nicholas Baker）来到洛斯阿拉莫斯视察，其子奥格·玻尔（Aage Bohr）则化名“吉姆·贝克”（Jim Baker）。在基地各大顶级科学家眼中，玻尔犹如科学界的“上帝”。初次全体会议时，由于众人争相瞻仰玻尔的风采，资历尚浅的费曼只能缩在会议室的角落里，穿过层层人群的间隙远远望着他。
然而在玻尔第二次来访的清晨，费曼意外接到了奥格·玻尔打来的电话，邀请他早上八点整在众人上班前与老玻尔闭门会面。一头雾水的费曼赶到办公室后，玻尔便开门见山地拿出关于提高炸弹效率的新设想。费曼生来只关注物理本身，从不把对方的头衔身份放在心上。面对科学泰斗，费曼毫不留情地当面驳斥糟糕的想法，直言“这行不通”、“太糟了”，若是合理的想法便认可其可行性。两人就这样激烈争辩了整整两个小时，其间玻尔不停地点燃他那总是熄灭的烟斗，用含混不清的声音阐述观点，奥格则在一旁负责帮费曼厘清玻尔的话。直到争论尘埃落定，玻尔吸着烟斗说道：“好啦，我想现在可以把那些大人物叫进来了。”
事后奥格向费曼透露了内情：初次来访后，老玻尔就特意叮嘱儿子记住角落里那个小家伙的名字。玻尔指出，基地里那些大人物见到他只会恭敬地说“是的，玻尔博士”，根本无法进行真正的批判性学术探讨；只有那个不知畏惧的年轻人敢于当面指出自己的愚蠢想法，因此在拿出方案面对大众之前，必须先找费曼“过招”。在这段波澜壮阔的技术准备告一段落后，首次核试验的时刻终于来临。当时费曼因料理妻子后事正在家休整，突然接到一封暗语电报，通知他“婴儿预计将在某日降生”。费曼火速乘机赶回前线，恰好在全员登上前往三位一体核试验场的大巴时赶到了现场。
- 尼尔斯·玻尔与其子奥格·玻尔在曼哈顿工程期间分别使用化名“尼古拉斯·贝克”和“吉姆·贝克”。
- 费曼面对科学权威秉持纯粹求真的科学态度：“若想法糟糕，我就说它糟糕；若想法精妙，我就说它精妙。”
- 玻尔之所以私下单独召见费曼，正是看中了他不迷信权威、敢于当面批驳错误的求实品格。
- 临近核试验时，内部以“婴儿预计降生”作为核装置即将起爆的秘密代号，紧急召回了身处后方的费曼。
### [01:03:07](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3787s) 三位一体核试验：唯一用肉眼直视核爆的人
在三位一体（Trinity）核试验当天，现场撤离工作极其紧张，理查德·费曼甚至连自己的房间都回不去就直接赶往了观测点。费曼所在的观测点距离爆心约20英里（约32公里），而其他更核心的人员则在距离仅6英里的较近位置。他们配有一台用于接收倒计时广播的收音机，但那台收音机一直故障失灵，导致大家完全不知道爆破现场的进展；直到爆炸前几分钟，收音机才奇迹般地恢复工作，播报还剩最后大约20秒。现场给远距离观测者们分发了极深的暗色护目镜，声称必须戴上它才能观看。费曼对此完全不以为然，他心想人距离爆心足足有20英里，透过那种极黑的镜片根本什么都看不见。
费曼凭借物理学常识敏锐地判断出：“真正会伤害眼睛的绝不是可见强光，而是紫外线（The only thing that would really hurt your eye, bright light never can hurt your eye. It's ultraviolet light that does）。”由于普通玻璃可以彻底阻挡紫外线，他立刻钻进了一辆卡车的驾驶室，躲在挡风玻璃后面直视爆心。时间一到，极度耀眼的强光瞬间照亮天地，强烈的刺激让费曼本能地迅速低了一下头。他在卡车底板上看到了一大块紫色的光斑，立刻意识到“那不是真正的景象，那是视网膜上的余辉（That ain't it. That's an after image）”，于是迅速抬头重新看向天空。
他亲眼见证了这幕人类历史上前所未有的奇景：纯白色的刺眼强光逐渐演变成明黄，随后又转为橙色；冲击波引发的空气剧烈压缩与膨胀，导致周围的云朵瞬间成型又迅速消散；爆心最刺眼的极亮中心随后化作一个巨大的橙色火球，开始翻滚着腾空升起，外围边缘泛起黑色，最终演变成一团翻腾的浓烟巨球，内部依然闪烁着往外喷涌的火焰与毁灭性的热浪。整整约一分钟的时间里，天地经历了一场从极昼般的炫目到阴暗烟尘的演变。费曼感慨自己大概是全世界唯一用肉眼看清这场核试验的人——因为6英里近处的人全被命令趴在地上死死闭着眼睛，而远处的其他人全戴着深黑护目镜，眼前只有漆黑一片。
在全场鸦雀无声地静默观看了大约一分半钟之后，一声撕裂天地的巨响“砰”地炸开，紧接着是滚滚如雷鸣般的轰响。在此之前没有任何人敢说一句话，直到这声沉闷而坚实的巨响传来，所有人紧绷的神经才瞬间释放。对费曼而言，这声音意味着物理计算完全正确，原子弹真的成功起爆了。站在费曼身旁的是《纽约时报》著名科学记者威廉·L·劳伦斯（William L. Laurence），他被巨响吓了一跳，脱口问道：“那是什么？”费曼平静地回答道：“那就是原子弹。”
---
### [01:05:48](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3948s) 钚球与黄金门碰
威廉·L·劳伦斯原本受邀前来撰写报道以记录曼哈顿计划的始末，费曼原本负责带他参观，但后来军方发现技术细节对他而言过于深奥复杂，之后便由亨利·德沃尔夫·史密斯（Henry DeWolf Smyth，著名的《史密斯报告》执笔人）前来考察，改由费曼带史密斯参观实验室。
在带史密斯参观时，他们走进了一间房间。屋里的基座上放着一个小球，表面镀银，拿手去摸会感觉它是温热的——这是一颗具有强放射性的钚球。费曼与史密斯站在门口久久凝视着它：
> “这是一种人类制造的全新元素，除了在宇宙极早期可能极短暂存在过之外，地球上此前从未有过这种东西；如今它被提炼分离出来，带有如此特殊的性质，而且价值连城。”
两人一边交谈，史密斯一边像平常人闲聊时那样漫不经心地用脚踢着门后的一块门碰（doorstop）。费曼见状打趣道，这个门碰与这扇门里存放的东西相当相称。原来，那个半球形的黄色金属门碰居然是纯金打造的。在设计核武器时，物理学家们需要测试各种不同材料对中子的反射效率，以便尽量节省极度珍贵的核裂变材料；实验室测试了铂、锌、黄铜和黄金等多种材质。测试结束后留下了这些金块，有人灵机一动，顺手把其中一个沉甸甸的黄金半球拿来当成了存放钚核心房间的门碰。
---
### [01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) 狂欢中的沉思与战后的虚无感
核试验成功后，整个洛斯阿拉莫斯基地陷入了彻底的狂欢。所有人都在举办庆祝派对，到处奔走呼号；费曼甚至兴奋地坐在吉普车车尾使劲敲打邦戈鼓狂欢。然而，在喧闹的庆祝人群中，有一个人却格格不入地独自坐在一旁生闷气——正是当初劝说费曼加入曼哈顿计划的罗伯特·R·威尔逊（Robert R. Wilson）。费曼走上前问他为什么闷闷不乐，威尔逊神色沉重地说：“我们制造出了一个多么可怕的怪物。”费曼反问他：“可是当初是你发起这一切、是你把我们带进来的啊！”
费曼后来反思道，当时包括自己在内的大多数人都陷入了一种盲目：
> “你看我和其他人身上发生了什么：我们最初是为了一个正当崇高的理由开始这项研究的；但随后你为了完成目标拼命工作，一旦达成它，随之而来的是狂喜与兴奋，你根本不会停下来思考——在最初的思考之后，你就停止了思考；只有威尔逊在那个狂欢的时刻，依然在清醒地反思。”
战争结束后不久，费曼离开洛斯阿拉莫斯回到文明社会，前往康奈尔大学任教。刚回到繁华都市时，他产生了一种极其诡异且强烈的虚无感。坐在纽约的一家餐馆里，他看着窗外的摩天大楼，脑海里会本能地根据广岛原子弹的破坏半径进行计算：如果一颗原子弹扔下来，破坏范围能一直延伸到第34街，周围林立的高楼将瞬间化为齑粉。走在街上，看到工人们在建造跨河大桥、修筑崭新的公路，费曼心里会觉得不可理喻：“这些人真是疯了，他们根本不明白，这一切全都是徒劳的，造这些新东西有什么用？”
不过，费曼随后幽默而欣慰地表示，幸运的是，这种“徒劳感”已经过去了将近30年，人类并没有毁灭彼此，自己关于“修桥无用”的悲观想法被事实证明错了整整30年，他很高兴当年的建设者们没有被这种虚无感阻碍。
---
### [01:09:09](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4149s) 撬开机密：钻研莫斯勒保险柜的开锁秘籍
讲座正文结束后，台下听众提问希望费曼讲讲当年在洛斯阿拉莫斯开保险柜的传奇故事。费曼笑着答应花10分钟讲三个关于保险柜的故事。
费曼提到，自己年轻时曾学会过传统的挑锁技巧，从此对安全锁具产生了浓厚兴趣；他笑称自己人生中有一个“怪病”：只要是带密码或保密的东西，他就非要想方设法解开不可。在洛斯阿拉莫斯，为了保管绝密档案，军方后来给各个办公室统一配发了由莫斯勒保险柜公司（Mosler Safe Company）制造的带有机械密码盘的保密文件柜。这立刻激发了费曼强烈的挑战欲。
```mermaid
flowchart TD
accTitle: Mosler Safe Cracking Logic
accDescr: Feynman's deduction process for cracking Mosler filing cabinet combination locks.
subgraph Analysis["原理分析与容差测算"]
A["传统听音/手感摸索失败<br/>(新型防推压防护盘设计)"] --> B["测算密码盘机械容差<br/>(误差正负2刻度内仍可转动)"]
end
subgraph Reduction["穷举范围缩减"]
B --> C["每5个刻度取1个有效测试点<br/>(100个刻度缩减为20个测试点)"]
C --> D["三位密码理论组合数：<br/>20 × 20 × 20 = 8000 组"]
end
subgraph Optimization["拨盘手法优化"]
D --> E["巧妙设计拨盘顺序<br/>(旋转时不打乱已设定的前位轮片)"]
E --> F["8小时内可地毯式穷举全部组合"]
end
```
早期的老式保险箱可以通过转动转盘时用手指感受内部阻力、倾听棘爪咬合的声音来盲猜密码，但莫斯勒的新型设计采用了防护结构，在拨动转盘时根本没有任何部件压迫内部轮片，老派开锁法全部失效。费曼翻阅了大量锁匠写的书籍，发现那些书开头总爱吹嘘传奇开锁故事（比如保险箱掉入水底、女人即将淹死之类），但书末关于技术的讲解完全是语焉不详的鬼扯，甚至教人根据保险柜主人的心理去猜测密码。
费曼决定完全依靠自己的逻辑和实验来破解。在洛斯阿拉莫斯空闲的近两年时间里，他像研究科学问题一样不断摆弄这些柜子，逐步发现了关键规律：
- **机械容差（Tolerance）**：他发现密码锁存在机械间隙，每个密码数字其实并不需要精准对齐，只要在目标数值的正负2个刻度以内，锁舌机构就能顺利滑入。
- **组合数缩减**：密码盘上的100个数字刻度，每隔5个数测试一次即可覆盖全部范围（即每个轮片只需测试20个特定位置）。三位密码的总组合数从而从理论上的 $1,000,000$ 种大幅缩减至 $20 \times 20 \times 20 = 8,000$ 种。
- **拨盘序列优化**：费曼研究出了一套特殊的转盘拨动算法，在尝试后一位数字的同时，不会打乱前面已经对齐的轮盘。利用这套系统，一个人在8小时内就能完全穷举完所有可能的8,000种组合。
- **抽屉开启状态下的“反推捷径”**：费曼进一步发现，如果文件柜的抽屉处于拉开状态，只要转动密码盘观察锁栓升降的咬合点，配合一点简单的小手法，就能轻而易举地当场探出该保险柜密码的后两位数字。
### [01:11:37](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4297s) 练习手法与维持“神偷”声誉的心理战术
理查德·费曼（Richard Feynman）像扑克老手练习切牌一样，随时随地磨炼自己获取保险柜密码的手法：越来越快，也越来越不引人注意。每当他走进同事的办公室闲聊时，便会看似不经意地斜靠在对方的文件柜旁，双手随性把玩着刻度盘，就像无意识地摆弄手表一样，旁人根本察觉不出任何异样。通过这种方式，他悄悄摸清了三位密码中的后两位数字，随后返回办公室将它们工整记录在案。只要掌握了后两位数字，第一位数字一共只有 20 种可能性，仅需短短一到三分钟就能彻底打开一台保险柜。
凭借这一绝活，费曼在洛斯阿拉莫斯赢得了“开锁大师”的盛名。当某位研究员出差而同事急需柜中机密文件时，众人便会向他求助：
> “他们会对我讲：‘施马尔茨先生（Mr. Schmaltz）出差去了，我们需要他保险柜里的一份文件，你能打开它吗？’”
费曼深知维持“神秘感”对于安全声誉的重要性。如果让人知道开锁如此简单，大家就会意识到保险柜有多不可靠，这种知识在基地扩散开来会极其危险。因此，他总是故弄玄虚地宣称“得回办公室取工具”，其实只是回去翻查自己记下的小本子，并在后裤兜里插上一把普通的螺丝刀以掩人耳目。回到现场后，他会把所有人关在门外，甚至故意在屋里看上 20 分钟杂志消磨时间，有时先迅速开锁确认无误，再硬坐满 20 分钟。随后，他装出一副大汗淋漓、精疲力竭的模样走出房间宣布搞定，让所有人深信这是一项极高难度且耗费心力的技术活。
- 获取后两位密码后，第一位数字在刻度盘上仅剩 20 种可能区间，排查仅耗时 1 到 3 分钟。
- 费曼的办公室记录了基地几乎所有相关人员文件柜的后两位密码。
- 携带螺丝刀与闭门独处 20 分钟纯属心理伪装，目的是防止开锁技巧外泄并强化专业威望。
---
### [01:13:14](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4394s) 战后重返洛斯阿拉莫斯：直面最高机密与绝妙灵感
战争结束后，费曼重返洛斯阿拉莫斯完成未竞的学术报告。在此期间，他完成了一次堪称传奇的“盲开”（cold open）经历。那台保险柜内封存着全人类历史上最危险的机密——完整的原子弹绝密资料，包括铀裂变释放中子的速率、制造原子弹所需的铀临界质量临界值、全套核心理论与极其繁复的数值计算。
当时正值周六，费曼原本以为基地会像战争时期那样全员加班，于是径直赶往洛斯阿拉莫斯图书馆借阅报告。然而，图书馆主金库的大门紧锁，门上装有他从未研究过的巨大旋钮，门前甚至还有荷枪实弹的卫兵来回巡逻，强行破解绝无可能。费曼随即灵机一动，想到了负责解密事务的弗雷迪·德霍夫曼（Frederic de Hoffmann）。由于德霍夫曼经常需要核实哪些绝密文件可以降级解密，频繁往返图书馆令其疲惫不堪，于是他采取了一个极其大胆的举措：将图书馆内的所有文件完整复印了一份，整整齐齐地塞满了自己办公室两个房间内的九个大型文件柜。
费曼来到德霍夫曼的办公室借阅，发现房门虚掩、室内亮着灯，仿佛主人随时都会返回。无聊等待之中，费曼随手拨动刻度盘测试了诸如 `20-40-60` 之类的常规组合，毫无成效。枯燥之中，他决定放弃纯机械手段，转而践行开锁指南里记载的“开锁心理学”。
```mermaid
flowchart TD
accTitle: Feynman Accessing Atom Bomb Secrets Flowchart
accDescr: 费曼获取原子弹核心机密文件的决策与推导路径图
subgraph Vault["常规渠道受阻"]
A["周六前往洛斯阿拉莫斯图书馆"] --> B["巨型金库未接触过且门前有持枪卫兵巡逻"]
B --> C["无法进入，寻找替代途径"]
end
subgraph Office["转寻解密办公室"]
C --> D["想到弗雷迪·德霍夫曼办公室藏有全套绝密复印件"]
D --> E["九个大型文件柜封存原子弹全部核心机密"]
E --> F["办公室亮灯但无人，盲转基础数字失败"]
end
subgraph Psychology["心理学破译法"]
F --> G["应用开锁心理学：搜查秘书可能遗留的笔记"]
G --> H["撬开办公桌抽屉，发现印有希腊字母表的纸张"]
H --> I["纸顶潦草手写圆周率 π = 3.14159"]
end
Vault --> Office
Office --> Psychology
```
---
### [01:15:33](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4533s) 心理学破译：从圆周率到自然常数的顿悟
根据安全手册与开锁指南所述，秘书和行政人员往往极度担忧自己或长官遗忘复杂的保险柜密码，通常会秘密将其写在某个隐蔽角落。指南列举的典型藏匿处之一，便是抽屉外侧木质边缘上伪装成发票单号的一串数字。费曼迅速撬开了锁着的办公桌抽屉进行检查，边缘并没有字迹，但他在翻动杂乱的文件时，意外发现了一张供秘书打字校对用的希腊字母表（涵盖 $\alpha, \beta, \gamma, \delta$ 等符号，以便秘书准确辨识物理学家的专业手稿）。
引人注目的是，这张表格的顶端极其潦草地手写着一串算式：$\pi = 3.14159$。秘书日常工作中无需进行任何数学计算，记录圆周率数值显得格外突兀。费曼立刻断定这就是密码线索，兴奋地扑向文件柜尝试：
> “我走上前去输入 `31-41-59`——打不开；`13-41-95`——打不开；`95-41-13`——打不开；`41-31`……我花了整整二十分钟把圆周率翻来覆去倒腾，全无所获。”
挫败之余，费曼正准备离开办公室，但他停下脚步再次审视这一心理推断。他意识到逻辑本身并没有错：像德霍夫曼这样痴迷科学的年轻学者，极有可能会选择一个具有代表性的“数学常数”作为保密代码。既然圆周率行不通，科学界另一个至关重要的常数便是自然常数 $e$（约等于 $2.71828$）。
费曼重新折回文件柜前，将转盘依次对准 `27-18-28`。伴随着清脆的一声“咔哒”（Click），沉重的柜门应声开启。经过检查，德霍夫曼名下两个房间内的全部九台文件柜使用的是完全相同的密码组合，整座洛斯阿拉莫斯基地的最高机密至此全部敞开。费曼在笑声中总结道，关于开锁还有许多离奇趣事，但鉴于时间已晚，这段绝妙的经历正是最完美的收尾。
- 物理学家德霍夫曼选取的密码并非随机数字，而是基于数学常数 $e$ 前六位：`27-18-28`。
- 办公室内的九台绝密文件柜统一共用此密码，导致基地整套原子弹理论与计算资料完全向该密码开放。
- 费曼最终依靠“物理学家的思维心理学”成功破解了战后最关键的绝密档案柜。
## Notable quotes
> [01:07:15](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=4035s) 我们最初是为了一个正当崇高的理由开始这项研究的；但随后你为了完成目标拼命工作，一旦达成它，随之而来的是狂喜与兴奋，你根本不会停下来思考——在最初的思考之后，你就停止了思考 — 理查德·费曼
> [01:00:11](https://www.youtube.com/watch?v=uY-u1qyRM5w&t=3611s) 若想法糟糕，我就说它糟糕；若想法精妙，我就说它精妙。 — 理查德·费曼
## Mentioned
理查德·费曼, 罗伯特·奥本海默, 雅科夫·弗伦克尔, 鲍勃·威尔逊, 理查德·托尔曼, 亨利·德沃尔夫·史迈斯, 哈罗德·尤里, 伊西多·拉比, 阿瑟·康普顿, 保罗·奥卢姆, 莱斯利·格罗夫斯, 约翰·威廉姆斯, 罗伯特·瑟伯尔, 维克托·魏斯科普夫, 爱德华·泰勒, 汉斯·贝特, 罗伯特·克里斯蒂, 阿琳, 戴维·鲍姆, 埃米利奥·塞格雷, 朱利安·韦伯, 斯坦利·弗兰克尔, 恩里科·费米, 约翰·冯·诺伊曼, 尼尔斯·玻尔, 奥格·玻尔, 威廉·L·劳伦斯, 克劳斯·富赫斯, 弗雷迪·德霍夫曼, 普林斯顿大学, 洛斯阿拉莫斯国家实验室, 芝加哥大学冶金实验室, 橡树岭工厂, 斯通-韦伯斯特公司, IBM, Marchant, Mosler Safe Company, 三位一体核试验
## Video description
There are quite a few copies of this Feynman lecture floating around out there, but most end prior to the question from the audience.
After the lecture, a guy in the audience asks Feynman about his safe-cracking stories and Feynman goes on for about another ten minutes relating three different stories on his safe-cracking while at Los Alamos National Laboratories. Enjoy!
[Source](https://www.youtube.com/watch?v=uY-u1qyRM5w)
[60:12] Richard Feynman: I was always dumb about one thing: I never knew who I was talking to; I would only worry about the physics: if the idea looked lousy, I said it looked lousy; if it looked good, I said it looked good.
[58:10] Richard Feynman: The one thing that von Neumann gave me was an idea that he had, which is interesting: that you don't have to be responsible for the world that you're in.
[51:35] Richard Feynman: The disease with computers is you play with them.
Original transcript
[0:13] This joint may get raided by the police
[0:15] department, but anyhow, we're going to
[0:17] get started.
[0:19] Um, our speaker today, as you know, is
[0:23] Richard Phillips Fineman, Nobel Prize
[0:26] winner in physics and scientist
[0:30] extraordinary. The difference between
[0:33] Dick Fineman and other people is that he
[0:36] is more logical and has a greater
[0:40] imagination. Uh, this is a rare
[0:42] combination.
[0:44] Most people who are logical don't have
[0:47] much
[0:47] [Music]
[0:48] imagination. As the great Russian
[0:51] physicist Franco said, knowledge
[0:53] progresses by illogical steps.
[0:57] Fineman uses his imagination to probe
[1:01] the scientific frontiers beyond the
[1:04] bounds of logic in much the same manner
[1:07] as mathematicians use complex variables
[1:10] in order to make analytic
[1:14] continuations. Uh, at Los
[1:21] Almas at Los
[1:23] Almas, Robert Oppenheimer rated Dick as
[1:26] his most valuable
[1:29] player. Uh, besides being an extremely
[1:32] talented theoretical physicist, he is a
[1:35] super duper troubleshooter.
[1:38] he can go into an experimental lab and
[1:41] by asking the right questions find the
[1:44] reasons why the experiment doesn't work.
[1:47] Uh it's a great pleasure to have Dick
[1:50] here and I
[1:59] and thanks very much. How do you do?
[2:02] Hello down
[2:04] there and how to everybody else. The
[2:08] introduction is quite inappropriate for
[2:10] my talk which is Los Alamos from below.
[2:13] And I mean from below because although
[2:16] as Mr. Hersshfeld says in my field at
[2:19] the present time I'm a slightly famous
[2:21] man at the time I was not anybody famous
[2:24] at all. I did not even have my degree
[2:27] when I started to work on uh stuff
[2:31] associated with the Manhattan project.
[2:35] So I want to describe it from many of
[2:38] the other people who tell you about Los
[2:40] Alamos uh know something. They're up in
[2:42] some higher echelon of government of the
[2:44] organization or something or worried
[2:46] about some big decisions. I worried
[2:47] about no big decisions. I was always
[2:51] flittering about underneath somewhere. I
[2:53] wasn't the absolute bottom as it'll turn
[2:55] out. I did get sort of up a few steps
[2:58] but I wasn't one of the higher people.
[3:00] So I want you to try to put yourself in
[3:02] a different kind of condition than the
[3:04] introduction said and just imagine this
[3:06] young graduate student that hasn't got
[3:08] his degree yet is working on his thesis.
[3:10] And I'll start by saying how I got into
[3:12] the project and then what happened to
[3:14] me. That's all just what happened to me
[3:16] during the project. I was working at
[3:18] Princeton in my room one day when Bob
[3:20] Wilson came
[3:22] in. I was working. What the hell? It
[3:24] isn't funny yet. I got lots of
[3:29] [Applause]
[3:33] better. Bob Wilson came in and said that
[3:37] he had been funded to do a job that was
[3:40] a secret and was uh not supposed to tell
[3:42] anybody, but he was going to tell me
[3:44] because he knew that as soon as I knew
[3:45] what he was going to do, I'd see that I
[3:47] had to go along with it. So, he told me
[3:49] about the problem of separating
[3:50] different isotopes of uranium. He had a
[3:53] to make a ultimately and make a bomb. He
[3:55] had a process for separating the
[3:57] isotopes of uranium, which is different
[3:59] than the one that was ultimately used
[4:00] that he'd wanted to try to
[4:02] develop. and he told me about it and he
[4:05] said, "There's a meeting." I said, "I
[4:06] didn't want to do it." He says, "All
[4:08] right, there's a meeting at 3:00. I'll
[4:10] see you there." He said, I said, "It's
[4:11] all right that you told me the secret
[4:13] because I'm not going to tell anybody,
[4:14] but I'm not going to do it." So, I went
[4:16] back to work on my thesis for about 3
[4:19] minutes and then I began to pace the
[4:21] floor and to think about this thing. The
[4:23] Germans had Hitler and the possibility
[4:25] of developing an atomic bomb was obvious
[4:28] and the possibility that they would
[4:29] develop before we did was very much of a
[4:32] fright. So I decided to go to the
[4:34] meeting at
[4:35] 3:00. By 4:00 I already had a desk in a
[4:39] room and was trying to calculate whether
[4:42] this particular method was limited by
[4:43] the total amount of current that you can
[4:44] get in an ion beam and so on. I won't go
[4:46] into the details but I had a desk and I
[4:48] had paper. I'm working as hard as I
[4:50] could and as fast as I can. Whereas the
[4:52] fellas were building the apparatus to do
[4:54] the experiments right there. The other
[4:56] fellas who were joined and it was like
[4:58] those moving pictures that you see where
[4:59] you see a piece of equipment go
[5:03] Every time I'd look up, the thing was
[5:04] getting bigger. And what was happening,
[5:07] of course, is that all the boys had
[5:08] decided to work on this and to stop
[5:10] their research in science. All the
[5:11] science stopped during the war, except
[5:13] the little bit that was done in Los
[5:15] Alamos. It was not much science. It was
[5:17] a lot of engineering. And uh they were
[5:20] robbing their equipment from their
[5:22] research and all the equipment of
[5:23] different research was being put
[5:24] together to make the new apparatus to do
[5:26] the experiments to try to separate the
[5:28] isotopes of uranium. and I stopped my
[5:31] work also for the same reason. Uh it is
[5:35] true that I did take a six- week
[5:36] vacation off after a while from that job
[5:40] and finished writing my thesis. So I did
[5:42] get my degree just before I got to Los
[5:44] Alamos. So I wasn't quite as far down as
[5:47] I led you to believe. One of the first
[5:50] experiences that was very interesting to
[5:51] me in this project at Princeton was to
[5:55] meet great men. I had never met very
[5:58] many great men before, but there was a
[6:00] committee, evaluation committee that had
[6:03] to decide which way we're going and to
[6:04] try to help us along to ultimately
[6:06] decide which way we're going to separate
[6:08] the uranium. This evaluation committee
[6:10] had men like Tolman and Smi and Yuri and
[6:14] Robbie and uh Oppenheimimer and so forth
[6:18] on it. And there was Compton for
[6:20] example. One of the things that was a
[6:23] terrible shock to I'd sit there because
[6:24] I had understood the theory of this
[6:26] process that we're doing. And so they
[6:27] asked me questions and I'd be there and
[6:29] then we'd discuss it and then they'd
[6:30] start to discuss something and one man
[6:33] would make a point and then Compton
[6:34] would explain a different point of view
[6:36] and he would be perfectly right and it
[6:38] was the right idea and he said it should
[6:39] be this way. Another guy would say well
[6:42] maybe there's this possibility we have
[6:43] to consider against it. Another
[6:45] possibility we have to consider. I'm
[6:46] jumping. They should repeat Compton. He
[6:49] should say it again. he should say it
[6:50] again because everybody's disagreeing
[6:52] with him all the way around the table.
[6:53] So finally at the end the tomusa
[6:55] chairman says well having heard all
[6:57] these arguments I guess it's true that
[7:06] compens thinking of a new facet and
[7:09] remembering what the other fellow said
[7:11] having paid attention and so at the end
[7:14] the decision is made as to which idea
[7:16] was the best of summary it all together
[7:17] without having to say it three times.
[7:19] You see, so that was a shock and these
[7:20] were very great men
[7:23] indeed. Uh this project was ultimately
[7:27] decided not to be the way that they were
[7:29] going to separate uranium. And we were
[7:30] told then that we were going to stop and
[7:33] that there would be starting in Los
[7:34] Alamos in New Mexico, the project which
[7:36] would actually make the bomb and that we
[7:38] would all go out there to make it. There
[7:41] would be experiments that we would have
[7:42] to do, theoretical work that we'd have
[7:44] to do. I was in the theoretical work.
[7:45] All the rest of the fellas were in
[7:47] experimental work. The question was what
[7:51] to do because we had this heus of time
[7:53] where we would just been told to turn
[7:55] off but with the Los Alamos wasn't ready
[7:56] yet. So Bob Wilson tried to make use of
[7:59] his time by saying to send me to Chicago
[8:03] to find out all that we could find out
[8:04] about the bomb and the problems so that
[8:07] we could start to build in our
[8:08] laboratories equipment counters of
[8:10] various kinds and so on that would be
[8:11] useful when we got to Los Alamos. So no
[8:13] time was wasted. I was sent to Chicago
[8:16] with the instructions to go to each
[8:18] group, tell them I was going to work
[8:19] with them, have them tell me about a
[8:22] problem to the extent that I knew enough
[8:23] details that I could actually sit down
[8:25] and start working on the problem and as
[8:27] soon as I got that far, go to another
[8:28] guy and ask for a problem. And that way
[8:30] I would understand the details of
[8:32] everything. It was a very good idea. My
[8:34] conscience bothered me a little bit, but
[8:35] it turned out accidentally. I was very
[8:37] lucky in one of the guys explaining the
[8:39] problem. I says, "Why don't you do it
[8:40] that way?" And in a half an hour, he had
[8:41] it solved and they've been working on it
[8:42] for three months. So I did
[8:46] something. So I came back from Chicago
[8:49] and I described this situation, how much
[8:51] energy was released, what the bomb was
[8:53] going to be like and so forth to the
[8:54] fellas. I remember a friend of mine who
[8:57] was a worked with me, Paul Olm,
[8:59] mathematician came up to me after he
[9:01] says when they make a moving picture
[9:02] about this, they'll have the guy coming
[9:03] back from Chicago telling the Princeton
[9:05] men all about the bomb and he'll be
[9:07] dressed in a suit and carry a a
[9:08] briefcase inside and you're in dirty
[9:10] shirt sleeves and just telling us all
[9:12] about it. but it's very serious thing
[9:13] anyway and so he appreciated the
[9:15] difference between the real world and
[9:17] that in the
[9:19] movies. Well, there still seemed to be a
[9:22] delay and Wilson went to Los Alamos to
[9:25] find out what was holding things up and
[9:27] how they were progressing. And when he
[9:28] got there, he found that the
[9:29] construction company was working very
[9:31] hard and had finished the
[9:33] theater and a few buildings because they
[9:35] understood how but they hadn't gotten
[9:37] instructions clear on how to build the
[9:39] laboratories, how many pipes for gas,
[9:41] how much for water. So he simply stood
[9:42] around and decided how much water, how
[9:44] much gas and so on and told him to start
[9:46] building the laboratories and he came
[9:48] back to us. Uh we were all ready to go
[9:50] you see and Oenheimer was having some
[9:53] difficulties in discussing some problems
[9:54] with Groves and we were getting
[9:56] impatient. So what as far as I
[9:59] understand it from the position I was in
[10:01] Wilson called Manley in Chicago and they
[10:03] all got together and they decided we'd
[10:05] go out there anyway when it wasn't
[10:06] ready. So, we all went out to Los Alamos
[10:09] before it was
[10:11] ready. We were recruited, by the way, by
[10:13] uh Oenheimr and other people. And he was
[10:16] very patient with everybody. He paid
[10:18] attention to everybody's problems. He
[10:20] worried about my wife who had TB and
[10:22] whether there would be a hospital out
[10:23] there and everything. And it was the
[10:24] first time I met him in such a personal
[10:26] way. And he was such a wonderful man. We
[10:29] were told, among other things, to be
[10:31] careful not to buy our train ticket at
[10:33] Princeton because Princeton was a very
[10:35] small train station. And if everybody
[10:37] bought a train ticket to Albuquerque,
[10:39] New Mexico in print, there'd be some
[10:41] suspicion that something was
[10:44] up. And so everybody bought that ticket
[10:46] at somewhere else. Sick me because I
[10:48] figured if everybody bought that ticket
[10:50] somewhere else, it
[10:53] was so when I went to the trade station
[10:56] and I said, "I want to go to
[10:57] Albuquerque, New Mexico." He says, "Oh."
[10:59] He says, "So all the stuff is for you.
[11:05] We've been shipping out crates full of
[11:06] counters for weeks and expecting they
[11:09] didn't notice that the address was
[11:11] Albuquerque. So at least I explained why
[11:14] it was that we were shipping out crates.
[11:15] I was going out to
[11:18] Albuquerque. Well, when we arrived, we
[11:21] were ahead of time and the houses for
[11:23] the dormitories and things like that
[11:25] were not ready. In fact, the
[11:26] laboratories weren't quite ready. We
[11:28] were pushing them. We were driving them
[11:29] by coming down ahead of time. They went
[11:31] crazy at the other end and they rented
[11:33] ranch houses all around in the
[11:35] neighborhood and we stayed at first at a
[11:36] ranch house and would drive in in the
[11:38] morning. The first morning I drove in
[11:40] was tremendously impressive. The beauty
[11:42] of the scenery for a person from the
[11:44] east who didn't travel much was
[11:46] sensational. There's a great cliffs.
[11:47] You've probably seen the pictures. I
[11:49] won't go into it much detail. This thing
[11:51] was high on a me and you'd come up from
[11:52] below and see these great cliffs. And uh
[11:56] we were very surprised. And the most
[11:57] impressive thing to me was that as I was
[11:59] going up, I said, "Maybe there were
[12:00] Indians even living here." And the guy
[12:02] who was driving the car said just stop.
[12:04] He stops the car, walks around the
[12:05] corner and there are Indian caves that
[12:07] you can respect. So it was really very
[12:09] exciting in that
[12:13] respect. When I got up to the site the
[12:15] first time, I saw at the gate of the was
[12:19] a technical area where supposed to do
[12:21] the work with a fence. Supposed to have
[12:22] a fence around it ultimately, but
[12:24] because they were still building it was
[12:25] open. Then there was supposed to be a
[12:26] town and then a big fence further out
[12:29] around the
[12:31] town. Now they were still building and
[12:33] my friend Paul who was my assistant was
[12:35] standing with a clipboard checking the
[12:37] trucks coming in and out and telling him
[12:38] which way to go to deliver the materials
[12:40] in different places. When I went into
[12:42] the laboratory I would meet men that I'd
[12:44] heard of by seeing their papers in the
[12:45] physical review and so on. I never met
[12:47] him before. This is John Williams. They
[12:49] said the guy comes standing up from a
[12:51] desk which is covered with blueprints.
[12:53] his sleeves all rolled up and he's
[12:54] standing by some windows at one of the
[12:56] buildings ordering trucks and things
[12:58] going in different directions to build
[13:00] the things. In other words, we took over
[13:01] the construction company and finished
[13:03] the finished the job. The physicists, in
[13:05] other words, at the beginning, the
[13:07] experimental physicists particularly had
[13:09] nothing to do until their buildings were
[13:10] ready and apparassed ready. So, they
[13:12] just built the buildings or assisted in
[13:14] building the buildings. The theoretical
[13:16] physicists on the other was decided that
[13:18] they wouldn't live in the ranch houses,
[13:20] but they would live up at the site
[13:22] because they could start working right
[13:23] away. So, we started working
[13:26] immediately. And that meant we would
[13:28] each get a little we in some rooms.
[13:30] There were no blackboard except for one,
[13:32] a rolled blackboard, you know, that on
[13:34] wheels that you'd roll around and we'd
[13:35] roll it around and Serber would explain
[13:37] to us all the things that they thought
[13:38] of in Berkeley about the atomic bomb and
[13:41] nuclear physics and all these things.
[13:42] And I didn't know very much about it.
[13:44] I'd been doing other kinds of things and
[13:47] so uh I had to do an awful lot of work.
[13:50] Every day I'd study and read and study
[13:51] and read and it was a very hectic time.
[13:54] I had some luck. Uh all the big shots by
[13:59] some kind of accident. Everybody but
[14:00] Hans Beta happened to have left at the
[14:03] same time. Like Vice had to go back to
[14:05] fix something at MIT and Tella was away
[14:08] just at a certain moment. And what Beta
[14:10] need is somebody to talk to to push his
[14:12] ideas against. So he came into this
[14:14] little squirt in an office and he starts
[14:16] to explain this idea and I says no no
[14:18] you're crazy it'll go like this and he
[14:19] says just a moment and he explains that
[14:21] he's not crazy that I'm crazy and we
[14:24] keep on going like this but I turned out
[14:25] that although you see when I hear about
[14:27] physics I just think about physics and I
[14:28] don't know who I'm talking to so I say
[14:30] the dopiest things like no no that's
[14:31] wrong or you're crazy but it turned out
[14:33] that's exactly what he needed and he was
[14:35] so I got a notch up on account of that
[14:38] and so when
[14:39] I I ended up as a group leader with four
[14:42] guys under me in which is underneath
[14:44] beta. There was several groups but I
[14:46] wasn't quite as far below as the title
[14:48] might indicate but I wasn't anywhere up
[14:50] anywhere. I had a lot of interesting
[14:52] experiences with beta. The first day
[14:54] when he came in we had an adding machine
[14:56] at Marshant that you work with your
[14:57] hands. And so he says uh let's see he
[14:59] said the formula they've been working
[15:00] out he says involves a pressure squared.
[15:02] The pressure is 48. The square of 48. I
[15:04] reach for the machine. He says it's
[15:06] about 2300. So I plug it out just to
[15:08] find out. He says, 'You want to know
[15:10] exactly? It's
[15:12] 2304. And so it came out
[15:15] 2304. Said, how do you do that? He says,
[15:17] don't you know how to take squares and
[15:19] numbers near 50? If it's near 50, if
[15:22] it's say three below, then it's three
[15:23] below 25. Like 47 squared is 22. And how
[15:27] much is left over is the square of
[15:29] what's residual? For instance, three
[15:30] lost. So you get nine. 2209 for 47
[15:32] squared. Very nice.
[15:34] Okay. So we keep on he was very good in
[15:37] arithmetic. So we kept going and a few
[15:40] moments later we had to take the cube
[15:42] root of 2 and 1/2.
[15:45] Now to do cube roots there was a little
[15:47] chart that you take. You have some trial
[15:49] numbers that you try on the adding
[15:50] machine the Marshian company had given.
[15:52] So I open this takes him a little longer
[15:53] you see. I open the drawer I take out
[15:56] the chart. He says
[15:59] 1.35. So I figure there's some way to
[16:01] take cube roots numbers near two and a
[16:03] half but it turns out
[16:05] no. How do you do that? He says, well,
[16:08] he says, you see, the logarithm of two
[16:10] of a half is so and so. You divide by
[16:11] three to get the cube root of so and so.
[16:12] Now, the log of 1.3 is this, the log of
[16:14] 1.4 is I interpolate in
[16:17] between. I couldn't have divided
[16:19] anything by three, much
[16:21] less. So, he knew all his arithmetic and
[16:24] he was very good at it. And that was a
[16:25] challenge to me. I kept practicing. We
[16:26] used to have little contests. Every time
[16:29] we'd have to calculate anything, we we
[16:30] rush to the answer, he and I, and I
[16:32] would win. After several years, I began
[16:35] to be able to do it, you know, get in
[16:36] there once, maybe one out of four,
[16:38] because you'd notice something funny
[16:40] about a number like if you have to
[16:41] multiply 174 by 140, for example. You
[16:44] notice that's like 173 by 141, which is
[16:46] like the square of 3 * the square
[16:48] of two, which is the square of 6,
[16:49] which is
[16:55] 245. But you have to notice the numbers.
[16:58] You see, each guy would notice a
[16:59] different way around. We had lots of
[17:00] fun. Well, when I was first there, as I
[17:04] said, we didn't have the dormitories,
[17:06] and the theoretical physicist had to
[17:07] stay up there, and the first place they
[17:09] put us was in the old school buildings
[17:10] from the boys school that had been there
[17:13] previously. The uh first place I lived
[17:16] in was a thing called the mechanics
[17:17] lodge. We were all jammed in there in
[17:19] bunk beds and so on. And it turned out
[17:21] it wasn't organized very well. And Bob
[17:24] Christie and his wife had to go to the
[17:25] bathroom in the morning through our
[17:26] bedroom. So, you know, so that was very
[17:29] uncomfortable. The next place we moved
[17:31] to was a thing called the big house
[17:32] which had a big patio all the way around
[17:34] the outside on the second floor where
[17:36] all beds had been stuck next to each
[17:37] other all along along. And then
[17:39] downstairs there was a big chart that
[17:41] told you what your bed number was and
[17:42] which bathroom you change the clothes
[17:44] in. So under my name it said bathroom
[17:47] C. No bed
[17:51] number. As a result of this I was rather
[17:53] annoyed by this time. I was getting
[17:55] annoyed. At last the dormatory is built.
[17:58] I go down to the dormatory place to get
[18:00] the rooms
[18:01] assigned and they say uh these are you
[18:04] pick your room now. I tried to pick one.
[18:06] You know what I did? I looked to see
[18:08] where the girl's dormatory was and I I
[18:10] picked one that you could look out
[18:11] across. Later I discovered a big tree
[18:13] was growing right in front
[18:15] of but anyway I picked this room and uh
[18:19] they told me that temporarily there
[18:22] would be two people in a room but that
[18:23] would only be temporary. That the rooms
[18:25] were two rooms which would share a
[18:26] bathroom. There would be double deck of
[18:28] bunk in there and I didn't want two
[18:30] people in the room. So when I first got
[18:31] there the first night, nobody else was
[18:33] there. Now my wife was sick with TB in
[18:36] Santa Fe in Albuquerque. So I had some
[18:39] boxes of stuff of hers. So I opened the
[18:42] box and I took out a little night gown.
[18:46] I opened the top bed. I threw the night
[18:47] gown carelessly on the top bed. I took
[18:49] out the slippers. I threw some powder on
[18:51] the floor in the bathroom. I just made
[18:53] it look like somebody else is there.
[18:55] Okay. so that the other bed is occupied.
[18:56] Nobody's going to sleep there. Okay. So,
[18:59] what happened? Because it's the men's
[19:02] dormatory. Well, I came home that night
[19:04] and my uh pajamas are folded nicely and
[19:07] put under the pillow at the bottom and
[19:09] the slip is put nicely at the bottom of
[19:10] the bed and the lady's pajama is nicely
[19:13] folded. The bed has been all fixed up
[19:14] and made and it's put under the pillow
[19:16] and the slippers are put down nicely and
[19:18] the powder is clean from the bathroom
[19:20] and nobody is sleeping up there. I still
[19:22] have the room to myself. So next night,
[19:25] same thing. When I wake up, I smuggle up
[19:27] the bed at the
[19:28] top sloppily. I throw the night gown and
[19:31] powder in the bathroom and so on. And I
[19:33] went on like this for four nights until
[19:36] it settled down. Everybody was settled
[19:37] and there was no more danger that they
[19:39] would put a second person in the room.
[19:41] And each night perpetually each morning
[19:43] everything was set very neatly.
[19:45] Everything was all right. Even though it
[19:46] was a men's dormatory. So that's what
[19:48] happened there. In that
[19:51] situation, I got involved in politics a
[19:54] little bit because there was a thing
[19:56] called the town council and apparently
[19:58] there were certain things the army
[20:00] people would decide how the town was
[20:02] supposed to be run with some assistance
[20:04] from some governing board up there that
[20:06] I never knew anything about from the
[20:07] civilians or something. But there were
[20:10] all kinds of excitement like there isn't
[20:12] any political thing. And so there were
[20:13] factions, the housewife faction, the
[20:16] mechanics faction, you know, the
[20:18] technical people faction. on. Well, the
[20:20] bachelors and bachelor girls, the people
[20:22] who lived in the dormitories, felt they
[20:24] had to have a faction because a new rule
[20:27] had been promulgated. No women in the
[20:29] men's dorms. Well, this is absolutely
[20:32] ridiculous. All grown people, of course.
[20:33] Haha. And
[20:35] so, what kind of nonsense and so forth.
[20:38] So, we had to have political action and
[20:39] we discarded and we debated and all this
[20:42] stuff. You know how it is. And so, I was
[20:44] elected to represent the dormatory
[20:46] people, you see, in a town council. So,
[20:48] I was in the town council for a while.
[20:51] After I was in the town council about a
[20:53] year or so, a year and a half, I was
[20:55] talking once to Hunts Beta about
[20:56] something and uh he was up in the
[20:58] governing council during all this time.
[21:00] And I told him this story I've forgotten
[21:02] about that I did this trick one time
[21:04] with my wife's stuff on the upper bed
[21:07] and he started to laugh. He says, "Ah,
[21:08] that's how you got on the town council."
[21:11] Because it turned out that what happened
[21:14] was this. There was a report, a very
[21:16] serious report. The poor woman was
[21:18] shaking. The woman who cleans the rooms
[21:20] in the dormatory had it. It just opened
[21:21] and all of a sudden there's trouble.
[21:23] Somebody's sleeping with one of the
[21:24] guys. She shaking. She doesn't know what
[21:26] to do. So charwoman reports to the chief
[21:28] charwoman. The chief charwoman reports
[21:30] to the lieutenant. The lieutenant
[21:31] reports to the major. It goes all the
[21:32] way up. It goes all the way up through
[21:34] to the general to the governing board.
[21:36] What are they going to do? What are they
[21:37] going to do? You're still going to think
[21:39] about it. So in the meantime, what in
[21:42] the meantime, instructions go down bound
[21:44] through the captains, down through the
[21:45] major, through the lieutenant, through
[21:47] the chief, through the chw woman to just
[21:48] put things back the way they are, clean
[21:50] them
[21:51] up and see what
[21:54] happens. Okay. Next day report, same
[21:59] thing. Meanwhile, for 4 days, they
[22:02] worried up there what they're going to
[22:04] do. So they finally promulgated a rule.
[22:07] No women in the men's dormitories.
[22:10] And that caused such a stink down
[22:12] there. You can see that they had all the
[22:15] politics and they elected somebody to
[22:16] represent
[22:21] the
[22:24] So now I'd like to tell you about the
[22:26] censorship that we had. They decided to
[22:29] do something utterly illegal, which is
[22:31] to censor mail of people inside the
[22:33] United States in the continental United
[22:35] States, which they have no right to do.
[22:37] So, it had to be set up very delicately
[22:39] as a voluntary thing. We would all
[22:42] volunteer not to seal our envelopes that
[22:45] we would send the letters out with. We
[22:48] would accept it be all right if they
[22:50] would open letters coming into us that
[22:52] was voluntarily accepted by us and we
[22:55] would leave the letters open. They would
[22:56] seal them if they were okay. they
[22:58] weren't okay in their opinion. In other
[22:59] words, they said something that we
[23:00] shouldn't send out, they would send the
[23:02] letter back to us with a note that said
[23:04] it was a violation of such and such a
[23:06] paragraph of our understanding and so
[23:08] on. So, very delicately amongst all
[23:10] these liberal-minded scientific guys
[23:13] agreeing to such a proposition, we
[23:15] finally got the censorship set up with
[23:17] many rules about that we were allowed to
[23:20] comment on the character of the
[23:22] administration if we wanted to. So, we
[23:23] could write our senator and tell him we
[23:24] don't like the way things are run and
[23:26] things like that. So it was all set up
[23:28] and uh they said they would notify us of
[23:31] any difficulty. So the day starts the
[23:33] first day for censorship. Telephone
[23:35] bring me what? Please come down. I come
[23:38] down. What's this? It's a letter from my
[23:40] father. But what is it? There's lined
[23:44] paper and there's these lines going on
[23:46] with dots. Four dots under one dot
[23:47] above. Two dots under one dot above. Two
[23:49] dot dot. What's that? I said it's a
[23:52] code.
[23:57] They said, "Yeah, what's a Cobra? What
[23:58] does it say?" I said, "I don't know what
[24:00] it
[24:01] says." They said, "Well, what's the key
[24:03] to the code? How do you decipher it?" I
[24:04] said, "I don't
[24:06] know." Then they said, "And what's
[24:08] this?" It was a letter from my wife. It
[24:09] says TJXY WZ
[24:13] TWW1X3. What's this? I said, "Another
[24:16] code." What's the key to it? I don't
[24:18] know. They said, "You mean you're
[24:20] receiving codes and you don't know the
[24:21] key?" I said, "Precisely." I said, "I
[24:23] have a game." I challenge them to send
[24:26] me a code that I can't decipher. See?
[24:28] So, they're making up codes at the other
[24:30] end, and they're not going to tell me
[24:31] what the key is, and they're sending
[24:32] them in. Now, one of the rules of the
[24:34] censorship was that they aren't going to
[24:36] disturb anything that you would
[24:37] ordinarily
[24:40] do in the mail. So, they said, "Well,
[24:43] you're going to have to ask them,
[24:45] please, to send the key in with the
[24:47] code." I said, "I don't want to see the
[24:49] key." And he says, "All right, we'll
[24:50] take the key
[24:52] out." So, we had that arrangement. Okay.
[24:55] All
[24:56] right. Next day, I get a letter from my
[24:58] wife which says, "Uh, it's very
[25:00] difficult writing because I feel that
[25:02] the is looking over my shoulder and in
[25:04] the spot there's nicely eradicated a
[25:06] splotch with an ink eradicator." So, I
[25:09] went down to the bureau. I says, "You're
[25:11] not supposed to touch the incoming mail.
[25:12] If you don't like it, you can tell me,
[25:13] but you're not supposed to do anything
[25:15] to it that you should look at it, but
[25:16] you're not supposed to take anything
[25:17] out." They said, "Don't be ridiculous.
[25:19] That's not the way sensors worked with
[25:20] incoradicator. They cut things out with
[25:22] a
[25:23] scissor." Okay. So, I wrote a letter
[25:24] back to my wife. Did you use
[25:26] Incoradicator in your letter? She writes
[25:28] back, "No, I didn't use Incorator in my
[25:30] letter. It must have been the And
[25:31] there's a hole cut out with a
[25:34] scissor." So, I went
[25:40] back I went back to the guy in charge,
[25:43] the the major who was supposed to be in
[25:44] charge of all this and complain. This
[25:46] went on for several days. I felt I was
[25:48] sort of the representative to get the
[25:49] things straightened out. He tried to
[25:50] explain to me that these people who were
[25:52] the sensors had been taught how to do it
[25:55] and they didn't understand this new way
[25:56] that we had to be so delicate about. So
[25:58] I was trying to be the front to try to
[26:00] the one with the most experience. I was
[26:02] writing to my wife back and forth every
[26:04] day anyway. So he said, "What's the
[26:06] matter? Don't you think I have
[26:07] goodwill?" I says, "Yes, you have
[26:08] perfectly goodwill, but I don't think
[26:09] you have power." Cuz he had done three
[26:11] or four days. He said, "Oh, we'll see
[26:13] about that." He grabs the telephone.
[26:14] Everything was straightened out. So no
[26:16] more was the letter cut. However, there
[26:18] was a number of difficulties that arose.
[26:20] For example, one day I got a letter from
[26:22] my wife and a note from the census said
[26:24] there was a code enclosed without the
[26:26] key and so we removed it. So when I went
[26:28] to see my wife in Albuquerque that day,
[26:30] she says, "Well, where's all the stuff?"
[26:31] I said, "What stuff?" She says, "Lith,
[26:33] glycerin, hot dogs, laundry." I says,
[26:35] "Wait a minute. That was a list." She
[26:37] says, "Yes, that was a code." I said,
[26:40] "They thought it was a code. lithog
[26:42] glycerin
[26:44] list. Then one day I'm jiggling around.
[26:47] We in the first few weeks all this went
[26:49] on. It was a few weeks before we ever
[26:51] got it straightened out. But I'm
[26:52] piddling around with the adding machine,
[26:53] the computing machine. I notice
[26:54] something. So having writing every day,
[26:56] I lot of things I write. It's very
[26:59] peculiar. Notice what happens. If you
[27:01] take one divided by
[27:02] 273, you get
[27:06] 04115 226 337. It's quite cute. And then
[27:10] it goes a little cockey when the
[27:11] carrying occurs only for about three
[27:13] numbers. And then you can see how the 10
[27:16] 1013 is really equivalent to 114 again
[27:19] or 115 again and keeps on going. And I
[27:21] explained that how nicely it repeated
[27:24] itself after a couple of cycles. I
[27:25] thought it was kind of amusing. Well, I
[27:27] put that in the mail and it comes back
[27:28] to me. It doesn't go through. And
[27:30] there's a little note. Look at paragraph
[27:31] 17B. I look at paragraph 17B.
[27:34] say letters are to be written only in
[27:36] English, Russian, Spanish, Portuguese,
[27:40] uh, Latin, I God knows what, I can't,
[27:43] German, and so forth. Permission to use
[27:44] any other language must be obtained in
[27:46] writing. And then it said, no
[27:49] codes. So I wrote back to the sensor a
[27:51] little note included in my letter which
[27:53] said of course this cannot be a code
[27:55] because there's no more if you actually
[27:57] do divide one into two you do in fact
[27:59] get and I wrote the whole thing out and
[28:01] therefore there's no more information in
[28:03] the number one one
[28:04] 004114 than there is in the number 273
[28:07] which is hardly any information and so
[28:10] forth. I therefore ask for permission to
[28:13] write my letters in Arabic
[28:15] numerals. I like to use Arabic numerals
[28:18] in my letters. So I got that through.
[28:19] All
[28:21] right. There was always some kind of
[28:23] difficulty with the letters going back
[28:25] and forth. For example, at one time my
[28:28] wife kept insisting on mentioning the
[28:30] fact that she feels uncomfortable
[28:31] writing the feeling that the sensor is
[28:32] looking over. So and there was a rule.
[28:34] We aren't supposed to mention
[28:35] censorship. We are. But how can they
[28:37] tell her? So they keep sending me a
[28:39] note. Your wife mentioned censorship.
[28:40] Certain. My wife mentioned
[28:42] censorship. So finally they send me a
[28:44] note and say please inform your wife not
[28:45] to mention censorship in her letter. So
[28:47] I take my letter and I start. I have
[28:49] been instructed to inform you not to
[28:50] mention censorship in your letter. Boom.
[28:52] Boom. Comes right
[28:58] back. So I write, I have been instructed
[29:00] to inform her not to mention sens. How
[29:02] in the heck am I going to do it?
[29:04] Furthermore, why do I have to instruct
[29:05] her not to mention scripture? You're
[29:07] keeping something from me. It's very
[29:09] interesting. The sensor himself has to
[29:11] tell me to tell my wife not to tell me
[29:13] that
[29:14] she's But they had an answer. They said
[29:16] yes, that they're worried about the mail
[29:19] being intercepted on the way from
[29:21] Albuquerque and that they would find out
[29:22] that there was censorship if they looked
[29:23] in the mail. And would she please act
[29:25] much more normal? And so I went down the
[29:27] next time to Albuquerque and I talked to
[29:28] her and I said, "Now, Alex, let's not
[29:30] mention censorship." But we had had so
[29:32] much trouble that we had at last worked
[29:35] out a code, something illegal. We had a
[29:37] code. If I would put a dot at the end of
[29:38] my signature, it meant that we had had
[29:40] trouble again. and she would move on to
[29:43] the next of the moves that she had
[29:45] concocted. She'd sit there all day long
[29:47] because she was ill and she'd think of
[29:48] things to do. The last thing that she
[29:51] did was to send me, cuz she found
[29:54] perfectly legitimately an advertisement
[29:56] that said, "Send your boyfriend a letter
[29:59] on a jigsaw puzzle. Here are the blanks.
[30:02] We send you the blank. You write the
[30:03] letter on it, take it all apart, and put
[30:04] a little sack and mail it."
[30:08] So, I received that one with a note
[30:10] saying, "We do not have time to play
[30:11] games. Please instruct her. Please
[30:14] instruct your wife to confine ourselves
[30:16] to ordinary
[30:19] letters." Well, we're ready with the one
[30:21] more dot, but they got ready just in
[30:22] time. But the thing that we had ready
[30:24] for the next one was the letter would
[30:26] start. I hope you remembered to open
[30:28] this letter carefully because I have
[30:30] included the Pepto-Bismol for your
[30:31] stomach as we
[30:33] arranged. Would be a letter full of
[30:35] powder in their office. We expect they'd
[30:37] open it quickly. The powder would go all
[30:38] over the floor. They get all upset
[30:40] because it's not supposed to disturb
[30:42] anything. You see, have to gather this
[30:43] captive business. But we didn't have to
[30:46] use that one.
[30:48] Okay. As a result of all these
[30:51] experiences with the sensor, I knew
[30:52] exactly what could get through and what
[30:53] could not get through. Nobody else knew
[30:55] as well as I. And so I made a little
[30:57] money finally out of all this by making
[30:59] bets.
[31:01] One day on the outside fence, the
[31:04] outside fence, I had discovered that
[31:06] workmen who lived still further out and
[31:08] wanted to come in were too lazy to go
[31:10] around through the gate. And so they had
[31:12] cut themselves a hole some distance
[31:13] along. And so I went out the gate, went
[31:15] over to the hole and came in, went out
[31:17] again and so on until the guy, the
[31:20] sergeant at the gate begins to wonder
[31:22] what's happening. This guy's always
[31:23] going out and never coming in. And of
[31:26] course, his natural reaction was to call
[31:28] a lieutenant and try to put me in jail
[31:30] for doing this.
[31:31] I explained there was a hole. So I was
[31:33] always trying to straighten people out,
[31:34] point out there was a hole. And so I
[31:36] made a bed with somebody that I could
[31:37] tell where the hole in the fence was in
[31:39] a mail mail it out. And sure enough, I
[31:42] did. And the way I did it, I I said,
[31:43] "You should see the way they administer
[31:45] this place." You see, that's what we
[31:46] were allowed to
[31:48] say. There's a hole in the fence 71 ft
[31:51] away from such and such a place that's
[31:52] this size and that size you can walk
[31:54] through. Now, what can they do? They
[31:56] can't say to me that there's no such
[31:57] hole. I mean, what are they going to do?
[31:59] It's their own hard look that there's
[32:00] such a hole they should fix the hole.
[32:03] So, they got that one through. I also
[32:04] got through a letter which told about
[32:06] how one of the boys who worked in one of
[32:09] my groups had been wakened up in the
[32:10] middle of the night and grilled with
[32:12] lights in front of him by some idiots in
[32:14] the army there because they found out
[32:16] something about his father or something.
[32:17] I don't know. I think supposed to be a
[32:18] comm. His name was Key. He's a famous
[32:20] man.
[32:22] Well, there were also some other things
[32:24] I was always trying to straighten out
[32:26] like point out the holes in the fence
[32:28] and so forth, but always these things
[32:29] where I was always doing these point
[32:31] things out in a nondirect manner. And
[32:33] one of the things I wanted to point out
[32:35] was this that at the very beginning we
[32:37] had terribly important secrets. We'd
[32:39] worked out lots of stuff about uranium,
[32:40] how it worked, and all this stuff was in
[32:42] documents that were in filing cabinets
[32:44] that were made out of wood that had on
[32:46] them little ordinary common
[32:49] padlocks and various things made by the
[32:51] shop like a rod that would go down and
[32:53] then a padlock to hold it. But so that's
[32:55] the
[32:56] padlock. Furthermore, you could get the
[32:58] stuff out even without opening the
[32:59] padlock out of these wooden cabinets.
[33:00] You just tilt it over backwards and the
[33:02] bottom drawer, you know, has a little
[33:03] rod that's supposed to hold and there's
[33:05] a hole in the wood underneath. You can
[33:06] pull the papers out from below
[33:07] everything.
[33:08] So I used to pick the locks all the time
[33:10] and point out that it was very easy to
[33:12] do. And every time we had a meeting of
[33:15] everybody together, I'd get up and I'd
[33:16] say that we have important secrets and
[33:18] we need better locks. And so uh one day
[33:21] Teller got up at the meeting I got up at
[33:23] and he says to me, "Uh, well, I don't
[33:25] keep my most important secrets in my
[33:27] filing cabinet. I keep them in my desk
[33:29] drawer. Isn't that better?" I said, "I
[33:31] don't know. I haven't seen your desk
[33:32] drawer." Well, he's sitting near the
[33:35] front of the meeting and I'm sitting
[33:36] further back. So, the meeting continues
[33:39] and I sneak out of the meeting and I go
[33:42] down to see his desk drawer. Okay, I
[33:45] don't even have to pick the lock on the
[33:46] desk drawer. It turns out if you put
[33:48] your hand in the back underneath, you
[33:49] can pull out a paper like those toilet
[33:51] paper dispensers. You pull out one, it
[33:52] pulls another, pulls
[33:53] another. I emptied the whole damn
[33:56] drawer, took everything out and put it
[33:58] away to one side and then went up on the
[34:00] higher floor and came back and the
[34:03] meeting is just ending and everybody's
[34:04] coming out and I join the crew like
[34:06] this. You see, I'm walking along with it
[34:08] and run up to catch up with Tella and
[34:10] says, "Oh, by the way, let me see your
[34:11] desk draw." You see? So, he says,
[34:14] "Certainly," he says. So, we walk into
[34:16] his office and he uh shows me the desk
[34:18] and I look at I said, "That looks pretty
[34:20] good to me. That's pretty safe. It looks
[34:22] to me like it's pretty safe." I said,
[34:23] "Let's see what you have in there." So,
[34:26] he says, "I'd be very glad to show it to
[34:27] you." He says, "Putting the key, opens
[34:29] it, looks, and he says, "If you hadn't
[34:31] already seen it
[34:32] yourself." The trouble with playing a
[34:35] trick on a highly intelligent man like
[34:36] Mr. Teller is the time that it takes him
[34:39] to figure out from the
[34:41] moment that he sees that there's
[34:43] something wrong till he understands
[34:45] exactly what happened is too damn small
[34:47] to give you any pleasure.
[34:57] Well, I had a lot of other fun with the
[34:59] safes, but it has nothing to do with Los
[35:01] Alamos, so I won't discuss it
[35:08] further. Where I want to tell about some
[35:11] of the special problems that I had that
[35:12] are rather interesting. One of them had
[35:14] to do with the safety of the plant at
[35:16] Oakidge. Los Alamos is going to make the
[35:18] bomb. But at Oakidge, they're trying to
[35:20] separate the isotopes of uranium,
[35:22] uranium 238 and uranium 236, the latter
[35:25] of which is the 235, which is the
[35:27] explosive one. All
[35:29] right. So, they were just beginning to
[35:33] get infinite amounts from an
[35:35] experimental thing of the 235. And at
[35:38] the same time, they were practicing.
[35:40] There's a big plant there that was going
[35:41] to be we're going to have vats of this
[35:42] stuff and they're going to chemicals and
[35:44] re take the purified stuff and then
[35:46] reurify and get it ready for the next
[35:48] stage after purified in several stages.
[35:50] So they were practicing the chemistry on
[35:52] the one hand and they were just getting
[35:54] a little bit from one of the pieces of
[35:55] apparatus experimentally on the other
[35:57] hand and they were trying to learn how
[35:59] to assay it to determine how much
[36:01] uranium 235 there is in it and we would
[36:03] send instructions and they never got it
[36:06] right. So finally sig gray said that the
[36:09] only possible way to get it right is to
[36:10] go down there to see what they're doing
[36:12] to understand why the assay is wrong.
[36:14] The army people said no it's our policy
[36:17] to keep all the information loss in one
[36:19] place and that the people in Oakidge
[36:20] would not know anything about what it
[36:21] was used for. They just knew what they
[36:23] were trying to do. I mean the higher
[36:24] people knew they were separating uranium
[36:26] but they didn't know how powerful the
[36:27] bomb was or exactly how it worked or
[36:29] anything and the people underneath
[36:31] didn't know at all what they were doing
[36:33] and the army wanted to keep it that way.
[36:35] There was no information going back and
[36:37] forth, but Sige finally insisted on it
[36:39] that it was important. They could never
[36:40] get the assays right. The whole thing
[36:41] would go up in smoke. So, Sigra went
[36:44] down to see what they were doing. And as
[36:46] he was walking through, he saw them
[36:48] wheeling a tank car boy of water, green
[36:51] water. The green water is uranium
[36:54] nitrate. He says, "Uh, you're going to
[36:56] handle it like that when it's purified,
[36:58] too?" They said, "Sure, why not?" "Well,
[37:01] wouldn't it explode?" He says, "Huh?
[37:04] Explode? how and so the army said, "You
[37:07] see, we shouldn't have let any
[37:08] information go
[37:10] across." Well, it turned out that the
[37:14] army had realized how much stuff we
[37:15] needed to make a bomb, 20 kg or whatever
[37:18] it was. And they realized that that much
[37:20] material purified would never be in the
[37:22] plant. So, there was no danger. But they
[37:24] did not know that the neutrons are
[37:26] enormously more effective when they're
[37:27] slowed down in water. And so in water it
[37:31] takes less than a tenth no 20 h 100th
[37:34] very much less material you can make a
[37:36] reaction which makes radioactivity
[37:38] doesn't make a big explosion but makes
[37:39] radioactivity kills people around and so
[37:41] on. So it was very dangerous and they
[37:44] had not paid any attention to safety at
[37:46] all. So a telegram goes from up to gray
[37:49] go through the entire plant notice where
[37:51] all the concentrations are supposed to
[37:53] be with the process as they've designed
[37:54] it. we will calculate in the meantime
[37:57] how much material can come together
[37:58] before there's an explosion. And so two
[38:01] groups started working on it.
[38:02] Christiey's group worked on water
[38:03] solutions and I worked on dry powder in
[38:06] boxes. My group and we calculated about
[38:09] how much material and Christy was going
[38:11] to go down and tell them all at Oakidge
[38:13] what the situation was. The whole thing
[38:14] was broken down. We have to go down and
[38:16] tell them now. And so I happily give all
[38:19] my numbers to Christie and say you have
[38:22] all the stuff you go. Christy got
[38:24] pneumonia. I had a go. I never traveled
[38:27] on an airplane before. I traveled in an
[38:29] airplane. They strapped the secrets with
[38:32] a little thing with belt on my
[38:34] back. The airplane in those days was
[38:36] like a bus. You stop off every once in a
[38:39] while, except the stations were further
[38:40] apart. You stop off and wait. There's a
[38:42] guy standing next to me with a chain
[38:44] swinging like this saying something
[38:45] like, "It must be terribly difficult to
[38:47] fly without a priority on airplanes
[38:49] these days." I couldn't resist. I said,
[38:51] "Well, I don't know." Oh, I said I have
[38:53] a
[38:54] priority a little bit later. It looks
[38:56] like some generals are coming. They're
[38:58] going to put out some of us number
[38:59] threes. It's all right. I'm a number
[39:04] two. He probably wrote to his
[39:06] congressman if he wasn't a congressman
[39:08] himself saying, "What are they doing
[39:09] sending these little kids around with
[39:11] number two priorities in the middle of
[39:12] the
[39:14] wall?" At any rate, I arrived there and
[39:17] the first thing I did was have them take
[39:19] me through the plant and I said nothing.
[39:20] I just looked at everything. I found out
[39:22] that the situation was even worse than
[39:23] Sigra reported because he was confused
[39:25] the first time. He noticed certain boxes
[39:27] in big lots and he didn't notice another
[39:30] boxes in another room in a big lot but
[39:32] it was the same room on the other side
[39:33] and things like that. So if you have too
[39:35] much stuff together it goes up you see.
[39:37] So it was I went through the entire
[39:39] plant and I have a very bad memory but
[39:41] when I work intensively I have a good
[39:42] shortterm memory and so I could remember
[39:45] all kinds of crazy things like building
[39:46] 9207 VAT number so and so and so forth.
[39:49] You see? So I had all that stuff. I went
[39:50] home that night and I went through the
[39:53] whole thing explaining where all the
[39:54] dangers were, what you would have to do
[39:56] to fix it. Rather easy. You put cadmium
[39:58] in solutions to absorb the neutrons in
[40:01] the water. You separate the boxes so
[40:03] they're not too dense, too much uranium
[40:04] together and so on according to certain
[40:06] rules. And so I worked out all the
[40:07] examples and how it worked. I felt that
[40:10] you couldn't make the plant safe unless
[40:11] you knew how it worked and so forth. So
[40:13] the next day there's going to be a big
[40:14] meeting. I forgot to say before I left
[40:18] Oppenheimer said to me uh now he said
[40:20] when you go their following people are
[40:22] technically able down there at Oakidge
[40:24] Mr. Julian Webb is so and so and so I
[40:27] want you to make sure that these people
[40:28] are at the meeting that you tell them
[40:30] the safety problem that they really
[40:31] understand they're in charge of and I
[40:33] said suppose they're not at the meeting
[40:34] what am I supposed to do he said then
[40:37] you should say Los Alamos cannot accept
[40:39] the responsibility for the safety of the
[40:41] Oakidge plant unless these guys I said
[40:43] you mean me little Richard's going to go
[40:45] in there and say he says yes little
[40:46] Richard you go and do that said I really
[40:48] grew up fast you know so when I arrived
[40:51] sure enough I arrived there and the
[40:53] meeting was the next day And all these
[40:55] people from the company, the big shots
[40:57] in the company and the technical people
[40:58] that I wanted were there and the
[40:59] generals and so forth that were
[41:00] interested in the problems and
[41:02] organizing everything. It was a big
[41:03] meeting about this very serious problem
[41:06] of the safety because the plant would
[41:07] never work. It would have blown up. I
[41:08] swear it would have if nobody had paid
[41:10] attention to it. So there was a
[41:13] lieutenant who took care of me. He told
[41:15] me that some colonel says that I
[41:17] shouldn't tell him how the neutrons work
[41:18] and all the details because we want to
[41:20] keep the thing separate. I just tell
[41:21] them what to do to keep it safe. I said
[41:24] in my opinion it's impossible for them
[41:25] to understand or to obey a bunch of
[41:27] rules that they don't understand unless
[41:29] they understand how it works. And so
[41:30] it's my opinion that it's only going to
[41:32] work if I tell them and so cannot accept
[41:34] the responsibility for the safety of
[41:37] Oakidge plant unless they are fully
[41:39] informed as to how it works. He was
[41:41] great. So he goes to the colonel and he
[41:44] says it to the colonel is just five
[41:46] minutes he says. So he goes to the
[41:47] window and he starts and thinks and
[41:49] that's what they're very good at.
[41:51] They're good at making decision. I
[41:53] thought it was very remarkable how a
[41:55] problem of whether or not information as
[41:57] to how the bomb works should be in the
[42:00] Oak Ridge plant or not had to be decided
[42:02] and could be decided in 5 minutes. So I
[42:05] have a great deal of respect for these
[42:07] military guys because I never can decide
[42:08] anything very important in any length of
[42:10] time at all. So he looks five minutes he
[42:12] says all right Mr. Fman go ahead. So he
[42:15] sat down and I told him all about
[42:16] neutrons, how they worked, you know, da
[42:19] that too many neutrons together. You got
[42:20] to keep the material apart. Cadmium
[42:22] absorbs them. Slow neutrons are more
[42:23] effective than fast neutrons. Yak yak
[42:25] all stuff which was elementary primmer
[42:28] stuff at Los. But they had never heard
[42:30] of any of it. So I turned out to be a
[42:32] tremendous genius to them. I was a god
[42:35] coming down from the sky with all all
[42:38] these phenomena that were not
[42:39] understood, never heard of before. I
[42:41] knew all about it. I could give them
[42:42] facts and numbers and everything else.
[42:44] So by being rather primitive back there
[42:46] at Los Alamos, I was a super genius at
[42:48] the other
[42:49] end. Well, the result was that they made
[42:52] little groups to make their own
[42:54] calculations to learn how to do it. They
[42:56] started to redesign plants. The
[42:57] designers of their plants were there.
[42:59] The construction designers, engineers,
[43:01] the chemical engineers for the new plant
[43:03] that was going to handle the separated
[43:04] material were there and other people
[43:06] were there. And I went away again. They
[43:08] told me to come back in a few months.
[43:09] They were going to redesign their plant
[43:11] for the separation.
[43:13] So I came back in a month or so and
[43:15] Stony and Webster Company engineers had
[43:17] finished design the plant and now it was
[43:19] for me to look at the plant.
[43:21] Okay. How do you look at a plant that
[43:23] ain't built yet? I don't know. So I go
[43:25] into this room with these fellas take me
[43:26] into the room. There was always a
[43:28] Lieutenant Zumalt that was always coming
[43:30] around with me taking care of me. You
[43:31] know I had to have an escort everywhere.
[43:33] So he goes with me, takes me into this
[43:35] room and there are these two engineers
[43:36] and a long table, great big long table.
[43:40] tremendous thing covered with a
[43:42] blueprint that's as big as a table. Not
[43:44] one blueprint, but a stack of blueprints
[43:46] like this. I took mechanical drawing
[43:49] when I was in school, but I wasn't too
[43:50] good at reading blueprints. So, they
[43:52] start to explain it to me because they
[43:54] thought I was a genius. And they start
[43:56] out, Mr. Prim, would like you to
[43:58] understand the plant is so designed. You
[44:00] see, one of the things we had to avoid
[44:01] was accumulation problems like there's
[44:03] an evaporator working, which is trying
[44:05] to accumulate stuff. If a valve gets
[44:07] stuck or something like that and
[44:08] accumulate too much stuff, it'll
[44:09] explode. So they explain to me that this
[44:11] plan is designed so that no one valve if
[44:14] any one valve gets stuck nothing will
[44:16] happen. It needs at least two valves
[44:18] everywhere. Okay. So then they explain
[44:21] how it works. The car tetrachloride
[44:23] comes in here. The uranium nitrate from
[44:24] here comes in here. It goes up and down.
[44:25] It goes through the floor comes up
[44:26] through the pipes coming up in the
[44:27] second floor through the blueprints down
[44:30] up down up. Very fast talking explaining
[44:32] it. Very complicated chemical plant. I'm
[44:35] completely dazed.
[44:36] worse. I don't know what the symbols on
[44:39] the blueprint
[44:40] mean. There's some kind of a thing that
[44:43] at first I think it's a window. It's a
[44:44] square with a little cross in the middle
[44:47] like this all over the damp lines with
[44:49] this damn square. Lines with this damn
[44:50] square. I think it's a window. No, it
[44:52] can't be a window cuz it ain't always at
[44:54] the
[44:55] edge. I want to ask him what it is. Now,
[44:58] you must have been in a situation like
[45:00] this. You didn't ask them right away.
[45:01] right away it would have been okay but
[45:03] they've been talking a little bit too
[45:07] long. You hesitated too long. If you ask
[45:10] them they'll say what are you wasting my
[45:12] time all this time for. I don't know
[45:14] what to do. I think to myself I swear I
[45:18] often in my life I've been lucky. You
[45:19] are not going to believe this story but
[45:21] I swear it's absolutely true. It's such
[45:23] sensational luck. It's I think what am I
[45:25] going to do? What am I going to do? I
[45:27] got an idea. Maybe it's a valve. So in
[45:30] order to find out whether it's a valve
[45:31] or not, I take my finger and I put it
[45:33] down in the middle of one of the
[45:34] blueprints on page number three down in
[45:35] the unit and I say what happens if this
[45:36] valve gets stuck. Figuring they're going
[45:39] to say that's not a valve so that's a
[45:42] window. So one looks at the other says
[45:44] well if that valve gets stuck and they
[45:45] go up and down in the blueprints up and
[45:46] down the blueprint. Other guy up and
[45:47] down the blueprint back and forth back
[45:48] and forth back and for both look at each
[45:49] other and they ch me and they open them
[45:52] out like this. You're absolutely right
[45:54] sir.
[45:55] [Applause]
[46:05] So,
[46:06] uh, they rolled up the blueprints of the
[46:09] way they went and we walked
[46:11] out. And Mr. Zumwalt, who'd been
[46:13] following me all the way through, he
[46:15] says, "You're a genius." He says, "I got
[46:18] the idea you're a genius when you went
[46:20] through the plant once and you could
[46:21] tell them about evaporator C-21 and
[46:22] build it 9207 the next morning." He says
[46:26] and when you knew all about the neutron
[46:27] but what you have just done he said was
[46:30] so
[46:31] fantastic I want to know how how do you
[46:34] do something like that I told him you
[46:36] try to find out whether it's a valve or
[46:42] not well another kind of problem that I
[46:46] worked on was this we had to do lots of
[46:49] calculations and we did them on Marshant
[46:50] calculating machines by the way just to
[46:53] give you an idea what Losamos was Right.
[46:55] We had these Marant computers. I don't
[46:57] know if you know what they look like.
[46:58] Hand calculated with numbers and you
[46:59] push them and they multiply, divide,
[47:00] add, and so on. Not like they do easy
[47:02] now, but hard. They were mechanical
[47:04] gadgets. And they had to be sent back to
[47:06] the factory to be repaired. We didn't
[47:08] have a special man to do it. And that
[47:09] was the standard way to do it. And so
[47:11] they were always being sent to the
[47:12] factory and pretty soon you were running
[47:14] out of machines. So I and a few other
[47:16] fellas started to take the covers off.
[47:18] We weren't supposed to. There's a rule.
[47:19] You take the covers off, we cannot be
[47:20] responsible. So we took the covers off.
[47:23] We had nice series of lessons like the
[47:24] first one we took the cover off for
[47:26] there was a shaft with a hole in it and
[47:28] a spring which was hanging this way and
[47:29] obviously the spring went in the hole so
[47:31] it was easy. So we got like a series of
[47:33] lessons by God on how to fix them and we
[47:35] got better and better and we did more
[47:36] and more elaborate repairs. When we got
[47:39] something too complicated we send it out
[47:41] back to the factory but we do the easy
[47:42] ones and kept the things going. Uh I
[47:44] also did some typewriters. I ended up
[47:46] doing all the computers. The other
[47:48] fellas quit on me. I did a few
[47:50] typewriters, but there's a guy in the
[47:51] machine shop who was better than I was
[47:52] and he took care of typewriters. I took
[47:54] care of any machines. However, we
[47:57] decided that the big problem, which was
[47:59] to figure out exactly what happened
[48:00] during the bomb's explosion, when you
[48:02] push the stuff in by an explosion and
[48:04] then it goes out again, exactly what
[48:06] happened so we could figure out how much
[48:07] energy was released and so on required
[48:10] much more calculating than we were
[48:11] capable of. And a rather clever fellow
[48:13] by the name of Stanley Frankle realized
[48:15] that it possibly could be done on IBM
[48:17] machines. The IBM company had machines
[48:20] for business purposes which adding
[48:21] machines they're called tabulators for
[48:23] listing sums and multiplier just a
[48:26] machine big machine you put cards in and
[48:28] it would take two numbers from a card
[48:30] and multiply and print it on the card
[48:31] and then there were collators and
[48:33] sortters and so on. So he figured out a
[48:35] nice program if we got enough of these
[48:37] machines in a room we could take the
[48:38] cards and put them through a cycle.
[48:40] Everybody who does numerical calculation
[48:41] now knows exactly what I'm talking about
[48:43] but this is kind of a new thing mass
[48:45] production with machines. We had done
[48:47] things like this on adding machines.
[48:49] Usually you go one step across yourself
[48:51] doing everything. But this was
[48:53] different. We go first to the adder,
[48:54] then we go to the multiply, then we go
[48:55] to the adder and so on. So he designed
[48:58] this thing and ordered the machines from
[49:00] the IBM company. We realized it was a
[49:03] good way of solving our problems. And we
[49:06] found that there was somebody in the
[49:07] army that had IBM training. We needed a
[49:09] man to repair them to keep them going
[49:10] and everything. And they were going to
[49:11] send this fell, but it was delayed.
[49:13] Always delayed. Now we always were in a
[49:15] hurry. I have to explain that everything
[49:17] we did we tried to do as quickly as
[49:19] possible. So in this particular case we
[49:21] worked out all the numerical steps that
[49:23] the machines were supposed to do
[49:24] multiply this and then do this and
[49:25] subtract that and know we worked out the
[49:27] program but we didn't have any machines
[49:29] to test it on. So what we did is we all
[49:32] I arranged was a room with girls in it.
[49:35] Each one had a marsh but she was the
[49:36] multiplier and she was the air and this
[49:38] one cubed. So we had cards index cards.
[49:42] All she did was cube this number and
[49:44] send it to the next one. She was
[49:45] imitating the multiplier. The next was
[49:46] imitating the adder. And we went through
[49:48] our cycle. We got all the bugs out when
[49:49] we did it that way. And it turned out
[49:51] that the speed at which we were able to
[49:53] do it. We' never done mass production
[49:56] calculator. Everybody who ever
[49:57] calculated before, no single person did
[49:59] all the steps, but it's Ford had a good
[50:01] idea. The damn thing works a hell of a
[50:04] lot faster the other way. And we got
[50:06] speed with this system that was the
[50:08] predicted speed for the IBM machine. The
[50:11] same. The only difference is that the
[50:13] IBM machines didn't get tired and could
[50:15] work three shifts, but the girls got
[50:16] tired after a while. So anyway, we got
[50:19] the bugs out during that process and
[50:21] finally the machines arrived, but not
[50:23] the repair man. So we went down to put
[50:27] them together and one of the most
[50:28] complicated machines, the technology of
[50:30] those days, these computing machines,
[50:32] the big thing, they came partially
[50:35] disassembled with lots of wires and
[50:36] blueprints of what to do. We went down
[50:38] and we put them together. Stan Franklin
[50:39] and I and another fella and uh we had
[50:43] our trouble. Most of the trouble was the
[50:45] big shots coming all the time and say
[50:46] you're going to break something. You're
[50:47] going to break
[50:48] something. We put them together and
[50:50] sometimes they would work and sometimes
[50:52] they were put together wrong and they
[50:54] didn't work. And so he fiddled around
[50:56] and they got it to work. We didn't get
[50:58] them all to work. And I was last working
[51:00] on some multiply. I saw a bed part
[51:02] inside but I was afraid to straighten it
[51:04] because it might snap off. They were
[51:05] always telling us we're going to bust it
[51:07] irreversibly. And finally, the man from
[51:10] the IBM company came in time, as a
[51:12] matter of fact, according to schedule.
[51:14] But he came and he fixed the rest that
[51:15] we hadn't got ready and everything was
[51:17] going. We got the program going, but he
[51:18] had trouble with the one that I had had
[51:20] trouble with that I couldn't fix. So
[51:22] after 3 days, he was still working on
[51:23] that one last one. I went down and said,
[51:24] "Oh, I noticed that was bent." Oh, he
[51:26] says, "Of course, that's all there is to
[51:29] it. It was all
[51:31] right." So that was it.
[51:35] Well, Mr. Frankle started this program
[51:39] and began to suffer from a disease. The
[51:41] computer disease that anybody who works
[51:42] with computers now knows about. It's a
[51:44] very serious disease and it interferes
[51:46] completely with the work. It was a
[51:47] serious problem we were trying to do.
[51:49] The disease with computers is you play
[51:51] with them. They're so wonderful. You
[51:53] have these X switches that determine if
[51:55] it's a even number, you do this, and if
[51:57] it's an odd number, you do that. And
[51:58] pretty soon, you can do more and more
[51:59] elaborate things if you're clever enough
[52:01] on one machine. And so after a while it
[52:04] turned out the whole system broke down.
[52:06] He wasn't paying any attention. He
[52:07] wasn't supervising anybody. The system
[52:09] was going very very slowly. The real
[52:11] problem. Well, he was sitting in a room
[52:13] figuring out how to make one tabulator
[52:15] automatically print arc tangent x and
[52:19] then it would start and print columns
[52:20] and then and calculate the arc tangent
[52:23] automatically by integrating as it went
[52:25] along and make the whole table in one
[52:27] operation. Absolutely useless. We had
[52:28] tables of arc tangent.
[52:31] But if you've ever worked with
[52:32] computers, you understand the disease,
[52:34] the delight to be able to see how much
[52:35] you can do. But he got the disease for
[52:37] the first time. The poor fellow who
[52:38] invented the thing got the disease. And
[52:40] so I was asked to stop working on the
[52:42] stuff I was doing in my group and to go
[52:44] down and to take over the IBM group. And
[52:47] so I noticed the disease and I tried to
[52:49] avoid the disease. And uh although they
[52:51] did three problems in nine months, I had
[52:53] a very good group. The first problem was
[52:55] that they had never told the fellas they
[52:57] had selected all over the country thing
[52:58] called special engineered detachment.
[53:00] clever boys from high school to go into
[53:02] the army who had engineering ability and
[53:04] they collected them together special
[53:05] engineer detachment. They sent them up
[53:07] to Los Alamos. They put them in barracks
[53:09] and they would tell them nothing. Then
[53:11] they came to work and what they had to
[53:12] do is work on IBM machines punching
[53:14] holes numbers that they didn't
[53:16] understand. Nobody told them what it
[53:17] was. The thing was going very slowly. I
[53:19] said that the first thing has to be is
[53:20] that the technical guys know what we're
[53:23] doing. So up and went and talked to the
[53:26] security and got special permission. So
[53:27] I had a nice lecture which I told them
[53:28] what we're doing. They were all excited.
[53:30] We're fighting a war. We see what it is.
[53:32] They knew what the numbers meant. If the
[53:33] pressure came out higher, it meant was
[53:35] more energy release. It was so on and so
[53:37] on. They knew what they were doing.
[53:38] Complete
[53:40] transformation. They began to invent
[53:42] ways of doing it better. They improved
[53:44] the scheme. They worked at night. They
[53:46] didn't need supervisors in the night.
[53:47] They didn't need anything. They
[53:48] understood everything. They invented
[53:49] several of the programs that were used
[53:51] and so forth. So, my boys really came
[53:53] through and all they had to be done was
[53:55] to tell them what it was. That's all.
[53:57] They're just not telling them they're
[53:58] punching holes,
[54:00] please. As a result, although it took
[54:03] them nine months to do three problems
[54:05] before, we did nine problems in three
[54:07] months. Nearly 10 times as fast. But one
[54:09] of the secret ways that we did our
[54:11] problems was this. The problems
[54:13] consisted of a bunch of cards which had
[54:14] to go through a cycle. First add, then
[54:16] multiply. And so it went through the
[54:18] cycle of machines in this room slowly
[54:20] about as it went around and around. So
[54:22] we figured a way by taking a different
[54:24] colored set of cards to put them through
[54:26] a cycle too but out of phase. We do two
[54:28] or three problems at a time. You see
[54:30] this was another problem but while this
[54:31] one was adding it was multiplying on the
[54:33] other problem and such managerial
[54:35] schemes we got many more
[54:38] problems. Finally near the end of the
[54:40] war just before we had to make a test in
[54:42] Albuquerque the question was how much
[54:45] would be released. We had been
[54:47] calculating the release from various
[54:49] designs but the specific design which is
[54:52] ultimately used we hadn't computed. So
[54:54] Bob Christie came down and said we would
[54:56] like the result for how this thing is
[54:58] going to work in one month or some very
[55:00] short time less than that three weeks. I
[55:03] said it's impossible but he says look
[55:04] you're putting out so and so many
[55:05] problems a month. It takes only two
[55:07] three weeks per problem. I said I know
[55:09] it takes much longer to do the problem
[55:11] but we're doing them in parallel. they
[55:12] go through. It takes a long time and
[55:14] there's no way to make it go around
[55:16] faster. So he went out and I began to
[55:18] think, is there a way to make it go
[55:19] around faster? Well, if we did nothing
[55:22] else on the machine, so there was
[55:23] nothing interfering and so and so. I
[55:25] began to think I put on the blackboard a
[55:26] challenge. Can we do it to the boys?
[55:29] They all thought that yes, we'll work
[55:31] double shift. We'll work over time.
[55:33] We'll try it. We'll try it. And so the
[55:36] rule was all of the problems out only
[55:38] one problem and just concentrate on this
[55:40] thing and so forth. So they started to
[55:43] work and my wife died in Albuquerque and
[55:44] I had to go down. I borrowed Fuch's car.
[55:46] He was a friend of mine in the
[55:48] dormatory. He had an automobile. He was
[55:50] using the automobile to take the secrets
[55:52] away, you know, at Santa Fe. He was the
[55:54] spy. I didn't know that. I borrowed his
[55:56] car to go to Albuquerque. The damn thing
[55:58] got three flat tires on the way.
[56:01] I came back from there and I went into
[56:04] the room cuz I was supposed to be
[56:06] supervising everything but I couldn't do
[56:07] it for three days and it was in this
[56:09] mess, this big rush to get the answer
[56:11] for the test that was going to be done
[56:12] in the desert. I go into the room and
[56:15] there are three different color cards.
[56:18] There's white cards, there's blue cards,
[56:19] there's yellow cards. And I start to
[56:21] say, "But you're not supposed to do more
[56:22] than one problem. Only one problem."
[56:23] They said, "Get out. Get out. Get
[56:25] out. Wait. We'll explain everything." So
[56:27] I waited. And what happened was this. As
[56:30] they went through, sometimes the machine
[56:32] made a mistake or they put a wrong
[56:34] number in. What we used to have to do,
[56:36] we find something is go back and do that
[56:38] over again. But they noticed this that
[56:40] there's a deck of cards represented
[56:41] positions in depth in space or
[56:43] something. The mistake made here in one
[56:45] cycle only affects the nearby numbers.
[56:47] The next cycle affects the nearby
[56:48] numbers and so on. It only works its way
[56:50] through the pack of cards. You have 50
[56:52] cards, you make a mistake in card number
[56:53] 39, it affects 37, 38, and 39. The next
[56:56] time 36, 37, 38, 39, and 40. Next time
[56:59] it spreads like a disease, the error. So
[57:02] they found an error back a ways and they
[57:05] got an idea. They would only compute a
[57:08] small deck of 10 cards around the error
[57:12] and because 10 cards could be put
[57:13] through the machines faster than the
[57:14] deck of 50 cards, they would go with
[57:16] this other deck rapidly through while
[57:18] they continued with the 50 cards with
[57:20] the disease spreading. But the other
[57:22] thing was computing faster and they
[57:24] would seal it all up and correct it.
[57:25] Okay, very clever. That was the way
[57:28] those guys worked. really hard, very
[57:29] clever to get speed. There's no other
[57:31] way. If they had stopped to try to fix
[57:33] it, we'd have lost our part. We couldn't
[57:34] have got it recycle. That's what they
[57:36] were doing. Of course, you know what
[57:37] happened while they were doing that?
[57:39] They found an error in the blue deck.
[57:41] And so they had a yellow deck with fewer
[57:44] cards. It was going around faster. And
[57:45] the blue deck and just when they're
[57:47] going crazy, you know, because after
[57:49] they get it straight, now they got to
[57:50] fix the white one. They got to take the
[57:51] other cards out and replace it by the
[57:52] right ones and continue correctly. And
[57:54] it's rather confusing how those things
[57:56] always are. You don't want to make a
[57:57] mistake. And just at the time when they
[57:59] got these three decks going, they're
[58:00] trying to seal everything else. The boss
[58:01] comes walking in. Leave us alone, they
[58:04] said. So I left them alone and
[58:06] everything came out. We solved the
[58:07] problem in time and uh that's the way it
[58:10] worked. I would like to uh tell you just
[58:13] a few words about some of the people
[58:14] that I met. I was an underling at the
[58:16] beginning. I became a group leader, but
[58:18] I met some very great men beside the men
[58:20] on the evaluation committee. The men
[58:22] that I met in Los Al. There's so many of
[58:24] them that it's just one of my great
[58:26] experiences in life was to have met all
[58:28] these wonderful physicists men that I
[58:30] had heard of smaller and larger but the
[58:32] greatest ones were there also there was
[58:34] of course fairmy he came down once the
[58:38] first time that he came he came from
[58:39] Chicago to consult a little bit to help
[58:41] us if we had some problems and we had a
[58:43] meeting with him and I'd been doing some
[58:45] calculations and gotten some results and
[58:48] the calculation was so elaborate it was
[58:49] very difficult now usually I was the
[58:51] expert at this I could always tell you
[58:52] what the answer was going to look black
[58:54] when I got it I could explain why but
[58:56] this thing was so complicated I couldn't
[58:57] explain why it was like that so I said
[59:01] to fmy that I was doing this problem and
[59:02] I started to calculate to get he said
[59:04] before you tell me the result let me
[59:05] think it's going to come out like this
[59:08] he's right and it's going to come out
[59:10] like this because of so- and so and it's
[59:12] a perfectly obvious explanation of the
[59:13] thing that I so he was doing what I was
[59:15] supposed to be good at 10 times better
[59:18] so that was quite a lesson for me then
[59:21] there was van who was the great
[59:22] mathematician He suggested I won't go
[59:24] into the things here, some very clever
[59:26] technical observations. We had some very
[59:28] interesting phenomena in the computing
[59:30] the numbers uh the problem looked as if
[59:32] it was unstable and he explained why and
[59:34] so forth. He was very good technical
[59:36] advice. We used to go for walks often to
[59:39] get rest like on Sunday or something. We
[59:40] walk in the canyons in the neighborhood.
[59:42] We'd often walk with Bayer and Vonman
[59:44] Baka. It was great pleasure. And the one
[59:46] thing that von Noman gave me was an idea
[59:48] that he had which is interesting that
[59:50] you don't have to be responsible for the
[59:52] world that you're in. And so I have
[59:54] developed a very powerful sense of
[59:56] social irresponsibility as a result of
[59:58] van advice and I made me a very happy
[60:01] man since I but it was van who put the
[60:05] seed in which grew now into my active
[60:09] irresponsibility. I also met Neil's
[60:12] boore that was interesting. He came
[60:14] down. His name was Nicholas Baker in
[60:15] those days and he came with Jim Beggar
[60:17] his son whose name is really Oabore.
[60:19] They came from Denmark and they came to
[60:21] visit and there were very famous
[60:23] physicists as you all know and all the
[60:25] big shot guys to him there was even the
[60:28] great god. So they were listening to
[60:30] him. So he was talking about things and
[60:32] we were at a meeting and everybody
[60:33] wanted to see the great war. So there
[60:34] were a lot of people and I was back in
[60:36] the corner somewhere and they discussed
[60:38] the problems of the bomb.
[60:40] That was the first time he came in. He
[60:42] went away and all I could see him from
[60:43] between somebody's heads from the
[60:45] corner. Next time he's due to come in
[60:48] the morning of the day he's due to come.
[60:50] I got a telephone call. Hello Fineman.
[60:53] Yes. This is Jim Baker. It's his son. My
[60:57] father and I would like to speak to you.
[61:00] Me? I'm Fineman. I'm just That's right.
[61:03] Okay. So 8 o'clock in the morning before
[61:05] anybody right. I go down to the place.
[61:07] If you go into an office in a technical
[61:08] area and we started this stuff, he says,
[61:11] "We have been thinking how we could make
[61:13] the bomb more efficient and we think of
[61:15] the following idea." Said, "No, it ain't
[61:16] going to work. It's not good. It's not I
[61:17] don't this and that." He said, "How
[61:19] about so and so?" I said, "That sounds a
[61:21] little bit better." No, but I've got
[61:22] this damn full idea in it. So forth,
[61:25] back and forth. I was always dumb about
[61:27] one thing. I never knew who I was
[61:28] talking to. I would only worry about the
[61:30] physics. If the idea looked lousy, I
[61:31] said it looked lousy. If it looked good,
[61:32] I said it looked good. Simple
[61:34] proposition. I've always lived that way.
[61:35] It's nice. It's pleasant. If you can do
[61:36] it, I'm lucky. Just as lucky as I am
[61:38] with that blueprint, I'm lucky in my
[61:40] life that I can do that. So after this
[61:43] went on for about two hours of going
[61:45] back and forth over lots of ideas at the
[61:47] time back and forth
[61:49] arguing the great Neil's always lighting
[61:51] his pipe perpetually. It always went out
[61:55] and he talked in a way that was
[61:59] ununderstable. Hard to understand. I
[62:00] couldn't understand him. But his son, I
[62:02] could understand
[62:03] better. Finally, he said, "Well," he
[62:06] says, "Light lighting his pipe. I guess
[62:08] we can call in the big shots now." So,
[62:10] then they called all the other guys and
[62:12] had a discussion with them. And his son
[62:13] told me that what happened was the last
[62:16] time he was there, he said to his son,
[62:17] "Remember the name of that little fell
[62:19] in the back over there. He's the only
[62:21] guy that's not afraid of me." And we'll
[62:24] say when I got a crazy idea. So, next
[62:27] time when we want to discuss ideas,
[62:28] you're not going to be able to do it
[62:29] with these guys who all say everything
[62:31] is yes. Yes. Dr. Moore, get that guy
[62:34] first. We'll talk with him
[62:36] first. So that was
[62:44] it. Well, it's almost finished.
[62:47] Uh the next thing that happened was of
[62:49] course the test after we' made the
[62:51] calculations. We had to make the test. I
[62:53] was actually at home on a short vacation
[62:56] at that time because I guess because my
[62:58] wife died and so I got a message that
[63:01] said the baby was expected on such and
[63:03] such a day. So, I flew back and I just
[63:05] arrived on the site while the buses were
[63:07] leaving. I couldn't even get to my room
[63:09] and I got out to the site and we waited
[63:11] out there at the distance. We were 20 mi
[63:14] away and we had a radio and they were
[63:16] supposed to tell us when the thing was
[63:17] going to go off and so forth. The radio
[63:19] wouldn't work and we never knew what was
[63:21] happening. But just a few minutes before
[63:23] it was supposed to go off, the radio
[63:24] started to work and they told us there
[63:25] was 20 seconds or something to go. The
[63:28] people that were far away like we were.
[63:30] The others were closer, six miles away.
[63:33] And uh they gave out dark glasses that
[63:36] said you could watch us dark glasses.
[63:38] We're 20 miles away from the damn thing
[63:40] because it's dark
[63:41] glasses. I said you're going to see a
[63:43] damn thing through dark
[63:45] glasses. And so what I figured ultra the
[63:48] only thing that would really hurt your
[63:49] eye, bright light never can hurt your
[63:50] eye. It's ultraviolet light that does.
[63:52] So I got behind a truck windshield so
[63:55] that the ultraviolet can't go through
[63:56] glass and that would be safe and so I
[63:58] could see the damn thing. Other people
[63:59] were never going to see the damn thing.
[64:01] Okay, time comes and this tremendous
[64:04] flash out there. So bright. I quickly my
[64:08] head goes like this because it was so
[64:09] bright. And I see this purple splotch on
[64:12] the floor of the truck. I says, "That
[64:14] ain't
[64:15] it. That's an after image." So I turn
[64:17] back up and I see this this white light
[64:20] changing into yellow and then into
[64:21] orange. The clouds form and then they
[64:24] disappear again. That the compression
[64:26] and the expansion forms and makes clouds
[64:28] disappear. And I saw and then finally a
[64:31] big ball of uh orange. It become a
[64:34] center that was so bright became a ball
[64:35] of orange. It started to rise and billow
[64:38] a little bit and get a little bit black
[64:39] around the edges. And then you see it's
[64:41] a big ball of smoke with flashes on the
[64:43] inside of the fire going out. The heat
[64:45] and I saw all that and all this that
[64:47] I've just described in just a moment
[64:49] took about one minute. It was a series
[64:51] from bright to dark. And I had seen it.
[64:53] I'm about the only guy in the world who
[64:54] actually looked at the damn thing. The
[64:56] first Trinity test. Everybody else had
[64:58] dark glasses. The people at six miles
[65:00] couldn't see it because they were all
[65:01] told to lie on the floor with their eyes
[65:03] like this. So, nobody saw it. And the
[65:05] guys up there where I was, they all had
[65:07] dark glasses. I'm the only guy I saw
[65:08] with the human
[65:11] eye. Finally, after about a minute and a
[65:14] half, suddenly there's a tremendous
[65:16] noise.
[65:17] Bang. And then rumbles like thunder. And
[65:19] that's what convinced me. Nobody had
[65:21] said a word during this whole minute.
[65:22] We're all just watching quietly. But
[65:24] this sound released everybody. Reed me
[65:26] particularly because the solidity of the
[65:28] sound at that distance meant that it
[65:30] really worked. Uh the man who was
[65:33] standing next to me said when the sound
[65:35] off, what's that? I said that was the
[65:39] bomb. The man was William Lawrence who
[65:41] had come. He was going to write an
[65:43] article that was going to describe the
[65:44] whole situation. I had been the one who
[65:46] was supposed to have taken him around.
[65:48] Uh it was found that it was too
[65:49] technical for him. And so later Mr.
[65:51] Smite came around and I showed Mr. Smite
[65:54] around.
[65:55] When I was showing Mr. Smite around, if
[65:57] we went into a room and there on the end
[66:00] of a pedestal, a little narrower than
[66:02] that, was a small ball about so big,
[66:05] silverplated, you could put your hand on
[66:07] it was warm. It was radioactive. It was
[66:10] plutonium. And we stood at the door of
[66:12] this room talking about it. There was a
[66:15] new element that was made by man that
[66:17] had never existed on the earth before
[66:18] except for a very short period possibly
[66:20] at the very beginning. And here it was
[66:22] all isolated, radioactive and had these
[66:24] properties that we had made it. And so
[66:27] it was tremendously valuable. There's
[66:28] nothing more valuable and so forth and
[66:30] so on. Meanwhile, these sort of, you
[66:33] know, how people do when you talk, you
[66:34] kind of dibble around and jiggle and so
[66:36] forth. He's kicking the door stop, you
[66:38] see. I says, "Yes." And I said, "And the
[66:40] doors stop is appropriate to the door."
[66:42] The doors stop was a hemisphere,
[66:44] yellowish metal, gold. As a matter of
[66:46] fact, it was a gold hemisphere about so
[66:47] big. What had happened was we needed to
[66:50] do experiments to see how many neutrons
[66:52] were reflected by different materials in
[66:54] order to save the neutrons. So we didn't
[66:55] use so much material. And so we had
[66:57] tested many different materials. We had
[66:59] tested platinum, we had tested zinc, we
[67:01] tested brass, we tested gold. And so in
[67:03] making the tests with the gold, we had
[67:05] these pieces of gold. And somebody had
[67:07] the clever idea to use that great ball
[67:09] of gold for a doors stop for the door
[67:12] that contained the plutonium, which is
[67:14] quite
[67:15] appropriate. After the thing went off,
[67:18] there was tremendous excitement at Los
[67:19] Alamos. Everybody had parties. We all
[67:22] ran around. I sat on the end of a jeep
[67:24] and beat drums and so on except for one
[67:26] man that I remember. His name was Bob
[67:28] Wilson who got me into it in the first
[67:29] place. He's sitting there moping. He
[67:31] said, "What are you moping about?" He
[67:33] says, "This terrible thing we made." I
[67:34] said, "But you started. You got us into
[67:36] it." You see what happened to me or what
[67:38] happened to the rest of us is we started
[67:40] it for a good reason but then we're
[67:42] working very hard to do something and to
[67:44] accomplish it is a pleasure is
[67:46] excitement and you don't stop to think
[67:48] any you know you just stop after you
[67:50] thought at the beginning you stop
[67:51] thinking so he was the only one who was
[67:53] still thinking about it at that
[67:54] particular moment I returned to
[67:56] civilization shortly after that and went
[67:58] to Cornell to
[68:00] teach and my first impression was a very
[68:04] strange one and I can't understand
[68:05] anymore more. But I felt very strongly
[68:07] then I'd sat in a restaurant in New
[68:09] York, for example, and I looked out at
[68:11] the buildings and how far away I would
[68:13] think, you know, how much the radius of
[68:14] the Hiroshima bomb damage was and so on.
[68:17] How far down there was down at 34th
[68:19] Street, all those buildings all smashed
[68:21] up and so on. And I got a very strange
[68:24] thing. I would go along and I'd see
[68:25] people building a bridge or they're
[68:27] making a new road. And I thought,
[68:29] they're crazy. They just don't
[68:31] understand. They don't understand. Why
[68:33] are they making new things? It's
[68:34] useless.
[68:36] But fortunately, it's been useless
[68:39] for 30 years now, wasn't it? Almost.
[68:42] Maybe we'll make 30 years. So, I've been
[68:44] wrong for 30 years about it being
[68:46] useless to make bridges. And I'm glad
[68:47] that those other people were able to go
[68:49] ahead. But my first reaction after I was
[68:51] finished with this thing was it's
[68:53] useless to make anything. Thank you very
[68:55] much.
[69:09] Yes. I wonder a story about some safes.
[69:16] Yes. Well, so what's the question? Could
[69:19] you tell us that story? Well, there's a
[69:21] lot of stories about safes. If you give
[69:22] me 10 minutes, I'll tell you three
[69:24] stories about safes. All right.
[69:32] All right. Now, I'll tell you some more
[69:33] stories about safe. The fact that I was
[69:36] able to pick the locks uh interested me
[69:39] in the safety of the whole thing.
[69:41] Somebody had taught me how to pick
[69:42] locks. And then they got filing cabinets
[69:45] which had safe combinations. One of my
[69:47] diseases, one of my things in life is
[69:48] anything that's secret I try to undo.
[69:51] And so those filing cabinets were made
[69:53] by the Mosula lock company in which we
[69:55] put the documents after that and
[69:57] everybody had them. They represented a
[69:59] challenge to me how the hell to open
[70:01] them. So I worked on them and I worked
[70:04] on them. There's all kinds of stories
[70:05] about how you can feel the numbers and
[70:06] listen to things and so on. That's true
[70:08] and I understand it very well for
[70:10] oldfashioned safes. They had a new
[70:12] design so that nothing would be pushing
[70:14] against the wheels while you were trying
[70:15] them. I won't go into the technical
[70:16] details but none of the old method would
[70:18] work.
[70:19] I read books by locksmiths. Books by
[70:21] locksmiths always say in the beginning
[70:23] how the safe's under water and a woman
[70:24] is drowning or something. I think they
[70:26] open the safe. I don't know a crazy
[70:27] story. And then in the back they tell
[70:29] you how they do it and they don't tell
[70:30] you anything sensible. It doesn't sound
[70:31] like they could really open safes that
[70:33] way. Like guess the combination on the
[70:35] basis of the psychology of the person
[70:37] who owns
[70:39] it. So I always figured I'd keep it a
[70:41] secret. Anyway, I kept working and so
[70:43] like a kind of a disease I kept working
[70:45] on these things until I found out a few
[70:47] things. First, I found out how big a
[70:49] range you need to open the combination,
[70:50] how close you have to be. And then I
[70:52] invented a system by which you could try
[70:54] all the combinations that you'd have to
[70:56] try. 8,000 it turned out because you
[70:58] could be within two of every number.
[70:59] Then it turns out that's every fifth
[71:02] number out of 100 and 20* 20 8,000
[71:05] combinations. And I worked out a scheme
[71:07] by which I could try numbers without
[71:08] altering a number that I already set by
[71:10] correctly moving the wheels so that I
[71:12] could do it in eight hours to try all
[71:14] the combinations.
[71:16] And then I discovered still fur. This
[71:17] took me about two years of researching
[71:19] because I had nothing to do up there.
[71:20] You see, and I was fiddling. Finally, I
[71:23] discovered a way that it's easy to take
[71:25] the last two numbers of the combination
[71:26] off the safe if the safe is open. If the
[71:29] draw is pulled out and you can turn the
[71:31] number and see the bolt go up and play
[71:32] around and find out what makes it what
[71:34] number it comes back at and stuff like
[71:35] that. With a little trickery, you can
[71:37] get the combination off. So, I used to
[71:39] practice it like a card shop practices
[71:41] cards, you know, all the time, all the
[71:43] time. all the time, quicker and quicker,
[71:45] more and more unobtrusively. I would
[71:47] come in and I talk to some guy and I
[71:48] kind of lean against his filing cabinet
[71:49] just like I'm playing with this watch
[71:50] now. You wouldn't even notice I'm doing
[71:52] anything. I'm not doing anything. I just
[71:54] play with the dial. That's all. Just
[71:55] play with the dial. But I was taking the
[71:56] two numbers off and I go back to my
[71:59] office and I write the numbers
[72:01] down. The last two numbers of the three.
[72:03] Now, if you have the last two numbers,
[72:05] it takes a minute to try the first
[72:06] number. There's only 20 possibilities
[72:08] and it's open. Okay? It takes three
[72:09] minutes to open a safe if you know the
[72:11] last two numbers. So, I got an excellent
[72:13] reputation for safe cracking. They would
[72:15] say to me, "Mr. Schmaltz is out of town.
[72:18] We need a document from his safe. Can
[72:19] you open
[72:21] it?" I'd say, "Yes, I can open it. I
[72:23] have to get my tools. I don't need any
[72:25] tools." I'd go to my office and I'd look
[72:28] at the number of his safe. I had the
[72:29] last two numbers. I had everybody's safe
[72:31] in my office. I put a screwdriver in my
[72:33] back pocket cuz I would account for the
[72:36] tool I claimed I needed. I'd go back to
[72:38] the room and I would close the door with
[72:42] the attitude is that this business about
[72:44] how you open safe is not something that
[72:45] everybody should know because it's very,
[72:47] you know, it makes everything unsafe is
[72:48] very dangerous to have everybody know
[72:50] how to do this. So I close the door and
[72:53] I sit down and read a magazine or do
[72:55] something. I'd average about 20 minutes
[72:58] of doing nothing and then I'd open it,
[73:00] you see, or I open it right away, make
[73:01] sure everything was all right and I'd
[73:02] sit there for 20 minutes to give myself
[73:04] a good reputation that it wasn't too
[73:05] easy. There was no trick to it. No trick
[73:07] to it. And then I come out, you know,
[73:09] sweating a bit. It's
[73:11] open. There you are. And so forth.
[73:14] Okay. I also at one particular moment I
[73:17] did open a safe purely by accident and
[73:19] that helped to enforce my reputation. It
[73:21] was a sensation. It was pure luck. The
[73:22] same kind of luck I had with the
[73:24] blueprints. But uh after the war was
[73:27] over, I went back to Los Alamos to
[73:29] finish some papers. And there I did some
[73:31] safe opening which I could write a safe
[73:33] cracker book better than any safe
[73:35] cracker book. It would start in the
[73:36] beginning and it would explain how I
[73:38] opened the safe absolutely cold without
[73:39] knowing the combination which contained
[73:41] more secret thing than any safe has ever
[73:44] been opened. I opened the safe that
[73:45] contained behind it the secret of the
[73:47] atomic bomb. All the secrets, the
[73:49] formulas, the rates at which neutrons
[73:51] are liberated from uranium, how much
[73:52] uranium you need to make a bomb, all the
[73:53] theories, all the calculations, the
[73:55] whole damn
[73:57] thing. This is the way it was done. All
[74:01] right. I was trying to write this
[74:03] report. I needed this report. It was a
[74:05] Saturday. I thought everybody worked. I
[74:07] thought it was like Los Alamos used to
[74:09] be. So I got down to the library. The
[74:11] library at Los Alamos had all these
[74:12] documents. And there was a great vault
[74:14] with a great knob of a different kind of
[74:16] knob. I didn't know anything about a
[74:18] filing cabinets. I understood, but I was
[74:19] only an expert on filing cabinets. Not
[74:21] only that, but there were guards walking
[74:23] back and forth in front with guns. So
[74:25] you can't get that one open. Okay. So I
[74:28] didn't get that. But I think,
[74:30] wait, old Freddy de Hoffman in the
[74:32] declassification section. He's in charge
[74:34] of declassifying document. Which
[74:36] documents now can be declassified? And
[74:38] so he had to run down to the library and
[74:39] back so often. He got tired of it. He
[74:41] got a brilliant idea. Did he get a copy
[74:44] made of every document in the Los Alamos
[74:46] library? Right. And he stick it in his
[74:47] filing cabinet and he had nine filing
[74:50] cabinets, one right next to the other in
[74:51] two rooms full of all the documents of
[74:54] Los Alamos. And I knew he had that. So
[74:56] I'll go up to the hop. I'll ask him to
[74:58] borrow the document from him. He's got a
[74:59] copy. So I went up to his office and the
[75:02] office door is open and it looks like
[75:04] he's coming back. The light is lit.
[75:05] Looks like he's coming back any minute.
[75:07] So I wait and as always when I'm waiting
[75:10] I didd it
[75:11] thems 20 30 didn't work. 20 40 60 didn't
[75:14] work. Try everything. I'm waiting.
[75:18] Nothing to do. Then I begin to think you
[75:22] know those locksmith people. I ain't
[75:25] never been able to figure out how to
[75:27] open them cleverly. Maybe they don't
[75:29] either. Maybe all the stuff they're
[75:30] telling me about psychology is right.
[75:33] I'm going to open this one by
[75:35] psychology. First thing, first thing in
[75:38] the book says the secretary is very
[75:40] often nervous that she will forget the
[75:42] combination and she's been told the
[75:43] combination she might forget and the
[75:45] boss might forget. She has to know. So
[75:47] she nervously writes it somewhere. List
[75:49] of places where secretaries write
[75:51] combinations. Okay. Starts out with, you
[75:53] know, the most clever thing. starts
[75:55] right out with you open the drawer and
[75:56] the wood along the side of the drawer on
[75:58] the outside is written carelessly a
[75:59] number like as if it was a invoice
[76:00] number. That's the combination number.
[76:03] So it's on the side of the desk. Okay. I
[76:06] I remember that in the book. Desk draws
[76:08] locked. I can pick the lock easy. I open
[76:09] the lock right away. Pull out the
[76:11] drawer. Look along the wood. Nothing.
[76:13] It's all right. Sorry. There's a lot of
[76:15] papers in the drawer. I fish around
[76:17] among the papers and finally I find
[76:19] there's a nice little piece of paper
[76:21] which is the Greek alphabet. Alpha,
[76:22] beta, gamma, delta, and so forth.
[76:23] carefully printed. The secretaries have
[76:25] to know how to make those letters and
[76:26] how to call them when they're talking
[76:28] about them. Right? So, they all had each
[76:30] one had a copy of this thing. But
[76:33] carelessly scrolled across the top is pi
[76:37] is equal to
[76:40] 3.14159. Well, why does she need the
[76:42] numerical value of pi? She's not
[76:44] computing anything. So, I go up to the
[76:47] safe. It's honest. It's like any other
[76:48] book, only I'm just telling you how it
[76:49] was done. I walk up to say 3141 59
[76:52] doesn't open a 134 95 doesn't open 95
[76:57] 1413 doesn't open 4131 I 20 minutes I'm
[77:00] turning pie upside down nothing
[77:04] happened so I start walking out of the
[77:06] office and I remember the book about the
[77:08] psychology and I said you know but it's
[77:10] true psychologically I'm right Freddy de
[77:14] Hoffman is just the kind of a guy to use
[77:16] a mathematical constant for his faith
[77:18] combination
[77:20] So the other important mathematical
[77:21] constant is E. So I walk back to the
[77:24] safes. 271 1828. Click. It
[77:32] opens. I checked, by the way, that all
[77:35] the combinations were the
[77:40] same. Well, there's a lot of stories
[77:42] about it, but it's getting late. And
[77:44] that's a good one. So we'll let it go at
[77:46] that. Heat. Heat.
