Feynman was not only a super genious (yes, my favoite scientist ever), he was also a notorious prankster. The man lived life to the fullest, all the while.making beautiful contributions to science as well as to his students and colleages. Amazing man. I think this guy sells him a little short. He was a tremendous collaborator and manager of people as his duties on the Manhattan Project reveal, and the young unknown guy on the team almost immeditely drew respect from many of the most brilliant and highly recognized men in their fields (and many famous). That says a lot!
But he was often pranking people who didn't deserve it.. like waiters or other people who were just doing a job serving him. that made him deeply flawed as a human being.
That's kind of a common thing with people in charge. They will get angry at the person pointing out a problem and not the problem itself or the person who caused it initially. I think it has something to do with them now being required to take action, since now that they have been made aware of the issue they are expected to deal with it. Until someone pointed it out, they didn't need to do anything. Now they need to invest time, money and work because they must fear/expect that their superiors might also hear of the problem and will hold them responsible for not doing anything about it. So they get mad at the messenger for making them work more.
@@ziadkhayat7299 - I don't remember exactly. But he enjoyed puzzles, and I think for him this was a diversion. I thought he only pointed out the problem after discovering how many of the safes were not modified from their factory settings.
In combination in most modern schools, you cannot set the combination, but you can easily find the first digit due to a flaw in the lock (apply pressure while rotating dile. Add five to the one number it gets stuck on.) then from there, after determining whether it is even or odd, you know both the second and third numbers are either even or odd (but all even, or all odd.) If there are 40 numbers, that gives you 400 combinations. But the third number doesn't require trying a new number everytime. Just set the first number, your guess for the second, then spin it around trying all numbers for the third one. Now, this is more like 20 different options, but doesn't require seeing the lock before breaking in. I have never used this technique for bad (nor should anyone) but I find it handy when I see that someone has stuck a lock in an unnecessary/obnoxious spot. Or if I find a lock, I can learn the combination then give it to a friend who needs one.
Per his autobiography, it is my understanding that Feynman also reduced the permutations by merging the rotation for the third combination with the beginning of the rotation of the first combination of the next number tried. Since the dials required Right-Left-Right (or Left-Right-Left) turning, the first and third numbers were approached from the same rotation direction. Also the first number required the dial to rotate over 360 degrees. The third number's rotation was less than 360 degrees. Feynman deduced that if he got the first two numbers correct, ignored the the third number entirely, and began spinning the dial for the first number of the next attempt, the lock would make a distinctive 'Click' as it passed over the correct third number. So with a lock dial with 100 positions per turn of which he had already reduced to 20 positions per turn, he could cracking the lock in no more than 20 x 20 = 400 attempts (plus one extra spin) without any social engineering required. So at 5 seconds per turn, in a worse case scenario, he could crack the safe in about 2000 seconds (-32 min) and on average, -16 min).
@@JaakkoF feel free to read his autobiography "Surely You're Joking, Mr. Feynman!". Further if you ever had the traditional style MasterLock combination padlock, you can easily emulate the same technique. The dial had enough slop that the 60 digits could be covered by just 12. As a kid, we used to open them in a few minutes.
Close... What he did report doing was retaining the setting for the first two numbers while trying all the possibilities for the third, then nudging the second wheel just slightly to the next number along. This saves time because you don't have to dial the entire combination on each attempt. The same trick works on the position of the first wheel. You can either memorize where you've set the first two wheels, or feel the slight increase in friction as the extra wheel gets picked up by the one behind it. I've tried all this, and it works.
As I understand it, when a cabinet was ALREADY OPEN he was able to use a "by touch" method of deriving the final two numbers in the combination as he seemingly twiddled the dial casually while chatting in the office - he then wrote down those numbers in a notebook or something, so he had the last two numbers on record! I could have that wrong slightly!
The simplest trick he used is that some people didn't change the factory combinations, which were just a few. He started by trying them, and they often worked. Also, when he found some people left the cabinets open, and he could use the information, he advised them not to leave them open. That triggered a memo asking people not to leave the cabinets open when Feynman was around :-D
Amazing how a tiny mechanical flaw can make it that much easier. I remember I used to crack open my friend's lockers at school , leave them in awe. Really all you had to do was if it wasn't cleared you could slowly turn it backwards (so if RLR opens then LRL). If you focused you could feel when the pins collided, which would give away the number for that dial.
Thanks for posting this video. I am not a scientist,or mathematician, but I became interested in Feynman from seeing him interviewed on programs such as NOVA. I didn't realize that an uncle of mine taught him in school, until after both Feynman in my uncle, Abram Bader, had died.
This is my favourite sixty symbol scientist. He is the most articulate and efficient of all the sixty symbol profs. Please do videos with him more often
Loved reading the books about Feynman and watching interviews with him and his lectures. And look at that face, you can just see the mischief lurking in there.
If you enjoy stuff like this(bypassing security) you will almost certainly like the lecture on UA-cam called "I'll Let Myself In: Tactics of Physical Pen Testers". I should probably warn everyone how unsettling it is that seemingly secure things are quite often not secure at all. Watch at your own risk of anxiety.
I didn't get the part where you said "people leave their safe open and there's a drawer in the bottom and he fiddled and remembered the last two number" etc etc, how did that work?
As I understand it, he could feel the mechanism inside the safe and tell when the first two numbers lined up. So while talking, slowly spin the knob until the first number engaged, remember the number. Then spin the knob until the second number engaged, remember the number.
There is a CD of an entertaining talk ("Los Alamos from Below") that Feynman gave about his time at Los Alamos. I got it with the book "Classic Feynman." It goes through his whole story and ends with some safecracking stories. I would recommend both the book and the talk.
I was just reading Feynman's book "Surely you're joking Mr. Feynman", and just finished the chapter where he was describing his passion for safe-breaking. I put the book aside and entered my UA-cam account, and lo and behold - a new numberphile's video about.. now, what's the odds for that? :) Amazing book by the way, highly recommended to anyone who would like to have a glimpse into the mind of this fascinating man.
On my second ship in the Navy, my office was a classified materials vault. Whenever I went in to do some work, I would lock the door behind me both for privacy and to prevent anyone from doing Feynman's "Open safe" trick.
I have never been able to work these locks very well , and I have lived in USA for 15 years , they are everywhere . I now feel more confident in tackling them again , without the bolt cutter or grinder
I didn't understand the last one, what do you mean he fiddled with the knobs when the safe was open: 1) on your video it says 2 of the numbers became known, how so? 2) were the correct numbers in position when you opened the safe?
I read "Surely you're joking Mr. Feynman", it had stories about his life and growing up. It also included Los Alamos and the safe-cracking business; the way he wrote it was simply hilarious. I recommend it. :)
It's in that Feynman book but I can remember what it was. It did something like the bolt moved if it was already open on the last digit and then you dialed the previous digit.
Several years back I managed to crack an inexpensive Sentry safe just by starting somewhere in the middle and brute forcing every combo in increments of 4 or 5, just like described here. I reasoned that the last number didn't really matter because I could just repeatedly yank on the handle while rapidly turning the dial. This means I really only had to try about 20^2 permutations max (and in practice, only about two-thirds that number). It only took about 10 minutes to get it open.
He's using birthdays; the first digit has to be between 0 and 30 (for the days of the month), the second will be 1 - 12 (months) and the last could be easily guestimated by looking at or knowing the safe-owner. And as the safes had quite a margin for error (as explained), that leaves few possibilities.
"By fiddling with his own safe he learned that when a door was open he could find the last number of a combination by turning the dial and feeling when the bolt came down. Given some time, he could find the second number that way, too. He made a habit of absently leaning against his colleagues' safes when he visited their offices, twirling the dials like the perpetual fidgeter he was, and thus built up a master list of partial combinations."
He also tweaked a pompous colonel in Oak Ridge who claimed to have the most secure safe in the country; it turned out to be a large armored cabinet with the same Mosler lock on it. He could get the contents out of any key-locked desk by getting under it and and reaching into drawers that were closed and locked from the outside. He was the bane of the security people at Los Alamos.
Duh-doy! :P And Feynman, being the ultimate simplifier in physics, (Feynman diagrams are a beautiful way of representing sub-atomic particles), his ability to distill problems down to it's simplest concepts, really allowed him to achieve much. He and Einstein had the same gift of seeing complex things in simple ways. Everyone should read Feynman's books, and a few of his biographies written by others, his life was eventful, and the love story with his wife made me shed tears
The logic behind the US format of dates is just from how we say the actual date over here. Is spoken language, in the US, we say "September 18, 2005" instead of "The 18th of September, 2005" so it just makes perfect sense to us over here.
+yohanes ronald To open the door, the three tumblers will need to line up so the latch can slide back into a hole. I'm guessing that this particular lock would prevent the dial from spinning freely. Perhaps it could only turn around between the position of the 2nd and 3rd numbers. Or if the dial could spin the whole way around, it probably felt different.
yeah he doesn't explain that but from the book Feynman found that you could pick the first two numbers off an OPEN safe by turning the dial both ways in in increments and testing the bolt of the safe (pushing it) until it would no longer retract.
When the safe was open, you could see the bolt that holds the door closed in the "open" position. By spinning the wheel, the bolt would eventually return to the "closed" position. The last number before the bolt "closed" would be the correct number. Then you would rotate the wheel in the other direction to get the 2nd number. On a completely unrelated note, I recommend the book "Surely You're Joking, Mr. Feynman!"
Can anyone explain that last devious trick he did, didnt quite understand :/ (The part where he would fiddle with an already open safe) What Im confused about is how the last 2 numbers would be know (and how) but not the first one.. Thanks!
I realized the construction tolerances on padlocks when I was assigned a locker in middle school. That, by far, is the main problem with padlocks and safes.
One of my favourite stories was when he was at MIT (or maybe Princeton). External lecturers would come in to give seminars and he used to sit in like all the other students ... except he would break into the lecture hall before and read the lecturers notes and more importantly, the conclusion. Then at the end of the seminar, just before the guest speaker summed up, Feynman would raise his hand and say things like: "But doesn't that mean we can conclude ... X" As if he figured it out right then.
The locker combinations today when they are opened, show you the last digit. Even if you don't know the last you can save time and do them in about 10 minutes by knowing the internal wheels. eg: spin + to 20, around once and - to 15, then + to 25 but then you can just do 30,35,40,45,50,55,00,05,10 all in a few seconds without doing the 1'st 2 again. Then set only the 2'nd wheel by - to 10. I used to open them driving home, just in the time stopped at red lights.
I do work which require me handling a lot of locks and keys (many buildings, many employees both permanent and temporary). Every time I encounter a new lock system I like to spend some down time figuring out how it works. Comes in handy and I got a bit of a reputation at this point at work for being able to get most things open if needed, to the point where they call me in instead of calling in a locksmith. Most of the time I just use a screwdriver/piece of plastic/whatever to force open things or unscrew something and it will open, most locks weakest points are the surroundings they are placed in.
From the stories, the safes used didn't have a lever. You had to set the wheel a 4th time to the number 10, Then it opens. However, he did try spinning it and quickly turning to 10, hoping friction would help
my father had a safe in his office, I taught myself how to crack it by turning the dial and feeling and listening for a slight detent then reverse, it worked on those old ww2 era safes.
Hey Brady where are you guys recording now that Professor Bowley isn't in his office? Does he have another office at the school or is this at his home now?
Not forgetting with almost any combination lock, the LAST tumbler (the third number), all possibilities can usually be tried without re-entering the first two numbers, because the first two tumblers are already set and the last is physically locked to the number dial, this means you really only need to solve for the first two tumblers 'slowly' and the last number is turn,try,turn,try,turn.try...etc. Doesn't work for all locks because they have a randomizer on unsuccessful attempts.
Surely You're Joking, Mr. Feynman! spends a decent amount of time on Richard Feynman's safecracking adventures, but also recounts a number of other events throughout his life, always with a good dose of humor (he was by all accounts a natural-born prankster, and it shows in the book); it makes for a very good read.
I agree about the year dial: you can't assume in 1942 that nobody will use a date in the 19th century. Checking American instead of international dates is the same number of combinations. If you want to check both, it's a little slower but not as slow as you suggest: you don't need to try 8x8 because, e.g., 30/30 isn't a date in either system. So you just need 3x8+5x3 (not even twice 3x8 because, e.g., 3/3 is a date in both systems and you only need to try it once).
Fyi, there is the story told in his own words on youtube... along with a number of other story details... definitely worth the listen (no video) search for "Los Alamos from below".
And with the locks in use in that facility, the final turn generated a pop and stopped the dial.. so if you went 9, then 13, then back to 5, on your way to the five, from the 13, you'd feel the pop and would be hard-pressed to go beyond it... So all you need are the first two numbers really -- then turning back the lock would "give" you the last one.
no, he reaches into the mechanics and then turns the knob and he would feel a metal bar move as he hits the right number. once he has the first number, he again puts his fingers into the mechanics, resets, dials to the first number he now knows, and turns the knob into the other direction, and would again feel something when he hits the second number. and he did this whenever he had a chance to do it.
That is all great assuming that everyone used dates for safe code. How long did it take him to crack one that was setup illogically to his logic? How did he "show off" then? How did he explain, "ooops can't crack it 12 min!"
+Robin Meade worst case 10 hours, that was the 20x20x20 one, which mean 5 hours average. and as they said he wouldn't show off that he did it in 12 minutes, just that he could do it, so he would sit at read for at bit after opening it, just make it look harder.
I have a question. Let's say you're going through every possible combo out of 1,000,000and it takes 5 seconds to try each combo. As you go through them, the more likely it is the next combo will be the right one, as the # of possible combos decrease as previous ones are eliminated, and can continue until you reach the very last one. I know how you could calculate the max or min time it might take, but is there a way to calculate the average amount of time it would take to get it?
... and "What Do You Care What Other People Think?" is pretty good too. From the book it sounded like he (Feynman) didn't have a roommate. He said he borrowed Fuchs car .. didn't know he shared a room (according to Bowley).
I did that in high school. My friend's lock had a ±5 margin on the mechanism. A few minutes at a time over weeks, and one day I opened his locker for him. Good times.
It's only more secure against a brute-force attack (that looks at every combo of characters). It's LESS secure against dictionary attacks which try combinations of real words first, because that's more likely to find your password sooner than if it were non-words.
he could always use the method explained in the video to open any random safe in under 12min. well,not any random safe,it would have to be using a date as the pin, which alot of safes do read his autobiographies if you are interested in the guy,they are soooo goood
If, like they say at the start of the vid, there's only one knob on a normal safe, then how could he possibly remember the last 2 numbers on the knobs? Since you only could only see that LAST number (which was the number it was at when the safe opened?)
Feynman was not only a super genious (yes, my favoite scientist ever), he was also a notorious prankster. The man lived life to the fullest, all the while.making beautiful contributions to science as well as to his students and colleages. Amazing man. I think this guy sells him a little short. He was a tremendous collaborator and manager of people as his duties on the Manhattan Project reveal, and the young unknown guy on the team almost immeditely drew respect from many of the most brilliant and highly recognized men in their fields (and many famous). That says a lot!
But he was often pranking people who didn't deserve it.. like waiters or other people who were just doing a job serving him. that made him deeply flawed as a human being.
@@stijnvandamme76 He seems like a guy that would make up for it after coming clean (the best part of the prank) - like an extra large tip. I do that.
Agree
I wouldn't worry about angry safe enthusiasts. They don't get angry, they tend to keep things locked up inside.
Larry Allen
You beautiful beautiful pun master human being
Eeeeyyyyyyy :D
I simultaneously love and hate that this is the top comment
so punny...lol
They didn't mention that he would tell the various people in charge of this security issue and they would get mad at HIM rather than fix the problem.
That's kind of a common thing with people in charge.
They will get angry at the person pointing out a problem and not the problem itself or the person who caused it initially. I think it has something to do with them now being required to take action, since now that they have been made aware of the issue they are expected to deal with it. Until someone pointed it out, they didn't need to do anything.
Now they need to invest time, money and work because they must fear/expect that their superiors might also hear of the problem and will hold them responsible for not doing anything about it. So they get mad at the messenger for making them work more.
losthor1zon as I recall the only reason he started cracking them was because they wouldn't listen when he reported how bad the security was.
You always shoot the messenger
losthor1zon I live his lectures
@@ziadkhayat7299 - I don't remember exactly. But he enjoyed puzzles, and I think for him this was a diversion. I thought he only pointed out the problem after discovering how many of the safes were not modified from their factory settings.
In combination in most modern schools, you cannot set the combination, but you can easily find the first digit due to a flaw in the lock (apply pressure while rotating dile. Add five to the one number it gets stuck on.) then from there, after determining whether it is even or odd, you know both the second and third numbers are either even or odd (but all even, or all odd.) If there are 40 numbers, that gives you 400 combinations. But the third number doesn't require trying a new number everytime. Just set the first number, your guess for the second, then spin it around trying all numbers for the third one. Now, this is more like 20 different options, but doesn't require seeing the lock before breaking in. I have never used this technique for bad (nor should anyone) but I find it handy when I see that someone has stuck a lock in an unnecessary/obnoxious spot. Or if I find a lock, I can learn the combination then give it to a friend who needs one.
+SwatTeamExit Well, yeah! But for those that do...
My favorite story is when he pulled the General's top secret filing cabinet away from the wall and unscrewed the back off of it.
Per his autobiography, it is my understanding that Feynman also reduced the permutations by merging the rotation for the third combination with the beginning of the rotation of the first combination of the next number tried.
Since the dials required Right-Left-Right (or Left-Right-Left) turning, the first and third numbers were approached from the same rotation direction.
Also the first number required the dial to rotate over 360 degrees. The third number's rotation was less than 360 degrees. Feynman deduced that if he got the first two numbers correct, ignored the the third number entirely, and began spinning the dial for the first number of the next attempt, the lock would make a distinctive 'Click' as it passed over the correct third number.
So with a lock dial with 100 positions per turn of which he had already reduced to 20 positions per turn, he could cracking the lock in no more than 20 x 20 = 400 attempts (plus one extra spin) without any social engineering required. So at 5 seconds per turn, in a worse case scenario, he could crack the safe in about 2000 seconds (-32 min) and on average, -16 min).
All of that is bollocks if you know how a combination lock operates.
they don't make em like they used to
@@JaakkoF feel free to read his autobiography "Surely You're Joking, Mr. Feynman!". Further if you ever had the traditional style MasterLock combination padlock, you can easily emulate the same technique. The dial had enough slop that the 60 digits could be covered by just 12. As a kid, we used to open them in a few minutes.
Close... What he did report doing was retaining the setting for the first two numbers while trying all the possibilities for the third, then nudging the second wheel just slightly to the next number along. This saves time because you don't have to dial the entire combination on each attempt. The same trick works on the position of the first wheel. You can either memorize where you've set the first two wheels, or feel the slight increase in friction as the extra wheel gets picked up by the one behind it. I've tried all this, and it works.
@@AlanCanon2222 yes, that is correct. It's been over 30+ years since I read the autobiography. Thanks for clarifying the details.
As I understand it, when a cabinet was ALREADY OPEN he was able to use a "by touch" method of deriving the final two numbers in the combination as he seemingly twiddled the dial casually while chatting in the office - he then wrote down those numbers in a notebook or something, so he had the last two numbers on record!
I could have that wrong slightly!
The simplest trick he used is that some people didn't change the factory combinations, which were just a few. He started by trying them, and they often worked. Also, when he found some people left the cabinets open, and he could use the information, he advised them not to leave them open. That triggered a memo asking people not to leave the cabinets open when Feynman was around :-D
Amazing how a tiny mechanical flaw can make it that much easier. I remember I used to crack open my friend's lockers at school , leave them in awe. Really all you had to do was if it wasn't cleared you could slowly turn it backwards (so if RLR opens then LRL). If you focused you could feel when the pins collided, which would give away the number for that dial.
I love Feynman. His lectures, his life story, its all very inspiring. A great man, and an inspiring one. Always interesting.
Thanks for posting this video. I am not a scientist,or mathematician, but I became interested in Feynman from seeing him interviewed on programs such as NOVA. I didn't realize that an uncle of mine taught him in school, until after both Feynman in my uncle, Abram Bader, had died.
I've been reading "Surely You're Joking, Mr. Feynman!" Awesome read, and I definitely recommend it to all audiences!
Surely you're joking.
I see what you did there!
I'm not joking and don't call me Shirley
+Ryan Freedman This is actually a wrong reference.
+indefinity I know
Absolutely not! I'll have my tea with both lemon and milk
Richard Feynman is one of my favorites. In happy he has been getting more recognition lately on UA-cam videos so other people can appreciate him.
they exist - I spent much of the morning reading about them!
Sweet, time to find rich guys with ancient safes!
Feynman was an amazing human.
This is my favourite sixty symbol scientist. He is the most articulate and efficient of all the sixty symbol profs. Please do videos with him more often
Love that you used those pots to demonstrate turning the dials on this video. Very simple substitution for a dial and they just look good.
"I never commit to memory anything that can easily be looked up in a book" - Albert Einstein
He would practically have Alzheimer's today with the internet..
Loved reading the books about Feynman and watching interviews with him and his lectures. And look at that face, you can just see the mischief lurking in there.
If you enjoy stuff like this(bypassing security) you will almost certainly like the lecture on UA-cam called "I'll Let Myself In: Tactics of Physical Pen Testers".
I should probably warn everyone how unsettling it is that seemingly secure things are quite often not secure at all. Watch at your own risk of anxiety.
I didn't get the part where you said "people leave their safe open and there's a drawer in the bottom and he fiddled and remembered the last two number" etc etc, how did that work?
As I understand it, he could feel the mechanism inside the safe and tell when the first two numbers lined up. So while talking, slowly spin the knob until the first number engaged, remember the number. Then spin the knob until the second number engaged, remember the number.
@@thadtheman3751 And nothing is stopping him from getting the third number either.
Agreed! The descriptions of his time in Alamos and the safecracking is my favourite part.
"I'm having fun" - precisely why you're my biggest inspiration!
There is a CD of an entertaining talk ("Los Alamos from Below") that Feynman gave about his time at Los Alamos. I got it with the book "Classic Feynman." It goes through his whole story and ends with some safecracking stories. I would recommend both the book and the talk.
One of the most intereseting videos on the Numberphile channel!
I have now - so few people actually go into the description it seems!!!
I was just reading Feynman's book "Surely you're joking Mr. Feynman", and just finished the chapter where he was describing his passion for safe-breaking. I put the book aside and entered my UA-cam account, and lo and behold - a new numberphile's video about.. now, what's the odds for that? :)
Amazing book by the way, highly recommended to anyone who would like to have a glimpse into the mind of this fascinating man.
On my second ship in the Navy, my office was a classified materials vault. Whenever I went in to do some work, I would lock the door behind me both for privacy and to prevent anyone from doing Feynman's "Open safe" trick.
I have never been able to work these locks very well , and I have lived in USA for 15 years ,
they are everywhere . I now feel more confident in tackling them again , without the bolt cutter or grinder
I didn't understand the last one, what do you mean he fiddled with the knobs when the safe was open:
1) on your video it says 2 of the numbers became known, how so?
2) were the correct numbers in position when you opened the safe?
Feynman may not have been my favourite scientist before, but he sure as hell is now! That's awesome!
OMG! my birthday is on the 21/4 too! Happy birthday Professor Bowley.
Happy birthday!
I was just thinking yesterday "we haven't been seeing enough of Professor Bowley lately." Good timing.
read about this in Richard Feynman's autobiography, hilarious anecdotes about lock-breaking he talks about. Ah gota love Feynman!!
Finally a new video with Prof. Roger Bowley. He's my biggest inspiration in Sixty Symbols!
Hi our excellent subtitles have always been done by a third party - I have no control over when they are done or added!
I read "Surely you're joking Mr. Feynman", it had stories about his life and growing up. It also included Los Alamos and the safe-cracking business; the way he wrote it was simply hilarious. I recommend it. :)
It's in that Feynman book but I can remember what it was. It did something like the bolt moved if it was already open on the last digit and then you dialed the previous digit.
Several years back I managed to crack an inexpensive Sentry safe just by starting somewhere in the middle and brute forcing every combo in increments of 4 or 5, just like described here. I reasoned that the last number didn't really matter because I could just repeatedly yank on the handle while rapidly turning the dial. This means I really only had to try about 20^2 permutations max (and in practice, only about two-thirds that number). It only took about 10 minutes to get it open.
Happy belated birthday Professor Bowley!
Geeky or not, I LOVE this numbrphile series. They’ve hit on really nice programmed format, imho!
I have many videos on the go - Zeno is coming!
Gets a a bit boring if you always know what is coming next, doesn't it!? ;)
Cool concept. Richard Feynman was such a cool guy. Who knew safes could be so interesting lol
This episode was so good.
Feynman was a bigger showman than he was a scientist. And he was one of the best scientists the 20th century had seen.
I just finished all Numberphile videos! Now I'm officially a nerd!
He's using birthdays; the first digit has to be between 0 and 30 (for the days of the month), the second will be 1 - 12 (months) and the last could be easily guestimated by looking at or knowing the safe-owner.
And as the safes had quite a margin for error (as explained), that leaves few possibilities.
I was jumping with delight when I saw the thumbnail + title for this video!
Feynman's such a mischievous genius :d
"By fiddling with his own safe he learned that when a door was open he could find the last number of a combination by turning the dial and feeling when the bolt came down. Given some time, he could find the second number that way, too. He made a habit of absently leaning against his colleagues' safes when he visited their offices, twirling the dials like the perpetual fidgeter he was, and thus built up a master list of partial combinations."
He also tweaked a pompous colonel in Oak Ridge who claimed to have the most secure safe in the country; it turned out to be a large armored cabinet with the same Mosler lock on it. He could get the contents out of any key-locked desk by getting under it and and reaching into drawers that were closed and locked from the outside. He was the bane of the security people at Los Alamos.
Duh-doy! :P
And Feynman, being the ultimate simplifier in physics, (Feynman diagrams are a beautiful way of representing sub-atomic particles), his ability to distill problems down to it's simplest concepts, really allowed him to achieve much. He and Einstein had the same gift of seeing complex things in simple ways.
Everyone should read Feynman's books, and a few of his biographies written by others, his life was eventful, and the love story with his wife made me shed tears
The logic behind the US format of dates is just from how we say the actual date over here. Is spoken language, in the US, we say "September 18, 2005" instead of "The 18th of September, 2005" so it just makes perfect sense to us over here.
Love these auto-biography episodes.
William Tutte has a really interesting story that relates to cryptography and WWII.
I dont understand int he end, how do Feynman know 2 other combination beside first one?
+yohanes ronald Same here
+yohanes ronald To open the door, the three tumblers will need to line up so the latch can slide back into a hole. I'm guessing that this particular lock would prevent the dial from spinning freely. Perhaps it could only turn around between the position of the 2nd and 3rd numbers. Or if the dial could spin the whole way around, it probably felt different.
yeah he doesn't explain that but from the book Feynman found that you could pick the first two numbers off an OPEN safe by turning the dial both ways in in increments and testing the bolt of the safe (pushing it) until it would no longer retract.
Glen Beestone
Thank you!
But if you need to have it open to know, then how do you know?
When the safe was open, you could see the bolt that holds the door closed in the "open" position. By spinning the wheel, the bolt would eventually return to the "closed" position. The last number before the bolt "closed" would be the correct number. Then you would rotate the wheel in the other direction to get the 2nd number.
On a completely unrelated note, I recommend the book "Surely You're Joking, Mr. Feynman!"
I read his autobiography. The safe-cracking parts was a fascinating read.
Yet another reason to idolize Feynman. The guy is my hero.
i read this story in "surely you're joking My Feynman". Great book, great story and a great man!
I've been doing safe cracking for years. As long as you don't put your fingers between the arms of the nutcracker, it's safe.
I know that's nuts.
Can anyone explain that last devious trick he did, didnt quite understand :/ (The part where he would fiddle with an already open safe) What Im confused about is how the last 2 numbers would be know (and how) but not the first one.. Thanks!
Richard Feynman was arguably the most patient man in history
I realized the construction tolerances on padlocks when I was assigned a locker in middle school. That, by far, is the main problem with padlocks and safes.
Big fan of Feynman, thanks for that Brady! Appreciate it
When I was 12, I figured out how a luggage combination lock worked. I had a blast messing up peoples combination locks
I wish i could thumbs this up a million times.
It is such a good book!
One of my favourite stories was when he was at MIT (or maybe Princeton). External lecturers would come in to give seminars and he used to sit in like all the other students ... except he would break into the lecture hall before and read the lecturers notes and more importantly, the conclusion. Then at the end of the seminar, just before the guest speaker summed up, Feynman would raise his hand and say things like:
"But doesn't that mean we can conclude ... X" As if he figured it out right then.
I didn't understand the last part, what he did when the safes were open
The locker combinations today when they are opened, show you the last digit. Even if you don't know the last you can save time and do them in about 10 minutes by knowing the internal wheels. eg: spin + to 20, around once and - to 15, then + to 25 but then you can just do 30,35,40,45,50,55,00,05,10 all in a few seconds without doing the 1'st 2 again. Then set only the 2'nd wheel by - to 10. I used to open them driving home, just in the time stopped at red lights.
Brady is teasing us so we keep coming back for more :)
I do work which require me handling a lot of locks and keys (many buildings, many employees both permanent and temporary).
Every time I encounter a new lock system I like to spend some down time figuring out how it works. Comes in handy and I got a bit of a reputation at this point at work for being able to get most things open if needed, to the point where they call me in instead of calling in a locksmith.
Most of the time I just use a screwdriver/piece of plastic/whatever to force open things or unscrew something and it will open, most locks weakest points are the surroundings they are placed in.
Thanks for doing this video, was reading the book and now I understood some details of his story.
This was one of my favorite chapters from Feynman's autobiography.
I didn't get the last part, so how did he figure out the last two numbers so that the would always be correct?
Happy birthday professor Bowley!
From the stories, the safes used didn't have a lever. You had to set the wheel a 4th time to the number 10, Then it opens. However, he did try spinning it and quickly turning to 10, hoping friction would help
my father had a safe in his office, I taught myself how to crack it by turning the dial and feeling and listening for a slight detent then reverse, it worked on those old ww2 era safes.
Hey Brady where are you guys recording now that Professor Bowley isn't in his office? Does he have another office at the school or is this at his home now?
Not forgetting with almost any combination lock, the LAST tumbler (the third number), all possibilities can usually be tried without re-entering the first two numbers, because the first two tumblers are already set and the last is physically locked to the number dial, this means you really only need to solve for the first two tumblers 'slowly' and the last number is turn,try,turn,try,turn.try...etc. Doesn't work for all locks because they have a randomizer on unsuccessful attempts.
Is it just me or is the last part of this missing some information? How did he find/know the last two settings?
I just want to point out that year-month-day is the global standard numeric date format set in 1988 (see ISO 8601).
Surely You're Joking, Mr. Feynman! spends a decent amount of time on Richard Feynman's safecracking adventures, but also recounts a number of other events throughout his life, always with a good dose of humor (he was by all accounts a natural-born prankster, and it shows in the book); it makes for a very good read.
I agree about the year dial: you can't assume in 1942 that nobody will use a date in the 19th century. Checking American instead of international dates is the same number of combinations. If you want to check both, it's a little slower but not as slow as you suggest: you don't need to try 8x8 because, e.g., 30/30 isn't a date in either system. So you just need 3x8+5x3 (not even twice 3x8 because, e.g., 3/3 is a date in both systems and you only need to try it once).
Thanks for making such great videos Brady!
Fyi, there is the story told in his own words on youtube... along with a number of other story details... definitely worth the listen (no video) search for "Los Alamos from below".
this was the first numberphile I saw years ago
we are the same people
And with the locks in use in that facility, the final turn generated a pop and stopped the dial.. so if you went 9, then 13, then back to 5, on your way to the five, from the 13, you'd feel the pop and would be hard-pressed to go beyond it...
So all you need are the first two numbers really -- then turning back the lock would "give" you the last one.
no, he reaches into the mechanics and then turns the knob and he would feel a metal bar move as he hits the right number.
once he has the first number, he again puts his fingers into the mechanics, resets, dials to the first number he now knows, and turns the knob into the other direction, and would again feel something when he hits the second number. and he did this whenever he had a chance to do it.
That is all great assuming that everyone used dates for safe code. How long did it take him to crack one that was setup illogically to his logic? How did he "show off" then? How did he explain, "ooops can't crack it 12 min!"
+Robin Meade worst case 10 hours, that was the 20x20x20 one, which mean 5 hours average.
and as they said he wouldn't show off that he did it in 12 minutes, just that he could do it, so he would sit at read for at bit after opening it, just make it look harder.
Feynman was a beast, there are many biographies about him - Laurence Krauss has a good one.
I have a question. Let's say you're going through every possible combo out of 1,000,000and it takes 5 seconds to try each combo. As you go through them, the more likely it is the next combo will be the right one, as the # of possible combos decrease as previous ones are eliminated, and can continue until you reach the very last one. I know how you could calculate the max or min time it might take, but is there a way to calculate the average amount of time it would take to get it?
... and "What Do You Care What Other People Think?" is pretty good too. From the book it sounded like he (Feynman) didn't have a roommate. He said he borrowed Fuchs car .. didn't know he shared a room (according to Bowley).
Just hit the CC button next to the quality control on the player. You can turn them back on
true, we kind of Anglicised it there, didn't we!?
I did that in high school. My friend's lock had a ±5 margin on the mechanism. A few minutes at a time over weeks, and one day I opened his locker for him. Good times.
It's only more secure against a brute-force attack (that looks at every combo of characters). It's LESS secure against dictionary attacks which try combinations of real words first, because that's more likely to find your password sooner than if it were non-words.
Same birthday as him... so honored
Happy birthday!
he could always use the method explained in the video to open any random safe in under 12min.
well,not any random safe,it would have to be using a date as the pin, which alot of safes do
read his autobiographies if you are interested in the guy,they are soooo goood
If, like they say at the start of the vid, there's only one knob on a normal safe, then how could he possibly remember the last 2 numbers on the knobs? Since you only could only
see that LAST number (which was the number it was at when the safe opened?)