WEBVTT 1 00:00:00.080 --> 00:00:02.759 Have you ever stumbled upon something and felt like, I 2 00:00:02.759 --> 00:00:05.040 don't know, like it was whispering a secret, maybe an 3 00:00:05.080 --> 00:00:08.119 old journal up in the attic, or a weird symbol 4 00:00:08.240 --> 00:00:10.640 carved on some furniture at a flea market. Yea, even 5 00:00:10.720 --> 00:00:13.320 like a cryptic message on a tombstone. 6 00:00:13.800 --> 00:00:17.199 Absolutely, that feeling that urged to figure it out. It 7 00:00:17.239 --> 00:00:21.079 taps into something really fundamental, doesn't it our drive to understand, 8 00:00:21.120 --> 00:00:24.920 to solve puzzles. It's just satisfying. 9 00:00:24.960 --> 00:00:28.640 It really is seeking patterns, finding meaning where others just 10 00:00:28.679 --> 00:00:29.879 see well junk. 11 00:00:30.079 --> 00:00:32.039 Maybe exactly, And that's exactly. 12 00:00:31.839 --> 00:00:34.479 What we're diving into today, the world of code breaking. 13 00:00:34.560 --> 00:00:38.399 It's fascinating stuff. Is this whole deep dive actually came 14 00:00:38.439 --> 00:00:41.000 from one of you, a listener who's really intrigued by 15 00:00:41.079 --> 00:00:44.439 hidden messages just like us. Ah, excellent, And our main 16 00:00:44.479 --> 00:00:46.399 guide for this is going to be the book Code Breaking, 17 00:00:47.240 --> 00:00:51.079 A Practical Guide by Ilonka Dunnin and Close schmah. 18 00:00:51.280 --> 00:00:54.520 Oh, that's a great choice, really accessible dun In and Schmid. 19 00:00:54.600 --> 00:00:58.079 They know their stuff, but they make it engaging, not 20 00:00:58.200 --> 00:00:58.719 too dry. 21 00:00:58.920 --> 00:01:03.280 Totally agree. I'm here, you know, always curious about intriguing subjects. 22 00:01:03.200 --> 00:01:05.799 And I'm here happy to delve into the history and 23 00:01:06.040 --> 00:01:09.640 the actual practice of encryption. It's a field I find 24 00:01:10.159 --> 00:01:11.200 endlessly fascinating. 25 00:01:11.519 --> 00:01:14.519 Great, so between us, hopefully we can unravel some of 26 00:01:14.560 --> 00:01:16.439 this for you. Let's give it a shot now. Our 27 00:01:16.560 --> 00:01:19.280 goal today. Look, we're not going to turn you into 28 00:01:19.319 --> 00:01:21.519 a master code breaker in what the next. 29 00:01:21.319 --> 00:01:24.519 Hour slight juckle, probably. 30 00:01:24.120 --> 00:01:26.400 Not no, but we want to give you a really 31 00:01:26.519 --> 00:01:30.879 solid grasp of the basic ideas. You know, what a codes? 32 00:01:30.920 --> 00:01:33.079 What are ciphers? Look at different types? 33 00:01:33.159 --> 00:01:36.480 You share some really cool stories, success stories, failures sometimes too, 34 00:01:36.599 --> 00:01:38.400 just to give you a real taste of it exactly. 35 00:01:38.439 --> 00:01:41.879 We want you to walk away appreciating the cleverness involved 36 00:01:42.439 --> 00:01:46.000 both in making the codes and breaking them. 37 00:01:45.760 --> 00:01:48.200 And maybe just maybe inspire you to try a simple 38 00:01:48.200 --> 00:01:51.400 puzzle yourself. It can be quite addictive, definitely, and. 39 00:01:51.359 --> 00:01:53.760 We should mention the authors again, Elanka Dunn and Cheese, 40 00:01:53.840 --> 00:01:57.920 known for working on huge unsolved codes like cryptos. 41 00:01:57.560 --> 00:01:59.680 Right, the CIA sculpture one. 42 00:01:59.519 --> 00:02:02.200 Yeah and she pops up in Dan Brown's the Last Symbol. 43 00:02:02.760 --> 00:02:06.239 And Klaus Schmah He's written loads, very respective and apparently 44 00:02:06.319 --> 00:02:08.000 quite entertaining in his presentations. 45 00:02:08.159 --> 00:02:10.840 They're a good team for this kind of practical guide 46 00:02:11.000 --> 00:02:13.400 deep knowledge, but they make it approachable. 47 00:02:13.680 --> 00:02:17.120 Okay, let's get started. Then, At its most basic level, 48 00:02:18.000 --> 00:02:21.680 what is encryption? Why do we even bother hiding messages? 49 00:02:21.879 --> 00:02:25.400 Well, fundamentally, it's about secrecy, concealing the meaning of a 50 00:02:25.400 --> 00:02:28.960 message so only the intended recipient can actually understand it. 51 00:02:29.039 --> 00:02:31.240 Privacy security exactly. 52 00:02:30.879 --> 00:02:35.319 Keeping information safe from prying eyes, whether that's military secrets, 53 00:02:35.360 --> 00:02:38.199 diplomatic messages, or you know, just a private note. 54 00:02:38.240 --> 00:02:41.240 Got it now. The book draws a distinction right away 55 00:02:41.240 --> 00:02:45.719 between two main things, cipher's and codes. What's the difference there. 56 00:02:45.800 --> 00:02:49.080 Right, that's a key distinction. So think of it like this. 57 00:02:49.719 --> 00:02:52.719 Ciphers generally work on the letters of a message. You're 58 00:02:52.759 --> 00:02:56.120 substituting or rearranging individual letters according. 59 00:02:55.719 --> 00:02:58.159 To some rule, okay, letter by letter, yeah. 60 00:02:58.199 --> 00:03:01.039 Whereas codes they work on whole words or phrases. You 61 00:03:01.080 --> 00:03:03.960 have a pre agreed list like a dictionary, where specific 62 00:03:04.000 --> 00:03:07.080 words or phrases are replaced by a symbol or another. 63 00:03:06.800 --> 00:03:08.120 Word, like a secret dictionary. 64 00:03:08.159 --> 00:03:11.360 Almost exactly. And the book uses the analogy that codebooks 65 00:03:11.360 --> 00:03:13.680 can get really huge and unwieldy. If you want to 66 00:03:13.719 --> 00:03:15.120 cover lots of words, I. 67 00:03:15.039 --> 00:03:18.360 Can imagine you'd need a massive book for every possible 68 00:03:18.360 --> 00:03:19.000 thing you might want. 69 00:03:18.919 --> 00:03:21.800 To say precisely, Cipher's because they work on the letters. 70 00:03:21.840 --> 00:03:25.680 The building blocks are often more flexible, more compact. You 71 00:03:25.719 --> 00:03:26.520 just need the method. 72 00:03:26.639 --> 00:03:29.280 The key makes sense, and the book mainly focuses on 73 00:03:29.319 --> 00:03:29.960 Cypher's right. 74 00:03:30.319 --> 00:03:32.960 Yes, for the most part. Codes and related systems like 75 00:03:33.000 --> 00:03:36.400 nomenclators come up later, but the bulk is on cipher techniques. 76 00:03:36.479 --> 00:03:39.919 Okay, so Cipher's it is mostly. Now, imagine you find 77 00:03:40.039 --> 00:03:43.280 this jumbled mess of letters. Where do you even start? 78 00:03:43.319 --> 00:03:45.919 What's the first thing a codebreaker does well. 79 00:03:46.000 --> 00:03:49.599 One of the absolute fundamental techniques, and the book introduces 80 00:03:49.639 --> 00:03:51.759 it early is frequency analysis. 81 00:03:51.840 --> 00:03:52.759 Ah, I've heard of that. 82 00:03:52.960 --> 00:03:57.280 Counting letters basically, yes, you literally count how many times 83 00:03:57.400 --> 00:04:00.680 each different letter or symbol appears in the ciphertext the 84 00:04:00.759 --> 00:04:01.800 encrypted message. 85 00:04:01.960 --> 00:04:03.080 Why what does that tell you? 86 00:04:03.560 --> 00:04:09.199 Because languages have characteristic frequencies. In English, for example, E 87 00:04:09.400 --> 00:04:13.199 is by far the most common letter, then TAOI, N, 88 00:04:13.319 --> 00:04:16.319 and so on. So the idea is, even when the 89 00:04:16.399 --> 00:04:20.600 letters are scrambled or substituted, those underlying frequencies might still 90 00:04:20.639 --> 00:04:24.439 show through in the ciphertext. If one symbol appears way 91 00:04:24.519 --> 00:04:27.839 more often than others, it might be the encrypted E exactly. 92 00:04:28.319 --> 00:04:31.680 It's an educated guess a starting point. The book gives 93 00:04:31.680 --> 00:04:34.759 an example ciphertext to show just this. You look for 94 00:04:34.839 --> 00:04:36.759 patterns and apparent chaos. 95 00:04:37.000 --> 00:04:39.879 Okay, so frequency analysis is your foot in the door 96 00:04:41.000 --> 00:04:42.759 looking for that disguised E. 97 00:04:43.399 --> 00:04:46.399 Makes sense, It's a cornerstone. Doesn't always work, especially with 98 00:04:46.439 --> 00:04:49.480 more complex ciphers, but for many classical ones it's vital. 99 00:04:49.759 --> 00:04:52.160 Right now, the book dies into a whole bunch of 100 00:04:52.160 --> 00:04:54.439 different cipher types. This is where it gets really interesting. 101 00:04:54.839 --> 00:04:58.720 Let's start with substitution ciphers. What's the basic idea. 102 00:04:58.279 --> 00:05:01.439 Substitution is pretty much what it sounds like. You substitute 103 00:05:01.439 --> 00:05:04.720 each letter in your original message the plaintext, with a 104 00:05:04.759 --> 00:05:07.240 different letter or symbol based on a fixed system. 105 00:05:07.360 --> 00:05:09.879 And the simplest version is just swapping one letter for 106 00:05:09.959 --> 00:05:13.079 another consistently, like every A becomes a Q, every B 107 00:05:13.319 --> 00:05:13.959 becomes an X. 108 00:05:14.240 --> 00:05:18.000 Precisely, that's a simple substitution cipher, or sometimes called a 109 00:05:18.040 --> 00:05:21.759 mono alphabetic cipher one to one mapping. The book mentions 110 00:05:21.759 --> 00:05:24.720 a fun example from Klaus's blog Oh Yeah, a challenge 111 00:05:24.720 --> 00:05:28.360 cipher from twenty fifteen. Someone cracked it really quickly using 112 00:05:28.360 --> 00:05:31.720 frequency analysis and figuring out The keyword used to create 113 00:05:31.759 --> 00:05:34.000 the substitution alphabet was Gordon Young. 114 00:05:34.160 --> 00:05:37.800 Wow, so the keyword determine the whole swap. M cool. 115 00:05:38.040 --> 00:05:40.360 What about the Caesar cipher. That's the famous one, right 116 00:05:40.399 --> 00:05:41.120 you and I've heard of that. 117 00:05:41.319 --> 00:05:44.560 Yeah. The Caesar cipher is a type of simple substitution, 118 00:05:44.680 --> 00:05:47.720 but a very specific one. You just shift every letter 119 00:05:47.839 --> 00:05:50.920 a fixed number of places down the alphabet, like three places. 120 00:05:50.920 --> 00:05:52.680 So A becomes dB becomes. 121 00:05:52.519 --> 00:05:55.399 E exactly, wrap around at the end, so x becomes a, 122 00:05:55.560 --> 00:05:57.680 hy becomes b, z becomes c if it's a shift 123 00:05:57.680 --> 00:06:02.000 of three. Super simple. The book mentions Herbert Yardley, an 124 00:06:02.079 --> 00:06:05.199 early US codebreaker, use it for his first ciphergram. 125 00:06:05.399 --> 00:06:07.360 Huh. Even the pros started. 126 00:06:07.040 --> 00:06:10.399 Simple, they did. And there's that funny newspaper example do Lo, 127 00:06:10.759 --> 00:06:13.720 which decrypts with a shift, but the result is still 128 00:06:13.759 --> 00:06:14.839 partly gibberish. 129 00:06:14.920 --> 00:06:17.000 Right April third, if you work here, it doesn't always 130 00:06:17.000 --> 00:06:18.480 pay perfect sense exactly. 131 00:06:18.720 --> 00:06:23.560 And the mug the inscription that decrypts to National Security Agency. 132 00:06:23.240 --> 00:06:25.759 With the typo. That's brilliant. Even spies make. 133 00:06:25.639 --> 00:06:28.120 Typos, apparently, so shows it happened. 134 00:06:28.240 --> 00:06:34.160 Okay, so simple substitution Caesar shifts, but they can get trickier. Right, 135 00:06:34.680 --> 00:06:38.879 What about homophonic ciphers. It sounds complicated. 136 00:06:38.920 --> 00:06:42.360 They are a step up. Homophonic means same sound, but 137 00:06:42.439 --> 00:06:45.240 here it means multiple Ciphertext symbols can stand for the 138 00:06:45.279 --> 00:06:46.480 same plaintext letter. 139 00:06:46.560 --> 00:06:49.120 Wait, so E might be represented by a seven, or 140 00:06:49.160 --> 00:06:49.959 a percent or. 141 00:06:49.959 --> 00:06:53.720 A k precisely. Especially for common letters like E or T, 142 00:06:54.199 --> 00:06:57.079 you give them multiple substitutes. The goal is to flatten 143 00:06:57.120 --> 00:06:59.560 out those telltale frequency counts we just talked about. 144 00:06:59.639 --> 00:07:02.680 Ah, so it makes frequency analysis much harder, much harder. 145 00:07:02.759 --> 00:07:05.279 If E can be five different things, Just counting symbols 146 00:07:05.279 --> 00:07:06.240 doesn't tell you as much. 147 00:07:06.279 --> 00:07:07.639 So how do you break those? Well? 148 00:07:07.680 --> 00:07:11.560 Today computers are the best bet, using techniques like hill climbing, 149 00:07:11.680 --> 00:07:14.839 basically trying small changes to a possible key and seeing 150 00:07:14.839 --> 00:07:16.560 if the output looks more like English. 151 00:07:16.680 --> 00:07:18.759 Okay, computer power, but manually. 152 00:07:18.879 --> 00:07:22.160 The book gives a geocaching example. If you have spaces 153 00:07:22.199 --> 00:07:26.000 between words, that helps enormously. You can still do some 154 00:07:26.040 --> 00:07:29.240 frequency counts on the symbols. Guess short words like at 155 00:07:29.680 --> 00:07:33.240 or the he, look for common pairs like thh. It's 156 00:07:33.360 --> 00:07:34.399 painstaking work. 157 00:07:34.519 --> 00:07:38.279 Geocaching makes sense they'd use tricky codes, okay. Another one 158 00:07:38.279 --> 00:07:39.160 that sounds weird. 159 00:07:39.560 --> 00:07:43.279 The pig pen cipher chuckles slightly. Yeah, the name's descriptive. 160 00:07:43.439 --> 00:07:46.720 It's geometric cipher. You use grids like a tic tac 161 00:07:46.800 --> 00:07:49.800 toe board, maybe with dots in some sections. Each letter 162 00:07:49.879 --> 00:07:52.040 is replaced by the shape of the pen or grids 163 00:07:52.040 --> 00:07:52.959 segment it sits in. 164 00:07:53.079 --> 00:07:55.199 Okay, so shapes instead of letters, right. 165 00:07:55.480 --> 00:07:58.240 The book shows that tombstone from seventeen ninety six. It 166 00:07:58.360 --> 00:08:02.920 uses pigpen and reads dia. Another typo should be deck. 167 00:08:03.000 --> 00:08:06.240 They were having trouble with spelling back then. Even in code. 168 00:08:05.959 --> 00:08:09.920 Seems like it. And there's the pirate Olivier Lavaser's cryptogram 169 00:08:09.920 --> 00:08:12.120 from the seventeen thirties. Many think of his pig pen 170 00:08:12.519 --> 00:08:15.519 possibly cracked using frequency analysis on the symbols. 171 00:08:15.279 --> 00:08:18.959 Pirate treasure maps with geometric codes. Yeah, fantastic. Okay. Playfair 172 00:08:19.000 --> 00:08:20.800 cipher sounds more serious. 173 00:08:20.879 --> 00:08:23.000 It is. Playfair is a step up again. It's a 174 00:08:23.040 --> 00:08:26.800 digraph cipher, meaning it encrypts letters in pairs, not one 175 00:08:26.839 --> 00:08:30.360 by one. It uses a five x five grid, usually 176 00:08:30.399 --> 00:08:32.679 filled with a keyword first than the rest of the alphabet, 177 00:08:32.720 --> 00:08:35.759 combining I and J. Then there are rules based on 178 00:08:35.799 --> 00:08:37.519 whether the pair of letters are in the same row, 179 00:08:37.679 --> 00:08:40.600 same column or form a rectangle in the grid. 180 00:08:41.000 --> 00:08:43.559 Wow, Okay, that sounds much more involved rules and a 181 00:08:43.600 --> 00:08:44.480 grid based. 182 00:08:44.200 --> 00:08:46.679 On a keyword, it is much harder to break than 183 00:08:46.720 --> 00:08:51.759 simple substitution. The book mentions Robert Thelis's nineteen forty eight cryptogram. 184 00:08:52.120 --> 00:08:55.720 Modern breaks often involve computers trying lots of potential keywords 185 00:08:55.759 --> 00:08:58.679 to generate the grid. The book gives an example using 186 00:08:58.759 --> 00:09:01.519 black and beauty as keywords in the analysis. 187 00:09:01.759 --> 00:09:05.960 So computers crunching keywords, got it? And general digraph substitution 188 00:09:06.120 --> 00:09:07.720 is at just any pair for pair swap. 189 00:09:07.799 --> 00:09:11.320 Pretty much Playfair has its specific rid rules. General digraph 190 00:09:11.399 --> 00:09:14.440 substitution just means swapping pairs of letters for other pairs, 191 00:09:14.440 --> 00:09:18.919 but without those specific Playfair constraints, potentially many more possibilities. 192 00:09:19.279 --> 00:09:21.759 Klaus apparently set a challenge on his blog with one 193 00:09:21.799 --> 00:09:22.120 of those. 194 00:09:22.440 --> 00:09:25.799 Okay, So substitution covers a lot, from simple shifts to 195 00:09:25.919 --> 00:09:30.360 complex paired letters. But the book also talks about transposition ciphers. 196 00:09:31.120 --> 00:09:31.960 How are they different? 197 00:09:32.200 --> 00:09:36.200 Totally different approach. Substitution changes what the letters are. Transposition 198 00:09:36.320 --> 00:09:37.120 just changes. 199 00:09:36.840 --> 00:09:38.759 Their order, like shuffling a deck of cards. 200 00:09:38.919 --> 00:09:42.080 Exactly all the original letters are still there, just jumbled 201 00:09:42.120 --> 00:09:45.080 up according to a specific secret rule or pattern. 202 00:09:44.879 --> 00:09:49.159 So it's like a complex anagram. The book mentions complete 203 00:09:49.159 --> 00:09:51.240 columnar transposition. How's that work? 204 00:09:51.399 --> 00:09:54.000 Right? You take your message, write it out in a grid, 205 00:09:54.279 --> 00:09:56.840 say row by row. The number of columns is usually 206 00:09:56.840 --> 00:09:59.279 based on the length of a keyword. Okay, then you 207 00:09:59.360 --> 00:10:02.159 read the letters out column by column to get the ciphertext. 208 00:10:02.519 --> 00:10:05.320 But here's the key The order you read the columns 209 00:10:05.320 --> 00:10:07.639 in depends on the alphabetical order of the letters in 210 00:10:07.679 --> 00:10:08.279 your keyword. 211 00:10:08.440 --> 00:10:12.840 Ah, so the keyword tells you the shuffled column order. Clever. 212 00:10:13.480 --> 00:10:15.679 The book uses Swiss cheesecake as an example. 213 00:10:15.919 --> 00:10:18.120 Yeah. With the keyword table, it lays it out clearly. 214 00:10:18.200 --> 00:10:20.279 A is the first letter and table. So you read 215 00:10:20.279 --> 00:10:22.480 the A column first, then the B column and so on. 216 00:10:22.879 --> 00:10:26.840 Got it? And incomplete Columner transposition is that just when 217 00:10:26.879 --> 00:10:27.679 the grid isn't. 218 00:10:27.480 --> 00:10:31.080 Full exactly, if your message length doesn't perfectly divide by 219 00:10:31.120 --> 00:10:33.879 the keyword length, the last row won't be full. The 220 00:10:33.919 --> 00:10:37.720 book mentions an IRA cryptogram one hundred and thirteen letters. 221 00:10:37.799 --> 00:10:38.720 That's a prime number. 222 00:10:38.759 --> 00:10:41.679 Oh right, so it can't make a need rectangle except 223 00:10:41.759 --> 00:10:43.720 one by one thirteen or one thirteen by one. 224 00:10:43.759 --> 00:10:49.159 Correct. Which makes it trickier. Breaking these sometimes involves multiple anagramming, 225 00:10:49.240 --> 00:10:53.279 basically trying to rearrange the columns computationally until fragments of 226 00:10:53.360 --> 00:10:55.039 words start appearing. 227 00:10:54.639 --> 00:10:57.559 Prime numbers, making life difficult for code breakers. Yeah, and 228 00:10:57.600 --> 00:11:01.080 then there's double column to transposition. Do it twice pretty much? 229 00:11:01.159 --> 00:11:03.840 Yeah, You transpose it once with one keyword, then you 230 00:11:03.879 --> 00:11:06.120 take that jumbled text and transpose it again with a 231 00:11:06.120 --> 00:11:09.480 second keyword. Makes it much much harder. About the book 232 00:11:09.519 --> 00:11:13.360 mentions Marzie's report on ciphers using keywords from an Italian poem. 233 00:11:13.639 --> 00:11:16.679 They were tough nuts to crack until Arman Krauss figured 234 00:11:16.679 --> 00:11:17.279 something out. 235 00:11:17.360 --> 00:11:17.799 What was that? 236 00:11:18.080 --> 00:11:20.399 He realized the first and last blocks of letters in 237 00:11:20.440 --> 00:11:22.600 the messages weren't part of the main text, but were 238 00:11:22.639 --> 00:11:27.799 like indicators metadata. Once you remove those, the double transposition 239 00:11:27.879 --> 00:11:29.200 underneath became solvable. 240 00:11:29.519 --> 00:11:34.480 Wow. Sometimes the clue isn't even in the main message. Fascinating. Okay, 241 00:11:34.519 --> 00:11:39.120 one more transposition type. The turning grill sounds very physical. 242 00:11:39.480 --> 00:11:41.559 It is. It uses a stencil like a piece of 243 00:11:41.600 --> 00:11:43.679 card stock with holes cut in it. That's the grill. 244 00:11:44.120 --> 00:11:46.960 You lay it over a grid, write letters through the holes, 245 00:11:47.440 --> 00:11:51.000 then you turn the grill. Usually ninety degrees, write more 246 00:11:51.080 --> 00:11:54.120 letters in the newly exposed spots. Repeat for four. 247 00:11:54.000 --> 00:11:57.840 Turns, so the grill reveals different spots with each turn exactly. 248 00:11:57.879 --> 00:11:59.919 Then the cipher text is just reading the whole grid 249 00:12:00.000 --> 00:12:02.399 filled with letters. The book has a Christmas card example, 250 00:12:02.879 --> 00:12:05.799 and there's a great historical case in eighteenth century German. 251 00:12:05.840 --> 00:12:09.000 One solved manually because the person writing the message was 252 00:12:09.000 --> 00:12:12.320 a bit sloppy didn't place the letters precisely in the openings. 253 00:12:12.440 --> 00:12:16.879 Huh. Human error strikes again even with gadgets. Okay, substitution 254 00:12:17.000 --> 00:12:21.240 changes letters, transposition shovels them. What about poly alphabetic ciphers? 255 00:12:21.519 --> 00:12:25.919 Poly alphabetic this is where things get significantly more secure 256 00:12:26.000 --> 00:12:30.559 classically speeding. Instead of one substitution alphabet, you use multiple. 257 00:12:30.159 --> 00:12:32.320 Alphabets a multiple alphabets. 258 00:12:32.240 --> 00:12:35.080 Usually controlled by a qword. Each letter of the keyword 259 00:12:35.120 --> 00:12:38.159 tells you which substitution alphabet to use for the corresponding 260 00:12:38.240 --> 00:12:42.240 letter of the plaintext. So the same plaintext letter, say E, 261 00:12:42.679 --> 00:12:45.159 could be encrypted to X one time and then K 262 00:12:45.440 --> 00:12:47.600 the next time it appears, depending on the keyword letter. 263 00:12:47.679 --> 00:12:50.279 Okay, so that really messes up frequency analysis. 264 00:12:49.840 --> 00:12:52.639 Right absolutely, because E doesn't always encrypted the same thing. 265 00:12:53.120 --> 00:12:56.440 The most famous poly alphabetic cipher is the Visienner cipher. 266 00:12:56.679 --> 00:12:57.639 Wait, I heard of that one too. 267 00:12:58.000 --> 00:13:01.000 The book explains the basics. Ngth really depends on the 268 00:13:01.039 --> 00:13:05.200 keyword length. Longer more random keywords make it much stronger, 269 00:13:05.399 --> 00:13:08.480 and the ultimate version is the one time pad. If 270 00:13:08.519 --> 00:13:12.120 your keyword is truly random, never repeats, and is as 271 00:13:12.200 --> 00:13:17.039 long as the message itself, theoretically unbreakable the gold standard. 272 00:13:17.200 --> 00:13:21.200 Wow and cryptos. The first two parts use these polyalphabetic types. 273 00:13:21.279 --> 00:13:24.159 That's right, K one and K two. The solved sections 274 00:13:24.159 --> 00:13:27.440 of the cryptos sculpture were polyolphabetic. K one use the 275 00:13:27.480 --> 00:13:31.960 keyword cryptos. K two use abscissa real world high profile 276 00:13:32.000 --> 00:13:33.399 examples cryptos. 277 00:13:33.440 --> 00:13:38.320 Again, that thing is everywhere in code breaking. Okay, we've 278 00:13:38.360 --> 00:13:42.200 hit cipher's hard letters letters, letters, but you mentioned codes earlier. 279 00:13:42.320 --> 00:13:45.840 Working on words and nomenclators. What are they? 280 00:13:46.120 --> 00:13:49.399 Nomenclators were historically super important. Think of them as a 281 00:13:49.480 --> 00:13:53.600 hybrid system. Hybrid how so they combined a codebook part 282 00:13:53.879 --> 00:13:57.639 lists of common or important words, names, maybe places with 283 00:13:57.720 --> 00:13:59.919 their secret code symbols, with a cipher part. 284 00:14:00.080 --> 00:14:02.480 Oh so code for common stuff, cipher for everything. 285 00:14:02.159 --> 00:14:06.039 Else exactly, Maybe a simple substitution alphabet to spell out 286 00:14:06.080 --> 00:14:08.720 unusual words or names not in the code list, or 287 00:14:08.759 --> 00:14:12.120 just to add another layer. Before maybe the nineteenth century, 288 00:14:12.120 --> 00:14:15.039 if you found a secret diplomatic message, chances were high 289 00:14:15.120 --> 00:14:16.240 it was a nomenclator. 290 00:14:16.879 --> 00:14:19.000 The standard for a long time. Then why did they 291 00:14:19.039 --> 00:14:19.759 fall out of favor? 292 00:14:20.000 --> 00:14:24.519 Well, partly because more complex mathematical ciphers became better understood, 293 00:14:24.799 --> 00:14:28.159 and then, especially with cipher machines like Enigma machines could 294 00:14:28.159 --> 00:14:32.279 handle complex polyophabetic stuff way faster and more securely. Right 295 00:14:32.360 --> 00:14:37.200 technology advanced, plus nomenclators had weaknesses. If someone captured the codebook, 296 00:14:37.240 --> 00:14:40.440 you were sunk. Or if codebreakers found a crib a 297 00:14:40.519 --> 00:14:43.000 known piece of plaintext within the message that could help 298 00:14:43.039 --> 00:14:46.559 unravel it. Or sometimes they were just poorly designed. Any 299 00:14:46.600 --> 00:14:49.919 famous examples, The book mentions the one used by Mary, 300 00:14:50.039 --> 00:14:53.200 Queen of Scott's in the sixteenth century. Thomas Philippus is 301 00:14:53.240 --> 00:14:56.639 thought to have broken it, likely using some frequency analysis 302 00:14:56.639 --> 00:15:00.200 on the cipher parts and recognizing patterns in the coded way. 303 00:15:00.639 --> 00:15:05.440 History turning on a broken code. Okay, and dictionary codes 304 00:15:05.480 --> 00:15:07.519 and book ciphers are they related? 305 00:15:07.679 --> 00:15:11.320 Yeah, there's specific types of codes that use well. Book 306 00:15:11.360 --> 00:15:14.399 says the key Dictionary codes use a specific dictionary. The 307 00:15:14.440 --> 00:15:16.559 code might be like page one, twenty three, word five. 308 00:15:16.720 --> 00:15:18.360 You need the exact same dictionary edition. 309 00:15:18.440 --> 00:15:22.159 Absolutely crucial. MI eight, a US code breaking unit, broke 310 00:15:22.200 --> 00:15:25.240 one by figuring out which dictionary was being used. Book 311 00:15:25.279 --> 00:15:28.399 ciphers are similar, but often point to individual letters within 312 00:15:28.440 --> 00:15:31.639 a specific book, like page fifty, line ten, letter three. 313 00:15:31.799 --> 00:15:33.399 Again, gotta find the book. 314 00:15:33.360 --> 00:15:37.159 Exactly, Stephen Machis. Finding Nicholas Trist's Key Book to the 315 00:15:37.159 --> 00:15:40.639 British Nation was a huge breakthrough, and that Hindu conspiracy 316 00:15:40.679 --> 00:15:43.120 code cracked when a blog reader tracked down the right 317 00:15:43.519 --> 00:15:48.159 Noah Webster eighteen fifty seven dictionary. Amazing detective work. 318 00:15:48.240 --> 00:15:50.720 Incredible finding the right book in a million Okay, So 319 00:15:50.799 --> 00:15:53.240 that covers a lot of the classical ground. But things 320 00:15:53.240 --> 00:15:58.360 didn't stop there. What about beyond classical methods. Machines come first. 321 00:15:58.960 --> 00:16:00.559 Enigma Enigma is the big one. 322 00:16:00.639 --> 00:16:00.879 Yeah. 323 00:16:01.240 --> 00:16:04.320 Electro mechanical machines like Enigma were a massive leap the 324 00:16:04.399 --> 00:16:09.799 use rotors wires creating an incredibly complex, constantly changing polyolphabetic 325 00:16:09.879 --> 00:16:11.440 cipher for every single letter. 326 00:16:11.600 --> 00:16:13.639 The WWII German. 327 00:16:13.320 --> 00:16:16.039 Machine famously used by them, Yes, okay, and breaking it 328 00:16:16.159 --> 00:16:19.440 was a monumental effort. The polls did crucial early work 329 00:16:19.639 --> 00:16:22.440 than the French British at Bletchley Park touring and his team. 330 00:16:22.600 --> 00:16:24.000 The Imitation game story a. 331 00:16:23.919 --> 00:16:26.320 Big part of it. Yes, and Elizabeth Friedman and the 332 00:16:26.399 --> 00:16:29.679 US also worked on Enigma variants. The book mentions recent 333 00:16:29.720 --> 00:16:33.080 successes too, like the M four project breaking original Navy 334 00:16:33.080 --> 00:16:35.159 messages and ongoing work on others. 335 00:16:35.200 --> 00:16:38.519 Just amazing the sheer brain power involved. What about encrypting voice? 336 00:16:38.519 --> 00:16:41.840 That sounds different, very different technical challenge, especially early on 337 00:16:41.919 --> 00:16:46.559 with analog signals. Early voice scramblers weren't always that secure. 338 00:16:46.360 --> 00:16:47.879 So how did they manage well? 339 00:16:47.960 --> 00:16:51.360 One method the book touches on is code talking using 340 00:16:51.360 --> 00:16:52.960 obscure languages. 341 00:16:52.600 --> 00:16:54.559 Like the Navajo code talkers exactly. 342 00:16:54.759 --> 00:16:57.879 The most famous example using the Navajo language as the 343 00:16:57.919 --> 00:17:01.279 basis for a code that Japanese couldn't break in WWII. 344 00:17:01.600 --> 00:17:05.079 It's a human form of encryption, incredibly effective. 345 00:17:04.759 --> 00:17:09.160 Using language itself as the lock. Brilliant. And lastly here 346 00:17:09.440 --> 00:17:13.039 staganography hiding the message itself right. 347 00:17:13.160 --> 00:17:16.559 Staganography isn't about making the message unreadable, It's about hiding 348 00:17:16.559 --> 00:17:19.000 the fact that there even is a secret message. 349 00:17:19.559 --> 00:17:21.079 How invisible ink. 350 00:17:20.960 --> 00:17:24.200 That's a classic example. Yeah, sympathetic ink right between the lines, 351 00:17:24.400 --> 00:17:27.000 but it can be more subtle. The book mentions Elonka 352 00:17:27.039 --> 00:17:30.039 Dunnan finding a hidden message on the Friedman's tombstone, on. 353 00:17:30.000 --> 00:17:33.440 Their tombstone, William and Elizabeth Friedman. 354 00:17:33.000 --> 00:17:37.000 The codebreakers, the very same, a fitting secret tribute and 355 00:17:37.160 --> 00:17:41.039 way back. Trithemiis's Staganographia book from the fifteen hundreds looked 356 00:17:41.160 --> 00:17:44.880 like magic but actually contained hidden text revealed centuries later 357 00:17:44.920 --> 00:17:46.839 by frequency analysis. 358 00:17:46.400 --> 00:17:51.640 Hiding in plain sight. So machines voice hidden messages. But 359 00:17:51.880 --> 00:17:53.880 computers must have changed everything again. 360 00:17:53.720 --> 00:17:57.200 Right, oh, completely. The role of computers is immense techniques 361 00:17:57.200 --> 00:18:00.440 like hill climbing simulated a kneeling. These are out rhythms 362 00:18:00.440 --> 00:18:03.519 where the computer makes guesses, evaluates how good the result 363 00:18:03.559 --> 00:18:07.599 looks like, how much it resembles English, adjusts and tries again, 364 00:18:07.880 --> 00:18:10.519 thousands millions of times a second, So tasks. 365 00:18:10.279 --> 00:18:12.319 That were impossible manually become. 366 00:18:12.079 --> 00:18:18.960 Feasible, absolutely breaking complex classical ciphers like homophonic visionnaire, columber transposition, 367 00:18:19.519 --> 00:18:22.640 even making inroads into things like the Zodiac killer ciphers. 368 00:18:22.680 --> 00:18:24.880 Computers when in revolutionary they do the heavy. 369 00:18:24.680 --> 00:18:27.680 Lifting, the ultimate code breaking tool. Now. Yeah, but even 370 00:18:27.720 --> 00:18:32.160 with computers, some things remain mysteries. The book mentions unsolved mysteries, 371 00:18:32.359 --> 00:18:34.400 cryptos K four has to be top of the list. 372 00:18:34.519 --> 00:18:37.000 Still unsolved. Yes, K four, the last part of the 373 00:18:37.000 --> 00:18:41.440 crypto sculpture remains stubbornly unbroken, despite clees from the sculptor himself. 374 00:18:41.519 --> 00:18:44.519 It's probably the most famous active puzzle. The book mentions 375 00:18:44.559 --> 00:18:49.319 others the Beal ciphers supposedly leading to buried treasure, the 376 00:18:49.559 --> 00:18:54.759 Somerton Man cryptogram found in Australia, the Action Line cryptogram. 377 00:18:54.880 --> 00:18:57.279 Puzzles that continue to baffle and intrigue. 378 00:18:57.319 --> 00:18:59.160 It's kind of cool that some secrets hold out, isn't 379 00:18:59.160 --> 00:19:01.480 It shows it's not just computation exactly. 380 00:19:01.559 --> 00:19:04.599 It still takes insight, maybe luck the right approach. There's 381 00:19:04.640 --> 00:19:07.119 still an art to it alongside the science. 382 00:19:07.599 --> 00:19:10.880 Okay, this has been absolutely fascinating. Let's try and wrap up. 383 00:19:11.200 --> 00:19:15.079 We've gone from simple letter swaps to enigma machines and 384 00:19:15.160 --> 00:19:18.720 computer analysis. What are the big takeaways for someone listening. 385 00:19:18.480 --> 00:19:22.160 Well, I think understanding the why encryption is for secrecy 386 00:19:22.319 --> 00:19:26.160 and what the difference between cipher's letters and codes words. 387 00:19:26.319 --> 00:19:28.640 We've seen how methods evolved getting more. 388 00:19:28.440 --> 00:19:32.200 Complex, substitution transposition, polyalphabetic. 389 00:19:31.640 --> 00:19:35.480 Right and then the counter effort cryptanalysis using tools like 390 00:19:35.519 --> 00:19:39.319 frequency analysis, pattern finding, and now massive computing power. It's 391 00:19:39.400 --> 00:19:43.319 really a story of ingenuity on both sides logic, language, maths, 392 00:19:43.599 --> 00:19:43.960 and for. 393 00:19:44.000 --> 00:19:47.599 Me, the aha moment is just seeing how even really 394 00:19:47.720 --> 00:19:50.599 complex looking jumbles can have an underlying logic that, with 395 00:19:50.759 --> 00:19:54.200 persistence and the right technique, secrets can be unlocked. It's 396 00:19:54.200 --> 00:19:56.200 a testament to human cleverness, really. 397 00:19:56.039 --> 00:19:59.039 It absolutely is. And understanding these older methods it's not 398 00:19:59.119 --> 00:20:03.039 just history. It helps appreciate modern crypto, the stuff protecting 399 00:20:03.079 --> 00:20:06.480 our emails and bank details. The principles often build on 400 00:20:06.519 --> 00:20:07.440 these foundations. 401 00:20:07.960 --> 00:20:10.839