WEBVTT 00:00:01.120 --> 00:00:04.419 So imagine you're up late, right? And you're 00:00:04.419 --> 00:00:08.160 trying to research the, I don't know, the profound 00:00:08.160 --> 00:00:11.439 cultural origins of the hip hop musical genre. 00:00:11.779 --> 00:00:13.779 Right, like a proper historical deep dive into 00:00:13.779 --> 00:00:16.019 music history. Exactly. So you type your query 00:00:16.019 --> 00:00:18.140 into a search engine, you click that very first 00:00:18.140 --> 00:00:21.059 Wikipedia link, and suddenly the internet decides 00:00:21.059 --> 00:00:22.899 you actually just want to watch Jimmy Fallon 00:00:22.899 --> 00:00:25.519 and Justin Timberlake perform a late night comedy 00:00:25.519 --> 00:00:28.339 bit. Which is, I mean, it's hilarious, but it's 00:00:28.339 --> 00:00:31.019 a real thing. It is. That literal collision of 00:00:31.019 --> 00:00:33.979 global history and late night television is happening 00:00:33.979 --> 00:00:36.200 right now. And it's completely out of sight. 00:00:36.500 --> 00:00:38.600 Yeah, it's all happening in the background architecture 00:00:38.600 --> 00:00:42.119 of the web. Right. So have you ever found yourself 00:00:42.119 --> 00:00:44.219 falling down one of those massive Wikipedia? 00:00:44.670 --> 00:00:47.009 rabbit holes, like you start out looking up the 00:00:47.009 --> 00:00:49.890 capital of Vermont and somehow three hours later, 00:00:50.270 --> 00:00:52.969 you were just deeply engrossed in the architectural 00:00:52.969 --> 00:00:56.229 nuances of 17th century naval tactics. Oh, absolutely. 00:00:56.229 --> 00:00:57.929 I mean, it happens to all of us. It's the classic 00:00:57.929 --> 00:01:00.990 Wikipedia experience. Yeah. But usually when 00:01:00.990 --> 00:01:04.030 we take those journeys, we're jumping from one 00:01:04.030 --> 00:01:07.010 giant sprawling encyclopedia entry to the next. 00:01:07.090 --> 00:01:09.489 Right. We kind of move from room to room in this 00:01:09.489 --> 00:01:11.969 endless information mansion. Right. Exploring 00:01:11.969 --> 00:01:15.010 all these massive topics. But today. We're doing 00:01:15.010 --> 00:01:17.689 something different. We're stopping and staring 00:01:17.689 --> 00:01:19.709 really, really closely at the doorway itself. 00:01:19.930 --> 00:01:21.969 Yeah. We're looking at the architecture of the 00:01:21.969 --> 00:01:23.810 portal rather than the room it leads to, which 00:01:23.810 --> 00:01:26.909 I find endlessly fascinating. It really is. Because 00:01:26.909 --> 00:01:30.370 today's deep dive is not into a massive, thousands 00:01:30.370 --> 00:01:32.969 of words long research paper. We are actually 00:01:32.969 --> 00:01:36.290 looking at a single, highly specific Wikipedia 00:01:36.290 --> 00:01:40.090 set index page titled, get this, lists of Jimmy 00:01:40.090 --> 00:01:42.709 Fallon games and sketches. It's so specific. 00:01:43.049 --> 00:01:45.569 So, so specific. And, you know, if you're tuning 00:01:45.569 --> 00:01:47.670 in right now expecting a retrospective on the 00:01:47.670 --> 00:01:50.909 comedic value of Jimmy Fallon's career or like 00:01:50.909 --> 00:01:52.870 a ranking of his best jokes. Yeah, you might 00:01:52.870 --> 00:01:54.709 be disappointed. Right. We're doing something 00:01:54.709 --> 00:01:57.189 a bit different. We're unpacking the fascinating 00:01:57.189 --> 00:02:00.349 ways human entertainment is categorized, archived 00:02:00.349 --> 00:02:03.189 and disambiguated online. We're basically going 00:02:03.189 --> 00:02:06.510 to use this one tiny skeletal directory page 00:02:06.510 --> 00:02:09.409 to understand how the Internet organizes pop 00:02:09.409 --> 00:02:12.129 culture. And, you know, it sounds incredibly 00:02:12.129 --> 00:02:15.469 niche, almost mundane on the surface, but the 00:02:15.469 --> 00:02:18.210 skeletal page acts as a perfect microcosm for 00:02:18.210 --> 00:02:20.629 how human knowledge is structured. Totally. We're 00:02:20.629 --> 00:02:23.270 basically analyzing the container that holds 00:02:23.270 --> 00:02:26.310 the career rather than the career itself. OK. 00:02:26.330 --> 00:02:28.090 So let's start with the primary function of this 00:02:28.090 --> 00:02:30.090 page. Right at the bottom, it categorizes itself. 00:02:30.129 --> 00:02:33.990 It says, category, set index articles. Right. 00:02:34.080 --> 00:02:36.560 And what this page essentially does is take Jimmy 00:02:36.560 --> 00:02:39.379 Fallon's entire late night sketch catalog and 00:02:39.379 --> 00:02:41.360 split it right down the middle into two completely 00:02:41.360 --> 00:02:44.479 distinct lists. You have the list of late night 00:02:44.479 --> 00:02:46.680 with Jimmy Fallon games and sketches, and then 00:02:46.680 --> 00:02:49.319 you have the list of the tonight show starring 00:02:49.319 --> 00:02:51.759 Jimmy Fallon games and sketches. Yeah, and in 00:02:51.759 --> 00:02:55.060 the taxonomy of online encyclopedias, a set index 00:02:55.060 --> 00:02:58.379 article isn't just a normal entry. And it's also, 00:02:58.460 --> 00:03:01.039 this is important, it's distinct from a standard 00:03:01.039 --> 00:03:04.080 disambiguation page. Wait, how so? Aren't they 00:03:04.080 --> 00:03:06.479 kind of doing the same thing? Well, sort of, 00:03:06.599 --> 00:03:08.419 but disambiguation pages are for things that 00:03:08.419 --> 00:03:11.099 just happen to share the exact same name by pure 00:03:11.099 --> 00:03:15.099 coincidence, like Apple the Fruit and Apple the 00:03:15.099 --> 00:03:17.020 Technology Company. Oh, right, because they have 00:03:17.020 --> 00:03:19.139 nothing to do with each other. Exactly. But a 00:03:19.139 --> 00:03:21.620 set index is designed for items that are intrinsically 00:03:21.620 --> 00:03:24.379 related. They share similar names or fall under 00:03:24.379 --> 00:03:27.099 a tightly bound umbrella. But collectively, they've 00:03:27.099 --> 00:03:30.500 just become far too unwieldy to exist on a single 00:03:30.500 --> 00:03:33.120 page. OK, I get it. It makes me think of, like, 00:03:33.449 --> 00:03:36.849 a musician's discography. Oh, that's a good comparison. 00:03:36.909 --> 00:03:39.449 Yeah, because fans will fiercely categorize a 00:03:39.449 --> 00:03:42.210 band's eras, right? You've got the gritty, experimental 00:03:42.210 --> 00:03:45.150 indie years, and then there's the highly polished 00:03:45.150 --> 00:03:47.830 stadium rock era. Yes, exactly. And you don't 00:03:47.830 --> 00:03:49.789 just mix all those tracks together in one playlist 00:03:49.789 --> 00:03:52.009 if you're a serious archivist. You separate them 00:03:52.009 --> 00:03:53.990 because the context of their creation matters. 00:03:54.389 --> 00:03:57.110 Right, it respects the distinct eat box of the 00:03:57.110 --> 00:04:00.169 creative output. And here's my pushback on this. 00:04:00.699 --> 00:04:04.340 Why couldn't this just be one massive chronological 00:04:04.340 --> 00:04:07.500 list? I mean, Jimmy Fallon is Jimmy Fallon. Sure. 00:04:08.360 --> 00:04:11.240 If he plays a game on late night and then he 00:04:11.240 --> 00:04:13.439 plays that exact same game years later on The 00:04:13.439 --> 00:04:16.980 Tonight Show, does shifting time slots really 00:04:16.980 --> 00:04:19.139 fundamentally change the categorization of the 00:04:19.139 --> 00:04:22.209 game? Or is Wikipedia just being, I don't know, 00:04:22.370 --> 00:04:24.589 overly bureaucratic? It's a fair question, but 00:04:24.589 --> 00:04:27.509 it isn't bureaucracy, actually. It's a structural 00:04:27.509 --> 00:04:30.209 necessity driven by the philosophy of broadcast 00:04:30.209 --> 00:04:32.829 history. OK, how so? Well, first, just consider 00:04:32.829 --> 00:04:35.050 the sheer volume of content. I mean, a daily 00:04:35.050 --> 00:04:37.370 late night show produces an astronomical amount 00:04:37.370 --> 00:04:39.410 of discrete segments over the years. Oh, yeah, 00:04:39.529 --> 00:04:41.310 thousands of episodes. Right. If you put that 00:04:41.310 --> 00:04:43.750 all on one page, the user experience breaks down 00:04:43.750 --> 00:04:46.149 completely. The browser struggles to load it. 00:04:46.529 --> 00:04:48.649 You know, the reader can't navigate it. It's 00:04:48.649 --> 00:04:50.689 just a mess. Nobody wants to scroll for 10 minutes. 00:04:51.189 --> 00:04:53.509 Exactly. But more importantly, Late Night and 00:04:53.509 --> 00:04:56.129 The Tonight Show are treated by television historians 00:04:56.129 --> 00:04:59.529 as entirely separate entities. So they aren't 00:04:59.529 --> 00:05:01.689 just different titles for the same guy sitting 00:05:01.689 --> 00:05:03.870 behind a desk. Not at all. They have different 00:05:03.870 --> 00:05:05.910 production codes, different legacy lineages, 00:05:06.110 --> 00:05:08.610 different network mandates. Think about it. Late 00:05:08.610 --> 00:05:11.750 Night was a franchise built by David Letterman, 00:05:12.110 --> 00:05:14.990 then Conan O 'Brien before Fallon even got there. 00:05:15.069 --> 00:05:18.949 Right. And The Tonight Show is like an institution. 00:05:19.100 --> 00:05:22.620 Yes. It carries the historical weight of Steve 00:05:22.620 --> 00:05:26.759 Allen, Jack Parr, and Johnny Carson. So by forcing 00:05:26.759 --> 00:05:29.579 this split, the set index article refuses to 00:05:29.579 --> 00:05:33.439 treat a single comedian's career as one continuous 00:05:33.439 --> 00:05:36.939 blurred timeline. It draws a hard line in the 00:05:36.939 --> 00:05:39.740 sand to preserve the institutional history of 00:05:39.740 --> 00:05:42.259 NBC's late night programming. OK, that actually 00:05:42.259 --> 00:05:44.060 makes a lot of sense. It's almost like a baseball 00:05:44.060 --> 00:05:46.660 player's statistics. Oh, yeah. Like, their stats 00:05:46.660 --> 00:05:49.370 for the minor leagues are kept completely separate 00:05:49.370 --> 00:05:51.430 from their stats in the major leagues, even though 00:05:51.430 --> 00:05:53.649 it's the exact same guy swinging the exact same 00:05:53.649 --> 00:05:55.610 bat. That's a great way to put it. So the set 00:05:55.610 --> 00:05:57.589 index is basically the manager making sure you 00:05:57.589 --> 00:05:59.550 know which ledger you're about to open before 00:05:59.550 --> 00:06:01.850 you start reading the numbers. Exactly. And this 00:06:01.850 --> 00:06:03.889 brings up the question of how users actually 00:06:03.889 --> 00:06:06.290 arrive at this ledger in the first place. Because 00:06:06.290 --> 00:06:09.110 the architecture of a doorway page is defined 00:06:09.110 --> 00:06:11.949 by the paths that lead to it. Right. And this 00:06:11.949 --> 00:06:13.970 is where the mechanics get wildly interactive. 00:06:14.290 --> 00:06:16.410 Yeah. I found this part so interesting. There 00:06:16.410 --> 00:06:18.750 are specific redirects that point directly to 00:06:18.750 --> 00:06:21.629 this page. Yes. Two of them in particular really 00:06:21.629 --> 00:06:24.310 stand out, first drafts of rock and history of 00:06:24.310 --> 00:06:27.290 rap. So if a user types the phrase first drafts 00:06:27.290 --> 00:06:30.050 of rock into the search bar, they don't get a 00:06:30.050 --> 00:06:32.410 page about the literal first drafts of rock music 00:06:32.410 --> 00:06:35.029 or unreleased studio tapes from the 1960s. Which 00:06:35.029 --> 00:06:37.629 you might expect them to. Right. Instead, the 00:06:37.629 --> 00:06:40.509 system automatically intercepts that query and 00:06:40.509 --> 00:06:42.750 redirects them to this Jimmy Fallon directory 00:06:42.750 --> 00:06:44.870 page, because that was the title of a popular 00:06:44.870 --> 00:06:47.490 sketch series he did. The encyclopedia is basically 00:06:47.490 --> 00:06:50.970 attempting to anticipate user intent. But that 00:06:50.970 --> 00:06:54.569 second redirect, history of rap, causes a massive 00:06:54.569 --> 00:06:57.629 multi -layered collision. Like, the disambiguation 00:06:57.629 --> 00:07:00.189 warning at the very top of this page is highly 00:07:00.189 --> 00:07:03.310 specific. It really is. It explicitly says... 00:07:03.420 --> 00:07:06.360 And I'm quoting here, history of rap redirects 00:07:06.360 --> 00:07:09.240 here. For the history of rap as a musical genre, 00:07:09.540 --> 00:07:11.680 see hip hop music. For the history of rapping 00:07:11.680 --> 00:07:14.860 as a vocal technique, see rapping section history. 00:07:15.040 --> 00:07:17.779 It is such a brilliant piece of traffic control. 00:07:17.899 --> 00:07:20.000 It's honestly absurd when you think about the 00:07:20.000 --> 00:07:22.439 real world implications of it. Oh, totally. The 00:07:22.439 --> 00:07:25.100 information ecosystem has to actively intercept 00:07:25.100 --> 00:07:29.829 people. who are trying to research the profound 00:07:29.829 --> 00:07:34.029 historical origins of a global music genre, just 00:07:34.029 --> 00:07:36.310 in case they accidentally stumbled into a directory 00:07:36.310 --> 00:07:38.410 of Jimmy Fallon comedy sketches. Yeah, it has 00:07:38.410 --> 00:07:40.410 to catch them before they get lost. It's like, 00:07:40.410 --> 00:07:42.250 I don't know, it's like a comedian naming a sketch 00:07:42.250 --> 00:07:45.329 The Industrial Revolution, and suddenly the entire 00:07:45.329 --> 00:07:47.230 encyclopedia has a panic attack trying to route 00:07:47.230 --> 00:07:49.329 traffic. Yeah. I mean, how often does pop culture 00:07:49.329 --> 00:07:51.810 hijack actual historical terminology like this? 00:07:51.889 --> 00:07:54.569 Honestly, it happens constantly. And it represents 00:07:54.569 --> 00:07:57.410 one of the greatest on -go - challenges in information 00:07:57.410 --> 00:07:59.990 architecture. Oh yeah, you have a late -night 00:07:59.990 --> 00:08:04.600 comedy bit. that decided to use the exact nomenclature, 00:08:04.680 --> 00:08:07.699 the literal phrase, history of rap, that describes 00:08:07.699 --> 00:08:10.439 a major cultural movement. And the thing is, 00:08:10.519 --> 00:08:14.199 the database is entirely blind to human motivation. 00:08:14.399 --> 00:08:15.800 Right. It doesn't know why you're typing it. 00:08:16.019 --> 00:08:18.939 Exactly. When you type history of rap into a 00:08:18.939 --> 00:08:21.639 search bar, the system doesn't know your context. 00:08:21.939 --> 00:08:24.139 It doesn't know if you're, say, a high school 00:08:24.139 --> 00:08:27.160 student writing a musicology paper on the 1970s 00:08:27.160 --> 00:08:29.660 Bronx, or if you're just a bored office worker 00:08:29.660 --> 00:08:32.340 looking for a vibe. from last night's broadcast. 00:08:32.580 --> 00:08:35.080 So it has to assume every possible state of the 00:08:35.080 --> 00:08:37.679 user simultaneously. Yes, exactly. And notice 00:08:37.679 --> 00:08:40.240 how surgically precise that warning text is. 00:08:40.500 --> 00:08:42.059 It doesn't just say, not what you're looking 00:08:42.059 --> 00:08:43.679 for, click here. Right, it doesn't leave you 00:08:43.679 --> 00:08:46.600 hanging. No, it acts as an educational tool right 00:08:46.600 --> 00:08:48.840 there in the warning banner. It cleanly separates 00:08:48.840 --> 00:08:51.019 the sketch from the broader historical reality. 00:08:51.480 --> 00:08:53.539 But then, and this is the cool part, it goes 00:08:53.539 --> 00:08:55.940 a step further and separates the musical genre 00:08:55.940 --> 00:08:58.659 from the specific vocal technique. Oh, wow. It 00:08:58.659 --> 00:09:00.860 directs you to the hip. pop music article for 00:09:00.860 --> 00:09:04.100 the genre, but directs you to a specific subsection 00:09:04.100 --> 00:09:06.460 of the rapping article for the vocal technique. 00:09:06.679 --> 00:09:10.080 That is so wild. It defines the nuanced difference 00:09:10.080 --> 00:09:12.879 between rap the genre and rapping the technique, 00:09:13.399 --> 00:09:15.419 all while you're literally just standing on the 00:09:15.419 --> 00:09:17.200 doorstep of a late night television directory. 00:09:17.360 --> 00:09:19.259 It really highlights how much work goes into 00:09:19.259 --> 00:09:21.840 this stuff. For sure. Because if this traffic 00:09:21.840 --> 00:09:24.539 cop didn't exist, the collision of pop culture 00:09:24.539 --> 00:09:27.500 and actual history would create a deeply confusing 00:09:27.500 --> 00:09:32.200 user experience. Human entertainment could inadvertently 00:09:32.200 --> 00:09:36.259 obscure actual historical facts in the digital 00:09:36.259 --> 00:09:38.399 record. And the directory ensures that doesn't 00:09:38.399 --> 00:09:41.509 happen. It manages the overlap of exact nomenclature 00:09:41.509 --> 00:09:43.929 without letting one concept cannibalize the other. 00:09:44.370 --> 00:09:46.210 OK, but intercepting traffic and routing users 00:09:46.210 --> 00:09:48.289 is only half the battle, right? Right. Because 00:09:48.289 --> 00:09:51.210 if we look under the hood of this specific page, 00:09:51.809 --> 00:09:53.870 there's a whole secondary layer of invisible 00:09:53.870 --> 00:09:56.950 labor keeping these links alive. Oh, the backend 00:09:56.950 --> 00:09:58.889 metadata. This is where it gets really good. 00:09:59.129 --> 00:10:01.190 Yeah. We have the visible structure, the links 00:10:01.190 --> 00:10:03.889 you slick, the warnings you read. But the backend 00:10:03.889 --> 00:10:06.009 metadata tells a completely different story. 00:10:06.200 --> 00:10:08.940 about how information is maintained. And there 00:10:08.940 --> 00:10:13.019 are two specific pieces of hidden data here that 00:10:13.019 --> 00:10:16.500 really highlight this. First, this page was edited 00:10:16.500 --> 00:10:19.460 very recently. It was last edited on March 5, 00:10:19.639 --> 00:10:23.440 2025. And second, There's a hidden category note 00:10:23.440 --> 00:10:26.259 attached to the page that simply reads, short 00:10:26.259 --> 00:10:28.559 description is different from wiki data. And 00:10:28.559 --> 00:10:30.940 those two tiny details are incredibly telling 00:10:30.940 --> 00:10:33.419 about the nature of knowledge curation. Right. 00:10:33.740 --> 00:10:36.460 Because, OK, if this page is literally just a 00:10:36.460 --> 00:10:39.039 doorway pointing to two lists, if its entire 00:10:39.039 --> 00:10:40.919 function is just to say, go left for late night 00:10:40.919 --> 00:10:43.379 and go right for the Tonight Show, what is there 00:10:43.379 --> 00:10:46.240 left to edit in March of 2025? It's not like 00:10:46.240 --> 00:10:48.580 the basic premise of a directory changes over 00:10:48.580 --> 00:10:51.860 time. Are they just fixing typos, or is the architecture 00:10:51.860 --> 00:10:53.740 of the site itself just shifting under our feet? 00:10:53.840 --> 00:10:55.919 Well, it's the latter, really. We tend to think 00:10:55.919 --> 00:10:58.559 of encyclopedias as these static objects, like 00:10:58.559 --> 00:11:01.080 a book gets printed, put on a shelf, and the 00:11:01.080 --> 00:11:03.000 text stays the same forever. Right, it's done. 00:11:03.299 --> 00:11:06.200 But digital curation is a living, breathing, 00:11:06.460 --> 00:11:09.899 continuous process. The fact that a simple directory 00:11:09.899 --> 00:11:13.220 requires an edit in 2025 proves that maintenance 00:11:13.220 --> 00:11:16.399 never stops. I mean, an editor might be fixing 00:11:16.399 --> 00:11:18.759 a broken link or maybe a new redirect needed 00:11:18.759 --> 00:11:22.360 to be added, but more likely, the global categorization 00:11:22.360 --> 00:11:25.080 standards of the entire platform might have updated. 00:11:25.279 --> 00:11:27.980 Like a site -wide update. Exactly. And that would 00:11:27.980 --> 00:11:30.639 require a manual tweak to the backend code of 00:11:30.639 --> 00:11:33.279 this specific page just to keep it compliant 00:11:33.279 --> 00:11:35.570 with the wider ecosystem. It makes me think of 00:11:35.570 --> 00:11:38.809 Wikidata and these hidden categories as like 00:11:38.809 --> 00:11:41.470 the backstage crew of a massive theater production. 00:11:41.549 --> 00:11:43.570 Oh, absolutely. You know, you as the audience 00:11:43.570 --> 00:11:45.450 member, you're sitting in the dark watching the 00:11:45.450 --> 00:11:47.490 actors, which are the articles themselves, just 00:11:47.490 --> 00:11:49.610 doing their thing in the spotlight. Right. You 00:11:49.610 --> 00:11:51.830 never see the crew in all black pulling the ropes, 00:11:52.049 --> 00:11:54.409 adjusting the lighting, moving the sets. But 00:11:54.409 --> 00:11:57.269 if they stop working, the entire show just collapses 00:11:57.269 --> 00:12:00.370 on stage. That scaffolding is exactly what holds 00:12:00.370 --> 00:12:03.259 the visible Internet up. And that hidden category 00:12:03.259 --> 00:12:05.259 note you mentioned, short description is different 00:12:05.259 --> 00:12:09.139 from Wikidata, reveals a really fascinating friction 00:12:09.139 --> 00:12:12.019 in how that scaffolding operates. OK, explain 00:12:12.019 --> 00:12:15.720 that. Because what even is Wikidata? Well, for 00:12:15.720 --> 00:12:17.659 anyone unfamiliar with the back end of these 00:12:17.659 --> 00:12:20.279 platforms, Wikidata is a separate but deeply 00:12:20.279 --> 00:12:23.759 interconnected database. Wikipedia is meant to 00:12:23.759 --> 00:12:27.720 be read by humans. It has prose, sentences, paragraphs. 00:12:27.980 --> 00:12:29.940 Right, it's written in English. Whatever language. 00:12:30.480 --> 00:12:33.000 Exactly. But Wikidata is meant to be read entirely 00:12:33.000 --> 00:12:36.240 by machines. It is pure structured data organized 00:12:36.240 --> 00:12:40.179 into relationships. Oh, OK. So Wikipedia is like 00:12:40.179 --> 00:12:43.460 the text of a novel. And Wikidata is the barcode 00:12:43.460 --> 00:12:45.879 on the back of the dust jacket that tells the 00:12:45.879 --> 00:12:48.100 scanner what genre the book is. That is a perfect 00:12:48.100 --> 00:12:50.399 analogy. And every single article has a short 00:12:50.399 --> 00:12:52.220 description attached to it, which is just a little 00:12:52.220 --> 00:12:54.419 tag of a few words that explains what the page 00:12:54.419 --> 00:12:57.710 is at a glance. Ideally, the human -written description 00:12:57.710 --> 00:13:00.129 perfectly matches the machine -readable data 00:13:00.129 --> 00:13:02.889 on Wikidata. But here, they're out of sync. They 00:13:02.889 --> 00:13:04.830 don't match. So what are the actual consequences 00:13:04.830 --> 00:13:07.570 of that mismatch? Why does it matter if a few 00:13:07.570 --> 00:13:09.590 words are different? Does the internet break? 00:13:09.710 --> 00:13:12.750 It doesn't break, but it matters immensely because 00:13:12.750 --> 00:13:15.529 of how modern technology interacts with the semantic 00:13:15.529 --> 00:13:19.110 web. See, Wikidata is the backbone feeding information 00:13:19.110 --> 00:13:21.269 to search algorithms, artificial intelligence 00:13:21.269 --> 00:13:24.149 models, and virtual voice systems. Oh, wow. I 00:13:24.149 --> 00:13:26.789 didn't realize that. Yeah. So if you ask your 00:13:26.789 --> 00:13:29.409 smart speaker to pull up information on Jimmy 00:13:29.409 --> 00:13:32.149 Fallon's sketches, the assistant is actually 00:13:32.149 --> 00:13:35.789 querying Wikidata, not reading the prose on Wikipedia. 00:13:36.009 --> 00:13:38.409 Oh. So if the human description says, you know, 00:13:38.549 --> 00:13:41.230 list of bits by Jimmy Fallon, but the machine 00:13:41.230 --> 00:13:43.269 readable data is coded slightly differently. 00:13:43.710 --> 00:13:46.769 Say it's coded as television series episode list. 00:13:47.259 --> 00:13:49.940 The algorithm gets confused. Oh, I see. It might 00:13:49.940 --> 00:13:52.279 try to play a full episode on your TV instead 00:13:52.279 --> 00:13:54.120 of returning the text directory you actually 00:13:54.120 --> 00:13:56.220 wanted to look at. That makes total sense. So 00:13:56.220 --> 00:13:59.340 that hidden category tag is basically like an 00:13:59.340 --> 00:14:01.399 automated flare going up in the back end. Yes. 00:14:01.559 --> 00:14:04.820 It's signaling to the volunteer archivists, hey, 00:14:05.200 --> 00:14:07.899 we have a database disagreement here. The human 00:14:07.899 --> 00:14:10.019 categorization and the machine categorization 00:14:10.019 --> 00:14:12.279 have drifted apart, and someone needs to come 00:14:12.279 --> 00:14:15.679 in and manually harmonize how we're defining 00:14:15.679 --> 00:14:18.309 this piece of pop culture. Exactly. It proves 00:14:18.309 --> 00:14:20.769 that categorizing human entertainment is not 00:14:20.769 --> 00:14:24.009 an exact science at all. It is highly subjective. 00:14:24.710 --> 00:14:26.809 The terminology one person uses to summarize 00:14:26.809 --> 00:14:29.950 a page can easily conflict with the rigid formatting 00:14:29.950 --> 00:14:32.250 required by an algorithm. It's just so messy. 00:14:32.440 --> 00:14:35.220 It is. And the invisible labor of thousands of 00:14:35.220 --> 00:14:37.759 editors is constantly working to resolve those 00:14:37.759 --> 00:14:40.480 micro -frictions so the end user never experiences 00:14:40.480 --> 00:14:42.759 a glitch. That completely changes how you view 00:14:42.759 --> 00:14:44.940 a simple internet search. Doesn't it? Yeah. I 00:14:44.940 --> 00:14:47.159 mean, you type in first drafts of rock, and boom, 00:14:47.259 --> 00:14:49.600 the page loads instantly. You just assume it 00:14:49.600 --> 00:14:51.639 works by magic. You don't realize there are traffic 00:14:51.639 --> 00:14:54.340 cops and stagehands and entire databases arguing 00:14:54.340 --> 00:14:56.379 with each other in the background just to deliver 00:14:56.379 --> 00:14:58.919 that one specific link to your screen. It truly 00:14:58.919 --> 00:15:01.620 is a monumental achievement of collaborative 00:15:01.620 --> 00:15:05.120 organization. Even the smallest, seemingly most 00:15:05.120 --> 00:15:08.519 insignificant pages are inextricably tied into 00:15:08.519 --> 00:15:12.000 this massive pulsating web of continuous curation 00:15:12.000 --> 00:15:14.259 and metadata. It's incredible. I mean, we started 00:15:14.259 --> 00:15:17.120 today with what looked like the most mundane 00:15:17.120 --> 00:15:19.960 technical page imaginable. Just a total skeleton 00:15:19.960 --> 00:15:22.299 of a page. Right. A set index article pointing 00:15:22.299 --> 00:15:24.759 to late night television bits. But by looking 00:15:24.759 --> 00:15:27.620 closely at its architecture. we discovered an 00:15:27.620 --> 00:15:31.100 incredibly rich intersection of forces. We saw 00:15:31.100 --> 00:15:33.820 how pop culture taxonomy respects the shifting 00:15:33.820 --> 00:15:37.100 eras of broadcast history, meticulously splitting 00:15:37.100 --> 00:15:39.659 late night from The Tonight Show to preserve 00:15:39.659 --> 00:15:42.159 institutional legacy. And we witnessed the highly 00:15:42.159 --> 00:15:44.899 necessary collision of historical disambiguation, 00:15:45.059 --> 00:15:47.399 where a comedy bit forces the platform to implement 00:15:47.399 --> 00:15:49.659 surgical traffic control so it doesn't accidentally 00:15:49.659 --> 00:15:52.700 overwrite the history of the entire hip hop genre. 00:15:52.799 --> 00:15:54.240 Which is still my favorite part of this whole 00:15:54.240 --> 00:15:57.299 thing. It's so good. And we've peeled back the 00:15:57.299 --> 00:15:59.860 curtain to look at the invisible labor, the ongoing 00:15:59.860 --> 00:16:02.679 maintenance required in 2025, and the database 00:16:02.679 --> 00:16:05.200 conflicts that prove organizing human knowledge 00:16:05.200 --> 00:16:09.039 is just a messy, subjective process of harmonizing 00:16:09.039 --> 00:16:12.139 human language with machine -readable data. It 00:16:12.139 --> 00:16:14.159 really makes you appreciate the doorway itself. 00:16:14.460 --> 00:16:17.639 Once you see the scaffolding holding the information 00:16:17.639 --> 00:16:19.860 ecosystem together, you just can't unsee it. 00:16:19.899 --> 00:16:22.059 You really can't. So here's something to mull 00:16:22.059 --> 00:16:23.679 over the next time you're surfing the web, clicking 00:16:23.679 --> 00:16:26.259 from link to link. We build these incredibly 00:16:26.259 --> 00:16:28.399 complex encyclopedias, and we assume they're 00:16:28.399 --> 00:16:30.200 for us right now in the present moment. Right. 00:16:30.360 --> 00:16:32.700 But imagine if this specific archive somehow 00:16:32.700 --> 00:16:35.539 survives intact for hundreds of years. Think 00:16:35.539 --> 00:16:38.580 about future historians centuries from now trying 00:16:38.580 --> 00:16:41.320 to interpret these specific disambiguation pages. 00:16:41.399 --> 00:16:44.399 Oh. Right. Like, will they understand the cultural 00:16:44.399 --> 00:16:46.320 context? Will they look at the architecture of 00:16:46.320 --> 00:16:49.159 the database and assume that the history of rap 00:16:49.159 --> 00:16:52.659 late night sketch was just as historically significant 00:16:52.659 --> 00:16:56.389 as the actual music? simply because the system 00:16:56.389 --> 00:16:59.830 had to treat them as equals, requiring disambiguation. 00:16:59.990 --> 00:17:03.049 That's a wild thought. What do our virtual doorways 00:17:03.049 --> 00:17:05.150 say about the world we are living in right now? 00:17:06.089 --> 00:17:07.890 The next time you find yourself stuck in a rabbit 00:17:07.890 --> 00:17:10.289 hole, don't just look at where you ended up. 00:17:11.009 --> 00:17:12.769 Pay attention to how the internet guided you 00:17:12.769 --> 00:17:13.049 there.