WEBVTT 00:00:00.000 --> 00:00:04.379 I have a confession to make. Last night at approximately 00:00:04.379 --> 00:00:07.299 1 .30 in the morning, I was lying in bed, you 00:00:07.299 --> 00:00:09.580 know, blue light blasting into my retinas, and 00:00:09.580 --> 00:00:11.580 I saw a link. It was a picture of a child star 00:00:11.580 --> 00:00:14.500 I haven't thought about since maybe 1998, and 00:00:14.500 --> 00:00:16.820 the headline was just three words, you won't 00:00:16.820 --> 00:00:19.879 believe. And let me guess, you clicked. I clicked. 00:00:19.940 --> 00:00:22.140 Of course I clicked. I mean, I had to know. And 00:00:22.140 --> 00:00:25.140 was the result believable? Yes. It was completely, 00:00:25.239 --> 00:00:27.179 utterly mundane. They just, you know, grew up 00:00:27.179 --> 00:00:29.320 and became a real estate agent. But in that moment. 00:00:29.850 --> 00:00:33.689 I was possessed. Yeah. Welcome back to the Deep 00:00:33.689 --> 00:00:35.929 Dive. I'm your host. And today we are tackling 00:00:35.929 --> 00:00:38.369 the very thing that stole, I don't know, a good 00:00:38.369 --> 00:00:40.789 10 minutes of my sleep last night. We are talking 00:00:40.789 --> 00:00:43.530 about the Internet's favorite trap, clickbait. 00:00:43.710 --> 00:00:46.409 It's the mosquito of the digital age. It really 00:00:46.409 --> 00:00:48.729 is. We all claim to hate it. We roll our eyes 00:00:48.729 --> 00:00:51.429 at it. We mock the one weird trick ads. And yet 00:00:51.429 --> 00:00:53.210 the numbers, which we're definitely going to 00:00:53.210 --> 00:00:56.030 get into, suggest that we are all collectively. 00:00:56.700 --> 00:00:58.899 addicted to it. Well, it's a fascinating paradox, 00:00:59.079 --> 00:01:01.219 isn't it? We say we want high quality, nuanced 00:01:01.219 --> 00:01:04.900 information, but our thumbs, our thumbs vote 00:01:04.900 --> 00:01:08.099 for the sensational. So today isn't just about 00:01:08.099 --> 00:01:10.239 making fun of bad headlines, though. Let's be 00:01:10.239 --> 00:01:12.180 honest. We will definitely be doing some of that. 00:01:12.299 --> 00:01:14.719 It's a deep dive into the mechanics of attention. 00:01:15.299 --> 00:01:17.859 We're going to look at the psychology that makes 00:01:17.859 --> 00:01:21.400 that curiosity gap so. So painful to ignore. 00:01:22.099 --> 00:01:24.900 We're going to trace the history back way further 00:01:24.900 --> 00:01:27.480 than I expected, like long before the Internet 00:01:27.480 --> 00:01:30.099 even existed. And we really have to look at the 00:01:30.099 --> 00:01:32.400 economics. Because that headline you clicked 00:01:32.400 --> 00:01:34.620 last night wasn't an accident. It was the tip 00:01:34.620 --> 00:01:37.680 of a very large, very expensive iceberg designed 00:01:37.680 --> 00:01:40.799 to monetize your, well, your cognitive vulnerabilities. 00:01:41.180 --> 00:01:43.760 That makes me feel slightly better about my lack 00:01:43.760 --> 00:01:46.500 of willpower. It's not me. It's the system. Well, 00:01:46.640 --> 00:01:48.799 let's not let you off the hook entirely just 00:01:48.799 --> 00:01:50.480 yet. We'll definitely get to the question of 00:01:50.480 --> 00:01:53.260 agency later. But to start, we really need to 00:01:53.260 --> 00:01:55.560 define what we are even talking about, because 00:01:55.560 --> 00:01:57.739 in preparing for this, I realized the definition 00:01:57.739 --> 00:02:00.920 itself is like the subject of a massive war in 00:02:00.920 --> 00:02:02.819 the media industry. Yeah, I saw that in the sources. 00:02:02.920 --> 00:02:05.760 There's this whole debate. Is a headline clickbait 00:02:05.760 --> 00:02:08.319 just because it's interesting or does it have 00:02:08.319 --> 00:02:11.479 to be a lie? Exactly. Is it hype or is it fraud? 00:02:11.680 --> 00:02:14.080 OK, let's unpack this. So if we look at the word 00:02:14.080 --> 00:02:17.000 itself, clickbait. It's incredibly visceral. 00:02:17.099 --> 00:02:19.280 It's an aggressive term. It comes, obviously, 00:02:19.400 --> 00:02:21.819 from fishing. Which, when you really stop and 00:02:21.819 --> 00:02:24.439 think about it, is a pretty dark analogy for 00:02:24.439 --> 00:02:27.020 how we consume content. It is. I mean, think 00:02:27.020 --> 00:02:29.379 about the mechanics of fishing. You have a hook. 00:02:29.539 --> 00:02:32.680 That hook is sharp, it is dangerous, and its 00:02:32.680 --> 00:02:35.900 sole purpose is to capture the fish. And no fish 00:02:35.900 --> 00:02:38.240 in its right mind is going to bite a bare metal 00:02:38.240 --> 00:02:40.819 hook. Right. If I see a bear link that just says, 00:02:40.900 --> 00:02:43.620 we want to show you ads so we make money, I'm... 00:02:43.819 --> 00:02:46.319 You know, I'm swimming right past that. Precisely. 00:02:46.319 --> 00:02:48.719 So you have to disguise the hook. You put a worm 00:02:48.719 --> 00:02:51.860 on it. You put a shiny lure on it. You create 00:02:51.860 --> 00:02:54.699 the illusion of value a meal. The fish bites 00:02:54.699 --> 00:02:57.520 the meal, but the reality is the hook. And in 00:02:57.520 --> 00:03:00.780 this analogy, the meal is the headline or like 00:03:00.780 --> 00:03:03.060 the thumbnail image, and the hook is the actual 00:03:03.060 --> 00:03:05.900 page view. Yes. The publisher needs you to swallow 00:03:05.900 --> 00:03:08.060 the hook to follow the link so they can serve 00:03:08.060 --> 00:03:10.419 that impression to the advertiser. Once you are 00:03:10.419 --> 00:03:12.879 on the page, you're caught. Whether you enjoy 00:03:12.879 --> 00:03:15.379 the content is, well, it's actually secondary 00:03:15.379 --> 00:03:17.939 to the fact that you arrived. That's the core 00:03:17.939 --> 00:03:21.419 cynicism of it all. But there's a specific psychological 00:03:21.419 --> 00:03:25.259 trigger that makes the bait work so well. It's 00:03:25.259 --> 00:03:28.159 not just, ooh, shiny. It's something called the 00:03:28.159 --> 00:03:30.699 curiosity gap. And I was reading up on this. 00:03:30.900 --> 00:03:33.199 It's actually rooted in some pretty heavy behavioral 00:03:33.199 --> 00:03:36.159 science. It is. Yeah. This concept was largely 00:03:36.159 --> 00:03:38.539 popularized by George Lowenstein, who's a behavioral 00:03:38.539 --> 00:03:41.240 economist at Carnegie Mellon. He developed what's 00:03:41.240 --> 00:03:43.819 called the information gap theory. And the theory 00:03:43.819 --> 00:03:45.919 is basically that not knowing something physically 00:03:45.919 --> 00:03:48.900 hurts. Right. Is that it? It's close. Lowenstein 00:03:48.900 --> 00:03:51.300 argued that curiosity isn't just a mental state. 00:03:51.400 --> 00:03:55.710 It's a form of. cognitive deprivation. When you 00:03:55.710 --> 00:03:58.030 perceive a gap between what you know and what 00:03:58.030 --> 00:04:00.569 you want to know, it creates this feeling of 00:04:00.569 --> 00:04:03.150 deprivation. It's like an itch. That is exactly 00:04:03.150 --> 00:04:04.990 what it feels like. When I saw that headline 00:04:04.990 --> 00:04:07.009 about the child star, it wasn't that I genuinely 00:04:07.009 --> 00:04:08.990 cared about that specific actor. It was that 00:04:08.990 --> 00:04:11.250 the headline created a hole in my knowledge. 00:04:11.430 --> 00:04:13.770 You won't believe what happened. Well, now I 00:04:13.770 --> 00:04:15.449 have to know what happened or I can't go to sleep. 00:04:15.879 --> 00:04:18.740 And that's the dopamine loop. The headline creates 00:04:18.740 --> 00:04:21.220 a tension. That's the gap. Clicking the link 00:04:21.220 --> 00:04:23.579 resolves the tension. It gives you a tiny little 00:04:23.579 --> 00:04:26.800 hit of dopamine relief. It's not about gaining 00:04:26.800 --> 00:04:29.220 knowledge. It's about scratching the itch. So 00:04:29.220 --> 00:04:31.920 the classic clickbait structure. This man found 00:04:31.920 --> 00:04:33.819 a box in his wall and you won't believe what 00:04:33.819 --> 00:04:36.379 was inside. That's just weaponizing that sense 00:04:36.379 --> 00:04:38.920 of deprivation. Perfectly. I mean, if the headline 00:04:38.920 --> 00:04:42.160 said man finds old coins and wall, the gap is 00:04:42.160 --> 00:04:44.620 closed. You have the information. You don't need 00:04:44.620 --> 00:04:47.519 to click. By withholding that one key piece of 00:04:47.519 --> 00:04:50.000 information. The what? They hold your attention 00:04:50.000 --> 00:04:52.720 hostage. But okay, here is where I get stuck. 00:04:53.120 --> 00:04:55.480 Because if the definition of clickbait is just 00:04:55.480 --> 00:04:58.779 withholding information to create suspense, then 00:04:58.779 --> 00:05:01.610 isn't every mystery novel clickbait? Isn't the 00:05:01.610 --> 00:05:03.649 trailer for the new Marvel movie technically 00:05:03.649 --> 00:05:06.610 clickbait? This is the absolute crux of the debate. 00:05:06.689 --> 00:05:09.329 Where is the line between good storytelling and, 00:05:09.370 --> 00:05:11.730 you know, manipulation? And this isn't just an 00:05:11.730 --> 00:05:13.829 academic question. This was the central conflict 00:05:13.829 --> 00:05:16.370 between two giants of the Internet age, BuzzFeed 00:05:16.370 --> 00:05:18.470 and Facebook. Right. This was back around 2014, 00:05:18.810 --> 00:05:21.420 wasn't it? The golden age of lists. The era of 00:05:21.420 --> 00:05:24.420 the listicle, exactly. And Ben Smith, who was 00:05:24.420 --> 00:05:26.300 the editor in chief of BuzzFeed at the time, 00:05:26.399 --> 00:05:29.000 made a very famous, very public defense of his 00:05:29.000 --> 00:05:32.579 company. He argued that BuzzFeed did not do clickbait. 00:05:33.500 --> 00:05:37.240 Which, that sounds insane. I mean, BuzzFeed was 00:05:37.240 --> 00:05:39.000 the factory that built this stuff. Listen to 00:05:39.000 --> 00:05:41.600 his logic for a second. Smith defined clickbait 00:05:41.600 --> 00:05:45.670 very, very strictly as dishonesty. He said, clickbait 00:05:45.670 --> 00:05:47.810 is when you promise one thing and deliver another. 00:05:48.110 --> 00:05:51.089 So the bait and switch. Exactly. He used a specific 00:05:51.089 --> 00:05:54.350 example. Imagine a headline that says, a five 00:05:54.350 --> 00:05:56.850 -year -old girl raised enough money to take her 00:05:56.850 --> 00:05:59.329 father, who has terminal cancer, to Disney World. 00:05:59.529 --> 00:06:01.810 Okay. Yeah, that grabs me. It's emotional. It's 00:06:01.810 --> 00:06:04.269 specific. I'd probably click. It's absolutely 00:06:04.269 --> 00:06:06.389 designed to go viral. But Smith's argument was 00:06:06.389 --> 00:06:08.790 this. Is the story actually about that? If you 00:06:08.790 --> 00:06:11.290 click the link and you read a true, well -reported 00:06:11.290 --> 00:06:13.589 story about that girl and her father, then the 00:06:13.589 --> 00:06:15.649 headline isn't bait. It's just a good headline. 00:06:15.829 --> 00:06:18.149 It delivered on its promise. I see his point. 00:06:18.269 --> 00:06:20.449 He's saying, don't blame us for packaging the 00:06:20.449 --> 00:06:23.000 truth in an interesting way. Exactly. He's basically 00:06:23.000 --> 00:06:25.300 saying that if the content itself is good, then 00:06:25.300 --> 00:06:27.800 the headline is justified no matter how emotional 00:06:27.800 --> 00:06:30.379 it is. But wait, hang on. I'm not sure I buy 00:06:30.379 --> 00:06:32.800 that completely. Because even if the story is 00:06:32.800 --> 00:06:34.660 true, isn't there something a little manipulative 00:06:34.660 --> 00:06:37.139 about the way it's framed? I mean, if you are 00:06:37.139 --> 00:06:39.500 explaining a personal tragedy just to get me 00:06:39.500 --> 00:06:44.540 to view an ad for car insurance, that feels slimy. 00:06:44.779 --> 00:06:47.360 You know, even if facts are straight. And that 00:06:47.360 --> 00:06:49.860 is precisely where Facebook came down on the 00:06:49.860 --> 00:06:52.670 issue. Facebook realized that their users felt 00:06:52.670 --> 00:06:54.750 manipulated, regardless of whether the story 00:06:54.750 --> 00:06:57.870 was technically true, so they defined clickbait 00:06:57.870 --> 00:07:00.670 differently. How so? For them, it was about headlines 00:07:00.670 --> 00:07:02.810 that encourage clicking but don't tell you what 00:07:02.810 --> 00:07:04.589 you will see. The you won't believe who just 00:07:04.589 --> 00:07:07.250 died style of headline. Exactly that. Facebook 00:07:07.250 --> 00:07:09.689 argued that withholding the core information 00:07:09.689 --> 00:07:12.689 is the problem. Even if the person who died is 00:07:12.689 --> 00:07:15.250 a major celebrity, making me click just to find 00:07:15.250 --> 00:07:17.689 out who it is creates a bad user experience. 00:07:17.829 --> 00:07:20.879 It wastes my time. It forces me to pay a click 00:07:20.879 --> 00:07:22.899 tax just to get the basic facts that should have 00:07:22.899 --> 00:07:25.240 been in the headline. Right. And the expert consensus 00:07:25.240 --> 00:07:27.720 now, well, it usually leans toward the disappointment 00:07:27.720 --> 00:07:31.620 factor. The most functional definition of clickbait 00:07:31.620 --> 00:07:34.620 today is a headline that intentionally over promises 00:07:34.620 --> 00:07:37.199 and under delivers. So it's about the ratio of 00:07:37.199 --> 00:07:39.699 promise to delivery. Right. If I promise you 00:07:39.699 --> 00:07:41.959 the secret to eternal life and the article says 00:07:41.959 --> 00:07:44.740 eat more kale, that is 100 percent clickbait. 00:07:45.000 --> 00:07:47.420 Because the gap between the promise, which is 00:07:47.420 --> 00:07:50.079 immortality, and the delivery, which is vegetables, 00:07:50.319 --> 00:07:53.579 is massive. And that disappointment, that feeling 00:07:53.579 --> 00:07:56.600 of being tricked, is what erodes trust over time. 00:07:56.800 --> 00:07:59.639 Yes. But, you know, talking about trust and sensationalism, 00:07:59.720 --> 00:08:02.180 we tend to treat this as a purely modern problem. 00:08:02.319 --> 00:08:04.000 Right. We act like the Internet came along and 00:08:04.000 --> 00:08:06.079 ruined the news. But when we looked into the 00:08:06.079 --> 00:08:09.079 history for this. Yeah. Wow. We are not the first 00:08:09.079 --> 00:08:10.800 generation to deal with it. Oh, not even close. 00:08:10.879 --> 00:08:13.420 Not by a long shot. If we connect this to the 00:08:13.420 --> 00:08:15.519 bigger picture, clickbait is just the latest 00:08:15.519 --> 00:08:17.759 technology for a very, very ancient practice. 00:08:18.279 --> 00:08:20.980 To really get it, we have to talk about the 1890s 00:08:20.980 --> 00:08:23.459 and the yellow journalism wars. This is one of 00:08:23.459 --> 00:08:25.759 my favorite parts of media history. This is Pulitzer 00:08:25.759 --> 00:08:28.480 versus Hearst, right? The big newspaper barons. 00:08:28.600 --> 00:08:31.399 The titans of the New York press, Joseph Pulitzer 00:08:31.399 --> 00:08:34.139 of the New York World, and William Randolph Hearst 00:08:34.139 --> 00:08:36.129 of the New York Journal. And it's so funny. We 00:08:36.129 --> 00:08:39.090 think of Pulitzer now as this highbrow prize 00:08:39.090 --> 00:08:42.090 for excellence in journalism. Which is deeply 00:08:42.090 --> 00:08:45.149 ironic because back then he was fighting a dirty, 00:08:45.250 --> 00:08:48.350 brutal war for circulation numbers. These papers 00:08:48.350 --> 00:08:51.009 were fighting for street sales. They needed people 00:08:51.009 --> 00:08:53.070 walking down the street to see a headline and 00:08:53.070 --> 00:08:55.330 buy the paper right then. So the front page was 00:08:55.330 --> 00:08:57.230 the thumbnail. The front page was the thumbnail. 00:08:57.289 --> 00:08:59.909 Exactly. And they didn't have algorithms. They 00:08:59.909 --> 00:09:02.570 had newsboys shouting the headlines on street 00:09:02.570 --> 00:09:05.679 corners. So the headlines had to be loud. They 00:09:05.679 --> 00:09:07.559 developed a style that focused almost entirely 00:09:07.559 --> 00:09:11.100 on crime, on scandal, disaster, and just wild 00:09:11.100 --> 00:09:13.179 exaggeration. The sources mentioned the Yellow 00:09:13.179 --> 00:09:15.019 Kid. That's where the name comes from, isn't 00:09:15.019 --> 00:09:17.200 it? Yeah, that's right. Hogan's Alley was this 00:09:17.200 --> 00:09:20.090 very early, very popular comic strip. It featured 00:09:20.090 --> 00:09:22.250 a kid in a yellow nightshirt. It was just wildly 00:09:22.250 --> 00:09:25.370 popular. And both papers fought over the cartoonist 00:09:25.370 --> 00:09:28.529 Richard Outcalt. The competition was so fierce 00:09:28.529 --> 00:09:31.070 that the papers themselves were just dubbed the 00:09:31.070 --> 00:09:33.450 yellow press. But it wasn't just cartoons. I 00:09:33.450 --> 00:09:35.169 mean, they were actually fabricating news stories 00:09:35.169 --> 00:09:37.610 from whole cloth. And this is where it gets really 00:09:37.610 --> 00:09:39.889 dangerous. The most famous example is the sinking 00:09:39.889 --> 00:09:43.509 of the USS Maine in 1898. The American battleship 00:09:43.509 --> 00:09:46.450 exploded in Havana Harbor, Cuba. Right. And at 00:09:46.450 --> 00:09:48.690 first, nobody knew why it happened. All the evidence 00:09:48.690 --> 00:09:51.409 now points to it likely being an accidental coal 00:09:51.409 --> 00:09:53.909 bunker explosion. But that's a boring headline. 00:09:54.250 --> 00:09:57.509 Coal dust accidentally ignites. Who's buying 00:09:57.509 --> 00:10:00.409 that paper? But Spanish mine blows up our ship. 00:10:00.789 --> 00:10:04.230 Now that sells. Hearst's paper immediately ran 00:10:04.230 --> 00:10:06.409 illustrations, completely fake illustrations 00:10:06.409 --> 00:10:09.070 of Spanish mines with cables attached to the 00:10:09.070 --> 00:10:11.710 ship. They ran the headline. Destruction of the 00:10:11.710 --> 00:10:13.990 warship Maine was the work of an enemy. They 00:10:13.990 --> 00:10:17.049 had zero proof, absolutely none. But it whipped 00:10:17.049 --> 00:10:19.929 the American public into such a frenzy that it 00:10:19.929 --> 00:10:22.210 arguably pushed the country directly into the 00:10:22.210 --> 00:10:24.529 Spanish -American War. That is, that's literally 00:10:24.529 --> 00:10:26.929 clickbait starting a war. It is the ultimate 00:10:26.929 --> 00:10:29.629 consequence of engagement optimization. They 00:10:29.629 --> 00:10:31.990 optimized for outrage and they got a war out 00:10:31.990 --> 00:10:34.309 of it. That really puts our current problems 00:10:34.309 --> 00:10:36.710 into perspective. At least, you know, a quiz 00:10:36.710 --> 00:10:40.059 titled, which potato are you? isn't launching 00:10:40.059 --> 00:10:43.799 any missiles. Not yet, anyway. Let's hope. Please 00:10:43.799 --> 00:10:45.899 don't jinx it. But there's another historical 00:10:45.899 --> 00:10:48.600 term that came up in the sources, checkbook journalism. 00:10:48.899 --> 00:10:51.700 Right. This is kind of a precursor to the modern 00:10:51.700 --> 00:10:54.799 celebrity gossip ecosystem. It's the practice 00:10:54.799 --> 00:10:57.379 of paying sources for information, often without 00:10:57.379 --> 00:11:00.000 any real verification. Which sounds fine on the 00:11:00.000 --> 00:11:02.139 surface, until you realize that if you pay for 00:11:02.139 --> 00:11:05.440 dirt, people will just manufacture dirt for you. 00:11:05.580 --> 00:11:09.190 Exactly. If I offer $10 ,000 for a story about 00:11:09.190 --> 00:11:11.850 a politician having an affair, suddenly everyone 00:11:11.850 --> 00:11:13.990 has a story about that politician having an affair. 00:11:14.250 --> 00:11:16.769 It completely removes the verification process. 00:11:17.190 --> 00:11:19.690 Howard Kurtz from the Washington Post wrote that 00:11:19.690 --> 00:11:22.429 this tabloid culture erased the old definitions 00:11:22.429 --> 00:11:25.490 of news. It prioritized the tawdry over the true 00:11:25.490 --> 00:11:28.049 simply because the tawdry was profitable. And 00:11:28.049 --> 00:11:30.549 that right there, that's the bridge to the Internet, 00:11:30.710 --> 00:11:34.029 the profit model. Because why do websites publish 00:11:34.029 --> 00:11:35.830 clickbait? It's not because they enjoy being 00:11:35.830 --> 00:11:38.169 annoying. It's because it pays. You have to follow 00:11:38.169 --> 00:11:40.710 the money. Always. So let's break down the economics 00:11:40.710 --> 00:11:43.549 of this. Because I think people generally understand 00:11:43.549 --> 00:11:46.070 ads, but they might not understand the sheer 00:11:46.070 --> 00:11:49.250 scale of it all. So the basic unit of the modern 00:11:49.250 --> 00:11:52.549 internet economy is the impression. When a page 00:11:52.549 --> 00:11:56.190 loads and an ad is shown to a user, tiny fraction 00:11:56.190 --> 00:11:58.980 of a cent changes hands. To make real money, 00:11:59.139 --> 00:12:02.700 you need millions or even billions of impressions. 00:12:02.960 --> 00:12:05.100 Which means you need millions and millions of 00:12:05.100 --> 00:12:07.860 clicks. And you need them cheaply. Think about 00:12:07.860 --> 00:12:10.320 it. If you spend three months investigating a 00:12:10.320 --> 00:12:12.879 complex financial scandal that costs thousands 00:12:12.879 --> 00:12:15.080 of dollars in reporter salaries, legal fees, 00:12:15.159 --> 00:12:18.090 and so on. If you write a list of 10 cats that 00:12:18.090 --> 00:12:20.730 look like Hitler, that takes 20 minutes. If both 00:12:20.730 --> 00:12:22.789 articles get the same number of clicks... The 00:12:22.789 --> 00:12:24.850 cats are infinitely more profitable. Infinitely. 00:12:25.309 --> 00:12:28.009 But it gets even more sophisticated when we look 00:12:28.009 --> 00:12:30.730 at video platforms like YouTube. The algorithm 00:12:30.730 --> 00:12:33.450 there has evolved in a very specific way, and 00:12:33.450 --> 00:12:35.870 it has completely shaped the content we see. 00:12:36.009 --> 00:12:38.570 The YouTube face, we've all seen it. The exaggerated 00:12:38.570 --> 00:12:41.490 open mouth, the wide eyes, the hands on the cheeks 00:12:41.490 --> 00:12:44.019 pointing at something with a big red arrow. That 00:12:44.019 --> 00:12:46.879 exists for a very specific reason, high arousal 00:12:46.879 --> 00:12:49.580 emotion. The sources highlight a really important 00:12:49.580 --> 00:12:52.240 statistic that was revealed at the Consumer Electronics 00:12:52.240 --> 00:12:55.519 Show. YouTube executives admitted that most watch 00:12:55.519 --> 00:12:57.500 time on the platform does not come from people 00:12:57.500 --> 00:12:59.559 actively searching for things. Wait, really? 00:12:59.700 --> 00:13:02.120 So it's not people typing in how to fix my sink? 00:13:02.379 --> 00:13:04.720 Well, that happens, but the vast majority of 00:13:04.720 --> 00:13:06.539 time people spend on the site is spent watching 00:13:06.539 --> 00:13:08.580 videos that the recommendation engines serve 00:13:08.580 --> 00:13:11.620 to them. It comes from the Up Next sidebar or 00:13:11.620 --> 00:13:15.250 the homepage. So the robot is the DJ. It's choosing 00:13:15.250 --> 00:13:18.250 the music for us. The robot is the DJ, the gatekeeper, 00:13:18.350 --> 00:13:20.850 and the kingmaker. And that robot looks at two 00:13:20.850 --> 00:13:23.669 main things above all else. Click -through rate, 00:13:23.769 --> 00:13:27.220 or CTR, and watch time. CTR is just the percentage 00:13:27.220 --> 00:13:28.919 of people who see the thumbnail and actually 00:13:28.919 --> 00:13:31.279 click on it. Right. If your video is shown to 00:13:31.279 --> 00:13:33.659 100 people on their homepage and 10 of them click, 00:13:33.879 --> 00:13:36.840 you have a 10 per CTR. If only one clicks, you 00:13:36.840 --> 00:13:39.460 have a 1 per CTR. The algorithm is absolutely 00:13:39.460 --> 00:13:42.559 ruthless. If your CTR is low, it just stops showing 00:13:42.559 --> 00:13:44.940 your video to people. It assumes the video is 00:13:44.940 --> 00:13:47.039 bad, or at least that the packaging is bad. So 00:13:47.039 --> 00:13:49.799 you have to be sensational. If you write a calm, 00:13:49.980 --> 00:13:52.580 accurate title like, My Thoughts on the New iPhone. 00:13:53.000 --> 00:13:55.860 And your competitor writes, the iPhone is dead. 00:13:56.000 --> 00:13:59.700 And they get the click. You lose. You basically 00:13:59.700 --> 00:14:02.360 vanish from the platform. It creates an intense 00:14:02.360 --> 00:14:05.519 evolutionary pressure. It's survival of the loudest. 00:14:06.100 --> 00:14:09.100 Creators aren't necessarily trying to be deceptive 00:14:09.100 --> 00:14:11.580 all the time. They are trying to survive in an 00:14:11.580 --> 00:14:14.940 ecosystem, an algorithm that fundamentally rewards 00:14:14.940 --> 00:14:17.919 hype. It's a constant arms race. And the weapon 00:14:17.919 --> 00:14:20.039 is the thumbnail. And this all spills over into 00:14:20.039 --> 00:14:22.179 what the platforms now call engagement bait. 00:14:22.480 --> 00:14:23.860 This is a little different from just getting 00:14:23.860 --> 00:14:25.879 you to click. This is about manipulating you 00:14:25.879 --> 00:14:28.480 once you are already there on the page. Oh, this 00:14:28.480 --> 00:14:30.159 is the stuff that drives me up the wall on Facebook. 00:14:30.379 --> 00:14:32.440 Let's talk about the taxonomy of bait. Yeah. 00:14:32.500 --> 00:14:34.179 We have a list here, and I feel like I need to 00:14:34.179 --> 00:14:36.480 complain about them one by one. Please, the floor 00:14:36.480 --> 00:14:40.340 is yours. Vent. Okay, first one. Vote baiting. 00:14:40.539 --> 00:14:42.480 This is when they post a picture and say, like 00:14:42.480 --> 00:14:45.340 for pizza, love for tacos. It seems so innocent, 00:14:45.519 --> 00:14:47.460 right? Just a fun little poll. But think about 00:14:47.460 --> 00:14:49.919 the algorithm. It just sees engagement. It sees 00:14:49.919 --> 00:14:51.960 you reacting. It doesn't know you're just voting 00:14:51.960 --> 00:14:54.720 for tacos. It thinks, wow, this user feels very 00:14:54.720 --> 00:14:57.139 strongly about this content. I should show this 00:14:57.139 --> 00:14:59.700 post to all of their friends. It hacks the social 00:14:59.700 --> 00:15:02.639 graph. It forces my friends to see my taco preference 00:15:02.639 --> 00:15:05.340 just so the page gets more reach. It feels so 00:15:05.340 --> 00:15:08.000 cheap. It is cheap and effective. Okay, next 00:15:08.000 --> 00:15:11.379 one. Tag baiting. Tag friend who looks like this 00:15:11.379 --> 00:15:14.399 potato. or tag a friend who would do this. This 00:15:14.399 --> 00:15:18.019 is arguably worse because it drags a non -consenting 00:15:18.019 --> 00:15:20.980 third party into the interaction. You are basically 00:15:20.980 --> 00:15:23.879 using your social capital, your friendship, to 00:15:23.879 --> 00:15:25.720 feed the page's reach. And then there's share 00:15:25.720 --> 00:15:28.220 baiting. This is the guilt trip one. Share this 00:15:28.220 --> 00:15:30.419 if you love your mom. Keep scrolling if you have 00:15:30.419 --> 00:15:33.139 no soul. That is just pure emotional manipulation, 00:15:33.460 --> 00:15:35.659 plain and simple. It exploits social anxiety. 00:15:35.860 --> 00:15:37.440 Nobody wants to be the person who hates their 00:15:37.440 --> 00:15:39.580 mom, so you share it. It weaponizes your own 00:15:39.580 --> 00:15:42.490 self -image against you. But the worst one, the 00:15:42.490 --> 00:15:44.309 one that I think is actually damaging to the 00:15:44.309 --> 00:15:47.929 world, is rage baiting. Ah, yes. The nuclear 00:15:47.929 --> 00:15:50.370 option of engagement. Explain how this works, 00:15:50.409 --> 00:15:52.649 because I feel like we are just drowning in this 00:15:52.649 --> 00:15:55.570 every single day. Well, psychologically, high 00:15:55.570 --> 00:15:58.870 arousal emotions travel much faster online than 00:15:58.870 --> 00:16:01.990 low arousal ones. Sadness is a low arousal emotion. 00:16:02.490 --> 00:16:06.389 Contentment is low arousal. But anger, anger 00:16:06.389 --> 00:16:09.830 is pure high octane energy. It makes you want 00:16:09.830 --> 00:16:12.409 to do something. And on the Internet, that something 00:16:12.409 --> 00:16:14.950 is usually commenting. I can't believe they said 00:16:14.950 --> 00:16:17.830 this or this is completely wrong. Exactly. So 00:16:17.830 --> 00:16:20.090 publishers will intentionally write a headline 00:16:20.090 --> 00:16:23.470 that is wrong or offensive or incredibly stupid 00:16:23.470 --> 00:16:25.950 specifically to get you to correct them in the 00:16:25.950 --> 00:16:28.409 comments. The typo strategy. I've seen this a 00:16:28.409 --> 00:16:31.159 million times. They spell a word wrong in the 00:16:31.159 --> 00:16:33.740 title on purpose so that 500 people will flood 00:16:33.740 --> 00:16:35.860 the comments with, you spelled their wrong, you 00:16:35.860 --> 00:16:38.519 idiot. And the algorithm doesn't care why they 00:16:38.519 --> 00:16:40.179 are commenting. It can't tell the difference 00:16:40.179 --> 00:16:42.340 between praise and outrage. It just sees 500 00:16:42.340 --> 00:16:44.899 comments and thinks, wow, this is a hot topic. 00:16:45.019 --> 00:16:46.940 Let's put it to the top of everyone's feed. Go 00:16:46.940 --> 00:16:50.500 viral. So we are literally rewarding stupidity. 00:16:50.750 --> 00:16:53.669 We are rewarding engagement, and it turns out 00:16:53.669 --> 00:16:56.590 that stupidity is very, very engaging. Okay, 00:16:56.649 --> 00:16:58.649 so talk to me about the aesthetic of all this, 00:16:58.769 --> 00:17:02.669 specifically the chum box. Ah, the bottom of 00:17:02.669 --> 00:17:04.900 the internet. The digital gutter. You know exactly 00:17:04.900 --> 00:17:06.400 what I'm talking about. You finish reading a 00:17:06.400 --> 00:17:08.700 legitimate news article, you scroll down past 00:17:08.700 --> 00:17:11.079 the end of the text, and suddenly it's a grid 00:17:11.079 --> 00:17:14.500 of the most grotesque, bizarre images imaginable. 00:17:14.599 --> 00:17:17.559 A 90 -year -old woman looks like this now. Doctors 00:17:17.559 --> 00:17:20.220 are begging you to throw out this one vegetable. 00:17:20.660 --> 00:17:23.059 And it's always a picture of, like, a diseased 00:17:23.059 --> 00:17:25.400 foot or something covered in holes. It's horrifying. 00:17:25.950 --> 00:17:28.670 They call that the chum box. And the name comes 00:17:28.670 --> 00:17:32.329 from chum, which is the fish bait made of chopped 00:17:32.329 --> 00:17:34.369 up fish guts and blood that you throw in the 00:17:34.369 --> 00:17:38.569 water to attract sharks. Gross, but very accurate. 00:17:38.789 --> 00:17:41.369 These are ad networks like Taboola or Outbrain. 00:17:41.369 --> 00:17:43.549 They use a very specific visual language. It's 00:17:43.549 --> 00:17:46.490 always designed to appeal to prurient interest 00:17:46.490 --> 00:17:50.569 or greed. Prurient meaning sex. Sex, bodily fluids, 00:17:50.829 --> 00:17:53.829 the gross out factor. get -rich -quick schemes, 00:17:54.130 --> 00:17:57.809 it appeals directly to the lizard brain. They 00:17:57.809 --> 00:18:00.670 use numbered lists and extreme information gaps. 00:18:00.970 --> 00:18:03.630 They're the absolute bottom feeders of the digital 00:18:03.630 --> 00:18:06.089 ecosystem, but they are incredibly profitable 00:18:06.089 --> 00:18:08.750 for the news sites that host them. It's amazing 00:18:08.750 --> 00:18:11.789 to me that a legitimate newspaper will have high 00:18:11.789 --> 00:18:14.230 -quality, Pulitzer -winning journalism up top, 00:18:14.329 --> 00:18:17.009 and then it'll sell the basement to the toe fungus 00:18:17.009 --> 00:18:19.750 people for ad revenue. It just shows you how 00:18:19.750 --> 00:18:21.990 desperate the revenue model for online news has 00:18:21.990 --> 00:18:23.670 become over the years. But there is a darker 00:18:23.670 --> 00:18:27.230 side to this curiosity gap than just bad ads 00:18:27.230 --> 00:18:29.950 and gross -out pictures. The sources we looked 00:18:29.950 --> 00:18:32.750 at mention fishing and malice. Yes, and this 00:18:32.750 --> 00:18:34.569 is important. We shouldn't think of clickbait 00:18:34.569 --> 00:18:36.750 as just an annoyance. It is a genuine vulnerability. 00:18:37.210 --> 00:18:39.829 And hackers use the exact same psychology to 00:18:39.829 --> 00:18:42.250 exploit us. The classic, check out this video 00:18:42.250 --> 00:18:44.650 of you, Scam. Exactly. If I send you an email 00:18:44.650 --> 00:18:47.170 that says... Attached is the quarterly financial 00:18:47.170 --> 00:18:49.269 report. You might open it, you know, eventually. 00:18:49.609 --> 00:18:51.589 But if I send an email with the subject line, 00:18:51.730 --> 00:18:54.769 urgent, is this you in this photoshopped image? 00:18:55.049 --> 00:18:57.490 Quo, your panic and your curiosity completely 00:18:57.490 --> 00:19:00.450 override your logical brain. You click the link 00:19:00.450 --> 00:19:02.829 and boom malware is installed on your machine. 00:19:03.049 --> 00:19:06.250 The curiosity gap is a security flaw in the human 00:19:06.250 --> 00:19:08.690 brain. And there is no software patch for it. 00:19:08.829 --> 00:19:11.029 That is a terrifying thought. But let's zoom 00:19:11.029 --> 00:19:15.049 out from the individual danger to the... The 00:19:15.049 --> 00:19:17.190 societal danger, we touched on rage bait a minute 00:19:17.190 --> 00:19:19.630 ago. Yeah. But when you scale that up to national 00:19:19.630 --> 00:19:21.930 politics, what happens? Well, here is where it 00:19:21.930 --> 00:19:23.750 gets really fascinating and deeply concerning. 00:19:24.269 --> 00:19:26.690 We enter the realm of what's now called post 00:19:26.690 --> 00:19:29.210 -truth politics. That phrase gets thrown around 00:19:29.210 --> 00:19:31.609 a lot. What does it actually mean in this specific 00:19:31.609 --> 00:19:34.210 context? It means that the emotional impact of 00:19:34.210 --> 00:19:36.329 a statement begins to matter more than its factual 00:19:36.329 --> 00:19:39.440 accuracy. Clickbait culture has trained us over 00:19:39.440 --> 00:19:42.039 years to react to the headline, not to the evidence 00:19:42.039 --> 00:19:44.539 within the article. Slate described the current 00:19:44.539 --> 00:19:47.079 media environment as an aggregation of outrage. 00:19:49.819 --> 00:19:52.599 That's a heavy concept. It means the entire ecosystem 00:19:52.599 --> 00:19:55.519 is designed to find, collect and amplify whatever 00:19:55.519 --> 00:19:58.539 makes people the maddest. Nuance doesn't go viral. 00:19:58.700 --> 00:20:01.420 A headline like politician proposes complex tax 00:20:01.420 --> 00:20:03.680 reform with potential pros and cons gets zero 00:20:03.680 --> 00:20:06.279 clicks. Politician wants to destroy your savings 00:20:06.279 --> 00:20:09.099 and steal your home. That gets a million clicks 00:20:09.099 --> 00:20:11.559 and shares. So the incentive structure itself 00:20:11.559 --> 00:20:14.500 eliminates the middle ground. It pushes everyone 00:20:14.500 --> 00:20:17.579 to the extremes. Katherine Viner, the editor 00:20:17.579 --> 00:20:19.720 in chief of The Guardian, wrote a really powerful 00:20:19.720 --> 00:20:22.480 piece about this. She said that chasing down 00:20:22.480 --> 00:20:25.259 cheap clicks at the expense of accuracy and veracity 00:20:25.259 --> 00:20:27.960 undermines the fundamental value of journalism. 00:20:28.240 --> 00:20:30.759 If the truth is boring and lies are profitable, 00:20:31.000 --> 00:20:33.420 the market will always favor the lies. And we 00:20:33.420 --> 00:20:35.160 see this across the political spectrum. This 00:20:35.160 --> 00:20:37.500 is not a left or right issue. Absolutely not. 00:20:37.660 --> 00:20:40.240 The sources are very clear on this. They list 00:20:40.240 --> 00:20:42.440 sites from Breitbart on the right to the Huffington 00:20:42.440 --> 00:20:45.779 Post on the left, from Salong to Town Hall. The 00:20:45.779 --> 00:20:48.519 business model is completely politically agnostic. 00:20:48.740 --> 00:20:51.519 Both sides discovered they could profit by producing 00:20:51.519 --> 00:20:54.079 these short, shareable pieces that offer simple, 00:20:54.200 --> 00:20:56.660 immediate moral judgments. It's confirmation 00:20:56.660 --> 00:20:59.970 bias. As a service. That's a perfect way to put 00:20:59.970 --> 00:21:02.049 it. They aren't trying to inform you. They are 00:21:02.049 --> 00:21:04.369 trying to affirm what you already believe and 00:21:04.369 --> 00:21:07.630 crucially make you angry at the other side. And, 00:21:07.730 --> 00:21:11.369 you know, I've noticed a new weird defense mechanism 00:21:11.369 --> 00:21:14.509 recently. People use the word clickbait as a 00:21:14.509 --> 00:21:17.250 shield now. What do you mean? Like if a news 00:21:17.250 --> 00:21:19.549 outlet publishes a story that is critical of 00:21:19.549 --> 00:21:21.890 a politician I happen to like, my first instinct 00:21:21.890 --> 00:21:24.210 and the first instinct of many people is to just 00:21:24.210 --> 00:21:26.319 say, oh, that's just clickbait. and dismiss it, 00:21:26.359 --> 00:21:28.940 even if it's a well -researched, thousand -word 00:21:28.940 --> 00:21:32.380 investigative piece. That is a very, very dangerous 00:21:32.380 --> 00:21:35.180 development because if clickbait just becomes 00:21:35.180 --> 00:21:38.940 a synonym for news I don't like, then we've completely 00:21:38.940 --> 00:21:41.799 lost the ability to distinguish truth from fiction. 00:21:42.019 --> 00:21:45.140 The real clickbait has flooded the zone with 00:21:45.140 --> 00:21:47.380 so much garbage that we've started to distrust 00:21:47.380 --> 00:21:50.460 the real thing, too. It's the boy who cried wolf, 00:21:50.619 --> 00:21:53.039 but the wolf is an algorithm. Precisely. And 00:21:53.039 --> 00:21:55.410 we're all tired of hearing the shouting. OK, 00:21:55.490 --> 00:21:57.549 this is getting a little depressing. I feel like 00:21:57.549 --> 00:22:00.490 we are trapped in this giant Skinner box, just 00:22:00.490 --> 00:22:03.230 pressing buttons for dopamine while the world 00:22:03.230 --> 00:22:06.269 outside burns. Is there any good news? Is there 00:22:06.269 --> 00:22:08.390 a backlash? Have people started fighting back? 00:22:08.609 --> 00:22:11.369 There is, yes. And delightfully, the resistance 00:22:11.369 --> 00:22:14.250 started with comedy. Satire is always the best 00:22:14.250 --> 00:22:16.029 weapon. Do you remember the website ClickHole? 00:22:16.170 --> 00:22:17.990 Oh, I love ClickHole. It's from the same people 00:22:17.990 --> 00:22:20.430 as The Onion, right? Yes. It launched around 00:22:20.430 --> 00:22:23.630 2014 at the peak of all this. It was a direct 00:22:23.630 --> 00:22:26.039 parody of... sites like Upworthy and BuzzFeed. 00:22:26.099 --> 00:22:29.319 They took the curiosity gap formula and pushed 00:22:29.319 --> 00:22:31.900 it to its absolute absurd limit. They had those 00:22:31.900 --> 00:22:34.539 amazing headlines like six kinds of hay that 00:22:34.539 --> 00:22:37.700 are horse food. Or my personal favorite. This 00:22:37.700 --> 00:22:39.640 stick of butter is left out at room temperature. 00:22:39.859 --> 00:22:42.539 You won't believe what happens next. And you 00:22:42.539 --> 00:22:44.660 click it and it's just a time lapse video of 00:22:44.660 --> 00:22:46.880 the butter melting. It was so brilliant because 00:22:46.880 --> 00:22:50.660 it was deconstructing the joke. It was holding 00:22:50.660 --> 00:22:53.079 a mirror up to all of us and saying, look how 00:22:53.079 --> 00:22:55.299 stupid this format is. Look what you're clicking 00:22:55.299 --> 00:22:57.380 on. And it really worked. It helped break the 00:22:57.380 --> 00:22:59.980 spell for a lot of people. It made us all realize 00:22:59.980 --> 00:23:02.279 how formulaic and predictable the manipulation 00:23:02.279 --> 00:23:04.559 had become. But we can't just laugh it away. 00:23:05.119 --> 00:23:07.119 The tech companies themselves had to step in, 00:23:07.140 --> 00:23:09.759 too, didn't they? And they did. We gave Facebook 00:23:09.759 --> 00:23:13.019 a bit of a hard time earlier, but in 2014, they 00:23:13.019 --> 00:23:15.519 made a massive change to the newsfeed algorithm 00:23:15.519 --> 00:23:18.160 that actually did help a lot. This was the big 00:23:18.160 --> 00:23:20.519 shift from just counting clicks to measuring 00:23:20.519 --> 00:23:23.440 time spent. Correct. Before 2014, a click was 00:23:23.440 --> 00:23:25.359 a click. If you click on a link, the algorithm 00:23:25.359 --> 00:23:28.829 considered the article. good. But Facebook realized 00:23:28.829 --> 00:23:31.250 that people were clicking, immediately realizing 00:23:31.250 --> 00:23:33.349 it was bait, and then bouncing right back to 00:23:33.349 --> 00:23:36.109 Facebook in a few seconds. The bounce rate, a 00:23:36.109 --> 00:23:38.990 sign of a dissatisfied user. So Facebook's engineers 00:23:38.990 --> 00:23:41.390 started measuring how long you stayed off site. 00:23:41.609 --> 00:23:44.250 If you clicked a link and stayed on that page 00:23:44.250 --> 00:23:46.390 for three minutes, they assumed you were actually 00:23:46.390 --> 00:23:48.869 reading the article. That's a signal of quality. 00:23:49.289 --> 00:23:51.650 If you came back in three seconds, that's a signal 00:23:51.650 --> 00:23:54.789 of bait. And they started algorithmically punishing 00:23:54.789 --> 00:23:57.619 the bait. That actually explains why we saw that 00:23:57.619 --> 00:24:00.380 big shift toward long form content for a while. 00:24:00.460 --> 00:24:03.660 You started seeing these huge in -depth articles 00:24:03.660 --> 00:24:06.059 everywhere. It forced publishers to actually 00:24:06.059 --> 00:24:08.980 put something of value on the page. You couldn't 00:24:08.980 --> 00:24:10.619 just trick people with the headline anymore. 00:24:10.720 --> 00:24:12.759 You had to entertain or inform them enough to 00:24:12.759 --> 00:24:15.059 keep them there. And then there were the ad blockers. 00:24:15.140 --> 00:24:17.240 That was the consumers fighting back with code. 00:24:17.440 --> 00:24:20.359 Yes. The rise of ad blockers completely destroyed 00:24:20.359 --> 00:24:22.460 the revenue model for the cheapest, lowest effort 00:24:22.460 --> 00:24:25.500 sites. If I can't show you the ads on my... page 00:24:25.500 --> 00:24:28.400 your click is literally worthless to me this 00:24:28.400 --> 00:24:31.440 pushed the entire industry toward native advertising 00:24:32.039 --> 00:24:34.420 And sponsored content, right? Which is a whole 00:24:34.420 --> 00:24:37.039 other can of worms, honestly. But in general, 00:24:37.079 --> 00:24:39.480 for a sponsored article to work, it has to be 00:24:39.480 --> 00:24:42.119 at least readable. A brand like Nike doesn't 00:24:42.119 --> 00:24:44.660 want their logo to be associated with a chum 00:24:44.660 --> 00:24:47.420 box style scam. They want high quality storytelling. 00:24:47.619 --> 00:24:50.220 So the money started slowly flowing back toward 00:24:50.220 --> 00:24:52.700 quality or at least better production value. 00:24:53.019 --> 00:24:55.160 And we also have to give a shout out to the Save 00:24:55.160 --> 00:24:57.660 You a Click heroes. on social media. The stop 00:24:57.660 --> 00:25:00.880 clickbait communities. I love these people. They 00:25:00.880 --> 00:25:03.519 see a headline that says why this celebrity is 00:25:03.519 --> 00:25:05.859 quitting Hollywood forever and they'll just comment 00:25:05.859 --> 00:25:08.460 right below it. She's not. She's taking a two 00:25:08.460 --> 00:25:10.960 month break to have a baby. Doing the Lord's 00:25:10.960 --> 00:25:13.720 work. It's a public service. It's like a form 00:25:13.720 --> 00:25:16.619 of community policing for information. It closes 00:25:16.619 --> 00:25:19.380 the curiosity gap for everyone else so we don't 00:25:19.380 --> 00:25:21.630 have to give the publisher the ad revenue. It's 00:25:21.630 --> 00:25:24.910 a small but powerful way of reclaiming our collective 00:25:24.910 --> 00:25:27.569 attention. So to recap, we've gone from fishing 00:25:27.569 --> 00:25:30.650 hooks to the yellow journalism wars to the psychological 00:25:30.650 --> 00:25:33.809 itch of the curiosity gap to the brutal logic 00:25:33.809 --> 00:25:37.329 of the YouTube algorithm. And now to this weird 00:25:37.329 --> 00:25:40.390 stalemate where we are all fighting the machines 00:25:40.390 --> 00:25:43.569 with satire and ad blockers. It's really a journey 00:25:43.569 --> 00:25:46.230 through the weirdest parts of human nature. Clickbait 00:25:46.230 --> 00:25:48.549 isn't just a technological annoyance. It is a 00:25:48.549 --> 00:25:51.529 mirror. It exploits the things that make us human. 00:25:51.650 --> 00:25:54.890 Our intense curiosity, our desire for social 00:25:54.890 --> 00:25:58.809 connection, and our unfortunate tendency toward 00:25:58.809 --> 00:26:01.470 outrage. It's humbling, really. We like to think 00:26:01.470 --> 00:26:03.650 we are these rational, logical beings making 00:26:03.650 --> 00:26:06.170 informed decisions. But show us a picture of 00:26:06.170 --> 00:26:08.670 a secret door found in a basement. And we just 00:26:08.670 --> 00:26:10.430 turn into curious little monkeys who have to 00:26:10.430 --> 00:26:12.549 click the button. We do. And if we connect all 00:26:12.549 --> 00:26:14.589 this to the bigger picture, I want to leave you, 00:26:14.690 --> 00:26:17.230 the learner, with a thought about agency. Go 00:26:17.230 --> 00:26:19.809 on. We complain about clickbait. We say we hate 00:26:19.809 --> 00:26:22.029 it. We say we want real news and deep, thoughtful 00:26:22.029 --> 00:26:24.710 content. But the algorithms are just prediction 00:26:24.710 --> 00:26:27.029 engines. At the end of the day, they are designed 00:26:27.029 --> 00:26:29.609 to give us what we want. Or what we click on. 00:26:29.730 --> 00:26:33.170 Which is their best proxy for what we want. If 00:26:33.170 --> 00:26:35.509 clickbait is still prevalent, and believe me. 00:26:35.960 --> 00:26:38.700 It is it means it is still working. It means 00:26:38.700 --> 00:26:40.720 that despite all our complaints, we are still 00:26:40.720 --> 00:26:44.039 clicking. So are you saying we're the villains 00:26:44.039 --> 00:26:47.359 of our own story? We're the enablers. The algorithm 00:26:47.359 --> 00:26:50.240 is just holding up a mirror to our own lack of 00:26:50.240 --> 00:26:53.180 impulse control. The provocative thought is this. 00:26:53.519 --> 00:26:57.039 Are we really mad at the beat or are we just 00:26:57.039 --> 00:26:59.359 mad at ourselves for continuing to bite every 00:26:59.359 --> 00:27:02.750 single time? That is. Deeply uncomfortable. And 00:27:02.750 --> 00:27:05.750 I feel personally attacked regarding my 1 .30 00:27:05.750 --> 00:27:08.250 a .m. browsing habits. As you should. But it's 00:27:08.250 --> 00:27:10.630 a necessary question. If we want a better media 00:27:10.630 --> 00:27:12.950 environment, we have to start by having better 00:27:12.950 --> 00:27:14.789 clicking habits. We have to vote with our attention. 00:27:14.950 --> 00:27:17.170 It's the only currency that matters online. Well, 00:27:17.230 --> 00:27:19.109 on that very heavy note, I want to thank you, 00:27:19.190 --> 00:27:21.329 the learner, for taking this deep dive with us. 00:27:21.769 --> 00:27:23.569 Hopefully the title of this discussion didn't 00:27:23.569 --> 00:27:25.549 overpromise. I think we delivered. I think so, 00:27:25.589 --> 00:27:28.029 too. Thank you for listening all the way to the 00:27:28.029 --> 00:27:29.990 end. You've proven that human attention spans 00:27:29.990 --> 00:27:32.029 aren't quite dead yet. And check your sources 00:27:32.029 --> 00:27:35.329 before you share. And please, think before you 00:27:35.329 --> 00:27:38.230 click. We'll see you in the next deep dive. Welcome 00:27:38.230 --> 00:27:42.210 to The Debate. So today we're going to be dissecting 00:27:42.210 --> 00:27:44.910 a mechanism that, well, it pretty much dictates 00:27:44.910 --> 00:27:47.549 the diet of our digital age. And that is, of 00:27:47.549 --> 00:27:51.349 course, clickbait. It's a term we all use, usually 00:27:51.349 --> 00:27:53.849 as an insult, to describe content that's just 00:27:53.849 --> 00:27:57.329 engineered to demand your attention. Right. But. 00:27:57.980 --> 00:27:59.819 You could also argue it's the very engine that 00:27:59.819 --> 00:28:02.839 keeps the free internet running. And that is 00:28:02.839 --> 00:28:05.180 exactly the tension we're exploring today. I'll 00:28:05.180 --> 00:28:08.859 be arguing that clickbait is inherently a form 00:28:08.859 --> 00:28:11.559 of digital deception. It's a bait and switch 00:28:11.559 --> 00:28:14.920 that prioritizes metrics over truth and ultimately 00:28:14.920 --> 00:28:17.720 erodes trust. And I'll be taking the dissenting 00:28:17.720 --> 00:28:20.339 view. I think the term is defined far too broadly. 00:28:20.539 --> 00:28:22.859 My position is that high engagement headlines 00:28:22.859 --> 00:28:25.960 are not actually clickbait if they're factually 00:28:25.960 --> 00:28:28.529 accurate. I think often what we dismiss as bait 00:28:28.529 --> 00:28:32.250 is simply a rational and effective response to 00:28:32.250 --> 00:28:34.730 the economics of attention. Well, let's look 00:28:34.730 --> 00:28:37.289 at the definitions, because the specific wording 00:28:37.289 --> 00:28:40.210 really reveals the problem here. If we use the 00:28:40.210 --> 00:28:43.069 frameworks from, say, Merriam -Webster or dictionary 00:28:43.069 --> 00:28:46.529 .com, clickbait isn't just about being catchy. 00:28:46.690 --> 00:28:49.529 It is defined by the quality of what you find 00:28:49.529 --> 00:28:52.420 when you get there. It's content designed to 00:28:52.420 --> 00:28:55.180 entice a click, specifically leading to material 00:28:55.180 --> 00:28:59.539 of, and this is the quote, dubious value. Dubious 00:28:59.539 --> 00:29:02.539 value. I mean, that's pretty subjective, isn't 00:29:02.539 --> 00:29:05.720 it? Okay, perhaps. But the mechanism itself isn't 00:29:05.720 --> 00:29:08.500 subjective. The term, it relies on that fishing 00:29:08.500 --> 00:29:12.099 analogy for a reason. Bait implies a hook that's 00:29:12.099 --> 00:29:14.319 been disguised by an enticement. You know, the 00:29:14.319 --> 00:29:16.720 fish thinks it's getting food, but it's actually 00:29:16.720 --> 00:29:19.200 getting caught. It's a breach of this sort of 00:29:19.200 --> 00:29:21.640 unspoken contract between the publisher and the 00:29:21.640 --> 00:29:24.819 reader. The goal isn't to inform. It's to maximize 00:29:24.819 --> 00:29:27.420 ad revenue by misrepresenting what's on the other 00:29:27.420 --> 00:29:31.720 side of that link. I think that analogy collapses 00:29:31.720 --> 00:29:35.400 when the food is actually real, though. I want 00:29:35.400 --> 00:29:37.859 to bring in the argument made by Ben Smith, who 00:29:37.859 --> 00:29:40.319 was the editor -in -chief of BuzzFeed back in 00:29:40.319 --> 00:29:44.440 2014. He drew a really hard line between curiosity. 00:29:45.319 --> 00:29:48.059 and deception. His point was that a headline 00:29:48.059 --> 00:29:51.460 is only clickbait if it lies. That feels like 00:29:51.460 --> 00:29:54.799 a pretty convenient distinction for the architect 00:29:54.799 --> 00:29:58.200 of BuzzFeed to make, don't you think? Well, maybe, 00:29:58.400 --> 00:30:01.279 but it is a functional one. Look, Smith used 00:30:01.279 --> 00:30:04.180 a specific example, a headline about a five -year 00:30:04.180 --> 00:30:06.119 -old girl raising money for her father's terminal 00:30:06.119 --> 00:30:09.279 cancer treatment. Now, that is sensational. It 00:30:09.279 --> 00:30:12.420 exploits emotion. It absolutely demands a click. 00:30:12.759 --> 00:30:15.599 But if the article actually tells that story, 00:30:15.880 --> 00:30:19.039 Smith argues it's not clickbait. It's just an 00:30:19.039 --> 00:30:22.200 eye -catching presentation of the truth. If we 00:30:22.200 --> 00:30:25.000 label every effective hook as bait, we're basically 00:30:25.000 --> 00:30:27.720 criticizing storytelling just for being interesting. 00:30:28.059 --> 00:30:30.420 But you're ignoring the psychological mechanism 00:30:30.420 --> 00:30:33.440 being used to get that click. And this brings 00:30:33.440 --> 00:30:36.619 us to the curiosity gap. This isn't just about 00:30:36.619 --> 00:30:39.519 whether the story is technically true. It's about 00:30:39.519 --> 00:30:42.519 how the user is being manipulated. You know, 00:30:42.559 --> 00:30:45.519 Facebook, in their own technical efforts to downrank 00:30:45.519 --> 00:30:48.339 clickbait, defined it as headlines that withhold 00:30:48.339 --> 00:30:51.720 crucial information. They create this artificial 00:30:51.720 --> 00:30:55.339 information gap that exploits a user's cognitive 00:30:55.339 --> 00:30:58.079 need for closure. You aren't clicking because 00:30:58.079 --> 00:31:00.579 you want the content. You're clicking to scratch 00:31:00.579 --> 00:31:02.980 a psychological itch that the headline forced 00:31:02.980 --> 00:31:06.160 upon you. I have to say, I find that curiosity 00:31:06.160 --> 00:31:10.259 gap argument incredibly restrictive. I mean, 00:31:10.259 --> 00:31:12.559 if we follow Facebook's logic that a headline 00:31:12.559 --> 00:31:14.680 has to tell you exactly what you're about to 00:31:14.680 --> 00:31:17.720 see, then a joke without a punchline is clickbait. 00:31:17.839 --> 00:31:19.859 A movie trailer that doesn't show the ending 00:31:19.859 --> 00:31:22.420 is clickbait. Withholding information is the 00:31:22.420 --> 00:31:25.500 basis of all suspense. If I write, you won't 00:31:25.500 --> 00:31:27.359 believe what happened, and then I show you something 00:31:27.359 --> 00:31:30.539 that is genuinely unbelievable, I haven't exploited 00:31:30.539 --> 00:31:32.880 you, I've entertained you. You may have entertained, 00:31:33.279 --> 00:31:36.500 yes, but at what cost to the ecosystem? This 00:31:36.500 --> 00:31:39.539 is the macro issue for me, the economic incentives. 00:31:39.819 --> 00:31:42.359 When you prioritize the click above everything 00:31:42.359 --> 00:31:45.500 else, you degrade the entire information environment. 00:31:46.559 --> 00:31:48.440 Katherine Viner, the editor -in -chief at The 00:31:48.440 --> 00:31:50.920 Guardian, she argued that chasing down cheap 00:31:50.920 --> 00:31:53.420 clicks at the expense of accuracy and veracity 00:31:53.839 --> 00:31:55.940 undermines the fundamental value of journalism. 00:31:56.200 --> 00:31:58.980 It just creates a race to the bottom where the 00:31:58.980 --> 00:32:00.799 only metric that matters is the click -through 00:32:00.799 --> 00:32:03.680 rate, not the quality of understanding. But you're 00:32:03.680 --> 00:32:06.720 blaming the players for the rules of the game. 00:32:06.880 --> 00:32:09.819 We have to view this within the attention economy. 00:32:10.259 --> 00:32:13.099 Look at the data that YouTube released at CES. 00:32:13.380 --> 00:32:17.000 They revealed that 70 % of watch time on their 00:32:17.000 --> 00:32:20.359 platform is driven by recommendations, by the 00:32:20.359 --> 00:32:22.980 algorithm, not by people actively searching. 00:32:23.450 --> 00:32:26.190 And the data also shows that videos with hyperbolic, 00:32:26.190 --> 00:32:29.390 high -energy titles, they objectively have higher 00:32:29.390 --> 00:32:31.730 click -through rates. So because the algorithm 00:32:31.730 --> 00:32:35.529 rewards it, we should just accept it? No, because 00:32:35.529 --> 00:32:38.890 it works. It becomes a survival mechanism. Creators 00:32:38.890 --> 00:32:41.730 are just rationally optimizing for an algorithm 00:32:41.730 --> 00:32:44.869 that favors engagement. If you write a dry academic 00:32:44.869 --> 00:32:47.710 headline in an ecosystem that demands sticky 00:32:47.710 --> 00:32:51.230 content, you will disappear. It isn't about degrading 00:32:51.230 --> 00:32:53.849 the ecosystem with malice. It's about navigating 00:32:53.849 --> 00:32:56.470 the reality that attention is a finite and very 00:32:56.470 --> 00:32:59.170 valuable commodity. But that survival mechanism 00:32:59.170 --> 00:33:02.430 has very deep roots in unethical practices. I 00:33:02.430 --> 00:33:04.730 mean, we can draw a direct line from modern clickbait 00:33:04.730 --> 00:33:07.150 right back to checkbook journalism and yellow 00:33:07.150 --> 00:33:09.750 journalism from the late 19th century. Those 00:33:09.750 --> 00:33:12.410 eras focused entirely on scandal mongering and 00:33:12.410 --> 00:33:15.640 sensation just to sell papers. Slate described 00:33:15.640 --> 00:33:17.900 the modern iteration of this as the aggregation 00:33:17.900 --> 00:33:20.880 of outrage. And we see this everywhere with rage 00:33:20.880 --> 00:33:22.960 baiting and political discourse. It's not just 00:33:22.960 --> 00:33:25.500 harmless fluff. It's a tool used to polarize 00:33:25.500 --> 00:33:27.359 people and profit from their moral judgments. 00:33:27.680 --> 00:33:30.380 I think you have to distinguish between the extremes 00:33:30.380 --> 00:33:33.420 and the norm. I mean, obviously, phishing or 00:33:33.420 --> 00:33:36.259 using clickbait to spread malicious code, what 00:33:36.259 --> 00:33:39.200 they call click fraud, is illegitimate. But the 00:33:39.200 --> 00:33:41.240 market has a way of correcting for the aggressive 00:33:41.240 --> 00:33:43.960 media practices you're worried about. We saw 00:33:43.960 --> 00:33:47.880 a massive backlash starting around 2014. Ad blockers 00:33:47.880 --> 00:33:51.099 became hugely popular. The algorithms were updated 00:33:51.099 --> 00:33:54.099 to filter out low -quality content. We even saw 00:33:54.099 --> 00:33:56.480 the rise of satirical sites like The Onion's 00:33:56.480 --> 00:33:59.259 Clickhole, which parodied these forms so effectively 00:33:59.259 --> 00:34:01.960 that it actually forced publishers to pull back. 00:34:02.200 --> 00:34:05.440 The audience isn't stupid. They eventually learn 00:34:05.440 --> 00:34:07.920 to identify the trap. They might learn, but the 00:34:07.920 --> 00:34:11.409 trap just keeps evolving. So to summarize my 00:34:11.409 --> 00:34:14.469 position, while I understand the economic pressures, 00:34:14.730 --> 00:34:18.090 the term clickbait correctly identifies a pattern 00:34:18.090 --> 00:34:21.250 of overpromising and underdelivering. Whether 00:34:21.250 --> 00:34:24.690 it's a curiosity gap or rage baiting, it treats 00:34:24.690 --> 00:34:27.849 the audience as a metric to be harvested, not 00:34:27.849 --> 00:34:31.510 as a public to be informed. And I maintain that 00:34:31.510 --> 00:34:34.250 while dishonesty is always wrong, dismissing 00:34:34.250 --> 00:34:37.090 all sensational or curiosity -inducing content 00:34:37.090 --> 00:34:40.130 as clickbait just ignores the reality of digital 00:34:40.130 --> 00:34:42.969 media economics. If the headline catches your 00:34:42.969 --> 00:34:45.429 eye and the content delivers on that promise, 00:34:45.670 --> 00:34:48.789 then the label of clickbait is, I think, misapplied. 00:34:48.929 --> 00:34:51.650 We should absolutely criticize fraud, but we 00:34:51.650 --> 00:34:54.429 shouldn't punish curiosity. It is a difficult 00:34:54.429 --> 00:34:57.769 line to walk in an internet built on sticky content. 00:34:58.130 --> 00:35:00.849 It is, indeed. And it leaves us all to wonder 00:35:00.849 --> 00:35:03.510 where the line between good marketing and outright 00:35:03.510 --> 00:35:04.989 deception truly lies.