WEBVTT 00:00:00.000 --> 00:00:02.839 Have you ever really stopped to think how much 00:00:02.839 --> 00:00:05.120 of your time online is actually spent interacting 00:00:05.120 --> 00:00:09.140 with, well not humans, beat? The answer, if you 00:00:09.140 --> 00:00:11.439 really look into it, it might genuinely surprise 00:00:11.439 --> 00:00:12.960 you. And it's leading a lot of people to ask 00:00:12.960 --> 00:00:15.400 this pretty profound question. Is the internet 00:00:15.400 --> 00:00:17.440 that we knew, you know, that big messy human 00:00:17.440 --> 00:00:21.440 place, is it maybe? Fading. Welcome to the Deep 00:00:21.440 --> 00:00:23.039 Dive. Today, we're really going to get into a 00:00:23.039 --> 00:00:26.219 concept that's quietly gaining some serious traction. 00:00:26.480 --> 00:00:28.420 The dead internet theory. This isn't just some 00:00:28.420 --> 00:00:31.300 fringe idea from obscure forums anymore. No, 00:00:31.300 --> 00:00:33.179 it's being talked about, even worried about, 00:00:33.280 --> 00:00:35.740 by industry folks, data scientists. So our job 00:00:35.740 --> 00:00:38.500 today is to try and separate the panic from the 00:00:38.500 --> 00:00:40.359 actual reality. We'll use the facts, the data 00:00:40.359 --> 00:00:42.780 we have, to see what's really going on. Exactly. 00:00:43.280 --> 00:00:45.299 Our mission here is to explore what the data 00:00:45.299 --> 00:00:48.719 is actually telling us about AI's, well, kind 00:00:48.719 --> 00:00:50.729 of silent takeover. We're going to look at the 00:00:50.729 --> 00:00:54.210 numbers, quantify this bot tsunami, then we'll 00:00:54.210 --> 00:00:56.829 unpack why spotting AI content isn't as easy 00:00:56.829 --> 00:00:58.929 as you might think. We'll look at the economic 00:00:58.929 --> 00:01:00.950 shifts, how the internet's business model is 00:01:00.950 --> 00:01:04.209 changing, and then, crucially, look at the proposed 00:01:04.209 --> 00:01:07.090 solutions. And yeah, there are tricky downsides. 00:01:07.390 --> 00:01:09.329 It's sort of a shortcut to understanding why 00:01:09.329 --> 00:01:11.209 the internet feels different lately and what 00:01:11.209 --> 00:01:13.909 that really means for you day to day. OK, let's 00:01:13.909 --> 00:01:16.569 unpack this bot tsunami then, these new AI models 00:01:16.569 --> 00:01:19.400 like ChatGPT. What was it that made them such 00:01:19.400 --> 00:01:21.500 a huge game changer? Well, it really comes down 00:01:21.500 --> 00:01:24.879 to how well they can mimic human chat. These 00:01:24.879 --> 00:01:27.140 models, especially the ones everyone can access, 00:01:27.200 --> 00:01:29.400 the large language models, they basically just 00:01:29.400 --> 00:01:31.780 smashed the Turing test. Ah, the Turing test. 00:01:31.819 --> 00:01:33.819 That's the classic one, right? Where an AI tries 00:01:33.819 --> 00:01:36.159 to fool you into thinking it's human. Exactly. 00:01:36.640 --> 00:01:39.159 Can a machine imitate a person so well you just 00:01:39.159 --> 00:01:40.859 can't tell the difference? That's the test, in 00:01:40.859 --> 00:01:44.760 a nutshell. And before these big LLMs, that was 00:01:44.760 --> 00:01:47.739 a pretty high bar. Now, it's almost normal. So 00:01:47.739 --> 00:01:50.040 they crossed some kind of threshold. They got 00:01:50.040 --> 00:01:52.579 really good at sounding human. And that's what 00:01:52.579 --> 00:01:55.140 kicked off this machine invasion. Pretty much. 00:01:55.340 --> 00:01:57.599 Yeah. And the scale of it. I mean, this is where 00:01:57.599 --> 00:01:59.239 it gets really interesting. The numbers are quite 00:01:59.239 --> 00:02:03.079 stark. Get this. The Imperva bad bot report for 00:02:03.079 --> 00:02:07.159 2025. It found that 51 % of all internet traffic 00:02:07.159 --> 00:02:10.560 in 2024 was automated. Wait, over half? Over 00:02:10.560 --> 00:02:12.659 half. Just machines talking to other machines, 00:02:12.680 --> 00:02:17.069 basically. And maybe more worrying, 37 % of that 00:02:17.069 --> 00:02:19.689 automated traffic was malicious. That's a record 00:02:19.689 --> 00:02:22.889 high. So it's a lot of automated stuff, and a 00:02:22.889 --> 00:02:26.770 big chunk of it isn't friendly. 51%. Wow. That's... 00:02:26.780 --> 00:02:28.439 That's incredible. It's like there's this whole 00:02:28.439 --> 00:02:30.219 other internet running underneath ours, just 00:02:30.219 --> 00:02:32.159 bots chatting. It kind of feels like that sometimes, 00:02:32.280 --> 00:02:34.539 yeah. And the speed it's growing is just wild. 00:02:34.599 --> 00:02:37.400 Look at Cloudflare's data. Perplexity bot traffic. 00:02:37.500 --> 00:02:41.939 It surged by something like 157 ,490%, almost 00:02:41.939 --> 00:02:45.680 unbelievable. 157 ,000%. Yeah. And the bots that 00:02:45.680 --> 00:02:48.539 LLMs use to grab real -time info, those retrieval 00:02:48.539 --> 00:02:51.319 bots, they grew by 49 % in just the first quarter 00:02:51.319 --> 00:02:54.419 of 2025. Just Q1. It's an explosion. That's not 00:02:54.419 --> 00:02:57.020 just growth. That's... Exponential isn't even 00:02:57.020 --> 00:02:59.680 the word. Absolutely. And think about the sheer 00:02:59.680 --> 00:03:02.400 power behind some of these. TechRadar said, open 00:03:02.400 --> 00:03:05.219 AI's crawlers alone send out over a billion requests 00:03:05.219 --> 00:03:08.340 a month. A billion? From one company's bots? 00:03:08.860 --> 00:03:11.020 Yep. And fastly found that those open AI bots 00:03:11.020 --> 00:03:15.300 made up, get this, 98 % of all the global Fetcher 00:03:15.300 --> 00:03:20.520 bot traffic in 2025. 98%. Whoa. Just... Imagine 00:03:20.520 --> 00:03:23.439 scaling to a billion requests a month. Just one 00:03:23.439 --> 00:03:25.620 AI. That really is a silent invasion, isn't it? 00:03:25.699 --> 00:03:28.080 But massive. It's like this unseen digital city 00:03:28.080 --> 00:03:30.080 humming away beneath us. These aren't just numbers, 00:03:30.139 --> 00:03:32.180 you know? It paints this picture of the internet 00:03:32.180 --> 00:03:34.300 becoming, well, mostly a machine conversation. 00:03:34.680 --> 00:03:36.060 That's a really good way to put it. And it's 00:03:36.060 --> 00:03:39.189 not just like... reading web pages. Over a third 00:03:39.189 --> 00:03:41.629 of all web traffic back in May 2025 came from 00:03:41.629 --> 00:03:44.389 APIs. APIs, remind us. That's like how different 00:03:44.389 --> 00:03:45.849 computer programs talk to each other, right? 00:03:45.889 --> 00:03:48.210 Exactly. That's the short version. So a third 00:03:48.210 --> 00:03:50.870 of traffic from APIs in these autonomous agents, 00:03:51.050 --> 00:03:53.669 not people clicking around, and all this activity, 00:03:53.689 --> 00:03:56.050 it puts a massive strain on the infrastructure. 00:03:56.710 --> 00:03:58.909 Wikimedia, for instance, they said bots are only 00:03:58.909 --> 00:04:01.909 35 % of their page views, but they eat up 65 00:04:01.909 --> 00:04:04.550 % of the resources for the complex stuff. So 00:04:04.550 --> 00:04:07.379 they're disproportionately resource -hungry. 00:04:07.819 --> 00:04:10.560 Hugely. Wikimedia saw their bandwidth use jump 00:04:10.560 --> 00:04:14.819 50 % in early 2025 just from AI scraping. Yeah. 00:04:14.860 --> 00:04:16.660 And there was an ARCIDIS study found that for 00:04:16.660 --> 00:04:19.660 many small servers like 80, 95 % of their traffic 00:04:19.660 --> 00:04:23.379 is now AI crawlers. 80 to 95%. Yeah. It basically 00:04:23.379 --> 00:04:26.079 makes it impossible or at least prohibitively 00:04:26.079 --> 00:04:28.629 expensive for them to run. chokes out the little 00:04:28.629 --> 00:04:31.149 human run sites. So, okay, with all this machine 00:04:31.149 --> 00:04:33.629 interaction, this overwhelming bot traffic, what 00:04:33.629 --> 00:04:35.949 does that actually mean for our experience, you 00:04:35.949 --> 00:04:37.970 know, when we just go online day -to -day? Well, 00:04:38.069 --> 00:04:40.069 fundamentally, you'll find less unique human 00:04:40.069 --> 00:04:42.910 stuff, just a lot more digital noise. Right. 00:04:43.350 --> 00:04:45.069 Now, I think a lot of us feel like, oh, I can 00:04:45.069 --> 00:04:48.730 spot AI content, you know. We look for those 00:04:48.730 --> 00:04:52.050 slightly weird phrases, maybe delve or tapestry 00:04:52.050 --> 00:04:55.310 or those super complex sentences. That chart 00:04:55.310 --> 00:04:57.370 showing the word delve just shooting up in research 00:04:57.370 --> 00:04:59.389 papers after ChatGPT came out, I mean, that's 00:04:59.389 --> 00:05:03.649 hard to ignore. But is that confidence really 00:05:03.649 --> 00:05:06.829 right? Or are we maybe fooling ourselves a bit? 00:05:07.509 --> 00:05:09.430 That's such a key point. Yeah. Because here's 00:05:09.430 --> 00:05:11.970 the twist, right? You're nacked for detecting 00:05:11.970 --> 00:05:14.519 AI. It isn't really some kind of superpower you 00:05:14.519 --> 00:05:16.459 developed. It's more that you've gotten really 00:05:16.459 --> 00:05:19.680 good at spotting one specific AI persona, the 00:05:19.680 --> 00:05:22.680 chat GPT persona. Ah, okay. So it's not all AI, 00:05:22.759 --> 00:05:24.459 just the one we see most often. Exactly. See, 00:05:24.480 --> 00:05:26.980 these models get trained on, like, the whole 00:05:26.980 --> 00:05:29.519 internet, but then they get fine -tuned. They're 00:05:29.519 --> 00:05:31.240 fed a smaller set of stuff the creators think 00:05:31.240 --> 00:05:33.920 is good writing. Yeah. And that process, it created 00:05:33.920 --> 00:05:36.360 a very specific, very recognizable voice or persona. 00:05:36.500 --> 00:05:38.899 So we're just recognizing that style. Precisely. 00:05:38.970 --> 00:05:41.069 You haven't magically become an AI detector. 00:05:41.509 --> 00:05:44.310 You've become a chat GPT detector. Other models, 00:05:44.730 --> 00:05:48.170 like Kimi for moonshot AI, huge in China. They 00:05:48.170 --> 00:05:49.829 have a totally different way of sounding. And 00:05:49.829 --> 00:05:51.430 they're probably just flying under our radar 00:05:51.430 --> 00:05:55.250 completely. And I gotta admit, even I still wrestle 00:05:55.250 --> 00:05:58.129 with prompt drift sometimes. Prompt drift. That's 00:05:58.129 --> 00:06:00.329 when the AI's output kind of wanders away from 00:06:00.329 --> 00:06:02.569 what you first asked it. Yeah, exactly. Its output 00:06:02.569 --> 00:06:04.629 subtly changes from your original intention. 00:06:04.850 --> 00:06:07.829 It's subtle, but yeah. Sometimes I feel like 00:06:07.829 --> 00:06:10.529 I'm writing to the AI, trying to guess its style, 00:06:10.990 --> 00:06:13.050 instead of just using it as a tool. It's a weird 00:06:13.050 --> 00:06:15.689 feeling. That raises a really important question 00:06:15.689 --> 00:06:18.269 then. What happens when the AI starts learning 00:06:18.269 --> 00:06:20.850 mostly from, well, from other AI content? I've 00:06:20.850 --> 00:06:23.189 heard this called model collapse, or sometimes 00:06:23.189 --> 00:06:26.170 the feedback loop of mediocrity. What does that 00:06:26.170 --> 00:06:28.329 actually mean for the quality of stuff online? 00:06:28.889 --> 00:06:31.649 That is exactly the worry. Model collapse is 00:06:31.649 --> 00:06:34.029 basically what happens when AI systems get trained 00:06:34.029 --> 00:06:36.490 on data that's full of their own outputs, stuff 00:06:36.490 --> 00:06:39.189 they generated before. It leads to this degradation. 00:06:39.670 --> 00:06:41.850 The quality, the diversity, the originality just 00:06:41.850 --> 00:06:44.110 goes down over time. So everything starts sounding 00:06:44.110 --> 00:06:47.110 the same? Pretty much. We risk creating this 00:06:47.110 --> 00:06:49.910 closed loop, this information echo chamber, where 00:06:49.910 --> 00:06:52.750 everything's just a rehash of a rehash. We lose 00:06:52.750 --> 00:06:55.449 that human spark, that unpredictability. Defined 00:06:55.449 --> 00:06:58.470 as... AI quality degrades when it learns from 00:06:58.470 --> 00:07:01.110 its own recycled outputs. About right. That's 00:07:01.110 --> 00:07:03.490 a good way to put it. Yeah. So thinking about 00:07:03.490 --> 00:07:06.910 human creativity, original thought, what's the 00:07:06.910 --> 00:07:10.649 real danger of this kind of uniform AI voice 00:07:10.649 --> 00:07:14.110 taking over? Well, the danger is that true human 00:07:14.110 --> 00:07:16.990 creativity just gets drowned out. Originality 00:07:16.990 --> 00:07:19.680 gets diluted. OK, so it's not just the content. 00:07:19.920 --> 00:07:21.959 The actual business model of the internet seems 00:07:21.959 --> 00:07:24.100 like it's under threat too. Yeah. We're heading 00:07:24.100 --> 00:07:26.579 towards this future where AI agents don't just 00:07:26.579 --> 00:07:29.019 write stuff. They do stuff for us. They execute 00:07:29.019 --> 00:07:31.660 tasks. Yeah. And what's really fascinating or 00:07:31.660 --> 00:07:34.019 maybe scary, depending on your view, is that 00:07:34.019 --> 00:07:36.639 optimizing your software just for humans, that 00:07:36.639 --> 00:07:38.579 might actually be a losing strategy soon. Really? 00:07:38.680 --> 00:07:41.560 How so? Because companies will increasingly need 00:07:41.560 --> 00:07:44.959 to build for AI agents. Think about advertising. 00:07:45.540 --> 00:07:47.240 AI agents, they're totally immune to all the 00:07:47.240 --> 00:07:49.920 psychological tricks marketers use. Like, an 00:07:49.920 --> 00:07:52.040 AI agent doesn't care at all if Sydney Sweeney 00:07:52.040 --> 00:07:55.480 wears those jeans in the ad. Huh. Right. No celebrity 00:07:55.480 --> 00:07:58.079 influence. None. It just optimizes for logic. 00:07:58.399 --> 00:08:01.579 Show me skinny jeans, under $30, locally sourced, 00:08:01.699 --> 00:08:04.939 rated 4 .5 stars or higher, best durability score, 00:08:05.360 --> 00:08:07.899 pure data. Marketers use celebrities because 00:08:07.899 --> 00:08:10.560 we humans tend to emulate people we admire. Right? 00:08:10.839 --> 00:08:13.269 And AI. Yeah, ignores that completely. So ads 00:08:13.269 --> 00:08:16.170 aimed at agents will just be spec sheets, basically. 00:08:16.269 --> 00:08:18.670 Essentially, yeah. Optimized for machine readable 00:08:18.670 --> 00:08:21.529 stuff. Right. Price, return policy clarity, shipping 00:08:21.529 --> 00:08:23.970 speed guarantees, super detailed product specs. 00:08:24.290 --> 00:08:26.230 So the whole attention economy grabbing human 00:08:26.230 --> 00:08:28.750 eyeballs that's giving way to this agentic economy, 00:08:29.110 --> 00:08:31.209 it's about efficiency, verifiable data. And who 00:08:31.209 --> 00:08:33.419 wings in that world? Well, obviously, the company's 00:08:33.419 --> 00:08:35.559 making the best base AI agents. Your open AI 00:08:35.559 --> 00:08:38.179 is Google's Entropix. But also businesses that 00:08:38.179 --> 00:08:41.620 have really clean, well -documented APIs. Because 00:08:41.620 --> 00:08:43.320 that makes it easy for the agents to find and 00:08:43.320 --> 00:08:45.679 use their products or services. APIs, again, 00:08:45.720 --> 00:08:47.980 the way programs talk. So having a good robot 00:08:47.980 --> 00:08:50.720 doorway becomes critical. Exactly. A good robot 00:08:50.720 --> 00:08:55.200 doorway. I like that. So this shift to an agentic 00:08:55.200 --> 00:08:59.159 economy, how does that fundamentally change how 00:08:59.159 --> 00:09:02.210 companies even reach us? Or maybe reach our agents 00:09:02.210 --> 00:09:05.210 effectively. Marketing totally shifts away from 00:09:05.210 --> 00:09:09.070 emotion towards pure, verifiable, machine -readable 00:09:09.070 --> 00:09:12.809 data. Mid -roll sponsor read. Okay, so the genie's 00:09:12.809 --> 00:09:14.980 out. We can't just, you know... turn off AI, 00:09:15.080 --> 00:09:17.419 that's not happening. But if the internet is 00:09:17.419 --> 00:09:19.320 filling up with bots, we definitely need ways 00:09:19.320 --> 00:09:22.019 to figure out who's human online. What are the 00:09:22.019 --> 00:09:23.720 main ideas being thrown around for making that 00:09:23.720 --> 00:09:26.299 distinction? How do we verify humanness? Well, 00:09:26.559 --> 00:09:29.240 one big concept is proof of humanhood. P -O -H, 00:09:29.299 --> 00:09:31.220 they call it. And there are a couple of main 00:09:31.220 --> 00:09:33.559 flavors. One is based on blockchain. People like 00:09:33.559 --> 00:09:36.000 OpenAI, Microsoft, Oxford University have proposed 00:09:36.000 --> 00:09:39.970 this. Blockchain. Like Bitcoin sort of yeah using 00:09:39.970 --> 00:09:42.769 that underlying tech for decentralized anonymous 00:09:42.769 --> 00:09:45.149 personhood IDs It relies on something called 00:09:45.149 --> 00:09:47.370 zero knowledge proofs zero knowledge proofs. 00:09:47.370 --> 00:09:49.809 Okay, what's that in plain English? It's basically 00:09:49.809 --> 00:09:52.250 a way to prove something is true Like your unique 00:09:52.250 --> 00:09:54.590 human without revealing the actual sensitive 00:09:54.590 --> 00:09:57.830 data about who you are. Ah Okay, so privacy is 00:09:57.830 --> 00:09:59.950 built in that's the idea it creates this kind 00:09:59.950 --> 00:10:03.490 of proof of human layer over the internet potentially 00:10:03.490 --> 00:10:06.370 interesting What about other approaches things 00:10:06.370 --> 00:10:11.070 based on? us, our bodies, biometrics. Yeah, that's 00:10:11.070 --> 00:10:13.149 definitely on the table too. WorldCoin is probably 00:10:13.149 --> 00:10:15.129 the most famous example. That's the Sam Altman 00:10:15.129 --> 00:10:18.769 project, right, from OpenAI. The very same. Their 00:10:18.769 --> 00:10:21.909 approach involves scanning people's irises. Unique 00:10:21.909 --> 00:10:24.769 identifier. Ooh, scanning eyeballs. That sounds 00:10:24.769 --> 00:10:27.850 intense. It does, yeah. And it raises huge privacy 00:10:27.850 --> 00:10:30.169 red flags for a lot of people. The idea of a 00:10:30.169 --> 00:10:32.830 massive global database of everyone's iris scans, 00:10:33.350 --> 00:10:36.080 potential for mass surveillance. It's understandable 00:10:36.080 --> 00:10:37.799 why people are worried. It's a big trade -off. 00:10:37.980 --> 00:10:39.460 OK, so that's proving you're human. But what 00:10:39.460 --> 00:10:41.559 about flipping it? Instead of humans proving 00:10:41.559 --> 00:10:43.200 they're human, what if the bots had to prove 00:10:43.200 --> 00:10:44.860 they're bots? That's the other side of the coin. 00:10:45.139 --> 00:10:47.759 Proof of bot. And yeah, it's a really interesting 00:10:47.759 --> 00:10:50.240 alternative. Cloudflare, they handle a huge chunk 00:10:50.240 --> 00:10:52.860 of internet traffic. They actually rolled out 00:10:52.860 --> 00:10:55.840 a system for AI agents to basically declare themselves, 00:10:56.120 --> 00:10:59.549 hi, I'm a bot. Oh, really? How's that going down? 00:10:59.789 --> 00:11:02.250 Well, one agent company, Prowzer Base, they already 00:11:02.250 --> 00:11:05.669 support it. So some are playing ball. But it's 00:11:05.669 --> 00:11:08.950 super controversial. Gary Tan, the CEO of Y Combinator, 00:11:09.929 --> 00:11:12.730 he like flat out called it an axis of evil. Wow. 00:11:13.300 --> 00:11:16.120 Strong words. Why such a strong reaction? The 00:11:16.120 --> 00:11:18.299 main worry is that it breaks the whole spirit 00:11:18.299 --> 00:11:20.799 of the open internet. It could create these powerful 00:11:20.799 --> 00:11:23.379 gatekeepers, right? Cloudflare, whoever runs 00:11:23.379 --> 00:11:25.340 these systems, gets to decide which bots are 00:11:25.340 --> 00:11:27.379 good and which are bad. Right. Who makes that 00:11:27.379 --> 00:11:29.980 call? That's a lot of centralized power. Exactly. 00:11:30.100 --> 00:11:33.279 It's a major sticking point. So if the main public 00:11:33.279 --> 00:11:36.440 internet gets... Well, noisier, more fragmented, 00:11:36.500 --> 00:11:39.279 full of bots. Where do you think people, actual 00:11:39.279 --> 00:11:41.759 humans are gonna go? Where will we find those 00:11:41.759 --> 00:11:45.580 real human connections? My feeling is that humans 00:11:45.580 --> 00:11:48.019 will naturally seek out, let's call them digital 00:11:48.019 --> 00:11:52.000 oases. More private, maybe curated spaces. Like 00:11:52.000 --> 00:11:54.950 gated communities online. Sort of, yeah. Social 00:11:54.950 --> 00:11:57.070 media might fragment more you could see more 00:11:57.070 --> 00:12:00.590 paid forums private discord slack servers Maybe 00:12:00.590 --> 00:12:02.870 even a comeback for personal newsletters and 00:12:02.870 --> 00:12:05.830 blogs, you know quality over quantity platforms 00:12:05.830 --> 00:12:07.809 like medium or sub stack They seem perfectly 00:12:07.809 --> 00:12:10.590 set up for this places offering high quality. 00:12:10.730 --> 00:12:12.950 Maybe longer form stuff away from all the noise 00:12:12.950 --> 00:12:15.889 It's like moving from the chaotic town square 00:12:15.889 --> 00:12:19.190 into smaller quieter cafes where you know the 00:12:19.190 --> 00:12:22.269 regulars. Mm -hmm So looking at all these paths 00:12:22.269 --> 00:12:25.309 proving we're human making bots identify themselves, 00:12:25.529 --> 00:12:28.210 retreating into these digital oases. Is there 00:12:28.210 --> 00:12:30.610 one that seems, I don't know, most aligned with 00:12:30.610 --> 00:12:33.289 keeping that genuine human connection alive online? 00:12:33.990 --> 00:12:36.129 Honestly, each one has pretty big trade -offs. 00:12:36.309 --> 00:12:37.929 It's always going to be a balance between open 00:12:37.929 --> 00:12:39.710 access, privacy, and who holds the control. There's 00:12:39.710 --> 00:12:41.730 no easy answer there. OK, so let's try and bring 00:12:41.730 --> 00:12:43.750 this all together. What's the big picture for 00:12:43.750 --> 00:12:46.269 us, the people using this network every day, 00:12:46.529 --> 00:12:50.490 this dead internet theory? It sounds grim, but 00:12:50.490 --> 00:12:52.610 it feels like it's not just a future prediction 00:12:52.610 --> 00:12:54.730 anymore. In many ways, it kind of feels spot 00:12:54.730 --> 00:12:56.509 on right now. The internet does feel different, 00:12:56.710 --> 00:12:59.309 sometimes broken. Less about finding unique voices, 00:12:59.370 --> 00:13:02.269 more about wading through an echo chamber. I 00:13:02.269 --> 00:13:04.370 think that's fair. Yeah. We've seen the numbers 00:13:04.370 --> 00:13:06.590 and the insane bot traffic. We've talked about 00:13:06.590 --> 00:13:08.549 how content is becoming more uniform because 00:13:08.549 --> 00:13:11.570 of how AI learns. And we've touched on how the 00:13:11.570 --> 00:13:14.029 basic economics of the online world are shifting 00:13:14.029 --> 00:13:16.629 under our feet. The real struggle I think is 00:13:16.629 --> 00:13:19.710 how do we preserve I don't know, the soul of 00:13:19.710 --> 00:13:22.090 the internet. How do we make sure we can still 00:13:22.090 --> 00:13:24.830 find and value that real human connection if 00:13:24.830 --> 00:13:26.750 it's all the machine noise? It's definitely not 00:13:26.750 --> 00:13:28.570 about killing AI. That's impossible. It's about 00:13:28.570 --> 00:13:31.129 building new systems, finding new norms so humans 00:13:31.129 --> 00:13:33.830 can thrive alongside these incredibly powerful 00:13:33.830 --> 00:13:36.230 new tools. That leaves us with a pretty big thought, 00:13:36.350 --> 00:13:39.029 doesn't it? If our digital world is increasingly 00:13:39.029 --> 00:13:42.610 becoming a conversation between machines, what 00:13:42.610 --> 00:13:45.590 actually happens to our voices? are authentic 00:13:45.590 --> 00:13:48.149 human voices? How do we make sure they don't 00:13:48.149 --> 00:13:50.649 just get lost in the noise? How do we ensure 00:13:50.649 --> 00:13:52.809 they're heard, and maybe more importantly, valued 00:13:52.809 --> 00:13:54.409 as something distinct from all the synthetic 00:13:54.409 --> 00:13:57.309 stuff? It's a huge question. And honestly, it's 00:13:57.309 --> 00:13:58.970 something for all of us to really think about 00:13:58.970 --> 00:14:01.870 as we figure out how to navigate this evolving 00:14:01.870 --> 00:14:03.809 landscape. Well, thank you for joining us on 00:14:03.809 --> 00:14:06.450 this really fascinating deep dive today. My pleasure. 00:14:06.769 --> 00:14:09.049 Until next time, keep exploring, and definitely 00:14:09.049 --> 00:14:10.750 keep questioning. Outtifro Music.