1 00:00:00,040 --> 00:00:03,200 I'm here with Benedict Evans. We worked together at a six and 2 00:00:03,200 --> 00:00:06,840 Z more than 10 years ago. Benedict is, you know, well 3 00:00:06,840 --> 00:00:08,920 known newsletter author, probably needs no introduction 4 00:00:08,920 --> 00:00:11,400 for people watching this. We're here in Singapore. 5 00:00:11,400 --> 00:00:13,120 We just came here for an AI conference. 6 00:00:13,280 --> 00:00:17,040 You do about 1/4 newsletter, 3/4 conference nowadays or speaking 7 00:00:17,720 --> 00:00:18,920 that that's brought out here, right? 8 00:00:19,200 --> 00:00:22,960 Yeah, and newsletters down like 175, something like that, you 9 00:00:22,960 --> 00:00:24,200 said? Yeah, something like that. 10 00:00:24,400 --> 00:00:27,560 Wobbles a bit from day-to-day. And it started out you, you 11 00:00:27,560 --> 00:00:29,680 started as a mobile analyst and you became like a broader tech 12 00:00:29,680 --> 00:00:31,760 analyst. Is that is that the evolution? 13 00:00:32,119 --> 00:00:34,680 Yeah, that's one way to put it. I mean, I think. 14 00:00:35,360 --> 00:00:37,520 You're at orange, is that right? A long time ago. 15 00:00:37,600 --> 00:00:39,480 Yes, long time ago. Yes, just when it was all 16 00:00:39,480 --> 00:00:42,920 becoming horribly French, there's there's like as we were 17 00:00:42,920 --> 00:00:45,160 chatting before this and I said, like the thing in tech is that 18 00:00:45,160 --> 00:00:47,480 the point that you understand something is, is often the point 19 00:00:47,480 --> 00:00:49,800 that you should be moving on to pay attention to something else. 20 00:00:50,320 --> 00:00:53,920 I started my career in the.com bubble as an equity analyst and 21 00:00:53,920 --> 00:00:57,080 I was covering mobile stocks. And at that time mobile was kind 22 00:00:57,080 --> 00:01:00,560 of dynamic and exciting and sexy and disruptive and they turned 23 00:01:00,560 --> 00:01:02,600 into water companies. They were going. 24 00:01:02,600 --> 00:01:04,519 Into water companies, utilities. Oh, utilities. 25 00:01:04,519 --> 00:01:04,959 Yeah. OK. 26 00:01:05,160 --> 00:01:07,120 They were going to connect to everybody in the world and then 27 00:01:07,120 --> 00:01:08,920 they did. And like now what they would 28 00:01:08,920 --> 00:01:10,840 like Marc Andreessen's phrase. They were like the dog that 29 00:01:10,840 --> 00:01:14,480 caught the truck, right? And I went and worked in 30 00:01:14,480 --> 00:01:16,760 strategy and a bunch of media and telecoms things. 31 00:01:16,840 --> 00:01:20,240 And yeah, I was analyzing, looking at smartphones because 32 00:01:20,240 --> 00:01:22,480 that was suddenly become the center of the industry and no 33 00:01:22,480 --> 00:01:26,120 one understood it. Now, like it happened, I'm time 34 00:01:26,160 --> 00:01:29,680 to look for different questions. Well, it's, it's funny because I 35 00:01:29,680 --> 00:01:32,720 think when we were overlapping, it was right in the middle of 36 00:01:32,720 --> 00:01:35,440 the, the smartphone dividend, the smartphone explosion. 37 00:01:35,440 --> 00:01:38,080 And just to, you know, we actually there's a few things. 38 00:01:38,080 --> 00:01:40,640 One is the smartphone dividend. That's a useful concept, right? 39 00:01:40,640 --> 00:01:44,200 Like that the rise of a billion smartphones meant that 40 00:01:44,200 --> 00:01:46,000 everything that went into them became cheaper. 41 00:01:46,000 --> 00:01:49,160 And that enabled VR headsets, that enabled drones, right? 42 00:01:49,160 --> 00:01:50,880 All this stuff. Yeah, all the components that 43 00:01:50,880 --> 00:01:55,360 came out of it, Yeah. So smartphone sales and now from 44 00:01:55,360 --> 00:01:59,960 memory like 1 and a quarter, 1 1/2 billion units a year and all 45 00:01:59,960 --> 00:02:03,160 the supply chain from that, all of those components is then 46 00:02:03,160 --> 00:02:05,880 available off the shelf if you want to buy 5000 of them or 47 00:02:05,880 --> 00:02:08,880 10,000 of them, all the Wi-Fi chips and the batteries and the 48 00:02:08,880 --> 00:02:10,800 cameras and and all the other bits. 49 00:02:11,360 --> 00:02:15,560 And before, if you wanted to put computer into something, you'd 50 00:02:15,560 --> 00:02:17,440 basically need to use PC components. 51 00:02:17,440 --> 00:02:21,200 So ATMs and so on, all basically PCs, like elevators are 52 00:02:21,200 --> 00:02:26,760 basically PCs, and that has size and power and cost constraints. 53 00:02:26,760 --> 00:02:29,840 And then smartphones become the thing and then all those 54 00:02:29,840 --> 00:02:32,640 components are available. And so that's what gets you 55 00:02:32,680 --> 00:02:36,960 drones and connected light bulbs and all the other bits and 56 00:02:36,960 --> 00:02:39,160 pieces around the edge of that. One of the things that I think 57 00:02:39,160 --> 00:02:42,040 people don't appreciate is they think, for example, like the 58 00:02:42,040 --> 00:02:45,280 consumer, they think like the military has like special gear 59 00:02:45,280 --> 00:02:47,440 and it's got its own kind of supply chain. 60 00:02:47,440 --> 00:02:50,840 And often the military supply chain is often just a subset of 61 00:02:50,840 --> 00:02:53,680 the consumer supply chain, because you sell a billion units 62 00:02:53,760 --> 00:02:56,120 into this and maybe you have 100,000 or 1,000,000 units of a 63 00:02:56,120 --> 00:02:58,160 military thing. It's, it's actually, it's almost 64 00:02:58,160 --> 00:03:00,600 kind of the reverse now in that it used to be. 65 00:03:00,600 --> 00:03:03,960 So that what the way I think about this is like in the past, 66 00:03:03,960 --> 00:03:07,600 like before we were born, the intelligence agencies would get 67 00:03:07,600 --> 00:03:10,920 the cool new stuff 1st and then the military would get it and 68 00:03:10,920 --> 00:03:13,520 then big corporations would get it and eventually consumers 69 00:03:13,520 --> 00:03:14,960 would get it like 30 years afterwards. 70 00:03:14,960 --> 00:03:16,600 So this is like the connection, yes. 71 00:03:16,600 --> 00:03:18,560 It's like microwaves were invented for NASA, right? 72 00:03:18,560 --> 00:03:20,240 And eventually consumers get them. 73 00:03:20,440 --> 00:03:23,320 Yeah, or like GPS was invented to guide missiles and now it's 74 00:03:23,320 --> 00:03:27,000 used for tagging cat photos. And the shift is like a 75 00:03:27,000 --> 00:03:30,240 combination of the stuff getting cheap enough that it can be for 76 00:03:30,240 --> 00:03:32,960 consumers instead of you needing a billion dollars to have one, 77 00:03:33,520 --> 00:03:36,560 and then the scale of consumers once it gets cheap enough. 78 00:03:37,040 --> 00:03:40,680 And so now the way it works is the consumers get the new stuff 79 00:03:40,680 --> 00:03:44,320 and the military gets it 10 years later because that's how 80 00:03:44,320 --> 00:03:46,720 long it takes to. The bureaucracy to similar to. 81 00:03:46,720 --> 00:03:51,080 That A, the bureaucracy B, to harden it and productize it and 82 00:03:51,080 --> 00:03:54,360 turn it into what you need if you're going to get shot out or 83 00:03:54,360 --> 00:03:56,520 it's going to be cold or hot or warm or whatever it is. 84 00:03:56,600 --> 00:04:00,920 Yeah, that's it's funny. Does it really improve through 85 00:04:00,920 --> 00:04:03,520 that process? I know people think it does, but 86 00:04:03,520 --> 00:04:07,600 I'm not sure it does relative to the the the cost of not using 87 00:04:07,920 --> 00:04:11,360 the pretty good product versus whatever improvements come from 88 00:04:11,360 --> 00:04:13,160 the delay to harden it. I'm not sure if it actually. 89 00:04:13,160 --> 00:04:14,520 I don't know. I think this is, but there's 90 00:04:14,520 --> 00:04:17,160 clearly there's a sort of a process if you have to put it 91 00:04:17,160 --> 00:04:19,480 into a fighter jet, yes. You don't replace the avionics 92 00:04:19,480 --> 00:04:21,160 in a fighter jet every six months, right? 93 00:04:21,600 --> 00:04:24,400 Well, but but yeah, you know, that's the kind of the core of 94 00:04:24,400 --> 00:04:26,920 it is the, the cutting edge of the innovation is for consumers 95 00:04:27,120 --> 00:04:29,560 and then that flows back through to everything else. 96 00:04:29,560 --> 00:04:30,880 That's right. Well, you know, I was going to 97 00:04:30,880 --> 00:04:34,240 say maybe in China you do maybe in China, like I think what 98 00:04:34,240 --> 00:04:35,640 happened with the consumer drones. 99 00:04:35,920 --> 00:04:40,400 They got good at quadcopters and that's LED them to their new 100 00:04:40,400 --> 00:04:41,520 form. Have you seen E hang? 101 00:04:41,840 --> 00:04:43,640 It's like the Chinese flying cars. 102 00:04:44,480 --> 00:04:48,040 And, you know, I, I played this clip like a year, year and a 103 00:04:48,040 --> 00:04:52,480 half ago and people said, you know, the TL1, like we want to 104 00:04:52,480 --> 00:04:54,280 fly in cars. We've got 140 characters. 105 00:04:54,280 --> 00:04:56,520 And I was like, a lot of people didn't riff on that, but I was 106 00:04:56,520 --> 00:04:58,480 like, we want to fly in cars. We got them in Chinese 107 00:04:58,480 --> 00:04:59,680 characters. OK. 108 00:05:00,240 --> 00:05:02,240 And The thing is when we put that up there, people are like, 109 00:05:02,960 --> 00:05:07,400 that's not a car. It's, you know, it's a right, 110 00:05:07,560 --> 00:05:09,760 that's not a car, it doesn't have wheels, but it's solved the 111 00:05:09,760 --> 00:05:12,360 problem differently, right? And actually I think was one of 112 00:05:12,360 --> 00:05:14,920 your lines. It's like unfair comparisons are 113 00:05:14,920 --> 00:05:16,480 often the best kind of comparisons, right? 114 00:05:17,120 --> 00:05:20,800 Yeah, I remember seeing a bunch of flying cars when we were at 115 00:05:20,800 --> 00:05:22,720 Andreessen Horowitz, I think Marc Andreessen, he said. 116 00:05:22,720 --> 00:05:25,040 It's like they were all like houseboats, and a houseboat is a 117 00:05:25,040 --> 00:05:27,040 crap house and a crap boat. Yes, that's right. 118 00:05:27,040 --> 00:05:32,080 And, you know, think of it, thinking of it as a flying car 119 00:05:32,080 --> 00:05:34,200 is like the wrong term. It's better to think of it as 120 00:05:34,200 --> 00:05:37,200 like a much a small, much better, much cheaper helicopter. 121 00:05:37,240 --> 00:05:40,760 Yes, maybe, but the point is that they now the other thing is 122 00:05:40,960 --> 00:05:43,840 it's been for short hops and like city city where you fly 123 00:05:43,840 --> 00:05:45,760 over the traffic and they've got this. 124 00:05:46,000 --> 00:05:49,320 Uber was going to do this by the way, before they decapitated 125 00:05:49,320 --> 00:05:51,440 Uber. Like the low altitude economy 126 00:05:51,440 --> 00:05:52,960 was something they were thinking about. 127 00:05:53,360 --> 00:05:56,080 And a lot of things get like cut off in the West and then they 128 00:05:56,360 --> 00:05:59,440 appear fully formed in China, like consumer drones, for 129 00:05:59,440 --> 00:06:04,520 example, you know, Chris Anderson, he was very early on 130 00:06:04,520 --> 00:06:06,720 drones and that got blocked by the FAA. 131 00:06:07,120 --> 00:06:09,480 And so consumer drones were hobbled in the US and that's why 132 00:06:09,480 --> 00:06:12,040 DJI arose in China. So lots of things get blocked in 133 00:06:12,040 --> 00:06:14,000 the US and they arise in China because of that. 134 00:06:15,160 --> 00:06:18,400 Anyway, coming back up. So smartphones, I mean, I think 135 00:06:18,400 --> 00:06:21,880 you and Horace did you of a simple. 136 00:06:22,240 --> 00:06:25,320 I was on his pod a while ago. I think you're two of the best. 137 00:06:25,360 --> 00:06:27,040 He's also like European or something like that. 138 00:06:27,280 --> 00:06:28,840 French he was. He was at Nokia. 139 00:06:28,840 --> 00:06:31,240 I mean, there's an interesting kind of like information. 140 00:06:31,240 --> 00:06:33,440 Do you know him? Yeah, yeah, he's a great guy. 141 00:06:34,920 --> 00:06:40,440 Part of it was it was like, and there was a moment in time when 142 00:06:40,760 --> 00:06:44,200 there weren't many people doing mobile who really understood 143 00:06:44,200 --> 00:06:49,320 this and were industry analysts and were able to talk in public, 144 00:06:50,240 --> 00:06:50,600 right? Yes. 145 00:06:50,600 --> 00:06:54,920 So there were people inside Nokia or Goldman's or or Bain or 146 00:06:54,920 --> 00:06:57,760 wherever who had all the data, but they couldn't publish the 147 00:06:57,760 --> 00:06:59,840 data and they weren't allowed to say stuff in public, right? 148 00:07:00,080 --> 00:07:02,800 Or if they were writing analysis, it was analysis for 149 00:07:03,560 --> 00:07:05,760 public markets, investors or something, right? 150 00:07:05,920 --> 00:07:08,480 And so there were very few people who were like, knew that 151 00:07:08,480 --> 00:07:11,280 you could go and take Apple's reports and make a chart of unit 152 00:07:11,280 --> 00:07:14,240 sales and make a chart of ASP and knew what ASP was, right? 153 00:07:14,320 --> 00:07:18,440 Or knew what AP was right. Now there's like an explosion of 154 00:07:18,440 --> 00:07:20,720 this. So there's huge numbers. 155 00:07:20,720 --> 00:07:23,760 And particularly if you look at AI now, there's like 10 people 156 00:07:23,760 --> 00:07:27,520 who do a really, really good 200 page deck of every possible AI 157 00:07:27,520 --> 00:07:28,520 chart. Is that right? 158 00:07:28,560 --> 00:07:29,160 Interesting. Yeah. 159 00:07:29,480 --> 00:07:33,200 And so that whole thing shifted. But at the time, yes, it was me 160 00:07:33,200 --> 00:07:37,600 and chorus and like Ben Baharin. Well, like the only and. 161 00:07:37,960 --> 00:07:39,520 Stratectory. Kind of, yeah, Yeah. 162 00:07:39,680 --> 00:07:41,680 Exactly. Well, like, you know, a handful 163 00:07:41,680 --> 00:07:45,080 of people who understood this and could do the charts and were 164 00:07:45,080 --> 00:07:48,480 allowed to do the charts. And so that was sort of, you 165 00:07:48,480 --> 00:07:50,640 know, being at the right place at the right time got me a lot 166 00:07:50,640 --> 00:07:52,000 of attention. Yeah, it's interesting. 167 00:07:52,000 --> 00:07:57,320 I think like you and Ben Thompson Stratechory and I'm not 168 00:07:57,320 --> 00:08:01,120 sure if Horace had a newsletter, but you guys were newsletters 169 00:08:01,120 --> 00:08:05,240 before Subsec productized. It's sort of like Rogan was 170 00:08:05,480 --> 00:08:08,440 podcast before that became productized as a category. 171 00:08:09,800 --> 00:08:12,120 And are you in Subsec? Were you on ghost? 172 00:08:12,120 --> 00:08:12,760 No. Yeah. 173 00:08:12,760 --> 00:08:14,120 No, I'm you. Got your own custom? 174 00:08:14,120 --> 00:08:17,720 I'm still on my old cobble together stack of MailChimp plus 175 00:08:17,720 --> 00:08:19,680 member full plus Squarespace. Why? 176 00:08:19,720 --> 00:08:21,240 Why don't you? You don't want to move to 177 00:08:21,240 --> 00:08:22,280 something. It's just a pain to move. 178 00:08:22,280 --> 00:08:26,360 It's a heavy lift to move platform and you sit and do the 179 00:08:26,360 --> 00:08:29,120 analysis and you're like, this is a good use of like a week of 180 00:08:29,120 --> 00:08:31,040 my time, right? Maybe. 181 00:08:31,120 --> 00:08:33,240 Maybe it might be at this point sub stack's pretty good, but I 182 00:08:33,240 --> 00:08:35,280 mean, it's you. Know you're well, there's 183 00:08:35,280 --> 00:08:38,000 there's a separate sub stack thing thing which is do you want 184 00:08:38,000 --> 00:08:40,200 it to be on your, your newsletter or your sub stack? 185 00:08:41,000 --> 00:08:42,799 Yes, that's true. Yeah, because it's a platform. 186 00:08:42,880 --> 00:08:46,080 And you get the advantage of I mean this is something we can 187 00:08:46,080 --> 00:08:47,760 talk about. It's Chris Dixon's line have 188 00:08:47,760 --> 00:08:49,200 come for the tool state for the network, right. 189 00:08:49,400 --> 00:08:51,680 You go on subset, they will get you new subscribers because it 190 00:08:51,680 --> 00:08:52,880 won't get you subscribers. Yes. 191 00:08:53,320 --> 00:08:56,200 On the other hand, they now they control who your readers are and 192 00:08:56,200 --> 00:08:59,040 you don't, which is always a thing of a network. 193 00:08:59,080 --> 00:09:01,040 Well. I mean used to still mail out. 194 00:09:01,040 --> 00:09:05,080 There that you do, but then them they're trying to get you to use 195 00:09:05,080 --> 00:09:07,800 their website and their algorithm to decide who reads 196 00:09:07,800 --> 00:09:09,560 and what that's true. So there's always these kind of 197 00:09:09,560 --> 00:09:12,480 questions, like, do you want to go with the people who will give 198 00:09:12,480 --> 00:09:14,280 you an audience? And in exchange for that, 199 00:09:15,000 --> 00:09:16,720 they're deciding that they'll give you the audience. 200 00:09:16,720 --> 00:09:18,080 It's always a trade off for the distribution. 201 00:09:18,120 --> 00:09:20,520 Yeah, I think Ghost is another option. 202 00:09:20,560 --> 00:09:21,680 Yeah, yeah. And I think. 203 00:09:22,280 --> 00:09:25,680 Ghost and Beehive are the 2. Others that people use ghost is 204 00:09:25,960 --> 00:09:29,400 like, you know, I saw ghost and it's very early and I just 205 00:09:29,400 --> 00:09:32,720 thought it was so good for for what it was like it was. 206 00:09:33,080 --> 00:09:35,640 I mean not even for it's just a very polished thought through 207 00:09:35,640 --> 00:09:37,880 for an open source product. It's unusually polished. 208 00:09:37,880 --> 00:09:42,520 John Nolan's very, very good. It's funny, you know, like on 209 00:09:42,520 --> 00:09:45,120 that, but there's a bunch of things we can talk about. 210 00:09:45,120 --> 00:09:48,840 But the, the whole newsletter thing, it's sometimes there's 211 00:09:48,840 --> 00:09:51,720 things that are like newsletters or podcasts that are what are 212 00:09:51,720 --> 00:09:54,360 considered lower case in technology before they become 213 00:09:54,360 --> 00:09:57,080 upper case. Like for example, Odeo, you 214 00:09:57,080 --> 00:10:00,200 know, like what Twitter was Twitter was a podcasting company 215 00:10:00,200 --> 00:10:02,520 before it became Twitter and the time constant. 216 00:10:02,520 --> 00:10:05,600 They just got the time constant wrong where you needed, which is 217 00:10:06,000 --> 00:10:08,600 hard to predict that microblogging would take off 218 00:10:08,600 --> 00:10:10,320 1st. And then it required like 219 00:10:10,880 --> 00:10:14,240 Airpods and everybody being online for a long time and 220 00:10:14,520 --> 00:10:18,960 maybe, you know, COVID before podcast really exploded and the 221 00:10:18,960 --> 00:10:22,000 term was around in lowercase. And even argue it's needed like 222 00:10:22,000 --> 00:10:24,760 5G or 4G, something like that. Yes, if you're if you're 223 00:10:24,760 --> 00:10:27,840 listening to it on the car, in the car, then you need a half 224 00:10:27,840 --> 00:10:30,440 fast enough network. Yes, bandwidth is a constraint, 225 00:10:30,440 --> 00:10:32,160 yes. And then the time works. 226 00:10:32,600 --> 00:10:34,320 So what do you think is lowercase today that's going to 227 00:10:34,320 --> 00:10:36,720 become uppercase? Like what's that? 228 00:10:36,720 --> 00:10:40,040 What's what exists in tech that people are like, Oh yeah, that 229 00:10:40,040 --> 00:10:41,640 exists. That that's going to go big. 230 00:10:42,440 --> 00:10:44,000 I, I have some ideas. I want to hear yours. 231 00:10:44,160 --> 00:10:49,600 Interesting question. I think there's probably the 232 00:10:49,600 --> 00:10:52,880 answer. If I was a consultant in trying 233 00:10:52,880 --> 00:10:56,880 to whiteboard this is I would be looking around AI because that's 234 00:10:56,880 --> 00:11:00,560 a new platform and you know, there's a lot all the white, the 235 00:11:00,560 --> 00:11:03,120 old white space got filled in and now you've got a whole bunch 236 00:11:03,120 --> 00:11:05,080 of new white space. So deterministically, there 237 00:11:05,080 --> 00:11:06,760 should be a bunch of those things here. 238 00:11:08,400 --> 00:11:11,120 AI, which I'm sure we'll talk about, does feel very sort of 239 00:11:11,120 --> 00:11:15,920 mid 90s in that you're like mid 90s Internet in that like, well, 240 00:11:15,920 --> 00:11:18,440 this is a browser, how do you use it? 241 00:11:18,440 --> 00:11:20,720 What's it for? How would you get to it? 242 00:11:20,720 --> 00:11:22,280 How does this work? Where's the value? 243 00:11:22,560 --> 00:11:24,440 Where's the value capture going to be? 244 00:11:25,880 --> 00:11:30,240 I'm not sure that there's like maybe 1 answer is like I'm too 245 00:11:30,240 --> 00:11:33,720 old and I'm not like spending too much time looking for like 246 00:11:33,720 --> 00:11:35,240 weird weird stuff around the edges. 247 00:11:36,120 --> 00:11:39,040 The last one of these that I spotted personally was Sheehan 248 00:11:39,520 --> 00:11:41,280 which. Is it Sheehan? 249 00:11:41,280 --> 00:11:43,480 Or I'm told it's Sheehan. Is that right to speak to people 250 00:11:43,480 --> 00:11:44,880 there? Like I haven't worked out her 251 00:11:44,880 --> 00:11:49,040 team yet. That was an interesting one that 252 00:11:49,040 --> 00:11:50,720 it was. Maybe you could also say it was 253 00:11:50,720 --> 00:11:52,240 the last of the ones that you could spot. 254 00:11:52,240 --> 00:11:54,320 Because suddenly, wait, what is this thing that's at the top of 255 00:11:54,320 --> 00:11:56,640 the iPod at the iPhone App Store charts all the time? 256 00:11:56,640 --> 00:11:57,280 Oh, I see. Yeah. 257 00:11:57,880 --> 00:12:00,960 Suddenly that thing exploded and that's like probably the same, 258 00:12:00,960 --> 00:12:03,480 probably the largest apparel that pure player apparel 259 00:12:03,480 --> 00:12:07,360 retailer on earth. Yeah, and like Sheehan and Temo 260 00:12:07,440 --> 00:12:09,800 and yeah, that's right, they're now getting hit with the tariff 261 00:12:09,800 --> 00:12:13,560 stuff and you. Know yeah, tariffs plus plus the 262 00:12:13,560 --> 00:12:16,360 de minimis rule in the US, but. That's only the US market and 263 00:12:16,360 --> 00:12:18,000 that's not, you know, I don't know what fraction of their 264 00:12:18,000 --> 00:12:18,680 sales that. Yeah, Yeah. 265 00:12:18,680 --> 00:12:20,440 It's like a third of their sales or a whole quarter of their 266 00:12:20,440 --> 00:12:22,240 sales or something. So that was like that was that 267 00:12:22,240 --> 00:12:23,520 was a thing that was interesting. 268 00:12:23,520 --> 00:12:25,320 I'm not sure. There's not like a new thing 269 00:12:25,320 --> 00:12:28,040 that I'm watching that I've noticed recently. 270 00:12:28,840 --> 00:12:31,160 I'm sure there will be. You know, I, I, I keep looking. 271 00:12:31,240 --> 00:12:33,080 So I have a few we we were talking about the glasses, 272 00:12:33,240 --> 00:12:35,080 right? Like I think smart glasses are 273 00:12:35,080 --> 00:12:38,200 sort of like the most predictable thing after the 274 00:12:38,200 --> 00:12:40,080 iPhone. Yeah, that's I put in a 275 00:12:40,080 --> 00:12:41,560 different category. I was sort of thinking like, 276 00:12:41,560 --> 00:12:43,720 what stuff that's being used now that people haven't quite 277 00:12:43,720 --> 00:12:46,440 noticed it's being used yet. I see so well I I guess that. 278 00:12:46,440 --> 00:12:48,360 Glasses glass is definitely your next thing. 279 00:12:48,480 --> 00:12:50,480 Sure. So so I guess I would, I would 280 00:12:50,480 --> 00:12:56,680 sort of bundle VR headsets, AR headsets, you know, like that 281 00:12:56,680 --> 00:12:59,920 with glasses and and say that that's just glasses are sort of 282 00:12:59,920 --> 00:13:03,160 the next version for goggles. But OK, so that's one that that 283 00:13:03,160 --> 00:13:05,240 that we agree on, except the question is, as you said, is it 284 00:13:05,240 --> 00:13:07,360 going to be watches or phones? How big does that get right? 285 00:13:08,240 --> 00:13:14,200 I think I think, you know, just like podcasts grew to mean like 286 00:13:14,200 --> 00:13:17,960 a video podcast and so on. The robot dogs are interesting. 287 00:13:19,040 --> 00:13:23,600 They are fun to play with and they're getting way cheaper now, 288 00:13:23,640 --> 00:13:24,640 right? They went from the Boston 289 00:13:24,640 --> 00:13:28,520 Dynamics kind of things. So the home robot as a toy, I 290 00:13:28,520 --> 00:13:30,440 think is probably going to become more and more popular 291 00:13:30,440 --> 00:13:32,480 like a Christmas present kind of thing at first, right? 292 00:13:33,320 --> 00:13:35,880 Because I see kids playing with them and they just love them 293 00:13:35,880 --> 00:13:39,200 just as a toy. And the, you know, it's kind of 294 00:13:39,200 --> 00:13:44,160 like the robot dog, the drone as like a starting to become a like 295 00:13:44,160 --> 00:13:45,400 a Christmas present kind of thing. 296 00:13:45,400 --> 00:13:48,280 I think that that becomes a thing and eventually, like we 297 00:13:48,280 --> 00:13:51,120 were talking about this at the Museum of the Future in the UAE, 298 00:13:51,480 --> 00:13:54,520 they clad these so that it doesn't, it's not just like a 299 00:13:54,520 --> 00:13:57,640 skeleton of a robot dog, but it actually looks like a animal. 300 00:13:57,800 --> 00:14:00,040 And that completely changes your perception of it, right? 301 00:14:00,800 --> 00:14:02,360 So I think that that'll be a thing. 302 00:14:03,840 --> 00:14:07,360 And with respect to AI, and so let's do, I mean, there's AI, 303 00:14:07,360 --> 00:14:09,320 there's Bitcoin, there's China, there's drones, there's biotech, 304 00:14:09,320 --> 00:14:10,840 there's actually several different areas that I'm 305 00:14:10,840 --> 00:14:12,120 tracking. I'm tracking eventually these 306 00:14:12,120 --> 00:14:15,560 very singularities, whatever. It's not all really actually 307 00:14:15,560 --> 00:14:17,720 singularities in the technical sense of going to Infinity, but 308 00:14:17,840 --> 00:14:19,240 ramps, curves, curves. That's right. 309 00:14:19,240 --> 00:14:24,920 Yeah. With AI, there's, you know, one 310 00:14:24,920 --> 00:14:29,320 way of thinking about it is like now we're 2 1/2 years in, let's 311 00:14:29,320 --> 00:14:31,120 say, let's call the ChatGPT moment, right? 312 00:14:31,680 --> 00:14:39,640 And it's interesting because it, I, I think what people really 313 00:14:39,640 --> 00:14:44,480 overestimated was how much it's agentic intelligence versus 314 00:14:44,480 --> 00:14:48,520 amplified intelligence like that say you still have to prompt it. 315 00:14:48,800 --> 00:14:50,800 So prompting is like higher level programming. 316 00:14:50,800 --> 00:14:53,040 Number one, you still have to verify the output. 317 00:14:53,440 --> 00:14:55,800 And that means you kind of need to know what it is you're 318 00:14:55,800 --> 00:14:57,200 looking for. For example, if it spits out a 319 00:14:57,200 --> 00:15:00,360 bunch of mathematical symbols in an area of math that you don't 320 00:15:00,360 --> 00:15:03,920 know, then you have to be Terrence Tao to verify it. 321 00:15:04,000 --> 00:15:06,480 It might be gibberish, it might be real, who knows, right. 322 00:15:06,880 --> 00:15:10,280 And so the prompting and verifying are actually the 323 00:15:10,280 --> 00:15:14,200 bottlenecks in many areas. Now Karpati and I, you know the 324 00:15:14,520 --> 00:15:16,760 Andrej Karpati, we were just having a discussion on this like 325 00:15:16,760 --> 00:15:21,000 a week or so ago. And the thing about verifying is 326 00:15:21,360 --> 00:15:24,240 if you're using the GPU's that we have built in and you're 327 00:15:24,240 --> 00:15:28,280 looking at images or video or front end code, right, like the 328 00:15:28,280 --> 00:15:31,400 like the user interface, your eye can just instantly pick out 329 00:15:32,200 --> 00:15:33,760 and you can verify pretty quickly. 330 00:15:34,160 --> 00:15:36,320 So for that side of things, AI is quite good. 331 00:15:36,320 --> 00:15:39,800 Anything that's images, video, your ear can also pick out 332 00:15:39,800 --> 00:15:42,160 audio, right? And front end. 333 00:15:42,640 --> 00:15:45,760 But when it's back end stuff, right, when it's like database 334 00:15:45,760 --> 00:15:49,960 code, when it's like crypto, when it's mathematical equations 335 00:15:50,560 --> 00:15:53,640 that you don't have like GPU's, you can't just like hit it with 336 00:15:53,640 --> 00:15:55,600 your eyes and and quickly detect it, right? 337 00:15:55,600 --> 00:15:57,560 That whether it's whether it's correct or not, you have to deep 338 00:15:57,560 --> 00:16:00,280 read it carefully, right? So it can generate reams of 339 00:16:00,280 --> 00:16:03,560 text, But then you have to verify it exactly. 340 00:16:03,560 --> 00:16:05,000 That's right. Maybe you have some thoughts on 341 00:16:05,000 --> 00:16:06,680 that? Well, so it's funny, I was 342 00:16:06,680 --> 00:16:08,520 talking to you, John Boltwick the other day and he said 343 00:16:08,520 --> 00:16:10,800 Benedict, you think in slides so. 344 00:16:11,160 --> 00:16:12,960 That we do too. We both think in slides. 345 00:16:12,960 --> 00:16:17,760 So I have a slide, yes. And, and maybe there's sort of 346 00:16:17,760 --> 00:16:19,880 a, I'll talk about the slide and there's sort of an observation 347 00:16:19,880 --> 00:16:21,560 around it. I think a lot of discussion of 348 00:16:21,560 --> 00:16:24,520 LLMS is sort of hunting for the like, what's the right, what 349 00:16:24,600 --> 00:16:26,600 would the right way to conceptualize this? 350 00:16:27,000 --> 00:16:29,040 It's like with machine learning, the right way to conceptualize 351 00:16:29,040 --> 00:16:30,360 it was this is pattern recognition. 352 00:16:30,360 --> 00:16:32,440 And we're still sort of hunting for the right way to 353 00:16:32,440 --> 00:16:36,200 conceptualize LLMS. The slide is that traditional 354 00:16:36,200 --> 00:16:38,600 software is deterministic and does things that are easy to 355 00:16:38,600 --> 00:16:41,480 explain to machines. In fact, automation machine 356 00:16:41,480 --> 00:16:44,160 tools, sewing machines, typewriters, adding machines, 357 00:16:44,560 --> 00:16:46,440 things that are easy to explain to a computer. 358 00:16:46,440 --> 00:16:49,360 There may be things that are very hard for people to to to 359 00:16:49,360 --> 00:16:51,920 do, but they're easy to explain. So it's hard for you to drill a 360 00:16:51,920 --> 00:16:55,000 hole 100 times or to calculate a mortgage in your head, but it's 361 00:16:55,000 --> 00:16:57,280 easy for you to write down the logical steps to explain how you 362 00:16:57,280 --> 00:16:58,680 do this. So that's traditional software 363 00:16:58,760 --> 00:17:01,960 like databases, data processing, the whole sixties, 70s mainframe 364 00:17:01,960 --> 00:17:04,599 thing. Machine learning is stuff that's 365 00:17:04,599 --> 00:17:09,000 hard to explain to a computer, so it's hard to explain why that 366 00:17:09,000 --> 00:17:10,200 credit card transaction is weird. 367 00:17:10,200 --> 00:17:11,319 It's hard. Or how to move your hands or 368 00:17:11,319 --> 00:17:12,680 something. Yeah, it's hard to explain why 369 00:17:12,680 --> 00:17:14,880 that's a picture of a dog and a cat, but you think it's easy 370 00:17:14,880 --> 00:17:17,400 until you try and do it. And then it's like you tried to 371 00:17:17,400 --> 00:17:20,240 make a mechanical horse. It always falls over until some 372 00:17:20,280 --> 00:17:22,680 robotics comes along. So that was machine learning. 373 00:17:23,119 --> 00:17:25,920 I also think that as a kind of quiz for you, do you think 374 00:17:25,920 --> 00:17:28,440 machine learning is still AI or is that now just software? 375 00:17:28,440 --> 00:17:31,360 Well, so the way that. I think there's a process right 376 00:17:31,360 --> 00:17:32,400 once it's been around for a while. 377 00:17:33,000 --> 00:17:34,080 It's AI anymore. It's funny. 378 00:17:34,080 --> 00:17:37,240 So I think like within the field, technically the division 379 00:17:37,240 --> 00:17:42,160 would be machine learning would be, you know, everything up to 380 00:17:42,160 --> 00:17:47,640 linear logist regression and you know, SVMS, all that kind of 381 00:17:47,640 --> 00:17:49,320 stuff. And then right at the point you 382 00:17:49,320 --> 00:17:52,360 start doing deep learning and you have large neural networks 383 00:17:52,640 --> 00:17:54,760 now, you start getting into what people would call modern AI. 384 00:17:55,120 --> 00:18:00,000 So ML is almost like the boundary of understandability, 385 00:18:00,000 --> 00:18:03,440 you might say, right where you can write clean equations and 386 00:18:03,440 --> 00:18:04,880 like really understand what's going on. 387 00:18:05,280 --> 00:18:08,760 And to me, the most surprising and confusing. 388 00:18:08,760 --> 00:18:13,240 I still don't feel like I know what the phenomenon is, but I 389 00:18:13,240 --> 00:18:15,880 still find it magical. It's something called the double 390 00:18:15,880 --> 00:18:17,160 descent problem. Do you know what that is? 391 00:18:17,720 --> 00:18:23,360 Basically, normally when you're fitting to data, you want to 392 00:18:23,360 --> 00:18:26,160 have the fewest possible parameters because you can 393 00:18:26,160 --> 00:18:29,840 overfit, right? And so your error goes down, and 394 00:18:29,840 --> 00:18:33,600 then your error starts going up on the holdout set. 395 00:18:33,600 --> 00:18:37,600 So you train your model in machine learning, and you want 396 00:18:37,600 --> 00:18:40,880 the minimum number of parameters to be able to explain the 397 00:18:42,280 --> 00:18:44,040 training data and predict the test data. 398 00:18:44,280 --> 00:18:46,600 And if you overfit, then you're no longer predicting out of 399 00:18:46,600 --> 00:18:50,040 sampled stuff. But double descent is when you 400 00:18:50,040 --> 00:18:53,920 do AI, you get actually a second wind when you start going to a 401 00:18:53,920 --> 00:18:57,080 very highly parameterized model and the error actually drops 402 00:18:57,400 --> 00:19:00,960 again, right? And which is just a really weird 403 00:19:00,960 --> 00:19:05,080 phenomenon that there's papers on this and so on. 404 00:19:05,560 --> 00:19:10,240 And it's one of the most counterintuitive things about 405 00:19:10,240 --> 00:19:13,000 the whole thing that just having these gigantically parameterized 406 00:19:13,000 --> 00:19:16,280 models would generalize. Well, right, Because it violates 407 00:19:16,280 --> 00:19:17,640 that. That's the biggest difference. 408 00:19:17,640 --> 00:19:19,000 Go ahead. There's other things people 409 00:19:19,000 --> 00:19:19,960 might say is the biggest. Difference. 410 00:19:19,960 --> 00:19:21,760 Well, that's why I mean, I think there's, there's one of the ways 411 00:19:21,800 --> 00:19:25,800 I, I sort of sort of think about the, the term AI is that people 412 00:19:25,800 --> 00:19:28,680 kind of use it like technology, the word technology. 413 00:19:28,760 --> 00:19:31,520 Yeah, that's right. That anything new is technology, 414 00:19:31,520 --> 00:19:34,000 anything your parents had. Is the technology I'm a stickler 415 00:19:34,000 --> 00:19:35,720 for precision? So there's, there's different 416 00:19:35,720 --> 00:19:38,760 ways that you can say, what do we mean by the word AII feel 417 00:19:38,760 --> 00:19:42,120 like AI has almost become like the word metaverse, where like 418 00:19:42,480 --> 00:19:44,520 you don't know what somebody means when they say it. 419 00:19:45,520 --> 00:19:47,560 Continue my slides. So the first point is there's 420 00:19:47,560 --> 00:19:49,680 deterministic software, which is stuff that's easy to explain, 421 00:19:49,680 --> 00:19:50,800 right? There's machine learning, which 422 00:19:50,800 --> 00:19:52,920 is stuff that was hard to explain, which basically machine 423 00:19:52,920 --> 00:19:55,720 learning solves this. And now an LLM is maybe stuff 424 00:19:55,720 --> 00:19:57,600 that's easy to explain to an intern. 425 00:19:58,960 --> 00:20:02,520 It's something where if you had to go away and have a have a 426 00:20:02,560 --> 00:20:06,040 like a kick off meeting and spend half an hour working out 427 00:20:06,040 --> 00:20:08,720 how we're going to do this project, then an LLM probably 428 00:20:08,720 --> 00:20:10,280 can't do that. But interesting. 429 00:20:10,280 --> 00:20:12,920 But if it's something that you could explain in 10 seconds or 430 00:20:12,920 --> 00:20:16,240 20 seconds, then an LLM is going to be able to do that. 431 00:20:16,520 --> 00:20:21,080 And part of the problem is, are you able to explain it even to 432 00:20:21,080 --> 00:20:23,720 yourself? Could you explain it to another 433 00:20:23,720 --> 00:20:26,120 human being? Can you actually kind of shut 434 00:20:26,120 --> 00:20:30,600 your eyes and conceptualize how is it that I'm going to explain 435 00:20:30,600 --> 00:20:32,120 what it is that I want this thing to? 436 00:20:32,120 --> 00:20:33,880 Do it's. So what you're saying is very 437 00:20:33,880 --> 00:20:39,280 important because there's several different angles I want 438 00:20:39,880 --> 00:20:43,600 to take off of that. You know, in one, since I had 439 00:20:43,600 --> 00:20:46,240 this tweet, we're living in the age of the phrase, right? 440 00:20:46,240 --> 00:20:52,520 So the prompt for the AI or the 140 character tweet or actually 441 00:20:52,560 --> 00:20:56,280 in crypto, like 14 words, 13 words, 12 words can be your, 442 00:20:56,480 --> 00:20:59,640 your crypto reset phrase, right? These like are phrases of power 443 00:20:59,840 --> 00:21:03,000 right in, in AI, in social and crypto, right? 444 00:21:03,000 --> 00:21:07,120 Like this strings of characters that do a lot, you know, and 445 00:21:07,120 --> 00:21:09,040 they're spells, they're spells, right? 446 00:21:09,400 --> 00:21:14,280 And the thing about it is the, the crisper you are as a 447 00:21:14,280 --> 00:21:17,320 manager, like, you know, if you're, if you're, if you're a 448 00:21:17,320 --> 00:21:20,320 really good engineering manager, you're great at prompting AI 449 00:21:20,840 --> 00:21:24,160 because crucially, you don't just say, Hey, code this, you 450 00:21:24,160 --> 00:21:28,360 say, hey, you know, try and use, you know, React for this. 451 00:21:28,360 --> 00:21:31,200 You can use React native for, you know, the iOS and Android 452 00:21:31,200 --> 00:21:33,080 interfaces. Use Tailwind, use it. 453 00:21:33,280 --> 00:21:36,000 The more in a sense vocabulary terms you have, the better you 454 00:21:36,000 --> 00:21:38,320 can prompt something with and if you use the vocabulary terms 455 00:21:38,320 --> 00:21:40,200 correctly. And what that meant is, for 456 00:21:40,200 --> 00:21:43,360 example, I realized with Dolly, you know, when that was first, 457 00:21:43,400 --> 00:21:46,160 you know, before the chat CPT moment, I was like, wow, art 458 00:21:46,160 --> 00:21:47,560 history is now an applied subject. 459 00:21:47,840 --> 00:21:52,000 Knowing like Cezanne and Picasso and what you know, these various 460 00:21:52,000 --> 00:21:53,920 kinds of obscure styles, Suddenly you can be like boom, 461 00:21:54,080 --> 00:21:55,560 style it like this, style it like this. 462 00:21:55,720 --> 00:21:57,000 And it'll do that, right? You could. 463 00:21:57,000 --> 00:21:59,920 Say the same thing for music, like what exactly is it that's 464 00:21:59,960 --> 00:22:02,160 being done there? And there is a word for that. 465 00:22:02,880 --> 00:22:04,600 So and you have to you have a word. 466 00:22:04,960 --> 00:22:05,760 That's right. Exactly. 467 00:22:05,760 --> 00:22:09,720 So you can upload a track and you can say what style is this? 468 00:22:09,720 --> 00:22:11,040 How do you caption this? Right. 469 00:22:11,840 --> 00:22:14,760 Ever seen, you know, like the restaurants with the fancy menus 470 00:22:14,760 --> 00:22:19,720 and they don't say. They don't say tomatoes. 471 00:22:19,720 --> 00:22:21,480 They seem like heirloom. There's a word. 472 00:22:21,480 --> 00:22:24,720 There's things that are theoretically subjective, yeah, 473 00:22:24,720 --> 00:22:26,800 but there are within the professions there is a 474 00:22:26,800 --> 00:22:29,040 particular term for doing that particular thing. 475 00:22:29,040 --> 00:22:30,240 Exactly. It's like the difference between 476 00:22:30,240 --> 00:22:34,240 like red versus burgundy and you know, Crimson and what have you. 477 00:22:34,240 --> 00:22:37,320 They've got precise words, which means something and then you can 478 00:22:37,320 --> 00:22:40,040 summon greater precision with those precise words. 479 00:22:40,400 --> 00:22:43,520 And So what a way of thinking about, you know, what you're 480 00:22:43,520 --> 00:22:46,680 saying is that, and I've written about this, AI is like 481 00:22:46,680 --> 00:22:51,080 undocumented AP is, right? So normal API, every function is 482 00:22:51,080 --> 00:22:53,280 like written out. And it's like, you can do this 483 00:22:53,280 --> 00:22:55,520 and you can do that. And so I've got 20 functions and 484 00:22:55,520 --> 00:22:56,640 here's everything is there, right? 485 00:22:57,040 --> 00:23:00,680 With AI, it can do lots of things that even the people who 486 00:23:01,400 --> 00:23:03,040 wrote it up. So, so there's, it's much more 487 00:23:03,040 --> 00:23:05,400 mysterious as to what it can do. You just have to try things 488 00:23:05,400 --> 00:23:06,960 right? So the way I was thinking about 489 00:23:06,960 --> 00:23:08,800 this from a different angle was to think about Guis. 490 00:23:09,720 --> 00:23:11,920 Oh yeah. Yeah, what a GUI is doing 491 00:23:11,960 --> 00:23:13,400 several things that a GUI is doing. 492 00:23:13,400 --> 00:23:15,600 One of them is, it's telling you all the features that the 493 00:23:15,600 --> 00:23:18,520 developers have created. And part of the reason that was 494 00:23:18,520 --> 00:23:20,720 a revolution is a, you knew what they were and you didn't need to 495 00:23:20,720 --> 00:23:23,680 memorize keyboard commands. But B, you can actually have 496 00:23:23,680 --> 00:23:25,720 more stuff because you're not constrained by the number of 497 00:23:25,720 --> 00:23:27,160 keyboard commands that you can write down. 498 00:23:27,320 --> 00:23:29,680 So you can have hundreds of functions instead of like, you 499 00:23:29,720 --> 00:23:31,520 know, you can just put them all. You can just add more shit to 500 00:23:31,520 --> 00:23:32,240 the menus. Yes. 501 00:23:33,200 --> 00:23:37,480 But the other part of it is that the GUI is telling the user a 502 00:23:37,480 --> 00:23:40,480 whole bunch of accumulated decision and institutional 503 00:23:40,480 --> 00:23:43,280 knowledge about what the right things to do at this point would 504 00:23:43,360 --> 00:23:45,560 be. And so if you're in a workflow 505 00:23:45,600 --> 00:23:48,720 as opposed to just a blank screen, you know, it's one thing 506 00:23:48,720 --> 00:23:50,160 if you're in like Photoshop or Excel. 507 00:23:50,240 --> 00:23:51,920 Yeah, it can. It can prompt you on the prompt. 508 00:23:51,920 --> 00:23:56,240 But if you're in a workflow in Salesforce, then there's a 509 00:23:56,240 --> 00:23:58,480 decision taken. This is I'm going to offer the 510 00:23:58,480 --> 00:24:01,960 use of these five options here and not 750 options. 511 00:24:02,240 --> 00:24:05,320 And with a, with a prompt, you don't have any of that. 512 00:24:05,640 --> 00:24:08,720 So you've got to shut your eyes and think for a minute of like, 513 00:24:09,120 --> 00:24:12,280 well, what would I do here? And you don't have that help. 514 00:24:12,280 --> 00:24:15,000 This is, you know, Carpathy has talked about this also, but I do 515 00:24:15,000 --> 00:24:17,360 think there's room for AIOS, right? 516 00:24:17,360 --> 00:24:21,240 Like in in a sense, and we can talk about crypto in a second, 517 00:24:21,240 --> 00:24:26,880 but I think and crypto are both actually operating system level 518 00:24:26,880 --> 00:24:30,280 innovations. And for example, maybe someone 519 00:24:30,280 --> 00:24:33,120 just does it as an app or like like a downloadable thing and 520 00:24:33,120 --> 00:24:34,680 just does it as a layer on top of the Mac. 521 00:24:35,000 --> 00:24:38,400 But if you have the full context of all the actions that are 522 00:24:38,400 --> 00:24:42,520 happening on your Mac, you can suggest which apps to use, 523 00:24:42,520 --> 00:24:45,760 suggest which apps to download, suggest, hey, you probably want 524 00:24:45,760 --> 00:24:47,120 to change these keyboard settings. 525 00:24:47,120 --> 00:24:49,920 And so like there's, you know, it's funny to put it this way, 526 00:24:49,920 --> 00:24:54,560 but Clippy is finally vindicated Clippy, but for everything, 527 00:24:54,560 --> 00:24:58,200 right? And because Clippy can now be 528 00:24:58,200 --> 00:25:00,160 really, really, really, really smart, right? 529 00:25:00,400 --> 00:25:02,320 Like, you know, it was Anderson's line. 530 00:25:02,320 --> 00:25:04,040 It's like everything in tech works. 531 00:25:04,040 --> 00:25:09,120 It's just when right. And, and even the thing that's 532 00:25:09,120 --> 00:25:12,040 interesting about the Clippy thing is somebody also made a 533 00:25:12,040 --> 00:25:17,360 point, which is that you actually want to put faces on 534 00:25:17,360 --> 00:25:19,720 your AI avatars, on your AI agents. 535 00:25:19,840 --> 00:25:22,480 So you could pick from Clippy or 10 other kinds of things. 536 00:25:22,480 --> 00:25:25,120 And the, and the reason you want to do that, this is counter 537 00:25:25,120 --> 00:25:30,200 intuitive, but people like you and I can use ChatGPT and, and 538 00:25:30,200 --> 00:25:32,400 you know, clawed and what have you, because we're familiar with 539 00:25:32,400 --> 00:25:35,120 interfaces, But the reason they're actually intuitive to 540 00:25:35,120 --> 00:25:38,040 100 million people is they're used to chatting with another 541 00:25:38,040 --> 00:25:40,960 human on the other side. So they're already modelling the 542 00:25:40,960 --> 00:25:43,680 chat box as being a human like response because they've been 543 00:25:43,680 --> 00:25:47,400 using WhatsApp or Facebook Messenger or Instagram chat or 544 00:25:47,400 --> 00:25:48,880 something like that for a long time, right. 545 00:25:50,360 --> 00:25:53,640 But when it's outside of that chat box environment and it's 546 00:25:53,640 --> 00:25:56,560 like suggesting on the screen, you kind of want to face to pop 547 00:25:56,560 --> 00:26:00,520 up so they can associate, OK, this person is suggesting this 548 00:26:00,520 --> 00:26:03,720 because that's who they are. And, and, and they kind of map 549 00:26:03,720 --> 00:26:05,520 that personality onto the AI agent. 550 00:26:05,960 --> 00:26:08,520 And so you can choose from different kinds of clippies that 551 00:26:08,520 --> 00:26:12,160 would give you prompts on what to do or, or it just does it for 552 00:26:12,160 --> 00:26:13,000 you. That's their possibility. 553 00:26:13,000 --> 00:26:14,640 But I don't think people like it when they does it. 554 00:26:14,640 --> 00:26:16,160 They want they want to be able to approve it. 555 00:26:16,160 --> 00:26:19,520 Before they do it, I think there's a sort of sense in here 556 00:26:19,720 --> 00:26:22,160 of how people people conceptualize what this thing is 557 00:26:22,160 --> 00:26:26,640 and how it works. I remember John Brotherow at 558 00:26:27,080 --> 00:26:31,680 Google showing me a chart, a Google Trends chart of best 559 00:26:31,680 --> 00:26:34,280 versus cheap. Best versus cheap? 560 00:26:34,800 --> 00:26:36,880 So the best does this and the cheap does that. 561 00:26:38,040 --> 00:26:39,920 And what are the axes? Crossing over time. 562 00:26:40,640 --> 00:26:43,200 So Google Trends. So what's the frequency of the 563 00:26:43,200 --> 00:26:45,160 word best? So it starts with like cheap 564 00:26:45,160 --> 00:26:46,920 phones and it goes to best phones, yes. 565 00:26:47,200 --> 00:26:50,440 And so the thesis was that this was shifting from the Internet 566 00:26:50,440 --> 00:26:53,560 as price comparison where you'd already knew what you wanted and 567 00:26:53,560 --> 00:26:55,840 it's at the top of the and that's the bottom of the funnel 568 00:26:56,200 --> 00:26:59,360 to the Internet as recommendation, curation, 569 00:26:59,360 --> 00:27:01,200 suggestion. Is that what you're looking for? 570 00:27:01,200 --> 00:27:02,640 Oh, that. Suggestion so interesting so let 571 00:27:02,640 --> 00:27:05,240 me see if I can understand the psychology so when they're. 572 00:27:05,280 --> 00:27:06,800 Used to it. So it starts from 20, it starts 573 00:27:06,800 --> 00:27:09,720 from 2004. So in 2004 you go on the 574 00:27:09,720 --> 00:27:13,200 Internet and you already know what you want and you look for 575 00:27:13,200 --> 00:27:16,640 the cheap, what is the cheap X? And then you put in a scoop or 576 00:27:16,640 --> 00:27:17,800 you put in a product or something. 577 00:27:18,480 --> 00:27:21,760 Whereas over time that goes down. 578 00:27:22,320 --> 00:27:24,520 And. Best goes up and crosses it. 579 00:27:24,520 --> 00:27:26,560 It's a perfect X on the chart. Unfortunate. 580 00:27:26,840 --> 00:27:30,280 And then the thesis is you're going further up the funnel. 581 00:27:30,520 --> 00:27:33,720 You're looking more and more for I want someone on the Internet 582 00:27:33,720 --> 00:27:37,360 to tell me the best X or Y, but previously you'd have got that 583 00:27:37,360 --> 00:27:39,280 from the magazine or newspaper or something. 584 00:27:39,280 --> 00:27:41,440 There's two or three. There's two things about that. 585 00:27:41,440 --> 00:27:44,400 The first is, you know, Andy Grossing about the paired 586 00:27:44,400 --> 00:27:48,280 metrics. So Andy Grove, whenever 587 00:27:48,280 --> 00:27:51,400 anybody's optimizing, like sales, for example, they will 588 00:27:51,560 --> 00:27:53,880 usually sort of recruiting, they'll start by optimizing 589 00:27:53,880 --> 00:27:55,880 quantity. But there's, you know, you can 590 00:27:55,880 --> 00:27:58,920 sometimes optimize quantity and then quality drops off, right? 591 00:27:59,240 --> 00:28:01,600 So quantity is easy to measure. It's like just the number of 592 00:28:01,600 --> 00:28:04,400 people we hired or whatever, but quality is how good were they, 593 00:28:04,480 --> 00:28:07,600 right? And so that's the the second 594 00:28:07,600 --> 00:28:10,280 paired metric is usually a quality metric that and so 595 00:28:10,720 --> 00:28:13,680 quantity is cheap, right? And people start with cheap and 596 00:28:13,680 --> 00:28:15,840 then quality is best and they go to best. 597 00:28:16,000 --> 00:28:18,400 So that's another lens on this 1/3 lens. 598 00:28:18,400 --> 00:28:21,240 What I thought my explanation, maybe it's different than what 599 00:28:21,240 --> 00:28:23,480 actually happened was when people are just trying out a 600 00:28:23,480 --> 00:28:26,320 space, they just want the cheap version to try it out. 601 00:28:26,840 --> 00:28:30,200 And once they've committed to a space at like for example, they 602 00:28:30,480 --> 00:28:33,480 cheap digital camera, cheap drone or something like that, I 603 00:28:33,480 --> 00:28:35,760 want to try it out, right? And they want to try it out at 604 00:28:35,760 --> 00:28:37,160 low cost. Try it before you buy. 605 00:28:37,160 --> 00:28:40,000 And then once they're committed to a space, then they're like, I 606 00:28:40,000 --> 00:28:42,200 want the best drone out there now because I'm, I want it. 607 00:28:42,200 --> 00:28:45,400 Well, the, the analysis then would be cheap drone versus best 608 00:28:45,400 --> 00:28:48,440 drone, right. But I think the that's. 609 00:28:48,440 --> 00:28:50,320 What I thought you were saying yes, but but you're saying 610 00:28:50,320 --> 00:28:53,120 cheapest, best overall. Yes, but I'd love to see a 611 00:28:53,120 --> 00:28:55,080 category by category. I wouldn't be surprised to see 612 00:28:55,360 --> 00:28:57,240 that happen category by category, but maybe not. 613 00:28:57,240 --> 00:28:59,240 Well, there's a. Different plot point there, 614 00:28:59,240 --> 00:29:01,480 which is sort of what I was talking about on in our in our 615 00:29:01,480 --> 00:29:04,080 panel this morning, which is this infinite product. 616 00:29:04,080 --> 00:29:07,000 So how do you know what to buy? And it used to be that you'd 617 00:29:07,000 --> 00:29:09,600 start with a magazine and then you'd go to the Internet to find 618 00:29:09,600 --> 00:29:12,200 the cheap place to buy it. Or you know what you wanted and 619 00:29:12,200 --> 00:29:14,200 now you go to the Internet to. Find. 620 00:29:14,680 --> 00:29:16,720 Like where's the, what's the right place to do this? 621 00:29:16,720 --> 00:29:18,920 So the Internet has become much more kind of a default. 622 00:29:19,200 --> 00:29:21,520 But actually the thing that the thought that prompted me to that 623 00:29:21,520 --> 00:29:24,240 was you can also go and play with Google trends and, and I 624 00:29:24,240 --> 00:29:30,240 did a chart played with like how, why, where what like more 625 00:29:30,240 --> 00:29:33,040 kind of basic questions. And you really need to be inside 626 00:29:33,040 --> 00:29:34,560 Google to do that analysis properly. 627 00:29:35,000 --> 00:29:37,640 But it's that sense of how much are people doing conversational 628 00:29:37,640 --> 00:29:42,240 queries into Google as a opposed to typing keywords into Google 629 00:29:42,800 --> 00:29:46,400 and things that are not really a Google query, like what is a is 630 00:29:46,400 --> 00:29:48,640 not. But it probably doesn't help 631 00:29:48,640 --> 00:29:50,480 Google. But that's still how people use 632 00:29:50,480 --> 00:29:52,200 it. People were trained for years to 633 00:29:52,200 --> 00:29:56,120 not to remove all prepositions, to remove all that stuff and 634 00:29:56,120 --> 00:29:59,760 just do keyword ease, and now were trained the opposite, to 635 00:29:59,760 --> 00:30:01,880 write full and complete English sentences. 636 00:30:01,880 --> 00:30:04,680 Like prompting is the new searching, but it's a completely 637 00:30:04,680 --> 00:30:06,440 different, you know, behavior, right? 638 00:30:07,240 --> 00:30:08,960 Go ahead. Well, so this is one of the, 639 00:30:09,960 --> 00:30:11,160 there's a sort of tangential point of that. 640 00:30:11,160 --> 00:30:15,000 Like one of the like the early, easy, obvious things that people 641 00:30:15,000 --> 00:30:19,120 have deployed with AI with LLMS on the Internet is different 642 00:30:19,120 --> 00:30:21,200 kinds of, is sort of natural language queries or different, 643 00:30:21,240 --> 00:30:23,600 not so much language, but like different kinds of query. 644 00:30:23,920 --> 00:30:26,800 So that the canonical one people talk about is Walmart saying now 645 00:30:26,800 --> 00:30:30,680 you can search for what should I buy to take on a picnic, which 646 00:30:30,680 --> 00:30:34,480 isn't a database query. And for Google, for Walmart or 647 00:30:34,480 --> 00:30:37,360 for Amazon, five years ago that search just wouldn't work 648 00:30:38,040 --> 00:30:40,760 because unless there's a product that's like tag with Picnic, 649 00:30:41,040 --> 00:30:45,280 it's not going to come up. Whereas now there's an LLM with 650 00:30:45,280 --> 00:30:47,920 a world model, they can pull it. Yes, that has some sense of what 651 00:30:47,920 --> 00:30:49,400 how you might answer that question. 652 00:30:49,600 --> 00:30:50,600 Yes. Is it a world model? 653 00:30:50,640 --> 00:30:52,200 It is. It's at least a web model. 654 00:30:52,200 --> 00:30:53,800 It's a different kind of query anyway. 655 00:30:53,800 --> 00:30:55,480 It's not. You're not doing a SQL query, 656 00:30:55,480 --> 00:30:57,640 you're doing something else. That's right. 657 00:30:57,720 --> 00:30:59,600 And I think, you know, one of the things that's interesting is 658 00:30:59,600 --> 00:31:05,840 the computers are we knew they were very good at that first 659 00:31:05,840 --> 00:31:08,000 kind of deterministic computation, the sequel query, 660 00:31:08,000 --> 00:31:09,400 the calculation. That's what they're built for, 661 00:31:09,400 --> 00:31:12,760 doing math, right? And now they've gotten good at 662 00:31:12,840 --> 00:31:14,120 probabilistic kinds of things, right? 663 00:31:14,120 --> 00:31:17,080 So this would be like system 1 and system 2 thinking, right? 664 00:31:17,080 --> 00:31:18,800 Probabilistic is like the quick impression. 665 00:31:18,800 --> 00:31:21,480 And then, you know, this is like the logical calculation. 666 00:31:21,680 --> 00:31:23,200 So it's actually good at the heart. 667 00:31:23,200 --> 00:31:27,560 The thing that's harder for humans is the, you know, like 668 00:31:27,880 --> 00:31:29,920 long involved mathematical because you can do that 669 00:31:29,920 --> 00:31:32,080 errorlessly. And now it can also do the other 670 00:31:32,080 --> 00:31:34,440 kind of thing. And so it does suggest that 671 00:31:34,440 --> 00:31:39,320 there would be some synthesis of that eventually where an AI can, 672 00:31:39,320 --> 00:31:41,360 I mean, this is like AI tool use or what have you. 673 00:31:41,400 --> 00:31:44,560 Like it detects that it needs to go to system 1 and and it starts 674 00:31:44,560 --> 00:31:48,080 invoking Python for that. And this is getting better, but 675 00:31:48,880 --> 00:31:54,120 it's surprisingly not amazing 2 1/2 years in right when when it 676 00:31:54,120 --> 00:31:57,640 needs to go deterministic. Well, so I, I wrote last long 677 00:31:57,640 --> 00:32:00,160 thing I read about this was about looking at deep research, 678 00:32:00,160 --> 00:32:05,280 which Open AI launched. And one of the kind of traps in 679 00:32:05,280 --> 00:32:08,480 looking at the new thing is to test it based on what was 680 00:32:08,480 --> 00:32:12,080 important to the old thing. So you know, to look at the 681 00:32:12,080 --> 00:32:14,680 Apple 2 and say, does this match the main the up time of a 682 00:32:14,680 --> 00:32:16,360 mainframe? No, So it's useless. 683 00:32:16,360 --> 00:32:18,560 Well, no, but that's not the right question. 684 00:32:19,480 --> 00:32:22,200 Can you write and build an Excel model on an iPhone? 685 00:32:22,200 --> 00:32:25,600 No, but that's not the point. It can still replace PCs and, 686 00:32:25,600 --> 00:32:28,400 and the reason I mentioned this is so, so, so, so deep research, 687 00:32:28,720 --> 00:32:31,800 open air launch that thing and it's whatever it was $100 a 688 00:32:31,800 --> 00:32:33,160 month or whatever. So, but then you look at the 689 00:32:33,160 --> 00:32:37,880 marketing page and the marketing page shows it doing a research 690 00:32:37,880 --> 00:32:41,480 project about mobile, which as we said, I know a lot about and 691 00:32:41,480 --> 00:32:45,280 it got the answers wrong and it's. 692 00:32:45,320 --> 00:32:47,200 That's verifying. See, you knew you could tell 693 00:32:47,200 --> 00:32:48,280 that it was wrong. Well, but it looked. 694 00:32:48,640 --> 00:32:49,600 It looked. Exactly. 695 00:32:49,600 --> 00:32:52,680 So this is the thing and it got stuff wrong in several levels. 696 00:32:53,440 --> 00:32:56,080 People think remembering now what I wrote like 2 months ago. 697 00:32:56,440 --> 00:32:59,600 And so there was a specific, it was make a table which shows 698 00:32:59,600 --> 00:33:02,480 mobile smartphone adoption in a bunch of countries and then the 699 00:33:02,480 --> 00:33:06,360 operating system market share. And then this is like an intern 700 00:33:06,360 --> 00:33:09,360 teaching moment because first of all, what does adoption mean? 701 00:33:09,880 --> 00:33:14,960 Does that mean unit sales share installed base App Store sales? 702 00:33:14,960 --> 00:33:15,760 Like what? What? 703 00:33:15,760 --> 00:33:18,360 What do you, what metrics specifically you asking me for? 704 00:33:18,440 --> 00:33:20,760 Yeah. Then it had given a source for 705 00:33:20,760 --> 00:33:23,600 the number it had come up with, which was Statista. 706 00:33:24,160 --> 00:33:26,760 And Statista is an aggregator that steals other people's data 707 00:33:26,760 --> 00:33:28,280 and who polishes it. Yeah. 708 00:33:28,440 --> 00:33:30,800 And when you jump through a bunch of registration hoops, you 709 00:33:30,800 --> 00:33:34,480 discover that the actual source was, I think Kantar. 710 00:33:35,360 --> 00:33:37,160 It's an it's an ad agency. It's OK. 711 00:33:37,400 --> 00:33:39,600 It's part of group. I thought it was part of one, 712 00:33:39,600 --> 00:33:41,360 that it's it's consumer survey data. 713 00:33:41,440 --> 00:33:43,480 So it's it's a proper, proper, proper company. 714 00:33:43,680 --> 00:33:45,360 Yeah. So it was actual proper consumer 715 00:33:45,360 --> 00:33:49,280 survey data, but the two things that so then when you go to the 716 00:33:49,280 --> 00:33:53,400 Canton chart page, you discover that deep research had got the 717 00:33:53,400 --> 00:33:56,360 numbers the opposite, so it'd flip percentages. 718 00:33:56,360 --> 00:33:59,400 I see. And then it had also said, 719 00:33:59,400 --> 00:34:00,720 right? Because it didn't have. 720 00:34:00,760 --> 00:34:02,800 Actually it just it didn't remove. 721 00:34:02,920 --> 00:34:04,040 It had copied them from the website. 722 00:34:04,040 --> 00:34:08,360 Well, I see and then the other source it gave was stat counter 723 00:34:08,960 --> 00:34:10,400 and stat. Counter was just using the same 724 00:34:10,440 --> 00:34:11,600 wrong. Data, which is a traffic 725 00:34:11,600 --> 00:34:13,320 measure. So that's not going to tell you 726 00:34:13,320 --> 00:34:16,639 an option because high end phones get used more and iPhones 727 00:34:16,719 --> 00:34:20,000 get used more So, and there's a bunch of things in here where 728 00:34:20,000 --> 00:34:21,560 you'd like. This is what I'd expect from an 729 00:34:21,560 --> 00:34:24,560 intern, right? I would go back and say, no, 730 00:34:24,560 --> 00:34:27,400 this is what I mean by adoption and this is a good data source. 731 00:34:27,400 --> 00:34:29,960 And that isn't right. And it's like a great first 732 00:34:29,960 --> 00:34:32,840 version. The problem is I had to copy the 733 00:34:32,840 --> 00:34:35,280 number out wrong, which is not what I would expect from an 734 00:34:35,280 --> 00:34:36,719 intern, or at least not a good intern. 735 00:34:37,480 --> 00:34:41,639 But secondly, I'd have to be a mobile analyst to know any of 736 00:34:41,639 --> 00:34:42,400 these. Things. 737 00:34:42,440 --> 00:34:43,760 And that's a verifying thing that I was. 738 00:34:43,760 --> 00:34:46,639 Getting at this is this is the kind of the core of it is all 739 00:34:46,639 --> 00:34:48,719 these people were looking at deep research and saying this is 740 00:34:48,719 --> 00:34:51,400 fantastic for researching things you don't know anything about. 741 00:34:51,400 --> 00:34:52,560 I was like. No, no it's not. 742 00:34:52,600 --> 00:34:55,480 Yeah, it's fantastic if you need a bunch of material about 743 00:34:55,480 --> 00:34:57,680 something you know a lot about. Exactly. 744 00:34:57,680 --> 00:35:00,440 So that's things that's why I think AI in its current 745 00:35:00,440 --> 00:35:03,040 incarnation is better thought of as amplified intelligence 746 00:35:03,480 --> 00:35:06,560 because the better the more you know about a field, the better 747 00:35:06,560 --> 00:35:09,040 you are at prompting because you got better vocabulary and the 748 00:35:09,040 --> 00:35:11,400 better you are at verifying because you know more facts 749 00:35:11,400 --> 00:35:13,280 about it and you have more cross cutting checks. 750 00:35:13,720 --> 00:35:16,640 And that is less true for the visual area. 751 00:35:16,960 --> 00:35:19,840 But just identifying that as a very important limitation where 752 00:35:19,840 --> 00:35:22,000 you have a completely different system you can use for the 753 00:35:22,000 --> 00:35:23,920 visual stuff, which is just your eyes, right? 754 00:35:24,720 --> 00:35:27,640 You don't have to use the you know, we have just different 755 00:35:27,640 --> 00:35:30,080 hardware for quickly seeing, you know, this way, the hands or 756 00:35:30,080 --> 00:35:32,240 something like that. Whereas if that. 757 00:35:32,320 --> 00:35:34,680 Was it's a monkey brain. It's a monkey brain, exactly 758 00:35:34,680 --> 00:35:36,040 right. So that's now an interesting 759 00:35:36,040 --> 00:35:37,480 question. This is, you know, Carpathy. 760 00:35:37,480 --> 00:35:40,840 I was discussing this is, is there some way to turn some or a 761 00:35:40,840 --> 00:35:46,960 subset of the non visual things into visual cues where you could 762 00:35:46,960 --> 00:35:49,400 see it was wrong immediately? So I'll give you a small and 763 00:35:49,400 --> 00:35:51,720 simple example. Let's say you generate an audio 764 00:35:51,720 --> 00:35:55,080 file, right? You know, like a spectrogram of 765 00:35:55,080 --> 00:35:57,160 an audio file, right? You could maybe immediately see 766 00:35:57,160 --> 00:35:58,480 if there's some artifact there, right? 767 00:35:58,480 --> 00:35:59,720 That's a trivial example. Right. 768 00:35:59,720 --> 00:36:02,960 So I think this it's a fascinating concept. 769 00:36:03,280 --> 00:36:07,600 I'm I would wonder whether that's the right split. 770 00:36:07,680 --> 00:36:10,160 OK, it's at least one split I found useful for now, but what 771 00:36:10,200 --> 00:36:11,640 are you thinking? Well, so the split I was 772 00:36:11,640 --> 00:36:16,320 thinking was that the natural language generation to make text 773 00:36:16,320 --> 00:36:20,160 is perfect, so the text is always grammatically correct. 774 00:36:20,200 --> 00:36:23,040 That is true, yes, but the model underneath, like the facts 775 00:36:23,040 --> 00:36:25,760 presented by the knot in the text, might be wrong. 776 00:36:25,960 --> 00:36:28,320 Yes. And that's sort of deceptive to 777 00:36:28,320 --> 00:36:30,960 us because we see the text is correct and it looks confident. 778 00:36:30,960 --> 00:36:36,920 Yeah, that's right. Whereas in an image, like you 779 00:36:36,960 --> 00:36:40,040 ask it for a picture of somebody and everything's perfect except 780 00:36:40,040 --> 00:36:45,040 the person's got six hands. I'm not sure conceptually what 781 00:36:45,760 --> 00:36:48,680 is it that that's flattened? Is that you're seeing two things 782 00:36:48,680 --> 00:36:53,400 in one layer or is it that? Do you see what I mean? 783 00:36:53,640 --> 00:36:54,960 I I see what you mean. I think it. 784 00:36:55,080 --> 00:36:57,040 Is or is it that it's a different level of? 785 00:36:57,080 --> 00:37:00,080 Well, maybe then maybe there's a different point here, which is 786 00:37:01,880 --> 00:37:03,320 if you ask for an image of a car. 787 00:37:03,520 --> 00:37:06,920 Yeah. And the car, like I actually did 788 00:37:06,920 --> 00:37:09,920 this ask for a fantasy 1960s French sports car, right? 789 00:37:10,080 --> 00:37:12,600 It will look French. It will look like a sports car. 790 00:37:12,600 --> 00:37:14,800 It will have 4 wheels. It might have two steering 791 00:37:14,800 --> 00:37:15,880 wheels. Yes, that's right. 792 00:37:15,920 --> 00:37:18,640 The two steering wheels is the equivalent of a grammatical 793 00:37:18,640 --> 00:37:21,000 mistake or spelling mistake in the text generator. 794 00:37:21,240 --> 00:37:25,120 Yes, because. However, it may also be that the 795 00:37:25,120 --> 00:37:28,160 balance of the car is all wrong and it would flip over if it 796 00:37:28,160 --> 00:37:30,240 tried to go round a corner. But you'd have to be an 797 00:37:30,240 --> 00:37:32,480 automotive expert to know that. Yes. 798 00:37:32,480 --> 00:37:35,200 So I'm saying that like there's. I'm sure you're saying levels of 799 00:37:35,200 --> 00:37:36,920 error. That's right. 800 00:37:36,920 --> 00:37:39,120 What you're saying is the 2 steering wheels is like a 801 00:37:39,120 --> 00:37:41,920 spelling error, but spelling errors are very rare for AI, 802 00:37:42,200 --> 00:37:44,160 whereas the two steering wheels is a common error, right? 803 00:37:44,520 --> 00:37:50,480 And I think that has to do with just the way diffusion models 804 00:37:50,480 --> 00:37:52,320 work versus how Transformers work. 805 00:37:52,320 --> 00:37:55,560 That'd be like 1 high level answer I'd give where it's doing 806 00:37:55,560 --> 00:38:02,040 like kind of it's more local with the diffusion model and you 807 00:38:02,040 --> 00:38:05,040 can be locally correct with the steering wheel but globally 808 00:38:05,040 --> 00:38:08,480 incorrect, whereas locally correct with spelling is usually 809 00:38:08,480 --> 00:38:10,480 correct. That's like maybe once that's 810 00:38:11,760 --> 00:38:12,760 useful. That's one answer. 811 00:38:14,600 --> 00:38:20,960 The second is that with there's only a small space. 812 00:38:21,000 --> 00:38:24,360 I think like, for example, we are optimized to recognize faces 813 00:38:24,360 --> 00:38:26,520 so we can detect very subtle differences in faces. 814 00:38:26,920 --> 00:38:29,760 But if I gave you like 5 different sheets of like static 815 00:38:29,760 --> 00:38:32,120 noise, right? Even if there are very clear 816 00:38:32,120 --> 00:38:34,520 patterns, like mathematically these are all like Fourier 817 00:38:34,520 --> 00:38:37,040 transforms of the same object and this is the one, they're 818 00:38:37,040 --> 00:38:39,720 just like total noise to you. A computer would be like these 819 00:38:39,720 --> 00:38:41,760 12 are the same and this one is odd one out, right? 820 00:38:42,200 --> 00:38:46,800 So in a sense, our eyes are optimized for a very low 821 00:38:46,800 --> 00:38:48,920 dimensional set of things, which are the things that occur in the 822 00:38:48,920 --> 00:38:50,520 real world. Like those are the things we can 823 00:38:50,520 --> 00:38:52,600 pick out, right? Which is also that our eyes are 824 00:38:53,320 --> 00:38:54,920 like, dogs are better at motion than us. 825 00:38:54,920 --> 00:38:56,360 Yeah. So even eyes are different 826 00:38:56,360 --> 00:38:57,720 depending on the species. That's right. 827 00:38:57,720 --> 00:39:02,640 So, so, so because of that, we actually have a like because 828 00:39:02,720 --> 00:39:04,320 they can't detect patterns in static. 829 00:39:04,320 --> 00:39:06,200 That's like too high dimensional space. 830 00:39:07,040 --> 00:39:10,880 I think text is kind of like that because it can describe one 831 00:39:10,880 --> 00:39:15,320 of the most, I mean surprising things to me about how AI has 832 00:39:15,320 --> 00:39:16,960 evolved. We were talking about this 833 00:39:16,960 --> 00:39:19,280 question before is I was surprised you could get so much 834 00:39:19,280 --> 00:39:21,800 mileage out of pure text. The reason is. 835 00:39:21,800 --> 00:39:24,480 So much Watson. So much mileage out of pure 836 00:39:24,480 --> 00:39:28,320 text, right? And the reason I was surprised 837 00:39:28,320 --> 00:39:32,080 by that is, you know, you'd think. 838 00:39:32,160 --> 00:39:34,920 You mean like reasoning and all stuff that looks like reasoning? 839 00:39:34,920 --> 00:39:37,800 Reasoning and also spatial manipulation like like picking 840 00:39:37,880 --> 00:39:41,360 like like having cameras, having eyes, seeing the world, 841 00:39:41,360 --> 00:39:43,640 reasoning about it like a baby and so and so forth. 842 00:39:44,280 --> 00:39:50,520 It is amazing how much of that world humans have assigned 843 00:39:50,800 --> 00:39:54,680 machine readable labels to with text And the way that you know, 844 00:39:55,640 --> 00:39:58,400 it's just, it's just very surprising how well that worked. 845 00:39:58,400 --> 00:40:02,600 Like language, what I'm trying to say is in a few, in like 40 846 00:40:02,600 --> 00:40:05,080 words, you can describe, it's like code. 847 00:40:05,080 --> 00:40:07,400 You can describe many, many, many different kinds of things 848 00:40:07,400 --> 00:40:12,640 in like 40 words, right? And, and it's just more general. 849 00:40:15,600 --> 00:40:17,920 It's one of those things where if you're sometimes when you're 850 00:40:17,920 --> 00:40:20,520 really close to a space, you're actually more surprised by a 851 00:40:20,520 --> 00:40:22,040 breakthrough than if you're farther away. 852 00:40:22,840 --> 00:40:25,680 And I, I should say like, you know, even looking seeing all 853 00:40:25,680 --> 00:40:28,640 the style transfer stuff in the mid twenty 10's and seeing image 854 00:40:28,640 --> 00:40:30,520 net and seeing the benchmarks and so on and so forth. 855 00:40:31,160 --> 00:40:37,080 I was surprised that it got you know, a markup chain is well, if 856 00:40:37,080 --> 00:40:42,160 you saw this stuff before GPD 3, right, it was like semi 857 00:40:42,160 --> 00:40:44,960 coherent, but it didn't look like it was converging on 858 00:40:44,960 --> 00:40:46,920 something. You know, it just looked like, 859 00:40:47,280 --> 00:40:49,640 you know, it repeat itself many times and what have you. 860 00:40:50,680 --> 00:40:53,760 And the fact that it broke through to what it did just 861 00:40:53,760 --> 00:40:57,360 based on language was so counter intuitive. 862 00:40:57,880 --> 00:41:00,280 And it's, I think it's because it's such a high dimensional 863 00:41:00,280 --> 00:41:02,560 thing. It captures so many different 864 00:41:02,560 --> 00:41:04,720 aspects of the world. Like anything you can perceive 865 00:41:04,720 --> 00:41:05,920 in the world, there's a word for it. 866 00:41:05,920 --> 00:41:09,280 There's many words for it. And then we also have billions 867 00:41:09,280 --> 00:41:12,040 of people who've been typing those words for two decades, 868 00:41:12,040 --> 00:41:13,920 right. So in a sense, like the entire 869 00:41:13,920 --> 00:41:15,760 Internet, the video games and social media were like this 870 00:41:16,080 --> 00:41:21,960 bootstrapper for for AI anyway. So on the on the other hand, AI 871 00:41:21,960 --> 00:41:24,120 is very bad at spatial stuff. You know this thing called ARC? 872 00:41:25,040 --> 00:41:27,760 Francois Shillet has this benchmark. 873 00:41:28,040 --> 00:41:30,560 Yeah, that I know, Francois. Yeah, and so he is. 874 00:41:31,320 --> 00:41:35,080 His benchmark actually got beaten by the recent, you know, 875 00:41:35,760 --> 00:41:39,480 ChatGPT release and he's got like a new one. 876 00:41:40,160 --> 00:41:43,560 And it's almost like a tetrisy kind of thing that's got some 877 00:41:43,560 --> 00:41:46,680 degree of logic and spatial type stuff that AI finds it hard, but 878 00:41:46,680 --> 00:41:48,760 humans still find it easy. It's kind of like maybe the next 879 00:41:48,760 --> 00:41:55,160 generation CAPTCHA and it's it's visual more than it is verbal, 880 00:41:55,680 --> 00:41:58,280 right? So for a reason. 881 00:41:58,280 --> 00:42:00,560 Is it something that would be hard to explain in words? 882 00:42:03,040 --> 00:42:07,680 Yes, I think kind of it's, it's about like this is here and and 883 00:42:07,800 --> 00:42:10,120 it's almost like minesweeper, you know minesweeper where you 884 00:42:10,280 --> 00:42:12,040 click and it expands and so on and so forth. 885 00:42:15,320 --> 00:42:20,200 I, I think AI because it started with words, it doesn't do well 886 00:42:20,200 --> 00:42:23,280 with the spatial side of things. Now on their hand, what the 887 00:42:23,280 --> 00:42:26,960 Chinese are working on in particular is physical robotics. 888 00:42:26,960 --> 00:42:28,800 Obviously Elon's working on it and so on and so forth. 889 00:42:28,800 --> 00:42:31,520 But China's way ahead on the physical supply chain. 890 00:42:31,520 --> 00:42:35,240 So like physical AI is robots and those definitely have 891 00:42:35,240 --> 00:42:38,720 cameras and XYZ and spatial and rotation and so on and so forth. 892 00:42:38,720 --> 00:42:41,840 So there's some eventual fusion, you know, like the the self 893 00:42:41,840 --> 00:42:44,280 driving cars have gathered hundreds of 1,000,000 billions 894 00:42:44,280 --> 00:42:47,640 of miles at this point. So there's some fusion of the 895 00:42:47,640 --> 00:42:52,160 web, which is words and the world, which is, you know, 896 00:42:52,200 --> 00:42:54,080 spatial. That will get you like a 897 00:42:54,080 --> 00:42:57,600 completely, you know, maybe a fusion set where it can reason 898 00:42:57,600 --> 00:43:00,200 about the world as it is. It knows how tall Everest is 899 00:43:00,200 --> 00:43:01,960 because someone, some robot has hiked it. 900 00:43:01,960 --> 00:43:03,040 You know, like Google Street View. 901 00:43:03,400 --> 00:43:06,640 You might eventually imagine a bunch of humanoids walking the 902 00:43:06,640 --> 00:43:10,320 world just like that, you know? I I wrote a thing years ago 903 00:43:10,320 --> 00:43:15,640 about Street View and Yahoo, and the sort of thing I was kind of 904 00:43:15,640 --> 00:43:19,040 poking away at is that basically every big Internet system is a 905 00:43:19,040 --> 00:43:21,920 mechanical Turk. And the question is, where do 906 00:43:21,920 --> 00:43:23,720 you put the people? Where the humans, Yes. 907 00:43:23,760 --> 00:43:28,640 And with Google search, the people are everybody, a 908 00:43:28,640 --> 00:43:31,400 everybody making a link on a web page and B everybody using 909 00:43:31,400 --> 00:43:32,840 Google. That's true. 910 00:43:32,880 --> 00:43:36,040 Whereas with Yahoo they tried to like, have a bunch of people in 911 00:43:36,040 --> 00:43:37,680 an office. Yeah, doing it in the middle. 912 00:43:37,680 --> 00:43:39,880 Making a hierarchical list of all the websites on the 913 00:43:39,880 --> 00:43:41,680 Internet, which was became impossible. 914 00:43:41,760 --> 00:43:43,120 Yeah. And with Street View, you just 915 00:43:43,120 --> 00:43:45,080 pay a bunch of people to drive down every street in the world, 916 00:43:45,080 --> 00:43:46,200 which is actually not impossible. 917 00:43:46,200 --> 00:43:46,960 It's just expensive. It's. 918 00:43:46,960 --> 00:43:48,440 Just expensive. It's really, it's an interesting 919 00:43:48,440 --> 00:43:50,000 computation. It's not obvious that it would 920 00:43:50,000 --> 00:43:52,080 be feasible to it. It's funny, you know, the Yahoo 921 00:43:52,080 --> 00:43:56,760 thing, Yahoo, you know, I think got started in like the early 922 00:43:56,760 --> 00:43:59,960 mid 90s, right? I think 94 ish, 93, something 923 00:43:59,960 --> 00:44:00,440 like that. Yeah. 924 00:44:01,000 --> 00:44:06,680 And the thing about it is it like Yahoo had to kind of get to 925 00:44:06,680 --> 00:44:10,200 its limit before it was obvious that you needed something like 926 00:44:10,200 --> 00:44:16,880 Google because like web pages had to be suffused with at the 927 00:44:16,880 --> 00:44:19,280 time they put on page spam and so on and so forth. 928 00:44:19,960 --> 00:44:22,920 You had to kind of top out. You had to get enough web pages 929 00:44:22,920 --> 00:44:25,120 that the hierarchical model broke down. 930 00:44:26,520 --> 00:44:28,880 You had to get enough economic value that people were really 931 00:44:28,880 --> 00:44:31,080 incentivized to game the system and so on and so forth. 932 00:44:31,440 --> 00:44:36,080 Before, you know, maybe Yahoo could have self disrupted, but 933 00:44:36,080 --> 00:44:38,480 before something like Google was there, Yahoo almost built out 934 00:44:38,480 --> 00:44:40,240 enough of the web economy to make something like Google 935 00:44:40,240 --> 00:44:46,480 necessary, you know anyway, so one thing I wanted to talk 936 00:44:46,480 --> 00:44:50,520 about, I want to go through various other areas, but what is 937 00:44:50,520 --> 00:44:52,440 AI disrupted? What is AI going to disrupt? 938 00:44:52,560 --> 00:44:53,760 Right. So what is it already disrupted? 939 00:44:54,040 --> 00:44:59,440 So search has taken points off of Google share, you know, like 940 00:44:59,440 --> 00:45:04,160 stack Overflow, you know, their queries are down image search 941 00:45:04,360 --> 00:45:07,200 because now image search is image generation, obviously 942 00:45:07,200 --> 00:45:14,640 video, obviously many different kinds of specialty apps will, 943 00:45:16,440 --> 00:45:21,720 you know, things that are, for example, like various sales 944 00:45:21,720 --> 00:45:24,800 tools that make templated emails and things like that. 945 00:45:24,800 --> 00:45:28,480 Those all, you know, change. I, I, I'm not sure Salesforce, I 946 00:45:28,480 --> 00:45:32,120 mean, Salesforce is, you know, certainly they're using AI, but 947 00:45:32,320 --> 00:45:34,520 the entire Salesforce model, like spamming people with 948 00:45:34,520 --> 00:45:37,800 e-mail, I'm not sure that's going to last in the age of AI 949 00:45:37,800 --> 00:45:39,960 because you can spam so many of them now, right. 950 00:45:41,520 --> 00:45:44,560 So, so those are some of the, you know, obviously robotics, 951 00:45:44,560 --> 00:45:46,160 obviously protein folding and and whatnot. 952 00:45:47,440 --> 00:45:49,840 What is it going to disrupt that people haven't thought about 953 00:45:49,840 --> 00:45:51,640 yet? And I can give some ideas. 954 00:45:51,680 --> 00:45:53,360 Well, one answer is we don't know. 955 00:45:53,400 --> 00:45:56,560 It's like trying to say that ask that question about the Internet 956 00:45:56,560 --> 00:45:57,800 in 1994. Sure. 957 00:45:57,800 --> 00:45:59,960 And the joke is always that newspapers thought the Internet 958 00:45:59,960 --> 00:46:01,520 would be great because they'd save on printing. 959 00:46:01,560 --> 00:46:03,440 And they. At first, probably was good for 960 00:46:03,440 --> 00:46:04,040 them. Yes, they did. 961 00:46:04,040 --> 00:46:07,120 Yes. I did a slide in my last 962 00:46:07,120 --> 00:46:09,880 presentation I did because it struck me that people would 963 00:46:09,880 --> 00:46:12,520 always say, well, you know, Uber didn't sell software to taxi 964 00:46:12,520 --> 00:46:14,760 companies and Airbnb didn't sell software to hotels. 965 00:46:14,760 --> 00:46:16,200 They redefined what those things were. 966 00:46:16,480 --> 00:46:19,080 So I went and did a chart of, well, what happened to taxis 967 00:46:19,080 --> 00:46:21,560 versus what happened to hotels. And that's actually the other 968 00:46:21,560 --> 00:46:22,960 one, surprisingly. What happened? 969 00:46:24,240 --> 00:46:26,360 Taxing down is. Uber demolishes taxis. 970 00:46:26,480 --> 00:46:30,000 Obviously Airbnb is mostly additive to hotels. 971 00:46:31,160 --> 00:46:32,960 Why is that? I think this is Airbnb is a 972 00:46:32,960 --> 00:46:34,480 different kind of experience in a hotel. 973 00:46:34,480 --> 00:46:36,360 It's not the substitutional experience. 974 00:46:36,360 --> 00:46:38,880 Yeah, it's complimentary. Yeah, half of business, half of 975 00:46:38,880 --> 00:46:41,240 hotels are business. There's another whole bunch of 976 00:46:41,280 --> 00:46:43,320 conferences. There's a bunch that's about 977 00:46:43,320 --> 00:46:47,480 like, I mean just OK, so 2 examples like my fiance works 978 00:46:47,480 --> 00:46:51,920 for goes to fly to Milwaukee. She arrives in town at 10:00 at 979 00:46:51,920 --> 00:46:53,400 night. She needs a gym. 980 00:46:53,400 --> 00:46:55,600 She's got a client meeting the next morning and then she's got 981 00:46:55,600 --> 00:46:57,880 a flying back to New York and she doesn't want to go and stay 982 00:46:57,880 --> 00:47:00,880 in some random strangers hotel which which you've got no idea 983 00:47:00,880 --> 00:47:02,680 what it's going to be like. She wants, you know, a very 984 00:47:02,680 --> 00:47:07,120 specific brand promise from you. Mean Brandon Strange Airbnb She 985 00:47:07,160 --> 00:47:08,840 won't stay in a hotel. Yeah, she will stay in a hotel. 986 00:47:08,880 --> 00:47:12,360 She's not staying in Airbnb. The other side of this is I 987 00:47:12,360 --> 00:47:14,200 think there's a more general point and same thing. 988 00:47:14,200 --> 00:47:16,160 I arrived in Singapore at 2:00 this morning. 989 00:47:16,160 --> 00:47:18,360 I'm not going to go and work out whether this Airbnb is any good. 990 00:47:18,360 --> 00:47:19,400 I'm going to stay in a hotel. Sure. 991 00:47:19,680 --> 00:47:22,040 I think there's a, there's a, there's a more general point, 992 00:47:22,080 --> 00:47:26,960 which is that like everything is probably disruptive to someone 993 00:47:27,360 --> 00:47:30,920 at some point in the value chain, but it kind of depends on 994 00:47:30,920 --> 00:47:33,960 the industry quite how much and in what sense. 995 00:47:34,280 --> 00:47:39,200 So like the iPhone demolished the existing cellular industry 996 00:47:39,720 --> 00:47:41,480 didn't really have any effect on telcos. 997 00:47:42,200 --> 00:47:44,240 Telcos kind of hoped that they were going to do all these 998 00:47:44,240 --> 00:47:45,880 services, but that was never going to happen. 999 00:47:46,240 --> 00:47:49,920 But telco mobile operators today are basically the same companies 1000 00:47:49,920 --> 00:47:52,240 that they were 20 years ago with more basically the same share 1001 00:47:52,240 --> 00:47:58,160 price because their business was not in anything that the iPhone 1002 00:47:58,160 --> 00:48:00,800 changed except that they're providing massively more data 1003 00:48:00,800 --> 00:48:03,440 than they were in the past. The business is basically owning 1004 00:48:03,440 --> 00:48:06,240 sites and owning spectrum and connecting them up and selling 1005 00:48:06,240 --> 00:48:08,800 that to consumers. The same thing with like online 1006 00:48:08,800 --> 00:48:11,400 travel booking completely demolished the travel agent 1007 00:48:11,400 --> 00:48:14,480 industry, didn't really change the airline business and 1008 00:48:14,480 --> 00:48:17,200 airlines had to do a bunch of stuff around loyalty and pricing 1009 00:48:17,200 --> 00:48:19,640 and maybe pricing became much more transparent and so on. 1010 00:48:19,880 --> 00:48:22,240 But at the end of the day, their business is owning or leasing 1011 00:48:22,240 --> 00:48:24,720 aeroplanes and. There's front end change. 1012 00:48:24,840 --> 00:48:28,840 Buying fuel and owning landing slots and maintaining the 1013 00:48:28,840 --> 00:48:32,800 aircraft. And so there now, of course, 1014 00:48:33,040 --> 00:48:35,640 the, the, the, the counter argument would be you could have 1015 00:48:35,640 --> 00:48:38,440 looked at taxis and say, well, clearly that's not going to get 1016 00:48:38,440 --> 00:48:41,160 changed by the Internet, except maybe you'll be able to book a 1017 00:48:41,160 --> 00:48:43,800 taxi more efficiently until it becomes long and changes it. 1018 00:48:44,160 --> 00:48:47,280 But the point is you can't, there's this sort of very naive 1019 00:48:47,280 --> 00:48:49,520 view that says, oh, well, there are, the software will just 1020 00:48:49,520 --> 00:48:50,840 destroy everything, right, right. 1021 00:48:50,920 --> 00:48:53,080 And it the answer is, well, it kind of depends. 1022 00:48:53,280 --> 00:48:54,680 It's path's moment that's true. Yeah. 1023 00:48:54,680 --> 00:48:57,800 And like one of the the ways that I sort of think about this 1024 00:48:57,840 --> 00:49:01,160 is that like the tech industry kind of it comes and changes 1025 00:49:01,160 --> 00:49:03,680 everything in the industry and resets how it works and then 1026 00:49:03,680 --> 00:49:07,040 leaves and goes off and works. And so you know the joke about 1027 00:49:07,040 --> 00:49:09,440 how consultants are seagulls. Yeah, they come, they fly. 1028 00:49:09,440 --> 00:49:11,480 In crap everywhere make lots of noise and fly out, right? 1029 00:49:11,600 --> 00:49:14,520 And so if you think about what happened to books or music, no 1030 00:49:14,520 --> 00:49:17,440 one in the tech industry cares about music anymore, right? 1031 00:49:17,520 --> 00:49:19,200 Like. Well, yeah, Spotify does, yeah. 1032 00:49:19,200 --> 00:49:20,040 Spotify but. It's not. 1033 00:49:20,040 --> 00:49:22,760 It's not the main main event. Yeah, recorded music is like $20 1034 00:49:22,760 --> 00:49:24,920 billion a year. It's like a rounding error in 1035 00:49:24,920 --> 00:49:27,520 the scale of the tech industry. It has no streaming means. 1036 00:49:27,520 --> 00:49:29,560 It has no strategic leverage for Apple or Google. 1037 00:49:29,640 --> 00:49:32,720 Suno is interesting though. So the so the AI, yeah. 1038 00:49:32,880 --> 00:49:36,320 Yeah but but but for the last 20 years, 20 years ago the Internet 1039 00:49:36,320 --> 00:49:38,920 completely screwed the music industry and since then it left 1040 00:49:38,920 --> 00:49:40,320 and doesn't care. Same thing in books. 1041 00:49:40,720 --> 00:49:43,320 Like all the conversations around books right now, some of 1042 00:49:43,320 --> 00:49:45,920 which were about Amazon, are book industry conversations. 1043 00:49:46,560 --> 00:49:49,320 I think there's something similar happening now with video 1044 00:49:49,320 --> 00:49:52,800 generation and Hollywood. Like everybody in Hollywood, 1045 00:49:52,800 --> 00:49:56,120 like all over the panic and now everyone is sitting and looking 1046 00:49:56,120 --> 00:49:58,320 at this and thinking, OK, well, this saves a bunch of second 1047 00:49:58,320 --> 00:49:59,680 unit stuff. So when we. 1048 00:49:59,720 --> 00:50:02,760 Think all like all the questions for What does this mean are 1049 00:50:02,760 --> 00:50:05,880 questions for people in LA. So one one way of thinking about 1050 00:50:05,880 --> 00:50:09,680 it is conversation is proportional to derivative 1051 00:50:09,680 --> 00:50:11,720 rather than absolute value. So let's say you have a, a 1052 00:50:11,720 --> 00:50:15,120 sigmoid that's going like like this and then it flattens out, 1053 00:50:15,160 --> 00:50:20,800 right? So when it's like a nullity or 1054 00:50:20,800 --> 00:50:24,040 ubiquity, you know, when it when it, when it doesn't exist or 1055 00:50:24,040 --> 00:50:27,240 when it's everywhere, when 0% or 100%, it's just not notable. 1056 00:50:27,240 --> 00:50:28,560 It's not worth talking about, right? 1057 00:50:28,800 --> 00:50:33,080 People use Uber or Dropbox a lot more today than when they were 1058 00:50:33,080 --> 00:50:35,760 talking about Dropbox and Uber a lot, right? 1059 00:50:36,080 --> 00:50:41,200 So the conversation is maximum at the time of maximum growth 1060 00:50:41,640 --> 00:50:43,920 and then it's just much less because now it's like not 1061 00:50:43,920 --> 00:50:46,560 notable, it's just a feature of the environment, right? 1062 00:50:46,640 --> 00:50:50,400 So you can do Google engrams? Yeah, that show exactly this, I 1063 00:50:50,400 --> 00:50:51,520 think. That'd be a great. 1064 00:50:51,520 --> 00:50:52,360 That'd be a. Great. 1065 00:50:52,360 --> 00:50:56,880 So you can do them for like steel or and some of. 1066 00:50:56,880 --> 00:50:58,480 These OK railroads, yeah, carload of. 1067 00:50:58,600 --> 00:51:01,680 Steel because it starts in 1800. And of course, some of them, you 1068 00:51:01,680 --> 00:51:04,080 look at it and you go, oh, I'm actually seeing a chart of World 1069 00:51:04,080 --> 00:51:08,440 War 2, yeah, where you see steel suddenly does that, or shipping 1070 00:51:08,440 --> 00:51:11,320 suddenly does that and. That's not obvious, right? 1071 00:51:11,320 --> 00:51:15,520 Because conversations or like attention is focused on change 1072 00:51:15,520 --> 00:51:18,240 rather than absolute value. Well, I always used to do AI, 1073 00:51:18,320 --> 00:51:20,400 always used to be fascinated by elevators. 1074 00:51:20,560 --> 00:51:23,400 I get these kind of autistic autism spectrum fascinations 1075 00:51:23,400 --> 00:51:27,800 about things, and there's a chart I did of the number of 1076 00:51:27,800 --> 00:51:30,120 people employed in the US as elevator attendance, which is a 1077 00:51:30,120 --> 00:51:32,080 perfect bell curve. Interesting. 1078 00:51:32,200 --> 00:51:33,720 Yeah, it's all curves up and down. 1079 00:51:34,000 --> 00:51:36,640 And this is because first half of the 20th century you. 1080 00:51:36,720 --> 00:51:38,480 Didn't have any. You deploy a lot of elevators, 1081 00:51:38,560 --> 00:51:41,520 right second-half of the 20th century, they become automatic 1082 00:51:41,640 --> 00:51:44,360 right button and you can go and find all this advertising. 1083 00:51:44,360 --> 00:51:46,440 Why were they at the beginning? Was it just like switch word? 1084 00:51:46,440 --> 00:51:47,360 Operators, I would tell you what. 1085 00:51:47,480 --> 00:51:49,600 Was it technical enough? There was no, but there was the. 1086 00:51:49,960 --> 00:51:52,160 Well, if you think about what it actually takes to have an 1087 00:51:52,160 --> 00:51:55,520 automatic, automatic elevator system in a building, you've got 1088 00:51:55,520 --> 00:51:59,520 to have all the dispatching. You've got to have the 1089 00:51:59,520 --> 00:52:01,680 dispatching and the queuing. I see. 1090 00:52:02,280 --> 00:52:04,320 There's an interim stage. We have an elevator attendant 1091 00:52:04,320 --> 00:52:06,080 who would just stand in the elevator and you would say I 1092 00:52:06,080 --> 00:52:08,320 want buff floor 5, please. And they'd press the button for 1093 00:52:08,320 --> 00:52:11,200 five, right? But if you get in, you know and 1094 00:52:11,200 --> 00:52:12,200 originally. Elevator. 1095 00:52:12,200 --> 00:52:13,800 What was it originally before the buttons? 1096 00:52:13,920 --> 00:52:16,360 There was a lever that's an accelerator and a brake. 1097 00:52:17,000 --> 00:52:18,160 Oh, so it's like a? Car. 1098 00:52:18,200 --> 00:52:21,120 Almost exactly. It's a streetcar elevator. 1099 00:52:21,120 --> 00:52:22,800 It's a vertical streetcar. I didn't know that. 1100 00:52:22,920 --> 00:52:26,480 So there's a fantastic book I have called The Cultural History 1101 00:52:26,480 --> 00:52:29,560 of Elevators, which is all about how weird this was so. 1102 00:52:29,640 --> 00:52:32,680 It was a vertical train. Yes, it's a vertical train. 1103 00:52:32,880 --> 00:52:33,840 Wow. And. 1104 00:52:33,840 --> 00:52:34,880 That's how people thought about it. 1105 00:52:35,240 --> 00:52:37,720 Yeah. And so an elevator attendant, 1106 00:52:37,720 --> 00:52:43,000 you can kill people. And there's this wonderful story 1107 00:52:43,000 --> 00:52:46,880 I tell everybody, which is that you you press the buzzer to 1108 00:52:46,880 --> 00:52:49,160 summon the elevator, but it's literally you're just ringing a 1109 00:52:49,160 --> 00:52:51,600 bell and a light goes on in the elevator car. 1110 00:52:52,160 --> 00:52:55,440 And there's this story from the the War Department. 1111 00:52:55,520 --> 00:52:58,160 It's like hailing a taxi. Yeah, there's a story from War 1112 00:52:58,160 --> 00:52:59,680 Department or ringing for a servant. 1113 00:53:00,320 --> 00:53:03,080 There's a story from the War Department in DC, which is that 1114 00:53:03,080 --> 00:53:05,360 you would buzz more based on how senior you work. 1115 00:53:06,560 --> 00:53:09,480 So imagine you're like a Lieutenant and you get into the 1116 00:53:09,480 --> 00:53:11,840 elevator on the 2nd floor and you want to go to the 10th 1117 00:53:11,840 --> 00:53:14,040 floor, but on the way the buzz rang. 1118 00:53:14,040 --> 00:53:17,640 It rings four times. So he has to stop on the 6th 1119 00:53:17,640 --> 00:53:20,720 floor and go down to the first floor and then a major gets in. 1120 00:53:21,240 --> 00:53:23,040 So now you're. So theoretically there's Paul of 1121 00:53:23,040 --> 00:53:24,680 10. It could be in a prior day in 1122 00:53:24,680 --> 00:53:26,160 the elevator going up and. Down. 1123 00:53:26,160 --> 00:53:28,520 So interesting. And we don't see any of this 1124 00:53:28,520 --> 00:53:30,200 now, which is your point about conversation. 1125 00:53:30,280 --> 00:53:32,000 Yeah, you don't get into an elevator now. 1126 00:53:32,000 --> 00:53:34,440 And so it's an, it's an electronic elevator, right? 1127 00:53:34,760 --> 00:53:37,120 It's automatic, right? It's just an elevator. 1128 00:53:37,680 --> 00:53:39,960 It's simply said something like there's a phrase which is 1129 00:53:39,960 --> 00:53:44,160 civilization advances as you can do more things without thinking 1130 00:53:44,160 --> 00:53:46,520 about them. Like the quote just work right 1131 00:53:48,320 --> 00:53:49,440 the. Classic one is light. 1132 00:53:49,440 --> 00:53:51,280 People with yeah, electricity. Light gets cheap. 1133 00:53:51,320 --> 00:53:55,480 Yes, that's right. And I think, you know, the age 1134 00:53:55,480 --> 00:53:59,120 of Internet now sometimes what happens is these things get 1135 00:53:59,120 --> 00:54:03,000 really ubiquitous and they're out of the conversation. 1136 00:54:03,000 --> 00:54:06,040 And then there's this. Now that you can treat them as 1137 00:54:06,040 --> 00:54:10,400 like at 100% adoption, then the new thing arises. 1138 00:54:10,400 --> 00:54:13,840 For example, all of the craziness of the last 10 years 1139 00:54:13,840 --> 00:54:17,160 is in part a function of the fact that social media got such 1140 00:54:17,280 --> 00:54:22,640 ubiquity in the early 20 tens such that it was no longer the 1141 00:54:22,640 --> 00:54:25,280 novelty was, oh, I'm on social media, I'm using it. 1142 00:54:25,280 --> 00:54:26,760 How do I use this Twitter app or whatever? 1143 00:54:27,040 --> 00:54:29,040 Everybody knows what Twitter is. Everybody knows how to use it. 1144 00:54:29,040 --> 00:54:30,680 They know what a like is, whatever, whatever. 1145 00:54:31,160 --> 00:54:34,200 And then you start getting. Then you get the 2nd order 1146 00:54:34,200 --> 00:54:35,440 effects the. 2nd order effects. That's right. 1147 00:54:35,440 --> 00:54:39,400 So it's almost like it's like installing a device driver and 1148 00:54:39,400 --> 00:54:41,320 then you can install the next one and the next one. 1149 00:54:41,320 --> 00:54:43,440 But it's like the device driver is the percentage of the 1150 00:54:43,440 --> 00:54:44,960 population that's adopted something. 1151 00:54:45,360 --> 00:54:49,320 And once it gets to 100% or 90 something, then you can like, 1152 00:54:49,320 --> 00:54:52,640 I'll give you an example, like during the pandemic, there's 1153 00:54:52,640 --> 00:54:54,480 just the assumption that everybody had a mobile phone, 1154 00:54:55,320 --> 00:54:58,360 right? And they could QR code scan this 1155 00:54:58,360 --> 00:55:00,400 and the Indonesia, that was a really big thing, right? 1156 00:55:01,480 --> 00:55:02,400 That's how you'd show your health. 1157 00:55:02,480 --> 00:55:04,600 That made QR codes work in the West as well. 1158 00:55:04,840 --> 00:55:07,280 Yeah, that's right. But basically, obviously 10 1159 00:55:07,280 --> 00:55:09,520 years ago, you know, 10 years beforehand, they wouldn't be 1160 00:55:09,520 --> 00:55:10,600 able to do that. They would have to have some 1161 00:55:10,600 --> 00:55:12,160 other paper system or something like that. 1162 00:55:12,160 --> 00:55:14,600 In 2010, you couldn't assume everybody on the planet had a 1163 00:55:14,600 --> 00:55:16,040 smartphone. It was, it was getting big, but 1164 00:55:16,040 --> 00:55:17,680 it wasn't yet there. It's certainly 15 years ago, 1165 00:55:18,320 --> 00:55:23,560 nobody would have it, right. So that was something where the 1166 00:55:23,560 --> 00:55:29,080 ubiquity of something maybe sometimes the next step is comes 1167 00:55:29,080 --> 00:55:32,200 from that ubiquity or or you, you could give two or three 1168 00:55:32,200 --> 00:55:34,320 things at the same time. Yeah, I mean, you could think 1169 00:55:34,320 --> 00:55:38,720 about TV and radio, all forms of mass media in the past. 1170 00:55:40,360 --> 00:55:43,760 And you know, the growth of pop music requires recorded music 1171 00:55:43,760 --> 00:55:47,360 and requires radio. And, you know, the great of mass 1172 00:55:47,360 --> 00:55:50,120 democracy kind of goes hand in hand with literacy and TV 1173 00:55:50,120 --> 00:55:54,040 newspapers, right, That you need newspapers before you can have 1174 00:55:54,960 --> 00:55:58,320 other stuff has to happen, right, For that, for that to 1175 00:55:58,320 --> 00:55:59,280 come. And then, of course, you have 1176 00:55:59,280 --> 00:56:01,680 backlash. It was sort of think there's 1177 00:56:01,680 --> 00:56:03,600 something interesting in looking at stuff like the Arts and 1178 00:56:03,600 --> 00:56:06,360 Crafts movement in the late 19th century because he's a bunch of 1179 00:56:06,360 --> 00:56:08,760 people who say we hate all this mass manufactured stuff That's, 1180 00:56:08,800 --> 00:56:10,720 yeah, handcrafted things. It's funny that that's not a 1181 00:56:10,720 --> 00:56:12,520 statement that would make any sense in 1800. 1182 00:56:12,600 --> 00:56:15,800 Well. Well, it's funny because there's 1183 00:56:15,800 --> 00:56:18,440 this what you're talking about, like people were farmers that 1184 00:56:18,440 --> 00:56:21,040 are artisans and are like, Oh my God, this automation is 1185 00:56:21,040 --> 00:56:22,480 disrupting us. We hate it so much. 1186 00:56:22,480 --> 00:56:25,240 We want to go back to the old ways and and now it's funny is 1187 00:56:25,480 --> 00:56:29,480 those manufacturing jobs that all these workers were so mad 1188 00:56:29,480 --> 00:56:31,840 about in the late age, hundreds and early 1900s, all the 1189 00:56:31,840 --> 00:56:33,200 strikes, all communism and so on. 1190 00:56:33,600 --> 00:56:36,240 Those are now the things are looked back on romantically by a 1191 00:56:36,240 --> 00:56:40,160 lot of mega types where they are like, oh, that was such a great 1192 00:56:40,160 --> 00:56:41,240 job. I wish I had that. 1193 00:56:41,400 --> 00:56:43,400 I hate this information job kind of thing. 1194 00:56:43,400 --> 00:56:46,720 I hate this, you know, these, these these desk jobs and so on 1195 00:56:46,720 --> 00:56:48,800 and so forth. So it's interesting because 1196 00:56:49,360 --> 00:56:52,240 there's a romanticization sometimes of the past thing, 1197 00:56:52,880 --> 00:56:55,800 even as millions of people are exiting that for the next thing. 1198 00:56:55,800 --> 00:56:57,960 Now, this is a little more complicated, obviously, by the 1199 00:56:57,960 --> 00:57:00,480 fact that China has a lot of those, quote, manufacturing 1200 00:57:00,480 --> 00:57:03,200 jobs, but yet a lot of them are being automated in China as well 1201 00:57:03,200 --> 00:57:06,680 with the robots. So it's funny, the thing that 1202 00:57:06,680 --> 00:57:09,040 people were so mad about that they were getting seemingly 1203 00:57:09,040 --> 00:57:11,040 pushed into, which was manufacturing out of farming 1204 00:57:11,040 --> 00:57:13,440 into manufacturing, are the things that at least some 1205 00:57:13,440 --> 00:57:15,080 fraction of this generation wants to go back to. 1206 00:57:15,080 --> 00:57:18,000 Or they think they do. You know, I think it's. 1207 00:57:18,000 --> 00:57:19,800 Interesting. Some of those things, I mean 1208 00:57:19,800 --> 00:57:22,480 the, the Luddites are those one of these sort of misunderstood 1209 00:57:22,480 --> 00:57:26,280 movements because a lot of what the Luddites are about is 1210 00:57:26,520 --> 00:57:32,400 self-employed high status artisans losing that status and 1211 00:57:32,400 --> 00:57:35,840 being pushed into low status commodity jobs. 1212 00:57:35,840 --> 00:57:37,880 So this is going to be the big thing with I think. 1213 00:57:38,400 --> 00:57:39,520 Have you seen the elephant graph? 1214 00:57:40,480 --> 00:57:44,000 So the elephant graph and some people dispute the graph, but I 1215 00:57:44,000 --> 00:57:46,560 think it it's probably gesturing at something that's right. 1216 00:57:47,040 --> 00:57:52,720 It shows percentiles or deciles of the world in terms of income. 1217 00:57:53,080 --> 00:57:56,640 And it shows over the last 20 something years, I think from 91 1218 00:57:56,640 --> 00:58:00,200 to 2008 or something like that, where the growth went like whose 1219 00:58:00,200 --> 00:58:03,480 incomes rose and basically most of the world. 1220 00:58:04,120 --> 00:58:07,800 So the, the lower 10% in Africa didn't gain that much, but like 1221 00:58:07,840 --> 00:58:13,240 maybe from the 10 to 20% through the 70 to 8% had huge growth. 1222 00:58:13,720 --> 00:58:16,880 Then it drops off and the 8 to 90% to almost zero and then it 1223 00:58:16,880 --> 00:58:18,760 picks up again at the very top, right. 1224 00:58:19,120 --> 00:58:23,880 And so that means is the like global, you know, elite in every 1225 00:58:23,880 --> 00:58:28,600 country did great and so did China, India, Vietnam, Eastern 1226 00:58:28,600 --> 00:58:30,920 Europe, all these countries are no longer socialist, communist, 1227 00:58:30,920 --> 00:58:33,640 etcetera, right? But the Western middle class 1228 00:58:33,640 --> 00:58:37,360 didn't. And that is a big part of, I 1229 00:58:37,360 --> 00:58:41,040 think the silence ability now when we're looking at it is, you 1230 00:58:41,040 --> 00:58:43,320 know, in America, they have, you know, obviously red versus blue. 1231 00:58:43,720 --> 00:58:48,040 But one way of thinking about it is starting in, you know, 1232 00:58:48,080 --> 00:58:51,640 certainly in 2008, there's a ramp where China flips US 1233 00:58:51,640 --> 00:58:54,080 manufacturing. And so China puts all this 1234 00:58:54,080 --> 00:58:56,720 pressure on red America and that leads to Trump and trade war. 1235 00:58:57,320 --> 00:58:59,840 And you've seen that that graph of print media disruption, 1236 00:59:00,040 --> 00:59:02,120 right? That's the Internet suddenly 1237 00:59:02,120 --> 00:59:07,000 rising after 2008 to flip blue America and it takes all the ad 1238 00:59:07,000 --> 00:59:08,680 revenue away and it's all not shattering. 1239 00:59:08,680 --> 00:59:10,640 It's also Craigslist, it's classified ads, a bunch of other 1240 00:59:10,640 --> 00:59:12,560 things. So the Internet disrupts blue 1241 00:59:12,560 --> 00:59:15,920 America and that leads to wokeness in in the 20 tens, I 1242 00:59:15,920 --> 00:59:18,760 think, and and also techlash, right, which is the anti tech 1243 00:59:18,760 --> 00:59:20,760 movement. So we look at it as red and 1244 00:59:20,760 --> 00:59:23,040 blue. There's also China and the 1245 00:59:23,040 --> 00:59:25,080 Internet over here where the Internet is disrupting blue and 1246 00:59:25,080 --> 00:59:26,920 China is destructing red. So the thing I think that's 1247 00:59:26,920 --> 00:59:31,840 coming next is AI disrupts blue America and robots disrupt red 1248 00:59:31,840 --> 00:59:34,680 America. And so Chinese robots and 1249 00:59:34,680 --> 00:59:38,600 Internet AI and so that artisan movement kind of thing is going 1250 00:59:38,600 --> 00:59:40,920 to accelerate where people are going to be mad about that 1251 00:59:40,920 --> 00:59:43,720 happening. I think on balance, there's 1252 00:59:43,800 --> 00:59:45,640 going to be a lot more productivity in the rest of the 1253 00:59:45,640 --> 00:59:47,040 world. But it's possible, for example, 1254 00:59:47,040 --> 00:59:51,400 that a job that's at let's say 200K or something like that in 1255 00:59:51,400 --> 00:59:57,080 the US and there's somebody in, in India or Mongolia or Vietnam 1256 00:59:57,080 --> 01:00:01,480 or something who's at $2000 a year that that equilibrates at 1257 01:00:01,480 --> 01:00:04,480 like 20 K, right? For like somebody supervising 1258 01:00:04,480 --> 01:00:06,080 medical results or something like that, right? 1259 01:00:06,240 --> 01:00:09,160 Where the licensure is no longer as important the the Western 1260 01:00:09,160 --> 01:00:11,960 licensure, the Western state doesn't can't really protect it 1261 01:00:11,960 --> 01:00:13,200 as much because it's all on the Internet. 1262 01:00:13,800 --> 01:00:16,680 And that's a boon for everybody who's a customer of that. 1263 01:00:16,680 --> 01:00:18,320 Like healthcare costs go down around the world. 1264 01:00:18,320 --> 01:00:20,200 You've got a great doctor on tap at any time. 1265 01:00:20,600 --> 01:00:25,920 Most people benefit from it. But those people who lost, you 1266 01:00:25,920 --> 01:00:28,400 know, relative status, relative money and that get super angry. 1267 01:00:28,800 --> 01:00:33,080 And I think the burning of the way MO's and like the extreme 1268 01:00:33,080 --> 01:00:39,000 anti AI sentiment that I see among some people is is kind of 1269 01:00:39,000 --> 01:00:41,680 a precursor to that. Let me know your thoughts. 1270 01:00:41,680 --> 01:00:47,280 So. I think this is a general 1271 01:00:47,280 --> 01:00:51,920 observation that like when Europeans live in Europe, 1272 01:00:52,400 --> 01:00:55,400 probably something similar in Asia, when Europeans live in 1273 01:00:55,400 --> 01:00:57,840 Europe, we all feel different. So like Germans are very 1274 01:00:57,840 --> 01:00:59,680 different to Italians and different to British people, 1275 01:00:59,680 --> 01:01:01,080 different French people and so on. 1276 01:01:01,680 --> 01:01:03,800 And when Europeans live in America, they all feel European 1277 01:01:03,800 --> 01:01:08,960 and America is buried is, is in a different place to the 1278 01:01:08,960 --> 01:01:11,560 aggregate of, of, of, of, of Europe. 1279 01:01:12,080 --> 01:01:17,720 And the US has its own sort of political culture and political 1280 01:01:17,720 --> 01:01:22,200 questions that are different to the questions in France or 1281 01:01:22,200 --> 01:01:25,640 Germany or Britain. I do think some of what's 1282 01:01:25,640 --> 01:01:29,200 happened and I don't wouldn't call myself political analyst, 1283 01:01:29,200 --> 01:01:32,920 but I think some of what's happened is that certainly in 1284 01:01:32,920 --> 01:01:37,160 the US to the some extent the UK, there were coalitions, 1285 01:01:37,760 --> 01:01:40,440 particularly there were on the progressive side or the left 1286 01:01:40,440 --> 01:01:43,840 side. There was a coalition of urban 1287 01:01:43,840 --> 01:01:46,720 upper middle class, highly educated people with a certain 1288 01:01:46,720 --> 01:01:50,840 set of social attitudes and working collar, blue collar 1289 01:01:50,840 --> 01:01:53,600 blood working class blue collar people. 1290 01:01:53,680 --> 01:01:55,400 Has been totally. Busted in a different part of 1291 01:01:55,400 --> 01:01:59,760 the country, often with rather different social and political 1292 01:01:59,760 --> 01:02:04,880 attitudes and the same thing, I think in the US and the 1293 01:02:04,880 --> 01:02:11,640 Republican Party on the right, you know, the coalition of sort 1294 01:02:11,720 --> 01:02:13,160 of. Wall Street Journal reading 1295 01:02:13,160 --> 01:02:16,360 Capitalist. Yeah, like Mitt Romney and. 1296 01:02:16,440 --> 01:02:19,040 Military. Guys that that is split apart 1297 01:02:19,040 --> 01:02:25,600 completely and all of those, you know, centre right, economically 1298 01:02:25,600 --> 01:02:29,120 conservative, socially liberal people who are Republicans kind 1299 01:02:29,120 --> 01:02:34,640 of don't have a political party anymore and equally people who 1300 01:02:34,640 --> 01:02:38,360 are sort of Bloomberg. Central, you know. 1301 01:02:38,400 --> 01:02:41,760 Sort of Bloomberg centralists, centrists kind of don't have a 1302 01:02:41,760 --> 01:02:44,000 political party anymore and there's a lot of those 1303 01:02:44,000 --> 01:02:45,840 coalitions have kind of broken apart now. 1304 01:02:46,120 --> 01:02:49,480 What you have in a bunch of European countries is partly 1305 01:02:49,480 --> 01:02:52,480 because of proportional representation is it's viable to 1306 01:02:52,480 --> 01:02:56,240 have half a dozen different parties and the US and the UK 1307 01:02:56,240 --> 01:02:58,600 because of the first past the post system you don't have 1308 01:02:58,640 --> 01:03:01,360 multiple. It's never been viable to have 1309 01:03:01,360 --> 01:03:03,680 five different political parties at different points in the 1310 01:03:03,680 --> 01:03:07,000 spectrum in the same way the US has got the UK has got this kind 1311 01:03:07,000 --> 01:03:09,800 of weird hangover sent Liberal Party, which is no one has ever 1312 01:03:09,800 --> 01:03:12,160 been quite clear what it was for sort of in the middle quote 1313 01:03:12,200 --> 01:03:15,520 called Liberal Party. It's there's an interesting sort 1314 01:03:15,520 --> 01:03:18,600 of sideline there, which is the Liberal Party in the UK in the 1315 01:03:18,600 --> 01:03:21,960 19th century was one of the two parties of government and it was 1316 01:03:21,960 --> 01:03:24,720 socially liberal and economically conservative. 1317 01:03:25,080 --> 01:03:27,560 But in the 19th century, what we now call economically 1318 01:03:27,560 --> 01:03:30,760 conservative in the 19th century meant pro free trade and against 1319 01:03:30,760 --> 01:03:34,120 regulation, right, Whereas now economically conservative is the 1320 01:03:34,120 --> 01:03:35,080 other way round. Yes. 1321 01:03:35,080 --> 01:03:38,320 So all of those labels kind of shift and move and change in the 1322 01:03:38,320 --> 01:03:40,840 different things at the time. It's interesting, Meg, I'd say 1323 01:03:40,840 --> 01:03:43,200 Meg is arguably against, certainly against free trade, 1324 01:03:43,200 --> 01:03:44,720 but they're also against sense regulation. 1325 01:03:44,720 --> 01:03:48,960 So it's like half right. But but it's finished what 1326 01:03:48,960 --> 01:03:49,960 you're saying. But I agree with you. 1327 01:03:49,960 --> 01:03:52,160 Of course the labels do change. Yeah, the labels change. 1328 01:03:52,160 --> 01:03:54,200 The coalition's broke apart, break apart. 1329 01:03:55,400 --> 01:03:59,760 I think there's always this tension in looking at 1330 01:04:00,120 --> 01:04:04,040 progressive ideas and saying, because if you look at the last 1331 01:04:04,040 --> 01:04:08,800 100 years, the social progress, progressive ideas have always 1332 01:04:08,800 --> 01:04:12,840 won. Like nobody today says like 1333 01:04:12,840 --> 01:04:16,600 being gay should be illegal. Like so you know, a little bit 1334 01:04:16,600 --> 01:04:19,720 like what we were saying about AIA while ago today you you, you 1335 01:04:19,720 --> 01:04:24,920 could deterministically say that what is woke today in 30 years 1336 01:04:24,920 --> 01:04:28,280 time will be what every far right conservative agrees with, 1337 01:04:28,800 --> 01:04:30,040 like. Yeah, people have said that 1338 01:04:30,080 --> 01:04:31,480 kind. Of thing theoretically in 50 1339 01:04:31,600 --> 01:04:34,040 years, you know, maybe, maybe not, but there's also you also 1340 01:04:34,080 --> 01:04:35,640 have these kind of overreaches around this. 1341 01:04:35,880 --> 01:04:39,280 It does strike me that one of the differences between the Usus 1342 01:04:39,280 --> 01:04:42,400 and UK politics is that what happened in the last in my 1343 01:04:42,400 --> 01:04:46,000 lifetime is that the right for want of a better term won the 1344 01:04:46,000 --> 01:04:50,680 economic argument that state ownership and government control 1345 01:04:50,680 --> 01:04:54,960 of the economy is bad right and the left one the social 1346 01:04:54,960 --> 01:04:58,640 arguments that like gay marriage is OK. 1347 01:04:59,000 --> 01:05:00,680 Well, it's. And and so on. 1348 01:05:00,960 --> 01:05:05,640 And in what happened in the UK was the the right embraced that 1349 01:05:05,640 --> 01:05:08,040 and the Conservative Party is the party that brought in gay 1350 01:05:08,040 --> 01:05:12,480 marriage in the UK, whereas in the left in the US it's kind of 1351 01:05:12,480 --> 01:05:15,080 the other way around, the Republicans kind of and Tony. 1352 01:05:15,080 --> 01:05:16,320 Blair sort of brought in, kind of. 1353 01:05:16,400 --> 01:05:20,080 Yeah, and he brought in level economics, whereas what happened 1354 01:05:20,200 --> 01:05:23,000 in the US is that the Republican Party in the US never kind of 1355 01:05:23,000 --> 01:05:25,360 accepted that it had lost the social arguments. 1356 01:05:26,520 --> 01:05:29,400 Well, it's interesting. I think from 1950, like the 1357 01:05:29,400 --> 01:05:34,520 moment of 1950, you do have something where because 1358 01:05:34,520 --> 01:05:38,160 communism fell basically because Nazis was defeated, the world 1359 01:05:38,200 --> 01:05:41,080 moved socially to the left. And then when communist as 1360 01:05:41,080 --> 01:05:43,760 communism was defeated, it moved economically to the right. 1361 01:05:44,200 --> 01:05:48,640 And so thus, for example, like the immigrant billionaire or gay 1362 01:05:48,640 --> 01:05:51,560 billionaire is like in a sense. Can be right wing. 1363 01:05:51,760 --> 01:05:55,920 Well, they're far, well, they're far to the left of 1950 socially 1364 01:05:56,200 --> 01:05:59,880 and they're far to the right in an economic right in a sense of 1365 01:05:59,880 --> 01:06:04,440 1950 economically because 1950, yes, the Soviet Union had 100% 1366 01:06:04,680 --> 01:06:08,440 taxes because it's communism, but the US had 90% marginal tax 1367 01:06:08,440 --> 01:06:10,360 rates. And you really couldn't get rich 1368 01:06:10,360 --> 01:06:12,880 mid century in the US. You could be a corporation man, 1369 01:06:12,880 --> 01:06:16,160 you could work for NASA or GM, General Motors, General Mills, 1370 01:06:16,160 --> 01:06:19,160 General Electric, but you're sort of funneled, channeled into 1371 01:06:19,160 --> 01:06:21,360 like these gigantic things. You had more freedom in the US 1372 01:06:21,360 --> 01:06:23,520 than other places, but you're still very stultified. 1373 01:06:23,520 --> 01:06:25,640 It was it was too capital intensive to be an entrepreneur 1374 01:06:25,640 --> 01:06:29,000 and so on. And then gradually with I think 1375 01:06:29,000 --> 01:06:31,640 the transition was the mirror moment where that's begun, A 1376 01:06:31,640 --> 01:06:34,880 decentralization arc and history is running in reverse since that 1377 01:06:34,880 --> 01:06:37,240 moment. But and so I think a lot of 1378 01:06:37,240 --> 01:06:39,760 things are happening this century that are like a reversal 1379 01:06:39,760 --> 01:06:42,560 of things in the past. I think it would be interesting, 1380 01:06:43,640 --> 01:06:46,080 and I have no opinion about this at all, but it would be 1381 01:06:46,080 --> 01:06:48,880 interesting to ask what is behind the growth in 1382 01:06:48,880 --> 01:06:51,720 billionaires? Oh, is this an unlocking of a 1383 01:06:51,720 --> 01:06:53,640 new kind? Is this a wave of company 1384 01:06:53,640 --> 01:06:55,920 creation? So I, I, you see what I mean? 1385 01:06:55,920 --> 01:06:59,760 Yeah, I do have this on this. Which is is is your point is why 1386 01:06:59,760 --> 01:07:01,960 are there new billionaires? Is that because there were a 1387 01:07:01,960 --> 01:07:04,600 bunch of new companies and they're first generation owners 1388 01:07:05,080 --> 01:07:09,720 and where did those come from? And certainly some of them came 1389 01:07:09,720 --> 01:07:13,240 from Google and you know, winner global winner takes all effects 1390 01:07:13,480 --> 01:07:15,200 and some of them didn't. I don't know. 1391 01:07:15,200 --> 01:07:19,080 I mean, I, I, I, I, I'm, I'm not sure how much value I can, I can 1392 01:07:19,080 --> 01:07:20,280 I kind of add to that conversation. 1393 01:07:20,280 --> 01:07:22,560 There's a bunch of kind of economic statistical question as 1394 01:07:22,560 --> 01:07:23,520 well. I just not spent the time 1395 01:07:23,520 --> 01:07:26,360 looking. I can I can give some thoughts 1396 01:07:26,360 --> 01:07:30,920 on that, which is that has AU curve right? 1397 01:07:31,160 --> 01:07:33,520 Where for example, like who is the richest guy in the Soviet 1398 01:07:33,520 --> 01:07:36,520 Union? Like didn't exist Communism, you 1399 01:07:36,520 --> 01:07:38,720 know, basically Stalin, you know, didn't need money because 1400 01:07:38,720 --> 01:07:40,000 you could just requisition anything I think. 1401 01:07:40,040 --> 01:07:41,840 The well, the Soviet Union is kind of a bad example of 1402 01:07:41,840 --> 01:07:43,640 creating billionaires. No, no, just cut the country up 1403 01:07:43,640 --> 01:07:44,760 and gave it to 20 people. Well. 1404 01:07:44,760 --> 01:07:46,160 No, no. Well, that's right, but that's 1405 01:07:46,160 --> 01:07:48,160 starting in the 90s, right? Then it wasn't, that was Russia 1406 01:07:48,160 --> 01:07:50,800 then, right? But basically the number of like 1407 01:07:50,880 --> 01:07:54,160 independently wealthy men who could do things in the US, for 1408 01:07:54,160 --> 01:07:56,960 example, a lot of the, the, the great fortunes, the robber 1409 01:07:56,960 --> 01:07:59,720 barons and and captains of industry were forced into 1410 01:07:59,720 --> 01:08:01,040 foundations. That's why you have the Ford 1411 01:08:01,040 --> 01:08:04,000 Foundation, Carnegie Foundation, Mellon Foundation, Rockefeller 1412 01:08:04,000 --> 01:08:07,240 Foundation, because in 1930s Roosevelt didn't want any other 1413 01:08:07,240 --> 01:08:09,680 powers besides him. So he, you know, went after 1414 01:08:09,680 --> 01:08:12,520 Andrew Mellon, all these people, Ida Tarbell went after 1415 01:08:12,520 --> 01:08:16,080 Rockefeller and those fortunes were corralled and basically 1416 01:08:16,080 --> 01:08:19,279 controlled by the state in these foundations in the Soviet Union, 1417 01:08:19,279 --> 01:08:21,000 in Communist China, they were just seized, right? 1418 01:08:21,319 --> 01:08:24,760 So basically, let me give the, the, the normal way of talking 1419 01:08:24,760 --> 01:08:26,399 about this is inequalities rising. 1420 01:08:26,399 --> 01:08:28,920 And that's terrible, right? Another way of thinking about it 1421 01:08:28,920 --> 01:08:34,120 is what is the state, right? The state is in a sense it's 1422 01:08:34,120 --> 01:08:37,200 like all the people who are it's citizens and they kind of crowd 1423 01:08:37,200 --> 01:08:41,800 fund the state, right? And the question is, are they do 1424 01:08:41,800 --> 01:08:43,120 they have a choice in doing that? 1425 01:08:43,160 --> 01:08:46,560 Can they opt out of that? Like what set are they part of? 1426 01:08:46,720 --> 01:08:48,840 You know, for example, if they're on the Franco German 1427 01:08:48,840 --> 01:08:50,760 border, can they call themselves part of the German side or the 1428 01:08:50,760 --> 01:08:52,760 French side? You know, how about the Polish, 1429 01:08:52,760 --> 01:08:54,520 the Polish German border with that kind of thing? 1430 01:08:55,040 --> 01:09:00,000 And how much does the state take and how powerful is it? 1431 01:09:00,000 --> 01:09:03,399 And mid century, because of mass media and mass production, the 1432 01:09:03,439 --> 01:09:05,840 the states were more centralized they've ever been in history. 1433 01:09:06,319 --> 01:09:07,399 I can show a bunch of graphs on that. 1434 01:09:07,399 --> 01:09:08,880 That's not just that's a quantitative thing. 1435 01:09:09,200 --> 01:09:11,560 So you had these Gaiga states, you had fewer sovereign units on 1436 01:09:11,560 --> 01:09:14,319 the planet than at any time before or since like only like 1437 01:09:14,319 --> 01:09:17,000 50 UN countries. So there's like 196. 1438 01:09:17,160 --> 01:09:18,600 So things have decentralized since then. 1439 01:09:18,840 --> 01:09:21,160 If you go backwards in time, you go to like Germany under 1440 01:09:21,160 --> 01:09:22,760 Bismarck, you've got all these principalities. 1441 01:09:23,000 --> 01:09:24,560 Go to France before the revolution, you have all these 1442 01:09:24,560 --> 01:09:27,399 things, Italy before Garibaldi, you have all of these little, 1443 01:09:27,560 --> 01:09:28,880 you know, city States and so on, right. 1444 01:09:29,040 --> 01:09:31,479 So you go backwards time and forth time is decentralized and 1445 01:09:31,479 --> 01:09:33,000 the same, the same thing happens where you've got lots of 1446 01:09:33,000 --> 01:09:36,319 fortunes, you've got lots of, you know, individual potentates 1447 01:09:36,319 --> 01:09:38,840 and what have you, right. So in a sense, like the world is 1448 01:09:38,840 --> 01:09:41,120 sort of returning to what it used to be, with a big exception 1449 01:09:41,120 --> 01:09:45,479 being China. I think China is the like the, 1450 01:09:45,640 --> 01:09:50,160 the 20th century centralized state that will keep scaling 1451 01:09:50,160 --> 01:09:52,920 into the, into the century. So anyway, the reason I just say 1452 01:09:52,920 --> 01:09:55,600 that is I, I think there is something real going on, which 1453 01:09:55,600 --> 01:10:00,720 is that the state is just take capturing less of the wealth of 1454 01:10:00,720 --> 01:10:02,520 its individuals. People are sort of breaking away 1455 01:10:02,520 --> 01:10:04,640 on the borders of it and then being able to do their own 1456 01:10:04,640 --> 01:10:07,000 thing. And so it's like Elon, not NASA, 1457 01:10:07,000 --> 01:10:10,640 It's like Travis not taxing medallions and so on. 1458 01:10:10,880 --> 01:10:12,960 And there's a good to that where there's a lot more room for 1459 01:10:12,960 --> 01:10:15,400 individual initiative, but there's a bad to that as well, 1460 01:10:15,400 --> 01:10:18,640 which is then people don't feel as bought in on the collective 1461 01:10:18,640 --> 01:10:20,560 project. And they're not like included in 1462 01:10:20,560 --> 01:10:21,520 it. It's some guy's thing. 1463 01:10:21,520 --> 01:10:23,320 It's not their thing. It's not like America lands on 1464 01:10:23,320 --> 01:10:26,000 the moon or it's Elon, OK, fine. You know, And they don't feel as 1465 01:10:26,000 --> 01:10:28,120 bought in, right? So it's complicated kind of 1466 01:10:28,120 --> 01:10:29,000 thing. I think we're going to have to 1467 01:10:29,000 --> 01:10:30,560 renegotiate all that stuff in the future. 1468 01:10:31,000 --> 01:10:33,640 I think there's a there's a lot of this outside, again outside 1469 01:10:33,640 --> 01:10:39,760 sort of US politics, which is that partly because the US, the 1470 01:10:39,760 --> 01:10:42,640 part of the nature of the US economy, partly because the US 1471 01:10:42,640 --> 01:10:45,960 is a big domestic market, partly because the successful Internet 1472 01:10:45,960 --> 01:10:49,240 companies are in the US and have global winner takes all effects. 1473 01:10:49,880 --> 01:10:53,000 People outside the US for the front of the first time think, 1474 01:10:53,000 --> 01:10:54,720 well, we've got all these giant company U.S. 1475 01:10:54,720 --> 01:10:59,320 Companies that are running stuff in our country and that was kind 1476 01:10:59,320 --> 01:11:02,000 of true for like General Motors or Coca-Cola. 1477 01:11:02,240 --> 01:11:03,600 That's much more direct. But not. 1478 01:11:03,600 --> 01:11:04,240 Really. Yeah. 1479 01:11:04,280 --> 01:11:06,160 Yeah, right. You know, General Motors sold 1480 01:11:06,160 --> 01:11:08,160 cars, but you had a lot of your own car companies as well, and 1481 01:11:08,160 --> 01:11:10,320 ITM didn't decide how you built roads or anything. 1482 01:11:10,680 --> 01:11:13,280 And there's certainly a sort of a you know, you go to European 1483 01:11:13,280 --> 01:11:15,800 events now and there's people saying, well, do we need our own 1484 01:11:15,840 --> 01:11:17,720 Google? And one level, those are like 1485 01:11:17,720 --> 01:11:20,720 dumb questions, but they're dumb questions about like a real 1486 01:11:21,320 --> 01:11:25,320 issue, which is you have this other layer of stuff that you're 1487 01:11:25,320 --> 01:11:29,400 using which didn't used to be globalized and used to be 1488 01:11:29,400 --> 01:11:31,240 subject to local democratic control. 1489 01:11:31,240 --> 01:11:33,760 And now, well, it's not quite clear how that works. 1490 01:11:33,760 --> 01:11:35,480 Yeah, yeah. So, so, so actually it's very 1491 01:11:35,480 --> 01:11:37,320 important. I mean, where you're hitting on 1492 01:11:37,320 --> 01:11:39,400 there is, I think one of the core questions, and I'll 1493 01:11:39,400 --> 01:11:42,480 actually ask it in reverse, which is, are those American 1494 01:11:42,480 --> 01:11:46,080 companies basically, is the Internet American right now? 1495 01:11:46,080 --> 01:11:48,360 On one level you'd say that's a weird question. 1496 01:11:48,360 --> 01:11:50,520 Of course, there's two parts to that is are they American? 1497 01:11:50,520 --> 01:11:53,520 But also is they're not. They're not in our country. 1498 01:11:53,520 --> 01:11:56,840 If you're Swedish or Italian, it's not a Swedish company. 1499 01:11:56,840 --> 01:12:00,880 That's right, that's right. So so like you know, my view is 1500 01:12:01,360 --> 01:12:03,560 the Internet is to America, but America was to Britain. 1501 01:12:04,360 --> 01:12:10,640 It is like the version 3 point O and because the early Americans 1502 01:12:10,640 --> 01:12:13,800 actually consider themselves as you know British right, all the 1503 01:12:13,920 --> 01:12:16,040 folkways and stuff came from Britain and the. 1504 01:12:16,040 --> 01:12:18,400 American War of Independence is essentially a civil war. 1505 01:12:18,560 --> 01:12:19,560 Yeah, exactly. That's right. 1506 01:12:19,560 --> 01:12:23,560 So they had a people and they had a land, but they didn't have 1507 01:12:23,560 --> 01:12:25,160 a government, right? Because the government was in 1508 01:12:25,160 --> 01:12:27,600 London, right. And when they had all three, 1509 01:12:27,600 --> 01:12:30,040 they became Americans. They had a sense of self and I 1510 01:12:30,040 --> 01:12:34,120 think with the Internet, we have actually a lot of tribes that 1511 01:12:34,120 --> 01:12:36,720 actually have a people and a government, but not land. 1512 01:12:37,000 --> 01:12:38,720 And the reason they have a government is they have a 1513 01:12:38,720 --> 01:12:42,440 blockchain, they have a social network they have with with 1514 01:12:42,440 --> 01:12:45,640 moderators or forums. And now increasingly they have 1515 01:12:45,880 --> 01:12:49,360 like an AI agent or like a central Oracle or something like 1516 01:12:49,360 --> 01:12:51,720 that where it almost takes the role of like a God, which they 1517 01:12:51,720 --> 01:12:54,600 all ask questions to, right. So you think of every large 1518 01:12:54,600 --> 01:12:57,640 enough online community that has its own social network, whether 1519 01:12:57,640 --> 01:13:00,200 it's a discord or a forum or something like that, its own 1520 01:13:00,200 --> 01:13:03,400 cryptocurrency, which has its, you know, smart contracts and, 1521 01:13:03,400 --> 01:13:06,800 and currency and its own AI, which is sort of like its Oracle 1522 01:13:06,800 --> 01:13:09,040 or search of all the community's knowledge, right? 1523 01:13:09,400 --> 01:13:11,920 And that's like a digital community that actually has a 1524 01:13:11,920 --> 01:13:14,600 fair amount of strength. And then because, you know, 1525 01:13:14,600 --> 01:13:15,840 where are your communications happening? 1526 01:13:15,840 --> 01:13:17,360 They're happening online, whereas your, where your 1527 01:13:17,360 --> 01:13:19,760 transactions are online, more and more of your wealth is 1528 01:13:19,760 --> 01:13:22,200 stored online like crypto's at trillions of dollars now. 1529 01:13:22,200 --> 01:13:24,680 It wasn't, it wasn't that 15 years ago is at 0 basically. 1530 01:13:25,360 --> 01:13:28,720 And so the significance of these cloud communities I think is 1531 01:13:28,720 --> 01:13:31,280 under appreciated. And eventually they're going to 1532 01:13:31,280 --> 01:13:33,480 be able to have enough money to crowd fund territory. 1533 01:13:33,760 --> 01:13:37,680 And so the because the tension between your primary identity is 1534 01:13:37,680 --> 01:13:40,840 online, your social network is online, your currency is online, 1535 01:13:41,080 --> 01:13:45,040 your your information is online and then not being grouped 1536 01:13:45,040 --> 01:13:47,480 offline. That'll resolve, in my view, in 1537 01:13:47,480 --> 01:13:49,280 terms of the descent of the cloud to the land. 1538 01:13:49,840 --> 01:13:52,400 So it's interesting. I mean, I probably take a sort 1539 01:13:52,400 --> 01:13:56,040 of more more prosaic view of this, but listening to you talk, 1540 01:13:56,040 --> 01:13:59,080 I am reminded of like distant memories of being at university 1541 01:13:59,080 --> 01:14:02,760 and looking at social history. And you know, there are a lot of 1542 01:14:02,760 --> 01:14:06,080 social history is about the kind of the, the joining into groups, 1543 01:14:06,120 --> 01:14:08,880 yes. And so the, the, the joining, I 1544 01:14:08,960 --> 01:14:11,320 think about why you joining and how, what is this sort of form 1545 01:14:11,320 --> 01:14:12,880 of sort of self, what direction? Yes. 1546 01:14:12,880 --> 01:14:15,440 Why do people want to fund monasteries? 1547 01:14:15,680 --> 01:14:19,000 Why do people form lay brotherhoods around the church? 1548 01:14:19,200 --> 01:14:22,040 Why do people like there's a whole 19th century British thing 1549 01:14:22,040 --> 01:14:24,320 of like all sorts of social joining. 1550 01:14:24,600 --> 01:14:28,240 Why do people want to join militias? 1551 01:14:28,680 --> 01:14:32,320 And you know, why do they want to form all these kind of former 1552 01:14:32,320 --> 01:14:34,680 Guild? Why do they want to form all of 1553 01:14:34,680 --> 01:14:37,560 these kind of different social groups and social clubs and ways 1554 01:14:37,560 --> 01:14:40,800 of getting together? And what are they trying to 1555 01:14:40,800 --> 01:14:42,280 achieve? And some of it is about, you 1556 01:14:42,280 --> 01:14:45,640 know, self-defense, you know, or pretty not, not in a kind of 1557 01:14:45,640 --> 01:14:48,600 military sense, but about, you know, forming your group to 1558 01:14:48,600 --> 01:14:51,280 protect your group's interests. Some of it is about establishing 1559 01:14:51,280 --> 01:14:53,000 status. Some of it is about, you know, 1560 01:14:53,280 --> 01:14:55,680 self-expression and self actualization, you know, kind of 1561 01:14:55,680 --> 01:15:00,320 classic Maslow hierarchy stuff. But it's not new to have lots of 1562 01:15:00,320 --> 01:15:03,160 communities. What is new is that they're not 1563 01:15:03,160 --> 01:15:07,040 necessarily kind of physically like Co located and they're not 1564 01:15:07,040 --> 01:15:09,600 necessarily centred around, I mean things like women's 1565 01:15:09,600 --> 01:15:11,520 suffrage, you know, they're not necessarily centred around a 1566 01:15:11,520 --> 01:15:14,400 movement or some. Specific, I think the objective, 1567 01:15:14,480 --> 01:15:16,400 I think they will be. I think they will be but. 1568 01:15:16,400 --> 01:15:18,160 Well, there may be, but we've had those in the past, you know 1569 01:15:18,160 --> 01:15:20,360 the Cornwall League or women's suffrage, all of those, yes, you 1570 01:15:20,400 --> 01:15:24,040 know veganism, slavery, anti slavery movements and so on. 1571 01:15:24,320 --> 01:15:28,640 So those senses of, you know, social organization and joining 1572 01:15:28,640 --> 01:15:32,640 and grouping in clubs in different forms, in different 1573 01:15:32,640 --> 01:15:34,560 aspects of society for different reasons. 1574 01:15:34,560 --> 01:15:37,800 It's kind of a, a recurrent pattern of human society. 1575 01:15:38,080 --> 01:15:40,560 And now it gets expressed, which is the sort of thing we always 1576 01:15:40,560 --> 01:15:43,840 talk about is, you know, the in the Internet is human behaviour 1577 01:15:44,160 --> 01:15:46,480 and it expresses and channels it in new ways. 1578 01:15:46,480 --> 01:15:49,000 And that's everything from, you know, people being horrible on 1579 01:15:49,000 --> 01:15:51,880 Twitter or doing terrible things on the Internet through to 1580 01:15:51,880 --> 01:15:55,920 people forming groups, clubs and societies on Discord or Reddit 1581 01:15:55,920 --> 01:15:57,120 or whatever it. Is that's right. 1582 01:15:57,280 --> 01:15:59,880 You know, by the way, I have, I have an explanation which you 1583 01:15:59,880 --> 01:16:03,600 might find funny as to I, I used to wonder why are people so 1584 01:16:03,600 --> 01:16:05,520 crazy on Twitter? Why are they so crazy on social 1585 01:16:05,520 --> 01:16:07,720 media? Because, you know, like starting 1586 01:16:07,720 --> 01:16:09,800 fights and stuff. Just as a sidebar, I was able to 1587 01:16:09,800 --> 01:16:12,200 explain it in the following way. You know, you know, the 1588 01:16:12,200 --> 01:16:14,040 Unabomber in the early 90s. Yeah. 1589 01:16:14,600 --> 01:16:18,240 So he blew up all these people. But, you know, the reason he did 1590 01:16:18,240 --> 01:16:20,560 that was to get an op-ed in the Washington Post, right? 1591 01:16:20,560 --> 01:16:22,560 So he killed those people for the distribution. 1592 01:16:23,040 --> 01:16:25,600 He killed all those people just to get his message out there. 1593 01:16:25,600 --> 01:16:27,720 So when you realize there's people like that, then it 1594 01:16:27,720 --> 01:16:29,840 actually makes it more understandable how many crazy 1595 01:16:29,840 --> 01:16:31,200 people there are on social media. 1596 01:16:31,480 --> 01:16:33,680 If, if someone is willing to kill all these people to get, 1597 01:16:33,800 --> 01:16:36,040 you know, just his message out there, a lot of other people are 1598 01:16:36,040 --> 01:16:37,840 willing to be very nasty on social media to get their 1599 01:16:37,840 --> 01:16:39,040 message. Out there, I always thought a 1600 01:16:39,040 --> 01:16:41,840 lot of it was about context collapse, which is sort of 1601 01:16:42,120 --> 01:16:44,320 actually, yeah, buzzy people doesn't mean anything. 1602 01:16:44,320 --> 01:16:47,240 I felt like some of it was you don't know who that person is 1603 01:16:47,240 --> 01:16:49,920 and you haven't understood what they've said and what else they 1604 01:16:49,920 --> 01:16:51,680 think, and you presume they think X. 1605 01:16:51,880 --> 01:16:52,960 It's like it's lossy compression. 1606 01:16:53,160 --> 01:16:55,000 You kind of compress 3 paragraphs. 1607 01:16:55,200 --> 01:16:56,920 There's no sub clause, there's no nuance. 1608 01:16:56,920 --> 01:17:01,600 You can't say of course I'm not a Nazi and you know some of it 1609 01:17:01,600 --> 01:17:03,840 is also which? And in fact, they can't take 1610 01:17:03,840 --> 01:17:05,160 that for granted, because you're not. 1611 01:17:05,320 --> 01:17:06,920 Maybe you are. Yeah, well, yeah. 1612 01:17:06,920 --> 01:17:10,800 Or basically they're like, you know, they have no context on 1613 01:17:10,800 --> 01:17:13,120 you, they can't read 5000 posts, they don't know where to trust 1614 01:17:13,120 --> 01:17:15,160 you and and so sort. Of yes, some of it is also just 1615 01:17:15,320 --> 01:17:17,280 morphs is Morgan Hassell I think. 1616 01:17:17,440 --> 01:17:18,280 Yeah, Morgan Hassell. Yeah. 1617 01:17:18,360 --> 01:17:21,120 You write a book that quoted me that gets endlessly requoted. 1618 01:17:21,440 --> 01:17:24,440 Well, I'd said something like like the Internet means that 1619 01:17:24,440 --> 01:17:26,640 basically you're confronted with people who disagree with you. 1620 01:17:26,880 --> 01:17:29,920 Yes, and you all the time. And you didn't realize there 1621 01:17:29,920 --> 01:17:32,560 were all these people who like, the particular thing I always 1622 01:17:32,560 --> 01:17:34,920 found was weird was there were people who were like very, very 1623 01:17:34,920 --> 01:17:36,000 far left. There were people who are 1624 01:17:36,000 --> 01:17:39,280 communists and they're like, you'll say something that isn't 1625 01:17:39,360 --> 01:17:41,240 communist and they'll be like amazed. 1626 01:17:42,080 --> 01:17:44,200 They were like, the thing was always, I always thought it was 1627 01:17:44,200 --> 01:17:46,280 weird. Is like, I can, I think it's 1628 01:17:46,280 --> 01:17:48,120 weird that you're a communist because at this stage you have 1629 01:17:48,120 --> 01:17:49,400 to be an idiot to be a. Communist. 1630 01:17:49,400 --> 01:17:51,320 Yeah, right. But it's even more weird that 1631 01:17:51,320 --> 01:17:53,320 you don't know that most people aren't. 1632 01:17:53,440 --> 01:17:55,080 Yeah, yeah, yeah, yeah. They're like shocked by it. 1633 01:17:55,080 --> 01:17:57,920 They're like amazed that anyone doesn't agree with their tiny 1634 01:17:57,920 --> 01:17:59,880 minority opinion. Yes, that's right. 1635 01:17:59,880 --> 01:18:03,000 And a lot of Twitter was that it was like, you're amazed that I 1636 01:18:03,000 --> 01:18:04,920 don't think everybody should own a car. 1637 01:18:05,120 --> 01:18:07,800 You're amazed that I don't agree with. 1638 01:18:07,800 --> 01:18:11,000 I'm not that I don't necessarily share your opinion on every. 1639 01:18:11,000 --> 01:18:12,080 Possible. That's right. 1640 01:18:12,120 --> 01:18:15,720 And I think the way that's gonna reconcile is you're gonna get a 1641 01:18:15,720 --> 01:18:18,760 lot more, I think smaller. I mean, in a sense, Twitter 1642 01:18:18,760 --> 01:18:21,480 doesn't exist anymore, right? X. 1643 01:18:22,320 --> 01:18:23,720 Is fragmented and. Exactly. 1644 01:18:23,720 --> 01:18:24,880 It's a tower available moment, right? 1645 01:18:24,960 --> 01:18:28,280 So Twitter no longer exists. There's X and there's Truth and 1646 01:18:28,280 --> 01:18:32,920 there's Gab and blue sky on the left and Mastodon and Threads 1647 01:18:33,280 --> 01:18:36,480 and and then the crypto ones like Farcaster Lens. 1648 01:18:36,480 --> 01:18:38,720 Noster a lot of stuff went to things that didn't look like 1649 01:18:38,720 --> 01:18:41,160 that, so stuff went to LinkedIn. TikTok. 1650 01:18:41,280 --> 01:18:43,800 Or it went to TikTok or it went to Instagram. 1651 01:18:43,960 --> 01:18:46,120 And people make fun of LinkedIn like there isn't a bunch of 1652 01:18:46,120 --> 01:18:50,360 bullshit on Twitter. But you know the I realized that 1653 01:18:50,360 --> 01:18:53,120 an awful lot of corporate people were sitting quietly using 1654 01:18:53,120 --> 01:18:55,200 LinkedIn when, yes, it didn't feel that they could use 1655 01:18:55,200 --> 01:18:56,640 Twitter. Yeah, because basically The 1656 01:18:56,640 --> 01:19:00,480 funny thing is it's interesting, something about LinkedIn means 1657 01:19:00,480 --> 01:19:04,360 people are artificially polite. And something about X or 1658 01:19:04,360 --> 01:19:07,920 Twitter, especially Twitter, I think even more than X in some 1659 01:19:07,920 --> 01:19:11,040 ways meant that they were artificially negative hostile, 1660 01:19:11,200 --> 01:19:13,280 right. And The funny thing about it is 1661 01:19:13,280 --> 01:19:18,080 artificially hostile reads to people as more sincere. 1662 01:19:18,800 --> 01:19:22,560 Like, that's to say, of the two, there's something about the 1663 01:19:22,560 --> 01:19:26,840 artificially polite, like, like, for example, a good review is 1664 01:19:26,840 --> 01:19:29,240 not a rave review. A good review is. 1665 01:19:29,800 --> 01:19:32,120 I love Ben's book. It was great, but he could 1666 01:19:32,120 --> 01:19:34,440 improve XY and Z. That's like, the best review 1667 01:19:34,440 --> 01:19:35,880 you'll get. Yeah, You know what I mean? 1668 01:19:35,880 --> 01:19:38,760 Usually. Whereas A hater will be like, 1669 01:19:39,160 --> 01:19:40,640 just complete crap on you. Right. 1670 01:19:40,840 --> 01:19:44,120 So the negative is generally much more negative than the 1671 01:19:44,120 --> 01:19:46,880 positive is positive. And so when you see a LinkedIn 1672 01:19:46,880 --> 01:19:49,600 style post, it's often like super positive and it feels fake 1673 01:19:49,600 --> 01:19:51,800 immediately. But people don't apply the same 1674 01:19:51,800 --> 01:19:53,520 filter. They think negative is real, but 1675 01:19:53,520 --> 01:19:55,160 they don't think negative could also be fake. 1676 01:19:55,560 --> 01:19:56,960 It's like. A mental, you know there. 1677 01:19:57,040 --> 01:19:59,480 Was a thing that went viral a while ago, Some surgeon who'd 1678 01:19:59,480 --> 01:20:02,760 got a, they'd got a review and it was like, he saved my life. 1679 01:20:02,760 --> 01:20:04,560 He's the most wonderful surgeon in history. 1680 01:20:04,560 --> 01:20:07,040 It's amazing. It's wonderful. 4 out of five 1681 01:20:07,040 --> 01:20:09,640 stars. Yeah, yeah, yeah, yeah, yeah, 1682 01:20:09,720 --> 01:20:11,920 exactly, exactly. Wow, what did I have to do to 1683 01:20:11,920 --> 01:20:13,200 get 5 stars? Yeah, exactly. 1684 01:20:13,200 --> 01:20:14,640 That's right. Like, you know, I forgot to give 1685 01:20:14,640 --> 01:20:16,360 the mint chocolate under the pillow or something. 1686 01:20:16,360 --> 01:20:17,720 Yeah, Yeah. OK. 1687 01:20:17,720 --> 01:20:20,840 So like, you know, let's do, let's change gears. 1688 01:20:20,920 --> 01:20:24,440 Let's talk about just survey of tech, just things, you know, you 1689 01:20:24,440 --> 01:20:26,040 can tell me you've been thinking about this, you have anything 1690 01:20:26,040 --> 01:20:31,120 about this. So we talked about like gadgets. 1691 01:20:31,120 --> 01:20:34,400 So we talked about, you know, the glasses we talked about we 1692 01:20:34,400 --> 01:20:35,560 haven't. Actually, did we talk about 1693 01:20:35,600 --> 01:20:37,360 glasses on the podcast or on the car in the? 1694 01:20:37,360 --> 01:20:38,600 Car. We talk about glasses a little 1695 01:20:38,600 --> 01:20:39,920 bit on the pod, but basically, well, tell me. 1696 01:20:39,920 --> 01:20:40,960 Tell me your thoughts on glasses. 1697 01:20:41,000 --> 01:20:43,560 Oh so. ARARVR Glasses. 1698 01:20:43,800 --> 01:20:44,920 Yeah, XR glasses. Yeah. 1699 01:20:45,280 --> 01:20:48,840 So I've made this point a while a bunch on online as far as I 1700 01:20:48,840 --> 01:20:51,640 can see, like you have the VR experience, you think it's 1701 01:20:51,640 --> 01:20:54,800 amazing. It's not clear to me that this, 1702 01:20:56,680 --> 01:20:59,760 my base case of VR is that it may end up like games consoles 1703 01:21:00,360 --> 01:21:02,320 in that you see a games console, it's amazing. 1704 01:21:02,320 --> 01:21:04,680 Most people don't buy it. There's a portion of people that 1705 01:21:04,680 --> 01:21:07,200 don't understand that games is actually quite a small industry 1706 01:21:07,200 --> 01:21:09,160 in terms of number of people. It's a lot of money, but a lot, 1707 01:21:09,160 --> 01:21:11,360 there's like 2 or 300 million people play games, console 1708 01:21:11,360 --> 01:21:14,200 games. And so it may be that VR, you 1709 01:21:14,200 --> 01:21:15,600 have the experience, it's amazing. 1710 01:21:15,600 --> 01:21:17,880 You put it down, you walk away. Most people don't buy it no 1711 01:21:17,880 --> 01:21:19,080 matter how good the hardware gets. 1712 01:21:19,880 --> 01:21:22,880 I think it's much easier to see something like what I'm wearing 1713 01:21:22,880 --> 01:21:28,440 now being a universal device at the level of a smartphone. 1714 01:21:28,800 --> 01:21:30,960 Clearly we don't have the optics for that yet. 1715 01:21:31,080 --> 01:21:32,960 We may it's. Improving every year though. 1716 01:21:32,960 --> 01:21:33,840 It is. Yeah, it is. 1717 01:21:33,920 --> 01:21:37,440 Yeah, the. Question is, it's is that next 1718 01:21:37,520 --> 01:21:39,960 five years time? Is that 2 years time? 1719 01:21:39,960 --> 01:21:43,200 Is that 10 years? It's not clear. 1720 01:21:43,240 --> 01:21:46,400 Yet, yeah, there's, there's there's a few people I know who 1721 01:21:46,400 --> 01:21:50,360 just like, they almost subscribe to the space in the sense of 1722 01:21:50,360 --> 01:21:52,520 they're constantly just getting the latest glasses, usually out 1723 01:21:52,520 --> 01:21:55,040 of China, and they're just trying them out, right? 1724 01:21:55,040 --> 01:21:56,800 Or getting prototypes. There's various prototypes 1725 01:21:56,800 --> 01:21:59,680 people are making. And this is something that I 1726 01:21:59,680 --> 01:22:04,560 feel there's some value in tracking because it's almost 1727 01:22:04,560 --> 01:22:06,240 being ignored by the world, right? 1728 01:22:06,240 --> 01:22:08,160 Now it is because it's like it's, it's, it's one. 1729 01:22:08,160 --> 01:22:09,840 Of the hit that Gartner Hype cycle things. 1730 01:22:09,840 --> 01:22:11,920 Yeah, curve that's bumping on the bottom and hasn't quite 1731 01:22:11,920 --> 01:22:13,880 happened. Yeah, or it's a through after 1732 01:22:13,880 --> 01:22:15,680 the hype of. Metaverse, And there's a subset 1733 01:22:15,680 --> 01:22:17,760 of that, which is OK. Clearly you want a wide field of 1734 01:22:17,760 --> 01:22:19,800 view. Do you need to have something 1735 01:22:19,800 --> 01:22:22,000 that looks like it's 3D, like it's really there? 1736 01:22:22,240 --> 01:22:25,560 So do I need to have glasses that could put something on the 1737 01:22:25,560 --> 01:22:27,840 table in front of us that looked like it was there? 1738 01:22:27,880 --> 01:22:31,240 And that's radically harder, Yeah, having a really good heads 1739 01:22:31,240 --> 01:22:34,320 up display that could put a public that could put and that 1740 01:22:34,320 --> 01:22:36,960 could put like an iPad display hovering in front of me. 1741 01:22:36,960 --> 01:22:41,680 I think it helps a lot with things like repair, like for 1742 01:22:41,680 --> 01:22:43,520 example you open the hood of a car. 1743 01:22:43,520 --> 01:22:46,440 And well, that, but that's still a Hut that's still like a 1744 01:22:46,480 --> 01:22:49,960 hovering label over the thing versus does it need to work in 1745 01:22:49,960 --> 01:22:51,880 broad daylight? Does it need to have black? 1746 01:22:51,880 --> 01:22:54,440 Does it need to be able to occlude a bright white table 1747 01:22:54,440 --> 01:22:55,440 like this? Right? 1748 01:22:55,720 --> 01:22:58,280 Maybe, maybe not. I think there's a range of 1749 01:22:58,280 --> 01:23:01,040 outcomes there where you maybe it ends up like a watch that 1750 01:23:01,040 --> 01:23:03,000 it's to be clearly to begin with, it'll be a smartphone 1751 01:23:03,000 --> 01:23:05,040 accessory just to have the. Computer in the battery, yes. 1752 01:23:05,200 --> 01:23:07,560 But does it end up like a watch where there's hundreds of 1753 01:23:07,560 --> 01:23:09,640 millions of people who have it, but the smartphone is the main 1754 01:23:09,640 --> 01:23:12,440 device, right? Or does it end up, no, actually 1755 01:23:12,440 --> 01:23:14,480 a couple of billion people? Are wearing this. 1756 01:23:14,480 --> 01:23:17,320 Let me ask you another question. Does does a watch top out? 1757 01:23:17,400 --> 01:23:20,760 And because The thing is wearables are another thing that 1758 01:23:20,760 --> 01:23:27,600 has huge traction and it's kind of like there's a lot, a lot, a 1759 01:23:27,600 --> 01:23:29,920 lot. We could fill this table, this 1760 01:23:29,920 --> 01:23:34,120 whole room now with IoT health stuff, right? 1761 01:23:34,400 --> 01:23:38,760 Because there's watches, there's rings like the ring, there's, 1762 01:23:40,080 --> 01:23:42,760 you know, wristbands, but. It depends. 1763 01:23:42,760 --> 01:23:46,280 It depends on the question. Is the the Mark Zuckerberg board 1764 01:23:46,280 --> 01:23:49,440 Oculus? Is that Mark Zuckerberg board 1765 01:23:49,480 --> 01:23:51,320 Oculus because he thinks this is the next smart thing? 1766 01:23:51,440 --> 01:23:54,200 He didn't buy it to be a games device or 100 million people 1767 01:23:54,200 --> 01:23:55,600 using it. He bought it because he thinks 1768 01:23:55,600 --> 01:23:58,240 this is an Xbox smart thing. Yeah, because also he had been 1769 01:23:58,240 --> 01:23:59,200 hit by the. Platform. 1770 01:23:59,200 --> 01:24:01,360 So hard yeah, he wants to own the platform for sure makes. 1771 01:24:01,360 --> 01:24:06,880 Sense so there's I I my base case is that VR might be might 1772 01:24:06,880 --> 01:24:10,880 crap out at 50 or 100 million people and I've struggled to see 1773 01:24:10,880 --> 01:24:13,880 it being 5 billion. I can see glasses being a couple 1774 01:24:13,880 --> 01:24:17,000 of 100 quite easily once it worked the optics are there I 1775 01:24:17,000 --> 01:24:20,560 can imagine it being 5 billion. I think that's harder, but. 1776 01:24:20,840 --> 01:24:26,680 ARAR AR/XR is probably bigger than VR, but as we were talking 1777 01:24:26,680 --> 01:24:28,760 about, VR is very, you know, the the thing new thing they're 1778 01:24:28,760 --> 01:24:31,040 doing with for controlling military drones like you. 1779 01:24:31,280 --> 01:24:32,880 Know there's loads of vertical stuff. 1780 01:24:32,880 --> 01:24:34,480 We're absolutely. That's gonna nail it, 1781 01:24:34,480 --> 01:24:35,920 Definitely, no question. That's right. 1782 01:24:35,920 --> 01:24:37,160 So all the telepresence have, yeah. 1783 01:24:37,360 --> 01:24:39,440 And you know, the guy up the telephone pole, the guy in the 1784 01:24:39,440 --> 01:24:41,440 oil wearing glasses. Yes, absolutely. 1785 01:24:41,440 --> 01:24:43,000 That will be a. That is a thing already. 1786 01:24:43,040 --> 01:24:44,400 That's right. And I think have you seen this 1787 01:24:44,400 --> 01:24:46,200 movie? It's called Surrogates. 1788 01:24:46,520 --> 01:24:48,760 It's actually, you know, pretty good sci-fi movie from like 1789 01:24:48,760 --> 01:24:52,160 almost 1015 years ago. And essentially like people are 1790 01:24:52,160 --> 01:24:56,720 like they stay at home and they pilot a good looking version of 1791 01:24:56,720 --> 01:24:59,160 themselves as a surrogate walking around outside. 1792 01:24:59,440 --> 01:25:00,920 So you can take more risks and so on. 1793 01:25:00,920 --> 01:25:03,920 Because if that thing gets in a car crash wherever, nobody cares 1794 01:25:04,280 --> 01:25:06,560 and then they could just do another surrogate and runner 1795 01:25:06,560 --> 01:25:09,960 inside, right. So I do think what are the use 1796 01:25:09,960 --> 01:25:15,000 cases for like a proper the VR, the VR control of a remote 1797 01:25:15,320 --> 01:25:16,920 thing. So it starts with I think 1798 01:25:16,920 --> 01:25:19,080 drones. And have you ever done a VR 1799 01:25:19,080 --> 01:25:21,160 headset with a drone? It's an experience. 1800 01:25:21,160 --> 01:25:23,400 You should definitely try it. It's a wild moment because it 1801 01:25:23,400 --> 01:25:26,640 really does feel like you're flying right, which is very cool 1802 01:25:26,640 --> 01:25:30,680 and an interesting experience. So yeah, I think it starts with 1803 01:25:30,680 --> 01:25:33,000 drones, but I think it eventually gets to something 1804 01:25:33,000 --> 01:25:36,680 where you've got gloves and maybe an omnidirectional 1805 01:25:36,680 --> 01:25:38,120 treadmill or something like that. 1806 01:25:38,120 --> 01:25:39,480 There's various kinds of things like that. 1807 01:25:39,840 --> 01:25:44,080 And you are able to control a humanoid anywhere, right? 1808 01:25:44,400 --> 01:25:47,040 So you control a humanoid and you can, you know, clamber up a 1809 01:25:47,040 --> 01:25:48,680 telephone pole and fix something. 1810 01:25:48,680 --> 01:25:51,200 You and you're training the AI as you're doing this, right? 1811 01:25:51,480 --> 01:25:54,760 You you could have a maintenance worker with skill in the art. 1812 01:25:55,360 --> 01:25:56,440 Yeah. You know, and we're not there 1813 01:25:56,440 --> 01:25:58,320 yet. It'll be years before we're 1814 01:25:58,320 --> 01:26:00,160 there. But eventually you have all 1815 01:26:00,160 --> 01:26:02,760 these humanoids around where you can just go into this, like 1816 01:26:02,760 --> 01:26:05,280 animate the suit and start doing things, you know? 1817 01:26:05,680 --> 01:26:09,040 So that's a pretty important use case for VR, like physical 1818 01:26:09,040 --> 01:26:11,080 telepresence. You have to nail a bunch of 1819 01:26:11,080 --> 01:26:12,920 technology for that. But I could go through the 1820 01:26:12,920 --> 01:26:14,760 gloves. I could go through the haptics. 1821 01:26:15,040 --> 01:26:16,880 A lot of those things are moving forward, Right. 1822 01:26:18,800 --> 01:26:21,640 And, you know, a lot of people are pouring money into this. 1823 01:26:21,720 --> 01:26:23,440 That's something I give a lot of credit to Zach for. 1824 01:26:23,440 --> 01:26:26,960 He's just, you know, he's just continuing this, you know, like, 1825 01:26:26,960 --> 01:26:28,840 I don't know how many 10s of billions of dollars have been 1826 01:26:28,840 --> 01:26:31,200 put into this. He's probably put the thick end 1827 01:26:31,200 --> 01:26:33,360 of 100 billion into that. Something along those. 1828 01:26:33,360 --> 01:26:36,240 Lines like it's 75 to 100. Yeah, I mean they are actually 1829 01:26:36,240 --> 01:26:37,520 selling a fair number of units now. 1830 01:26:37,520 --> 01:26:40,520 It just hasn't come close to keeping up with the the spend. 1831 01:26:40,520 --> 01:26:43,640 Yeah, the sales, the the sales are just bouncing along. 1832 01:26:43,640 --> 01:26:46,960 It's like it's not good enough to break out of VR enthusiasts. 1833 01:26:47,120 --> 01:26:48,840 Yeah. And it's it's funny you you go 1834 01:26:48,840 --> 01:26:50,120 back to what you said about Twitter. 1835 01:26:50,360 --> 01:26:53,800 There's almost like a test, which is if you say that 1836 01:26:53,800 --> 01:26:56,240 something probably isn't working yet and you get a bunch of 1837 01:26:56,240 --> 01:26:58,760 people shouting at you on social media there, and that proves 1838 01:26:58,760 --> 01:27:02,440 you're right because if it was working, they wouldn't care. 1839 01:27:02,440 --> 01:27:03,880 Yeah. Yeah, Yeah, that's right. 1840 01:27:03,880 --> 01:27:06,120 Well, so. If you went on social media and 1841 01:27:06,120 --> 01:27:08,840 said nobody uses TikTok, then people would just say this guy's 1842 01:27:08,840 --> 01:27:11,480 an idiot, you're on social media and say yeah, and there aren't 1843 01:27:11,480 --> 01:27:13,400 actually any consumer use cases for drones. 1844 01:27:13,600 --> 01:27:15,320 You'll get like the 10 people who love their. 1845 01:27:15,320 --> 01:27:17,800 Drones, OK, there's one exception which I will argue 1846 01:27:17,800 --> 01:27:19,840 with you on, which is crypto, yes, right. 1847 01:27:20,040 --> 01:27:22,280 So that is something where people will say there's no use 1848 01:27:22,280 --> 01:27:24,240 for crypto, you will say. Yes, but there there's just a 1849 01:27:24,240 --> 01:27:26,440 huge number of idiots on every. Side that's also true, yes, 1850 01:27:26,480 --> 01:27:29,000 right, that's right so so OK, so we did. 1851 01:27:29,000 --> 01:27:30,880 So I don't say there's no useful use case for crypto. 1852 01:27:30,880 --> 01:27:33,160 I have the most unpopular position possible, which as I 1853 01:27:33,160 --> 01:27:35,800 say, it's kind of useful but not completely useful, which means I 1854 01:27:35,800 --> 01:27:37,560 get both sides screaming. At me, yeah, that's right. 1855 01:27:37,600 --> 01:27:41,840 That's just perfect position. So actually, what has Ben Evans 1856 01:27:41,840 --> 01:27:43,960 on crypto then? I'll tell you biology on crypto. 1857 01:27:44,080 --> 01:27:45,680 There's several answers to that question. 1858 01:27:45,960 --> 01:27:48,640 One of them is, and this is sort of more an observation, which I 1859 01:27:48,640 --> 01:27:52,720 hope you you won't tell me I'm wrong, is like there's a bunch 1860 01:27:52,720 --> 01:27:55,960 of clever people working away building like all the tourists 1861 01:27:55,960 --> 01:27:58,200 left, like the whole line of tea thing was, was all nonsense. 1862 01:27:58,200 --> 01:28:00,800 And that all that that all all the tourists left the the 1863 01:28:00,800 --> 01:28:02,920 tourists and the grifters basically all moved on to AI. 1864 01:28:03,320 --> 01:28:05,720 Yeah, a lot of them, yes. And all the kind of people 1865 01:28:05,720 --> 01:28:08,080 trying to build content brands saying this is all wonderful, 1866 01:28:08,080 --> 01:28:09,960 all this is all bullshit, they all moved off to AI. 1867 01:28:10,600 --> 01:28:13,560 There's a bunch of people sitting and doing like abstruse, 1868 01:28:13,560 --> 01:28:15,200 very clever, very technical stuff. 1869 01:28:16,640 --> 01:28:20,240 There's a bunch of stuff working or being built that may work 1870 01:28:20,240 --> 01:28:24,040 around a financial, the finance industry, around finance rails, 1871 01:28:24,040 --> 01:28:27,200 around stable coins, various kinds of financial instruments, 1872 01:28:27,840 --> 01:28:31,520 most of which is storing money or speculating in money or 1873 01:28:31,520 --> 01:28:36,680 moving money around. Yeah, there is a thesis that you 1874 01:28:36,680 --> 01:28:40,320 could build Instagram on this, that this is sort of an open 1875 01:28:40,320 --> 01:28:44,160 source computer in which you could write software that 1876 01:28:44,160 --> 01:28:47,800 consumers would use. And I have a bunch of questions 1877 01:28:47,800 --> 01:28:52,640 about how that would work, whether that would work, whether 1878 01:28:52,640 --> 01:28:55,840 you would need to abstract the open, sort the crypto stuff away 1879 01:28:55,840 --> 01:28:57,560 so that the consumers didn't see it. 1880 01:28:57,840 --> 01:29:00,040 And if you did that, then why would they care? 1881 01:29:00,320 --> 01:29:03,160 Totally. And but none of that's kind of 1882 01:29:03,160 --> 01:29:06,160 there yet. Like they're all billion scale 1883 01:29:06,160 --> 01:29:08,880 consumer apps built on blockchain yet. 1884 01:29:10,040 --> 01:29:13,480 So there's a sort of watch this space around that. 1885 01:29:13,880 --> 01:29:17,480 And then there's the finance side, which I think is sort of 1886 01:29:17,600 --> 01:29:21,680 theoretically very interesting, but I struggle to get very 1887 01:29:21,680 --> 01:29:24,480 interested in it. Just personally, it's not what 1888 01:29:24,480 --> 01:29:27,840 I'm interested in and I struggle to see ways that I could add 1889 01:29:27,840 --> 01:29:32,800 value in talking about it. So I kind of pay attention to 1890 01:29:32,800 --> 01:29:34,320 it. And every now and then I point 1891 01:29:34,320 --> 01:29:39,080 out, like my newsletter on Sunday, I pointed to the Shopify 1892 01:29:39,080 --> 01:29:43,920 and Stripe and, and said, like, there's stuff happening here. 1893 01:29:43,960 --> 01:29:46,120 Yeah, and you should pay attention to this. 1894 01:29:46,120 --> 01:29:48,160 And there's people still interested in trying to build 1895 01:29:48,160 --> 01:29:49,720 things. So if you've just written this 1896 01:29:49,720 --> 01:29:51,360 off as all bullshit, you're kind of wrong. 1897 01:29:51,680 --> 01:29:54,560 Right. But as a writer and an analyst, 1898 01:29:55,440 --> 01:29:57,880 I haven't moved it on to something that I feel I should 1899 01:29:57,880 --> 01:29:58,960 write about. Totally. 1900 01:29:59,000 --> 01:30:01,840 So OK, so that's very helpful. It's always helpful for me to 1901 01:30:01,840 --> 01:30:05,560 kind of triangulate on an area, you know, on the. 1902 01:30:06,360 --> 01:30:10,000 So here is my basic view. You may have heard me say this 1903 01:30:10,000 --> 01:30:11,640 12 years ago. I think this is still true. 1904 01:30:12,960 --> 01:30:17,040 Crypto is good for transactions that are very large, very small, 1905 01:30:17,600 --> 01:30:24,440 very fast, very international, very automated, very complex, or 1906 01:30:24,440 --> 01:30:29,760 then to be very transparent. And the reason for that is like, 1907 01:30:29,760 --> 01:30:32,760 for example, a Starbucks swipe like of a credit card is none of 1908 01:30:32,760 --> 01:30:34,400 those things. It's not very large or very 1909 01:30:34,400 --> 01:30:35,960 small. It's like a mezzanine 1910 01:30:35,960 --> 01:30:38,160 transaction. It doesn't need to be very 1911 01:30:38,160 --> 01:30:40,880 automated because you can just talk to the, you know, cashier 1912 01:30:40,880 --> 01:30:42,800 and see your receipt. It's not international. 1913 01:30:42,800 --> 01:30:45,600 Both you and them are in the same room at the same time. 1914 01:30:46,360 --> 01:30:47,640 It doesn't need to be transparent. 1915 01:30:47,640 --> 01:30:49,880 You don't need a receipt on the blockchain for everybody to see 1916 01:30:49,880 --> 01:30:52,400 and so on and so forth, right? So the reason people think about 1917 01:30:52,400 --> 01:30:55,240 the coffee transaction and the thing about crypto is it's one 1918 01:30:55,240 --> 01:30:56,680 of the most common transactions people do. 1919 01:30:56,680 --> 01:30:58,280 They pay for their coffee every day, right? 1920 01:30:58,640 --> 01:31:02,240 So it's like, I don't know, 10% of your transactions, 20% or 1921 01:31:02,240 --> 01:31:04,360 maybe coffee because it's very few things you buy every day. 1922 01:31:04,360 --> 01:31:05,920 Coffee is one of those things people buy every day. 1923 01:31:06,600 --> 01:31:14,080 So where crypto really shines is the the alternative forms of 1924 01:31:14,080 --> 01:31:15,120 traffic. Actually, let me take your 1925 01:31:15,120 --> 01:31:20,720 mobile example, right? The Internet can do telephony, 1926 01:31:21,120 --> 01:31:23,440 but that was actually the thing that was best served by the 1927 01:31:23,440 --> 01:31:26,720 existing system, right? We still have like local 1928 01:31:26,720 --> 01:31:29,280 telephone calls, right? You can still use the telephone 1929 01:31:29,280 --> 01:31:32,320 network to place telephone calls where the Internet shine was. 1930 01:31:32,480 --> 01:31:34,960 And telephone calls were sort of like mezzanine amounts of 1931 01:31:34,960 --> 01:31:37,760 information, right? Especially local was like 1932 01:31:37,760 --> 01:31:39,360 between people in the same country, wasn't very 1933 01:31:39,360 --> 01:31:42,360 international where the Internet shine was, for example, moving 1934 01:31:42,360 --> 01:31:46,160 really large files like Dropbox or very small files like tweets, 1935 01:31:46,320 --> 01:31:48,240 right? Being very international, like 1936 01:31:48,240 --> 01:31:50,560 across borders, being very automated. 1937 01:31:50,560 --> 01:31:53,280 So it wasn't a human on both sides of the call, right? 1938 01:31:54,320 --> 01:31:57,280 It, it's shown for, you know, being very transparent. 1939 01:31:57,280 --> 01:32:00,160 You're broadcasting the web page here, but it's not a phone call 1940 01:32:00,160 --> 01:32:02,080 just between two people and so on and so forth, right. 1941 01:32:02,440 --> 01:32:05,560 So that I think is a good analogy where you like, yes, now 1942 01:32:05,560 --> 01:32:08,600 today, eventually the Internet took over long distance 1943 01:32:08,600 --> 01:32:12,160 telephony because that was Skype and then WhatsApp and what have 1944 01:32:12,160 --> 01:32:13,840 you. But even still today, telephony 1945 01:32:13,840 --> 01:32:15,880 as well captured by the current system, right? 1946 01:32:16,400 --> 01:32:18,880 And like the existing phone lines still exist. 1947 01:32:19,280 --> 01:32:22,320 That I think is a useful analogy for crypto where crypto, for 1948 01:32:22,320 --> 01:32:25,040 example, if you have, if you're a power user of money, right? 1949 01:32:25,360 --> 01:32:30,360 If I want to receive or send a wire to a startup in Japan, 1950 01:32:30,760 --> 01:32:34,720 USCC, I can do that in seconds and then I can refresh the page. 1951 01:32:34,720 --> 01:32:36,840 They can refresh the page and they can see it's cleared, 1952 01:32:36,880 --> 01:32:38,680 right? Yep, that is a real use case. 1953 01:32:38,680 --> 01:32:41,560 It's international wire transfers from anybody to 1954 01:32:41,560 --> 01:32:45,040 anybody with and by the way, the bank account set up also is 1955 01:32:45,040 --> 01:32:47,600 instant, right? So think about what we've done. 1956 01:32:47,640 --> 01:32:51,320 We've taken it from days to get AUS and Japanese bank account 1957 01:32:51,320 --> 01:32:54,360 set up to seconds. We've taken it from paying money 1958 01:32:54,360 --> 01:32:57,160 to do that to for the for the transfer itself to free. 1959 01:32:57,640 --> 01:33:00,160 We've taken it from taking multiple days for a wire 1960 01:33:00,160 --> 01:33:04,160 transfer to clear to seconds. And we also by the way, the up 1961 01:33:04,160 --> 01:33:06,880 time, it's not nine to five banking hours. 1962 01:33:06,880 --> 01:33:10,880 You can do it, you know, 24/7 and you can do it on any device, 1963 01:33:11,000 --> 01:33:12,960 right? That's a lot of improvements 1964 01:33:13,040 --> 01:33:15,920 just for the the important use case of international buyer 1965 01:33:15,920 --> 01:33:18,200 transfers, right? Then you also have the digital 1966 01:33:18,200 --> 01:33:22,400 gold use case that one you'll only believe in if I mean, I can 1967 01:33:22,400 --> 01:33:25,680 just point to the graph, Bitcoin is appreciated from .1 cents per 1968 01:33:26,560 --> 01:33:29,360 per Bitcoin to $100,000. So like there's enough people 1969 01:33:29,360 --> 01:33:31,360 who believe in it for gone up 100 million X, right? 1970 01:33:31,360 --> 01:33:33,240 Yeah. It's also, I mean digital gold. 1971 01:33:33,360 --> 01:33:36,640 I think it's also something that there's a kind of country 1972 01:33:36,640 --> 01:33:39,040 mapping here because some of what you're talking about is a 1973 01:33:39,040 --> 01:33:42,480 much bigger problem in say in the US than it is in countries 1974 01:33:42,480 --> 01:33:44,240 with less with different banking systems. 1975 01:33:44,840 --> 01:33:46,760 Some of it is also. Sepa, you guys have Sepa in 1976 01:33:46,760 --> 01:33:48,600 Europe and. It's like you send the money, it 1977 01:33:48,600 --> 01:33:51,280 arrives for free like it's also this is the point about PayPal. 1978 01:33:51,560 --> 01:33:53,240 But the SEPA is worse than Europe though. 1979 01:33:53,440 --> 01:33:56,400 SEPA would not work for a wire transfer to Brazil, for example. 1980 01:33:56,480 --> 01:33:57,960 So you saw the same issue, though I think there's. 1981 01:33:58,480 --> 01:34:03,160 Another point which is like I remember reading about people in 1982 01:34:03,160 --> 01:34:06,920 Argentina, yes, literally keeping their money in bricks. 1983 01:34:07,080 --> 01:34:08,240 Exactly. That's right. 1984 01:34:08,240 --> 01:34:11,800 So it's Argentina, Nigeria. Lebanon where you you actually 1985 01:34:11,800 --> 01:34:14,200 can't trust your government, Yes, and there are kind of. 1986 01:34:14,200 --> 01:34:15,400 Places. There's a lot of places like 1987 01:34:15,400 --> 01:34:17,920 that, unfortunately, yes. There's also a bunch of places 1988 01:34:17,920 --> 01:34:20,960 where nobody's worried about that they sent you for 100 1989 01:34:20,960 --> 01:34:22,040 years. Exactly, that's right. 1990 01:34:22,040 --> 01:34:26,840 So so the more middle class, stable and so on you are the 1991 01:34:26,840 --> 01:34:30,600 like basic crypto is for the power user of money and the 1992 01:34:30,600 --> 01:34:33,080 powerless, right? The purpose person who's like 1993 01:34:33,120 --> 01:34:35,920 reinventing what a bank account even is and the person who's 1994 01:34:35,920 --> 01:34:37,360 just trying to hang on to a bank account. 1995 01:34:37,360 --> 01:34:40,080 So it's like a U-shaped coalition, right, Similar to the 1996 01:34:40,080 --> 01:34:42,600 people who actually benefited most from the global economy. 1997 01:34:42,600 --> 01:34:44,720 Remember I said it was like the the elephant graph, right? 1998 01:34:45,200 --> 01:34:49,480 You had the basically 10th to 80th percentile of the world who 1999 01:34:49,480 --> 01:34:53,480 grew and you have the top 1% who grew and the the Western middle 2000 01:34:53,480 --> 01:34:56,920 class didn't, right. That coalition is actually also 2001 01:34:56,920 --> 01:34:59,560 the Cryptic coalition. It's like the people who are 2002 01:34:59,560 --> 01:35:06,120 just, you know, intranet as, as Tim Ferriss put it, James God, 2003 01:35:06,120 --> 01:35:09,840 what's, what's his name? Jason Bournes of the Internet, 2004 01:35:09,960 --> 01:35:11,640 right? Like just Internet hackers who 2005 01:35:11,640 --> 01:35:14,040 are just trying to move money. Like, for example, I'll give a 2006 01:35:14,040 --> 01:35:16,200 concrete example. Brian Armstrong, you know, my 2007 01:35:16,200 --> 01:35:19,800 friend, CEO of Coinbase, one of the reasons he got into crypto, 2008 01:35:19,800 --> 01:35:22,560 he had a few different life experiences that led him there. 2009 01:35:22,560 --> 01:35:24,280 One was actually lived in Argentina for a while. 2010 01:35:24,280 --> 01:35:26,440 So we saw like what a failed state would be like. 2011 01:35:27,000 --> 01:35:28,760 And the second, though, was actually being an Airbnb 2012 01:35:28,800 --> 01:35:31,760 engineer. So The thing is Airbnb even 2013 01:35:31,760 --> 01:35:35,000 still today has the problem of transactions that are very 2014 01:35:35,000 --> 01:35:40,880 large, very international, right and and also very one time low 2015 01:35:40,880 --> 01:35:43,240 trust, right? Because you've got like somebody 2016 01:35:43,240 --> 01:35:46,920 from Denmark staying with someone from Japan and it's, 2017 01:35:46,920 --> 01:35:50,720 it's a one time transaction of maybe on the order of $1000. 2018 01:35:50,720 --> 01:35:51,960 There's actually a fair amount of money. 2019 01:35:52,640 --> 01:35:57,480 And like the wire system is simply not set up for that 2020 01:35:57,480 --> 01:36:00,320 frequency of use between unrelated parties. 2021 01:36:00,320 --> 01:36:02,720 And there's a lot of friction on something like that. 2022 01:36:03,320 --> 01:36:08,960 And to a surprising extent, Airbnb had a lot of forex risk, 2023 01:36:09,000 --> 01:36:10,880 like, you know, because they had to hold currencies and all these 2024 01:36:10,880 --> 01:36:13,680 different things. And the thing you thought was a 2025 01:36:13,680 --> 01:36:16,080 solved problem, like just moving money from one country to 2026 01:36:16,080 --> 01:36:18,760 another, it's like, well, Airbnb has to do. 2027 01:36:18,760 --> 01:36:23,120 It's accounting in USD, but it's got income in, you know, if 2028 01:36:23,120 --> 01:36:24,880 they're, if they're an American company and they've got somebody 2029 01:36:24,880 --> 01:36:29,120 transferring money from Denmark to Japan there there's three 2030 01:36:29,120 --> 01:36:31,240 currencies in that transaction just right there, right? 2031 01:36:31,560 --> 01:36:33,760 So there's at least three currency pairs which fluctuate 2032 01:36:34,320 --> 01:36:37,520 and you've got at least two or three banking systems and all 2033 01:36:37,520 --> 01:36:40,400 the delays and fees, you start to see if people are like, wow, 2034 01:36:40,400 --> 01:36:43,080 this sucks so much. We need an Internet first 2035 01:36:43,080 --> 01:36:45,720 banking system, right? We need something which is 2036 01:36:45,720 --> 01:36:48,720 payments as packets, right? So that was the second thing 2037 01:36:48,720 --> 01:36:50,520 that motivated Brian to do it, right? 2038 01:36:51,200 --> 01:36:52,360 There's other things as well, right? 2039 01:36:52,680 --> 01:36:56,280 But so where, where would I put crypto today, right? 2040 01:36:56,280 --> 01:36:58,760 I'd say there's at least three applications. 2041 01:36:59,360 --> 01:37:01,800 There's more, but I'd say at least three that are at the 2042 01:37:02,200 --> 01:37:04,440 trillion or multi 100 billion range. 2043 01:37:05,320 --> 01:37:07,240 And those are a digital gold, right. 2044 01:37:07,240 --> 01:37:09,320 Just whether you believe in gold or not, like that's that's 2045 01:37:09,320 --> 01:37:11,400 there, people do, people do even if you just consider an 2046 01:37:11,400 --> 01:37:12,440 insurance. It's the thing that people are 2047 01:37:12,440 --> 01:37:13,320 doing. It's the thing that people are 2048 01:37:13,320 --> 01:37:13,800 doing. That's right. 2049 01:37:15,120 --> 01:37:18,800 B is it's like even if you didn't believe in luxury cars, 2050 01:37:18,800 --> 01:37:20,600 that's a market, right? So there's a market for it, 2051 01:37:20,600 --> 01:37:23,400 right, OK. B is international wire 2052 01:37:23,400 --> 01:37:24,920 transfers. I think stablecoins are now 2053 01:37:25,120 --> 01:37:29,080 there at this point. There are now one 2% is $250 2054 01:37:29,080 --> 01:37:32,040 billion silver coins have passed Visa, they passed MasterCard, 2055 01:37:32,040 --> 01:37:35,760 right And then third is actually crowdfunding, right. 2056 01:37:35,760 --> 01:37:38,320 So if you look at the largest crowdfundings of all time, most 2057 01:37:38,320 --> 01:37:40,800 of them are crypto. And the reason is that capital 2058 01:37:40,800 --> 01:37:44,160 formation online, like if you think about something like 2059 01:37:44,160 --> 01:37:46,360 Kickstarter or what have you, it's actually more 2060 01:37:46,360 --> 01:37:48,960 geographically limited and more limited by the credit card 2061 01:37:48,960 --> 01:37:50,320 rails. And you might think, for 2062 01:37:50,320 --> 01:37:54,560 example, it's not that easy for somebody in Brazil and Japan and 2063 01:37:54,880 --> 01:37:58,280 India to put 5000 bucks into your Kickstarter, right? 2064 01:37:58,720 --> 01:38:01,880 They the credit card rails being accepted might be fraud hit. 2065 01:38:01,880 --> 01:38:02,520 Go ahead. Yeah. 2066 01:38:02,520 --> 01:38:05,600 I was just to say, I wonder with some of the there's a certain 2067 01:38:05,600 --> 01:38:08,040 amount of swapping paper for paper in some of that. 2068 01:38:08,960 --> 01:38:12,120 Go ahead. Well, in the sense of here is a 2069 01:38:12,120 --> 01:38:16,400 new crypto project, Yes, a bunch of people who've speculated, oh, 2070 01:38:16,400 --> 01:38:20,200 totally made a bunch of crypto money for sure put their paper 2071 01:38:20,200 --> 01:38:22,280 gains in Bitcoin into this new crypto. 2072 01:38:22,320 --> 01:38:23,280 Project, that's right. That's right. 2073 01:38:23,320 --> 01:38:25,360 But but, but I'd say you're right. 2074 01:38:25,360 --> 01:38:28,840 A bunch of it is like that. Which is what a lot of NFTS was. 2075 01:38:28,880 --> 01:38:31,280 Yes, that's right. But but even if we even if 2076 01:38:31,280 --> 01:38:35,080 you're just totally right, what was funded off just the mechanic 2077 01:38:35,080 --> 01:38:38,840 of crowdfunding, Yeah, shows that that mechanic for capital 2078 01:38:38,960 --> 01:38:40,640 information, what they spent it on, I'd agree with you. 2079 01:38:40,640 --> 01:38:42,000 Many of those projects didn't go somewhere. 2080 01:38:42,000 --> 01:38:44,480 Some of them went really far. Like Ethereum was a really that 2081 01:38:44,480 --> 01:38:46,440 paid for all the rest in a sense, you know, if all the ones 2082 01:38:46,520 --> 01:38:47,880 went to zero, that was so successful. 2083 01:38:49,040 --> 01:38:53,400 But but just the mechanic of capital formation where you have 2084 01:38:53,640 --> 01:38:57,440 so, so that gets me to #4 right. If you look at now, you may, you 2085 01:38:57,600 --> 01:38:59,560 may, you may sort of disbelieve. So at least those three markets, 2086 01:38:59,880 --> 01:39:04,000 gold wire transfers, crowdfunding, those are very 2087 01:39:04,000 --> 01:39:05,840 large markets. Those are $100 billion trillion 2088 01:39:05,840 --> 01:39:08,680 dollar markets. Then you go to like other cases. 2089 01:39:09,000 --> 01:39:14,480 Now if I just look at trade volume, right, crypto today is 2090 01:39:14,480 --> 01:39:16,640 actually the number 4 Stock Exchange in the world in terms 2091 01:39:16,640 --> 01:39:21,280 of volume #1 I see #2 NASDAQ #3 Shenzhen #4 crypto and it's 2092 01:39:21,360 --> 01:39:24,280 rising fast. The thing that has held it back 2093 01:39:24,280 --> 01:39:28,840 for almost 15 years is the doing the obvious things was 2094 01:39:28,840 --> 01:39:33,440 pathologized, meaning it like literally yesterday or like like 2095 01:39:33,440 --> 01:39:38,040 a day or two ago, we finally fully legalized, very clearly 2096 01:39:38,040 --> 01:39:42,120 legalized putting a dollar on chain right now that we can put 2097 01:39:42,120 --> 01:39:45,160 a dollar on chain very clearly such to the point that Amazon 2098 01:39:45,160 --> 01:39:47,120 and Walmart are like OK, congressional legislation is 2099 01:39:47,120 --> 01:39:49,400 perfectly good. Let's go time right now. 2100 01:39:49,400 --> 01:39:52,320 We can finally put an equity on chain and we can put a fund 2101 01:39:52,320 --> 01:39:54,520 interest on chain. We can put every paper kind of 2102 01:39:54,520 --> 01:39:58,920 thing on chain. That is a very big deal, right? 2103 01:39:59,120 --> 01:40:02,880 That means that crowdfunding thing I talked about says that 2104 01:40:02,880 --> 01:40:06,080 an Internet company can issue Internet equity and anybody in 2105 01:40:06,080 --> 01:40:08,040 the world can be part of that cap table. 2106 01:40:08,040 --> 01:40:10,240 Whether you choose to accept them or not is another thing. 2107 01:40:10,560 --> 01:40:13,720 But the capital formation mechanism, it's now possible for 2108 01:40:13,720 --> 01:40:17,920 somebody in Japan or Brazil or or Mexico to invest in your 2109 01:40:17,920 --> 01:40:19,840 company. Once you have Internet equities, 2110 01:40:19,840 --> 01:40:22,200 Internet capital markets, that is now within sight. 2111 01:40:22,200 --> 01:40:24,400 Now that we have the Sablecoin thing, boom, done. 2112 01:40:24,960 --> 01:40:27,320 There's nothing, you know, now it's just a mechanical thing to 2113 01:40:27,320 --> 01:40:29,760 get the legal system going to make the on chain equities work 2114 01:40:29,760 --> 01:40:30,800 and there's already work on that. 2115 01:40:31,200 --> 01:40:35,720 So that is a big deal, right, Because the US doesn't want to 2116 01:40:35,720 --> 01:40:38,360 be the center of global financial empire anymore, right? 2117 01:40:38,360 --> 01:40:42,400 It's like it's, it's very conflict about this, but with 2118 01:40:42,400 --> 01:40:46,480 the tariffs and the trade war and you know, tourist visas, 2119 01:40:46,480 --> 01:40:50,600 work visas, student visa bans and so on, it like is very 2120 01:40:50,600 --> 01:40:52,840 conflict about whether he even wants foreign money coming in 2121 01:40:53,000 --> 01:40:55,480 to, to America, right. And they've got remittances 2122 01:40:55,480 --> 01:40:59,600 taxes coming up like one for 5%. So U.S. financial markets I 2123 01:40:59,600 --> 01:41:03,120 don't think are going to be there in the same way by 2035. 2124 01:41:03,880 --> 01:41:05,240 I think Chinese markets are rising. 2125 01:41:05,240 --> 01:41:07,160 Chinese stocks are rising. That's going to be one thing 2126 01:41:07,160 --> 01:41:08,680 that's there. But I think the Internet capital 2127 01:41:08,680 --> 01:41:11,400 markets will take over from American capital markets and 2128 01:41:11,400 --> 01:41:12,960 that's a very, very big application. 2129 01:41:14,080 --> 01:41:15,800 Let me go through a few more. Is this interesting? 2130 01:41:15,800 --> 01:41:17,800 So far, no. It's it's interesting. 2131 01:41:17,800 --> 01:41:20,040 I mean I, I, I mean sort of think about I. 2132 01:41:20,440 --> 01:41:21,600 Mean we've got numbers now, yeah. 2133 01:41:21,960 --> 01:41:24,560 Yeah, yeah, there was a thing that I was, so I'm getting away 2134 01:41:24,560 --> 01:41:26,680 from the microphone because I, we were chatting about it in the 2135 01:41:26,680 --> 01:41:28,960 car this morning. I have a sort of a mental Venn 2136 01:41:28,960 --> 01:41:34,520 diagram of like stuff I'm feel I can add something to. 2137 01:41:34,560 --> 01:41:37,080 Sure. Stuff that I feel I understand 2138 01:41:37,080 --> 01:41:38,640 and stuff where there's an audience. 2139 01:41:38,800 --> 01:41:40,880 Yes. And the challenge I always had 2140 01:41:40,880 --> 01:41:44,920 in writing about crypto, this is like a kind of a practical 2141 01:41:44,960 --> 01:41:50,400 question as an as an analyst is all AI, all all kind of crypto 2142 01:41:50,400 --> 01:41:51,960 questions. It felt like they were either 2143 01:41:51,960 --> 01:41:55,040 very, very technical conversations about it was kind 2144 01:41:55,040 --> 01:41:58,000 of like writing about Linux. So and I should always think 2145 01:41:58,000 --> 01:42:01,520 that like crypto reminds me a lot of open source and you are 2146 01:42:01,520 --> 01:42:04,240 either it is open source, but in just in the sense of the 2147 01:42:04,240 --> 01:42:05,320 general. Movement. 2148 01:42:05,840 --> 01:42:07,840 It was sort of. It reminded me a bit of like 2149 01:42:07,840 --> 01:42:11,880 either I write something about like the new kernel memory 2150 01:42:11,880 --> 01:42:15,560 management thing in Linux where I don't understand it and the 2151 01:42:15,560 --> 01:42:17,920 people who do aren't interested in what I'm going to say and no 2152 01:42:17,920 --> 01:42:19,040 one else cares. Yes. 2153 01:42:19,200 --> 01:42:25,280 It just it gets better or it was like, imagine what will happen 2154 01:42:25,360 --> 01:42:29,120 when it's like talking about open source in the early 90s. 2155 01:42:29,280 --> 01:42:31,880 Imagine what's what is going to happen when software is free and 2156 01:42:31,880 --> 01:42:34,840 there's not there was I've I've struggled and it's actually it's 2157 01:42:34,840 --> 01:42:37,480 a it's a, it's a, it's a thing. I've also had writing about AI 2158 01:42:37,480 --> 01:42:39,880 because I want to kind of it's not, it's not a specific about 2159 01:42:39,880 --> 01:42:43,280 what you think about this. It's what I'm most good at. 2160 01:42:43,280 --> 01:42:46,120 I think all the stuff that I write that people seem to like 2161 01:42:46,120 --> 01:42:50,440 most is kind of talking about the product strategy of how is 2162 01:42:50,440 --> 01:42:52,720 this going to work? Who's going to win? 2163 01:42:52,720 --> 01:42:56,080 Who's not going to win? How is a corporation or consumer 2164 01:42:56,080 --> 01:42:58,160 going to buy this? What would you do with it? 2165 01:42:58,280 --> 01:43:00,240 Right? And I struggled for a while to 2166 01:43:00,480 --> 01:43:03,400 to write about our lands on that point because it was either 2167 01:43:03,400 --> 01:43:07,760 like, what are the 30 new papers this year or like this is going 2168 01:43:07,760 --> 01:43:10,680 to transform humanity, right? And it was kind of hard to find 2169 01:43:10,680 --> 01:43:13,800 anything in the middle. Is in in the weeds or super 2170 01:43:13,800 --> 01:43:16,960 macro but the mezzo. Super kind of messianic, but not 2171 01:43:16,960 --> 01:43:18,560 much about like product strategy in the middle. 2172 01:43:18,920 --> 01:43:22,280 And I have the same challenge in writing about crypto in that 2173 01:43:22,280 --> 01:43:24,280 it's either very, very technical, OK, I've got 2174 01:43:24,280 --> 01:43:28,000 something for you or it's yeah, imagine in 30 years or it's 2175 01:43:28,000 --> 01:43:31,000 about finance where I don't you don't care that much about it. 2176 01:43:31,080 --> 01:43:32,680 Yeah, it's not just that I don't care. 2177 01:43:32,680 --> 01:43:35,680 It's like I would have to spend 6 months to get to the point 2178 01:43:35,680 --> 01:43:38,080 that I know what all the acronyms for moving money 2179 01:43:38,080 --> 01:43:40,000 between banks are. I have an opinion about them. 2180 01:43:40,240 --> 01:43:43,680 So I've never like seen Well, is that is it? 2181 01:43:43,680 --> 01:43:46,320 It's, it's in a kind of in a completely different analogy. 2182 01:43:46,440 --> 01:43:48,760 It's also like talking about chips, you know, should I, 2183 01:43:48,800 --> 01:43:51,640 should I get to the point that I understand what's going on in 2184 01:43:51,640 --> 01:43:54,440 chips? Is that a good use of my time? 2185 01:43:54,720 --> 01:43:57,040 Would I be able to say anything of value there? 2186 01:43:57,320 --> 01:43:59,800 And I so far I've kind of felt no, there's a bunch of people 2187 01:43:59,800 --> 01:44:01,960 who know way more about that. Like the semis analyst guys have 2188 01:44:01,960 --> 01:44:04,320 got it. So it, it, it so let me let me 2189 01:44:04,320 --> 01:44:07,880 actually empathize with you in a certain way, which is I was 2190 01:44:07,880 --> 01:44:11,640 actually a very late user of social media, right? 2191 01:44:12,040 --> 01:44:16,960 I only got on Twitter in like December 2013, OK, Which is like 2192 01:44:17,800 --> 01:44:19,280 like a decade. Hello Boomer. 2193 01:44:19,800 --> 01:44:20,960 Hello, Boomer. Exactly. 2194 01:44:20,960 --> 01:44:22,960 That's right. No, I mean, The thing is I got 2195 01:44:22,960 --> 01:44:26,120 onto Facebook very early because it just was like moving around 2196 01:44:26,120 --> 01:44:27,800 universities or what have you at the time. 2197 01:44:28,160 --> 01:44:31,960 But I didn't really use it. And the reason is that until 2198 01:44:31,960 --> 01:44:36,760 2013, I essentially believed that there was absolutely. 2199 01:44:37,000 --> 01:44:38,720 I was just a very private person, you know, I was just 2200 01:44:38,720 --> 01:44:42,440 like, you know, it's, it's weird because I now post a lot or what 2201 01:44:42,440 --> 01:44:44,520 have you. I was just a very private person 2202 01:44:44,600 --> 01:44:48,080 and I never, I didn't give any public talks until late 2013 and 2203 01:44:48,080 --> 01:44:51,640 so on. And I just thought social media 2204 01:44:51,640 --> 01:44:54,960 was a complete waste of time. And all that mattered was 2205 01:44:55,280 --> 01:44:59,840 genomics and math. And you know, like, like hard, 2206 01:44:59,960 --> 01:45:01,520 like what people call hard tech now. 2207 01:45:01,520 --> 01:45:03,480 Like I was doing genomics and robotics and I'm, you know, 2208 01:45:03,480 --> 01:45:05,640 proud of that work. I think is is important stuff. 2209 01:45:06,200 --> 01:45:10,400 And I didn't see the utility in tweeting my breakfast and I 2210 01:45:10,400 --> 01:45:13,000 didn't see the utility in just, you know, petting each other's 2211 01:45:13,000 --> 01:45:15,560 fur, which is a lot of what people do on Facebook or 2212 01:45:15,560 --> 01:45:18,800 whatever, you know, right. So I didn't see the value in any 2213 01:45:18,800 --> 01:45:24,640 of that. And it was only once all of that 2214 01:45:24,640 --> 01:45:27,680 was what bootstrapped the space. All of the fur petting got 2215 01:45:27,680 --> 01:45:30,480 hundreds of millions of people on there, all of the breakfast 2216 01:45:30,480 --> 01:45:34,200 tweeting and so on until, you know, what actually made it 2217 01:45:34,200 --> 01:45:38,800 useful and interesting to me was I saw somebody tweeting a 2218 01:45:38,800 --> 01:45:41,360 summary of a genomics conference at Cold Spring Harbor that I 2219 01:45:41,360 --> 01:45:44,080 didn't have the time to attend. And they gave a much better 2220 01:45:44,080 --> 01:45:47,360 account of it than any layman would have. 2221 01:45:47,560 --> 01:45:49,560 It's like, you know, like someone tweeting a mobile thing 2222 01:45:49,560 --> 01:45:51,680 and you're like, oh, that's those are really great details 2223 01:45:51,680 --> 01:45:53,160 and you're skilled in the art, right? 2224 01:45:53,640 --> 01:45:56,480 And and then I was like, oh, wow, I can get like really 2225 01:45:56,760 --> 01:45:57,840 detailed information. You're OK. 2226 01:45:57,840 --> 01:45:59,640 Now this is valuable to me as a reader, right? 2227 01:45:59,880 --> 01:46:02,080 What's my point? My point is, I think the 2228 01:46:02,080 --> 01:46:05,160 parameter that you want to track when you're looking at crypto is 2229 01:46:05,160 --> 01:46:06,920 block space. Have you heard that parameter 2230 01:46:06,920 --> 01:46:09,360 before? OK, That is the most important 2231 01:46:09,360 --> 01:46:11,680 parameter in crypto that people outside crypto don't realize 2232 01:46:11,680 --> 01:46:14,560 governs crypto. Block space is to crypto what 2233 01:46:14,560 --> 01:46:17,880 bandwidth is to the web. So if you think about the early 2234 01:46:17,880 --> 01:46:20,160 Internet or the early web, I should be more precise in the 2235 01:46:20,160 --> 01:46:22,760 90s like it was very bandwidth constraints. 2236 01:46:22,760 --> 01:46:26,320 It's 28857 six modems. And so that's why like Google 2237 01:46:26,320 --> 01:46:30,120 was 10 blue links and I think Amazon even had many images at 2238 01:46:30,120 --> 01:46:33,760 all. And in fact, you remember 6°, it 2239 01:46:33,760 --> 01:46:35,880 was a social network, right? So that was a text based social 2240 01:46:35,880 --> 01:46:37,160 network. It didn't take off because 2241 01:46:37,520 --> 01:46:39,080 without images, people didn't really. 2242 01:46:39,080 --> 01:46:39,920 Yeah. You got nothing to. 2243 01:46:39,920 --> 01:46:41,080 Share. You got nothing to share exactly 2244 01:46:41,080 --> 01:46:43,880 right. ICQ was a chat app that did 2245 01:46:43,880 --> 01:46:45,200 work. AOL Instant Messenger worked 2246 01:46:45,200 --> 01:46:47,680 because that was just text that could be sent on that low 2247 01:46:47,680 --> 01:46:50,560 bandwidth thing. It was only in the 2000s that 2248 01:46:50,560 --> 01:46:52,640 you started to get more graphical things when bandwidth 2249 01:46:52,640 --> 01:46:54,920 increase, like Facebook, the reason it took off at Harvard, 2250 01:46:54,920 --> 01:46:58,080 everybody had AT1 connection being at Harvard and they, they 2251 01:46:58,080 --> 01:47:00,240 finally had digital cameras so you could have photos. 2252 01:47:00,560 --> 01:47:03,240 And as digital cameras propagated out, so did Facebook, 2253 01:47:03,240 --> 01:47:06,760 right. And you go further and further 2254 01:47:06,760 --> 01:47:10,800 and like, you know, the Internet only or Internet Explorer only 2255 01:47:10,800 --> 01:47:13,520 got disrupted by Firefox in like the late 2000s, right? 2256 01:47:13,720 --> 01:47:17,520 It was only really by the early twenty 10s that you had the full 2257 01:47:17,520 --> 01:47:20,400 JavaScript stack of like jQuery and then only later for React 2258 01:47:20,400 --> 01:47:22,800 and what have you. So this concept that we have 2259 01:47:22,800 --> 01:47:25,720 today of like a mobile web app or you can download JavaScript 2260 01:47:25,720 --> 01:47:29,600 and run an app in the browser on a phone was a vision in the 90s, 2261 01:47:29,600 --> 01:47:31,600 but it took a long time together because bandwidth had to 2262 01:47:31,600 --> 01:47:33,440 increase for that, right? So what's the analogy here? 2263 01:47:33,880 --> 01:47:38,520 Block space, basically block space is the amount of storage 2264 01:47:38,640 --> 01:47:41,000 that you have on a blockchain. Like, think of a blockchain as 2265 01:47:41,000 --> 01:47:42,960 like an armored car for data, right? 2266 01:47:43,240 --> 01:47:45,480 Because this is data that people want to corrupt, right? 2267 01:47:45,760 --> 01:47:50,720 In a sense, if it's a file on disk, it's important to you. 2268 01:47:50,760 --> 01:47:53,800 If it's a file online, it's important to others. 2269 01:47:53,800 --> 01:47:56,760 And if it's a file on chain, it's really important to others. 2270 01:47:57,120 --> 01:47:59,200 And it's so important that they might try to screw with it. 2271 01:47:59,600 --> 01:48:04,360 And so Bitcoin came up with like an armored car for data where 2272 01:48:04,360 --> 01:48:07,400 you could guard the -1 or plus one of who had what Bitcoin. 2273 01:48:08,160 --> 01:48:11,520 And over time, that block space increased so that you could do 2274 01:48:11,520 --> 01:48:13,280 some basic smart contracts on Ethereum. 2275 01:48:13,880 --> 01:48:18,000 And now it's increased enough that you can blast millions of 2276 01:48:18,000 --> 01:48:22,360 stablecoin transactions a day on like base and Solana and so on 2277 01:48:22,360 --> 01:48:25,680 and so forth. And so you should conceptualize 2278 01:48:25,680 --> 01:48:27,920 it as why hasn't this happened yet? 2279 01:48:29,000 --> 01:48:31,800 And instead think of, OK, these applications are gated by the 2280 01:48:31,800 --> 01:48:34,840 amount of block space. And so they're coming online 2281 01:48:34,920 --> 01:48:36,200 similar to the amount of bandwidth. 2282 01:48:36,200 --> 01:48:38,840 You had like text only apps, then you had images, then you 2283 01:48:38,840 --> 01:48:40,480 had videos. And like Netflix only did 2284 01:48:40,480 --> 01:48:43,320 streaming video in like the early 20 tens, right? 2285 01:48:43,880 --> 01:48:45,560 I mean, we think about all that as recent. 2286 01:48:45,560 --> 01:48:46,760 No, I don't I don't I. Don't. 2287 01:48:46,760 --> 01:48:47,680 That's the way of thinking about it. 2288 01:48:47,680 --> 01:48:49,920 Yeah, I don't have a problem with the idea that you couldn't 2289 01:48:49,920 --> 01:48:52,800 build Instagram on this because the infrastructure isn't fast. 2290 01:48:52,840 --> 01:48:53,720 Block space wasn't. Current. 2291 01:48:53,720 --> 01:48:56,800 Yes. Yet I think there's a bunch of 2292 01:48:56,800 --> 01:49:00,440 like, interesting conceptual questions around, well, what 2293 01:49:00,440 --> 01:49:02,320 would happen when we got there? Yeah. 2294 01:49:02,320 --> 01:49:04,840 So here's a few things. Interesting. 2295 01:49:04,840 --> 01:49:07,160 There will be also kind of you're sort of speculating five 2296 01:49:07,160 --> 01:49:08,560 years in advance. Yeah. 2297 01:49:08,760 --> 01:49:13,680 So my view is, I'm not sure if it'll be exactly Instagram, you 2298 01:49:13,680 --> 01:49:16,520 know? On well, it would and I, I, I 2299 01:49:16,520 --> 01:49:18,760 think we can be sure it wouldn't be exactly Instagram, right, 2300 01:49:18,760 --> 01:49:20,000 right, right, but just kind of conception. 2301 01:49:20,000 --> 01:49:23,040 What is the app you could build? Consumer applications you could 2302 01:49:23,040 --> 01:49:24,360 use? I mean, this is the phrasing I 2303 01:49:24,360 --> 01:49:28,440 remember you using years ago that one should think of of 1 2304 01:49:28,440 --> 01:49:32,680 should think of a blockchain as a distributed virtual machine. 2305 01:49:33,080 --> 01:49:35,080 Yes, and it's another. Layer it is exactly that's. 2306 01:49:35,320 --> 01:49:37,000 Right. And every layer of abstraction 2307 01:49:37,000 --> 01:49:41,640 is always slower and crapper than running on the bare metal, 2308 01:49:41,960 --> 01:49:44,960 except that it allows you to do a bunch of stuff that you can't 2309 01:49:44,960 --> 01:49:46,040 do if you want on the bare metal. 2310 01:49:46,200 --> 01:49:47,680 That's exactly right. That's exactly right. 2311 01:49:47,680 --> 01:49:51,440 And The thing is, block chains are in a sense one of the 2312 01:49:51,440 --> 01:49:53,120 frontiers of operating systems research. 2313 01:49:53,120 --> 01:49:55,400 Like in the same way like there's operating system like 2314 01:49:55,400 --> 01:49:58,480 Windows, there's a browser which is itself an operating system 2315 01:49:58,480 --> 01:50:00,520 because you can run apps in it. It's got a full programming 2316 01:50:00,520 --> 01:50:02,240 language like that's how Chrome layered. 2317 01:50:02,320 --> 01:50:04,520 Were you at A16Z when Martin Cassado was our? 2318 01:50:04,840 --> 01:50:08,280 Yeah, we overlapped just a bit and we invested a bunch of 2319 01:50:08,280 --> 01:50:10,800 things together, Yeah. Well, Martin had this great 2320 01:50:10,800 --> 01:50:13,280 observation. You remember when YC said that 2321 01:50:13,280 --> 01:50:17,000 like for 1/4 of their the companies 90% of the code was 2322 01:50:17,040 --> 01:50:18,640 written with AI? Yeah. 2323 01:50:18,640 --> 01:50:21,280 And he responded to this by saying yes, but if you write an 2324 01:50:21,280 --> 01:50:23,360 iPhone out 90% of your code is written by Apple. 2325 01:50:23,600 --> 01:50:24,960 Yes. And so there were all those 2326 01:50:24,960 --> 01:50:27,280 levels of abstraction. Prompting is just a higher level 2327 01:50:27,280 --> 01:50:28,080 of programming. That's right. 2328 01:50:28,120 --> 01:50:32,800 Yeah, exactly. And so there's a, the, I suppose 2329 01:50:32,800 --> 01:50:36,400 the another way of answering your question is like the 2330 01:50:36,400 --> 01:50:38,080 finance stuff is there. I can see it. 2331 01:50:38,080 --> 01:50:39,760 I get it. I'm not sure I can add any value 2332 01:50:39,760 --> 01:50:41,440 to that. It's interesting and I will tell 2333 01:50:41,440 --> 01:50:43,840 people it's kind of interesting, but you pay attention to this. 2334 01:50:44,040 --> 01:50:46,000 I think you'll be. A leader, Go ahead, Sir. 2335 01:50:46,000 --> 01:50:49,840 The building more generalized consumer applications on it is 2336 01:50:50,000 --> 01:50:53,880 conceptually more more interesting to me is something 2337 01:50:53,880 --> 01:50:55,800 that I could make money telling other people about. 2338 01:50:56,000 --> 01:50:59,280 Yes, except that it isn't happening yet, and it probably 2339 01:50:59,280 --> 01:51:01,320 will. At a certain point the curve 2340 01:51:01,320 --> 01:51:03,920 will tell it curve up. The blocked space will expand. 2341 01:51:03,920 --> 01:51:06,280 The stuff will get faster and cheaper and can store more stuff 2342 01:51:06,520 --> 01:51:09,200 and people you will be people will be able to build stuff on 2343 01:51:09,200 --> 01:51:12,080 this deterministically. It won't be exactly Instagram. 2344 01:51:12,080 --> 01:51:14,920 I think that's just kind of a useful mental model for thinking 2345 01:51:14,920 --> 01:51:16,280 that you could build. Something. 2346 01:51:16,280 --> 01:51:19,200 Like that you could build consumer network apps like that 2347 01:51:19,240 --> 01:51:23,280 on this, this at that point, then I think you have a bunch of 2348 01:51:23,280 --> 01:51:26,440 kind of new interesting questions like, well, is it a 2349 01:51:26,440 --> 01:51:28,760 good idea to have a social network where all the users have 2350 01:51:28,760 --> 01:51:29,920 a vote? What would that look like? 2351 01:51:29,920 --> 01:51:31,200 What problems does that? Right, right, right, right. 2352 01:51:31,520 --> 01:51:33,040 Yes. And well, Dallas are that 2353 01:51:33,040 --> 01:51:35,440 already this. Is yeah, exactly, which struck 2354 01:51:35,440 --> 01:51:37,560 me the other day that all the arguments against that are 2355 01:51:37,560 --> 01:51:40,320 basically all the argument and saying no, you need ACEO in 2356 01:51:40,320 --> 01:51:41,880 charge. It basically all the same 2357 01:51:41,880 --> 01:51:44,240 arguments to say no, you don't want mass democracy, you need 2358 01:51:44,280 --> 01:51:46,200 you need a king. And you can have some balance 2359 01:51:46,200 --> 01:51:47,800 like representative democracy, right? 2360 01:51:47,800 --> 01:51:50,160 So you have the vote and they vote for somebody for mixed. 2361 01:51:50,280 --> 01:51:53,400 Constitutions which again like look at Africa to see how Latin 2362 01:51:53,400 --> 01:51:55,480 America to see how mixed constitutions. 2363 01:51:55,480 --> 01:51:58,120 Work out well, I'm saying representative actually, where 2364 01:51:58,120 --> 01:52:00,680 you have a leader, but they've got a fixed term and there's 2365 01:52:00,680 --> 01:52:03,040 both of them, for example. All of that stuff is 2366 01:52:03,040 --> 01:52:06,240 fascinating. I, it's like we don't have it 2367 01:52:06,240 --> 01:52:10,000 yet and I and no one's going to pay me to go to a conference and 2368 01:52:10,000 --> 01:52:11,120 give a presentation. Totally, totally. 2369 01:52:11,280 --> 01:52:12,880 So it's kind of tough for me to write about. 2370 01:52:12,920 --> 01:52:14,920 Yeah, totally. I will say all I just say is to 2371 01:52:14,920 --> 01:52:18,240 put on your radar if you go to likesnapshot.org or vote Agora, 2372 01:52:18,520 --> 01:52:21,880 there are actually very large treasuries where all that voting 2373 01:52:21,880 --> 01:52:24,240 stuff is happening on chain cryptographic voting. 2374 01:52:24,240 --> 01:52:26,400 And so and so that's that's growing like stablecoins kind of 2375 01:52:26,720 --> 01:52:28,440 people that ignored stablecoins for a while just kept 2376 01:52:28,440 --> 01:52:30,080 compounding. So the the on chain voting stuff 2377 01:52:30,080 --> 01:52:32,320 is there. But what I will say is that I 2378 01:52:32,320 --> 01:52:34,960 think just like I was like a late adopter of social media 2379 01:52:34,960 --> 01:52:37,520 since I just it, it had to get to a certain level of 2380 01:52:37,520 --> 01:52:39,840 significance before I cared about it for the kinds of things 2381 01:52:39,840 --> 01:52:42,600 I care about. Just I think the kinds of people 2382 01:52:42,600 --> 01:52:46,240 interesting in crypto are either A, they're engineers and they 2383 01:52:46,240 --> 01:52:48,080 just like the developers or the power users. 2384 01:52:48,520 --> 01:52:52,520 B, their financiers, right? Or, or in some sense financiers 2385 01:52:52,520 --> 01:52:54,640 or day traders, whatever it is, both the high and the lower. 2386 01:52:55,040 --> 01:52:57,400 And then C in the part we didn't say is just like they're 2387 01:52:57,400 --> 01:52:59,080 political, right? It's like a political 2388 01:52:59,080 --> 01:53:00,320 motivation. So I kind of being like being a 2389 01:53:00,640 --> 01:53:02,520 Protestant or a Catholic. They have a certain worldview. 2390 01:53:02,520 --> 01:53:04,560 Also very open source. Yeah, that's right. 2391 01:53:04,560 --> 01:53:07,080 Exactly. So, so like I have that, you 2392 01:53:07,080 --> 01:53:10,600 know, we both like enterprise SAS type stuff, product type 2393 01:53:10,600 --> 01:53:13,520 stuff, that kind of discussion. And but I also like a bunch of 2394 01:53:13,520 --> 01:53:16,920 other things and you like you like art museums and, and things 2395 01:53:16,920 --> 01:53:18,520 like that, which I'm like, OK, that's cool. 2396 01:53:18,520 --> 01:53:20,680 You know, go have fun, right? So we have, we have our own Venn 2397 01:53:20,680 --> 01:53:23,440 diagram kind of thing, right? So, OK, so switching gears, I 2398 01:53:23,440 --> 01:53:25,360 think you'll be more interested in crypto as block space 2399 01:53:25,360 --> 01:53:29,240 increases. And once crypto wallets, let me 2400 01:53:29,240 --> 01:53:31,360 actually give you an example of something which it's used for 2401 01:53:31,400 --> 01:53:33,600 useful for right now, where the block space increased enough, 2402 01:53:33,880 --> 01:53:37,760 you know, open router that that allows you to try a bunch of 2403 01:53:37,760 --> 01:53:41,040 different AI models and just use a crypto to pay for all of it. 2404 01:53:41,040 --> 01:53:44,120 OK, So this way you don't have to have 500 different accounts 2405 01:53:44,120 --> 01:53:46,800 at 500 different because there's so many different AI models, you 2406 01:53:46,800 --> 01:53:48,520 don't necessarily set up accounts and all that stuff, 2407 01:53:48,520 --> 01:53:50,120 right? So it just takes all that 2408 01:53:50,120 --> 01:53:52,480 account set of process and you just have one account, you pay 2409 01:53:52,480 --> 01:53:54,480 crypto and it settles it with all these other guys, right? 2410 01:53:54,920 --> 01:53:56,760 That isn't kind of completely transiental. 2411 01:53:56,840 --> 01:53:59,640 Thing that just occurs to me as you were speaking is, you know, 2412 01:53:59,680 --> 01:54:02,680 LM Arena as this distributed voting system. 2413 01:54:03,960 --> 01:54:06,360 The thing I always thought would be interesting would be to flip 2414 01:54:06,360 --> 01:54:09,440 that and say can you pass a double-blind test? 2415 01:54:10,360 --> 01:54:15,800 Yeah, if you take a model that's on the top 20, yeah, and give me 2416 01:54:15,800 --> 01:54:18,160 a bunch of responses, how many people would pass a double-blind 2417 01:54:18,160 --> 01:54:19,160 test? You know which is which. 2418 01:54:19,560 --> 01:54:21,640 Well, The thing is probably some kinds of question. 2419 01:54:21,640 --> 01:54:23,680 You would tell very easily, but an awful lot. 2420 01:54:23,680 --> 01:54:25,920 I bet most people probably. Wouldn't, so the most 2421 01:54:25,920 --> 01:54:30,040 fundamental one would be like what is the private key to this 2422 01:54:30,240 --> 01:54:33,000 or like a basically what is the private key to this wallet? 2423 01:54:33,320 --> 01:54:36,360 That's something that depending on how it's set up, we were 2424 01:54:36,360 --> 01:54:39,040 talking about this in the car, but basically another major use 2425 01:54:39,040 --> 01:54:41,320 case for crypto is AI makes everything fake. 2426 01:54:41,320 --> 01:54:44,640 Crypto makes it real again, because AI can fake all kinds of 2427 01:54:44,640 --> 01:54:47,000 stuff and give you this very convincing thing on like the, 2428 01:54:47,280 --> 01:54:49,880 the, the deep resourcing where it said 40% of the phones or 2429 01:54:49,880 --> 01:54:53,080 whatever you're saying, but it cannot fake the private key. 2430 01:54:53,080 --> 01:54:55,440 So it cannot show a nonzero Bitcoin balance or nonzero 2431 01:54:55,440 --> 01:54:57,840 Ethereum balance without actually having the the 2432 01:54:57,840 --> 01:54:59,520 cryptographic solution there. So. 2433 01:54:59,520 --> 01:55:01,560 But it could probably just tell you that the balance is 0, 2434 01:55:01,960 --> 01:55:03,920 because it might be. Yeah, sure, sure. 2435 01:55:04,520 --> 01:55:06,640 But what I mean about that is like, for example, all kinds of 2436 01:55:06,640 --> 01:55:08,920 let me give you, you know, Captchas, right, websites. 2437 01:55:09,320 --> 01:55:11,440 So AI can bust a lot of captchas. 2438 01:55:11,440 --> 01:55:13,800 Now it can get through. It can am IA robot, it can 2439 01:55:13,800 --> 01:55:18,000 figure it out, get through. But if you had to log in with a 2440 01:55:18,000 --> 01:55:22,400 crypto wallet that had $1.00 in it or $10.00 or $100, AI can't 2441 01:55:22,400 --> 01:55:24,280 fake that. It cannot fake the possession of 2442 01:55:24,280 --> 01:55:27,280 that cryptography, right? Like to give you 1. 2443 01:55:27,320 --> 01:55:29,840 Here's one motivating example for why crypto will get. 2444 01:55:30,120 --> 01:55:32,040 Maybe this argument will convince you, maybe not. 2445 01:55:32,040 --> 01:55:35,560 If it's fine, you know, Google login, you agree, Is it billions 2446 01:55:35,560 --> 01:55:38,800 of users, right? But Google login, when you log 2447 01:55:38,800 --> 01:55:41,960 into a website, you only can log in basically with your e-mail 2448 01:55:41,960 --> 01:55:43,640 address and the permissions to your Google account. 2449 01:55:43,920 --> 01:55:47,640 There's something very obvious that somehow even Google with 2450 01:55:47,640 --> 01:55:50,120 all of its strength has not been able to implement, which is an 2451 01:55:50,120 --> 01:55:53,080 international balance, a spendable balance, right? 2452 01:55:53,520 --> 01:55:57,240 Google login could not have, for their reason, a spendable 2453 01:55:57,240 --> 01:55:58,560 balance across different countries. 2454 01:55:58,560 --> 01:56:01,240 They've solved that for Google itself, where everybody can pay 2455 01:56:01,240 --> 01:56:03,240 Google and subscribe to Google with a zillion credit cards in 2456 01:56:03,240 --> 01:56:05,480 all these different countries, but somehow they couldn't make 2457 01:56:05,480 --> 01:56:07,680 it work so you could log into a third party site with a 2458 01:56:07,680 --> 01:56:10,000 spendable balance. Crypto did solve that. 2459 01:56:10,640 --> 01:56:14,240 Just that alone means that every Google and Facebook login will 2460 01:56:14,240 --> 01:56:18,520 eventually be either augmented or replaced by a crypto login. 2461 01:56:19,920 --> 01:56:23,200 So I'm going to pick up something you said, which I you 2462 01:56:23,200 --> 01:56:26,840 mentioned, which I mentioned in the car around what's fake and 2463 01:56:26,840 --> 01:56:27,640 what's real. Yeah. 2464 01:56:29,360 --> 01:56:33,400 So if you're buying an apartment and well, so going back a step, 2465 01:56:33,440 --> 01:56:36,680 I think most of what most people follow on Instagram is no longer 2466 01:56:36,680 --> 01:56:38,680 their friends. It's interest graph. 2467 01:56:38,680 --> 01:56:44,200 Yes, that's right. And so do you care if that photo 2468 01:56:44,680 --> 01:56:47,160 is a photo of a real thing or not? 2469 01:56:47,160 --> 01:56:49,240 Sometimes you really do, and sometimes you really don't. 2470 01:56:49,240 --> 01:56:49,920 Exactly. Yes. 2471 01:56:50,160 --> 01:56:51,880 And I think that's kind of interesting. 2472 01:56:51,880 --> 01:56:54,440 It's a sort of not so much generative search as generative 2473 01:56:54,440 --> 01:56:55,520 content. Exactly. 2474 01:56:55,520 --> 01:56:58,080 If you're decorating your apartment and you want a mood 2475 01:56:58,080 --> 01:57:01,480 board and you can specify some styles and you can say I like 2476 01:57:01,480 --> 01:57:04,920 this and this and this and this, and it gives you more and you 2477 01:57:04,920 --> 01:57:07,480 look and you say more, more like that or more like this. 2478 01:57:07,920 --> 01:57:11,560 It doesn't necessarily matter at all if those images are real. 2479 01:57:11,800 --> 01:57:14,640 It does if like maybe you want to buy that table and it that 2480 01:57:14,640 --> 01:57:17,560 table doesn't exist, it just looks like those kinds of tables 2481 01:57:17,600 --> 01:57:19,800 or it looks like those kinds of chairs or whatever. 2482 01:57:20,320 --> 01:57:25,080 But if what you're looking for is no, I want to be more like 2483 01:57:25,080 --> 01:57:27,920 this or more like that, and you keep going until you get a mood 2484 01:57:27,920 --> 01:57:31,120 board of exactly what you want, doesn't may not matter at all 2485 01:57:31,160 --> 01:57:33,640 whether those images are real. That's right. 2486 01:57:33,640 --> 01:57:36,840 So, so if it's Pinterest on the one hand, then just inspiration 2487 01:57:36,840 --> 01:57:38,640 or what have you. But if it is. 2488 01:57:38,640 --> 01:57:41,880 If it's shoppable, then maybe it does, unless you it's then 2489 01:57:41,880 --> 01:57:44,160 there's an extreme case here, which is they'll just send that 2490 01:57:44,160 --> 01:57:45,640 dress to she and she will make it for you. 2491 01:57:45,640 --> 01:57:47,160 That's right. Or, or let's say, you know, 2492 01:57:47,160 --> 01:57:49,840 there's some a photo of a fire somewhere, right? 2493 01:57:50,160 --> 01:57:53,280 And quite a lot of times people will post photos of fires and 2494 01:57:53,280 --> 01:57:56,280 it's from like some like a concrete example, the Brazilian 2495 01:57:56,280 --> 01:57:59,360 fires from a few years ago. There's like a fake photo like 2496 01:57:59,360 --> 01:58:04,600 from that that Macron tweeted out because he he was told it 2497 01:58:04,600 --> 01:58:05,880 was a photo of the Brazilian fires. 2498 01:58:06,280 --> 01:58:09,240 But someone was able to show that it was actually like a like 2499 01:58:09,240 --> 01:58:11,600 AI think it was like a Reuters image or something, but from a 2500 01:58:11,600 --> 01:58:14,000 photographer who died years ago. Yeah, it wasn't that image. 2501 01:58:14,240 --> 01:58:16,240 Well, this is The funny thing about people complaining about 2502 01:58:16,240 --> 01:58:17,440 deep fakes. It's like we don't. 2503 01:58:18,360 --> 01:58:20,440 The problem isn't the picture, the problem is the label. 2504 01:58:20,440 --> 01:58:23,320 This the label exactly that's right so so The thing is that 2505 01:58:23,560 --> 01:58:27,440 with with crypto you can do what I call chain of custody right 2506 01:58:27,440 --> 01:58:32,200 block chain of custody where you can have a camera and by the way 2507 01:58:32,200 --> 01:58:35,560 this is also important in scientific work as well there's 2508 01:58:35,560 --> 01:58:38,440 this huge replication crisis with all these labs and data and 2509 01:58:38,440 --> 01:58:39,280 you. Know fake the data. 2510 01:58:39,440 --> 01:58:41,440 Yeah, exactly or, or or something, right. 2511 01:58:41,720 --> 01:58:43,840 So you could have, you know, there's something called pre 2512 01:58:43,840 --> 01:58:46,560 registration of studies, where like if you're doing a study, 2513 01:58:46,560 --> 01:58:50,120 you have to describe in some places who you're doing it on 2514 01:58:50,120 --> 01:58:51,920 what you're doing. It's like monitored to make sure 2515 01:58:51,920 --> 01:58:53,720 that people report the results, whether they're positive or 2516 01:58:53,720 --> 01:58:56,240 negative, right? So let's say it's, you know, 2517 01:58:56,240 --> 01:59:00,440 it's a study or it's a camera. You can have like a like either 2518 01:59:00,480 --> 01:59:03,520 crypto software or hardware in there such that when the frames 2519 01:59:03,520 --> 01:59:08,120 of images are recorded, they're instantly hashed and put on 2520 01:59:08,120 --> 01:59:10,960 chain either directly or as a digest of some kind, right? 2521 01:59:11,640 --> 01:59:15,640 That basically is like tamper proofing such that before the 2522 01:59:15,640 --> 01:59:18,960 data is even like collected or analyzed, this Internet 2523 01:59:18,960 --> 01:59:20,240 connected thing is doing something. 2524 01:59:20,280 --> 01:59:22,360 Now it's possible maybe to hack the firmware and mess with that, 2525 01:59:22,720 --> 01:59:25,840 but it would be pretty hard to. Depending on how you do this, it 2526 01:59:25,840 --> 01:59:26,880 would be pretty hard to do that. Right. 2527 01:59:26,880 --> 01:59:29,760 You also have this on Google and so on in front of watermark. 2528 01:59:30,760 --> 01:59:34,080 Yeah, generated images. So the challenge is if the image 2529 01:59:34,080 --> 01:59:37,200 isn't water March that doesn't, that won't stop people believing 2530 01:59:37,200 --> 01:59:38,000 it. True. 2531 01:59:38,000 --> 01:59:41,200 That's right. But but I think over time this 2532 01:59:41,200 --> 01:59:44,720 type of stuff where it'll gain traction at first our crypto 2533 01:59:44,720 --> 01:59:48,200 oracles for prediction markets because if you're making a 2534 01:59:48,200 --> 01:59:50,320 financial decision, I don't know if you've seen that stuff. 2535 01:59:50,320 --> 01:59:51,840 Alex Tabrock has talked about this. 2536 01:59:52,920 --> 01:59:56,600 When people have money on the line, their partisanship reduces 2537 01:59:56,600 --> 01:59:58,520 and they actually get a different chip in their head. 2538 01:59:58,520 --> 01:59:59,960 Really. Like, is this true or not? 2539 01:59:59,960 --> 02:00:02,080 They're trying to dispassionately figure it out, 2540 02:00:02,080 --> 02:00:03,080 right? They're not just cheering my 2541 02:00:03,080 --> 02:00:06,600 tribe, your tribe, whatever. And the is this true chip? 2542 02:00:07,360 --> 02:00:09,160 It basically means, OK, I'm going to double click into this, 2543 02:00:09,160 --> 02:00:10,880 I'm going to verify this, I'm going to look at this. 2544 02:00:11,280 --> 02:00:12,720 And that's where like Oracle's come in. 2545 02:00:12,720 --> 02:00:14,600 They're like feeds of data that have some degree of 2546 02:00:14,600 --> 02:00:16,120 verification. And right now they're like 2547 02:00:16,360 --> 02:00:18,800 mostly price data, but people use it for weather data. 2548 02:00:18,800 --> 02:00:20,320 They use it for this, that and the other, right, All these 2549 02:00:20,320 --> 02:00:22,640 different feeds of information that people trade on. 2550 02:00:23,040 --> 02:00:26,840 And over time, I think those feeds, once you can guard price 2551 02:00:26,840 --> 02:00:29,960 data, weather data, you know, health data, etcetera, 2552 02:00:30,200 --> 02:00:31,920 eventually you can guard any kind of data. 2553 02:00:32,440 --> 02:00:35,520 And then now you've got like a chain of custody for data, like 2554 02:00:35,520 --> 02:00:40,480 the scientific data rough off of it anyway, why don't we we 2555 02:00:40,480 --> 02:00:42,680 should wrap, but this is actually awesome conversation. 2556 02:00:43,920 --> 02:00:45,520 Anything, what's your latest stuff? 2557 02:00:45,520 --> 02:00:46,680 What should people go and check out? 2558 02:00:46,680 --> 02:00:48,720 Anything. Well, I've been publishing A 2559 02:00:48,720 --> 02:00:51,960 newsletter every week since 2013, and I always welcome all 2560 02:00:51,960 --> 02:00:54,600 subscribers to that. You should write Where is there 2561 02:00:54,600 --> 02:00:56,720 gonna be a Bendick book? Google, Benedict Evans. 2562 02:00:56,800 --> 02:00:59,200 My parents had good SEO. Book is interesting. 2563 02:00:59,200 --> 02:01:01,720 I've had publishers approach me every now and then about doing a 2564 02:01:01,720 --> 02:01:03,080 book. I have to work out what it would 2565 02:01:03,080 --> 02:01:05,000 actually be and why it would be worth reading. 2566 02:01:05,080 --> 02:01:07,440 Honestly, if you just, I don't know, maybe a history of tech, 2567 02:01:07,640 --> 02:01:10,000 like because all your slide decks are very good, right? 2568 02:01:10,520 --> 02:01:14,480 And there's one of the things I learned from, you know, my 2569 02:01:14,480 --> 02:01:17,120 friend novel, like the Navalomanac, right, that sold 2570 02:01:17,120 --> 02:01:19,280 1,000,000 copies. Why did it sell million copies? 2571 02:01:19,640 --> 02:01:21,240 I was surprised, but he was surprised by that. 2572 02:01:21,560 --> 02:01:26,600 It was Eric Jorgensen went and curated novel's old content and 2573 02:01:26,600 --> 02:01:28,680 turned into a book. And I was really surprised. 2574 02:01:28,680 --> 02:01:30,760 I was like, wait a second, isn't that all available on Twitter 2575 02:01:30,760 --> 02:01:32,560 for free already? Didn't people already see it? 2576 02:01:33,160 --> 02:01:35,960 They did. However, if you say what is the 2577 02:01:35,960 --> 02:01:39,960 one work that represents like the best of nobles thought over 2578 02:01:39,960 --> 02:01:43,960 years, you know, just to see his latest tweets is not the entry 2579 02:01:43,960 --> 02:01:46,520 point for that. You want to kind of collect all 2580 02:01:46,520 --> 02:01:50,080 of them, sort them, filter them, organize them thematically, 2581 02:01:50,080 --> 02:01:51,120 style them, and so on and so forth. 2582 02:01:51,520 --> 02:01:52,960 And I think you could have a pretty good book. 2583 02:01:53,640 --> 02:01:55,480 If you do that, let me know. Well, that's one thing on the on 2584 02:01:55,480 --> 02:01:57,960 the list And yes, the other thing is I, I used to do an 2585 02:01:57,960 --> 02:02:00,320 annual presentation. I've now shifted my cadence. 2586 02:02:00,320 --> 02:02:03,600 So I did a, a new AI presentation last month that I 2587 02:02:03,600 --> 02:02:05,800 published, which I, I was just in town to present. 2588 02:02:06,200 --> 02:02:09,560 And then I will do another one in the autumn, the fall for 2589 02:02:09,560 --> 02:02:11,240 American listeners. Great. 2590 02:02:11,240 --> 02:02:15,120 On a sort of e-commerce advertising, marketing brand, 2591 02:02:15,160 --> 02:02:18,560 like all the other stuff that's being transformed by AI right 2592 02:02:18,560 --> 02:02:22,240 now. And in general, what do I do? 2593 02:02:22,240 --> 02:02:25,680 I try and work out what's going on and how to explain it and how 2594 02:02:25,680 --> 02:02:27,760 I can explain it. And then I go and do 2595 02:02:27,760 --> 02:02:30,520 presentations and speak at events and talk to companies and 2596 02:02:30,520 --> 02:02:32,640 I do slides for money. Well. 2597 02:02:32,680 --> 02:02:34,680 That is similar. I do a lot of slides too. 2598 02:02:34,720 --> 02:02:38,440 I do a lot of speak. So you know, I've mentioned the 2599 02:02:38,440 --> 02:02:40,280 cloud communities thing and materializing those cloud 2600 02:02:40,280 --> 02:02:41,720 communities. So that's what I'm working on 2601 02:02:41,720 --> 02:02:45,040 atns.com like network school. So if people are interested in 2602 02:02:45,040 --> 02:02:46,720 this kind of stuff, we we talk about that there. 2603 02:02:46,720 --> 02:02:51,040 So subscribe to bendixnewsletter@isthatbenedictevans.com. 2604 02:02:51,280 --> 02:02:54,040 Ben, Ben Dash. Evansben-evans.com OK, great. 2605 02:02:54,480 --> 02:02:57,000 And and then if you want to check out network school, come 2606 02:02:57,000 --> 02:02:58,040 to ns.com. Sure, there you. 2607 02:02:58,040 --> 02:03:01,520 Gobenedictevans.com is another Benedict Evans, no, who is a 2608 02:03:01,520 --> 02:03:04,120 photographer really. And so my profile picture is 2609 02:03:04,120 --> 02:03:06,120 taken by him because I used to get his e-mail. 2610 02:03:06,120 --> 02:03:07,760 This is an obviously, this is. Obviously this is a. 2611 02:03:08,000 --> 02:03:10,520 In this case, there's a contact form on my website. 2612 02:03:10,520 --> 02:03:12,480 And he won't sell it to you. And I redesigned it. 2613 02:03:12,480 --> 02:03:14,920 I don't even didn't ask. I redesigned my website recently 2614 02:03:14,920 --> 02:03:17,920 so it's clear of who I am. But it was quite generic. 2615 02:03:18,120 --> 02:03:20,480 And people would go to the contact form and they would say, 2616 02:03:20,680 --> 02:03:23,920 hey Benedict, we really liked your work photographing Harvey 2617 02:03:23,920 --> 02:03:25,600 Keitel. Would you like to go to Mexico 2618 02:03:25,600 --> 02:03:27,840 next week and take pictures of Robert De Niro? 2619 02:03:27,920 --> 02:03:29,720 And I would look at it that's. So funny. 2620 02:03:29,840 --> 02:03:31,480 Forward. Well, you know it's funny. 2621 02:03:32,120 --> 02:03:35,240 You know, it's funny, there's actually probably as maybe even 2622 02:03:35,240 --> 02:03:38,200 more biology stream of Austins that are, but because there's 2623 02:03:38,200 --> 02:03:41,840 like 12 people last I checked in like the SF Bay Area alone with 2624 02:03:41,840 --> 02:03:45,280 my, my first and last name, you know, so just I, I I feel your 2625 02:03:45,280 --> 02:03:46,920 pain. OK. 2626 02:03:47,080 --> 02:03:48,880 Well, this is great. Really great seeing you in a 2627 02:03:48,880 --> 02:03:50,080 while. And we should do this more. 2628 02:03:50,280 --> 02:03:50,600 Yeah. Great. 2629 02:03:50,920 --> 02:03:51,280 Thank you, Sir.