1 00:00:19,012 --> 00:00:22,492 S1: All right. Welcome to episode 486. This is Daniel Missler. 2 00:00:24,172 --> 00:00:26,612 S1: All right. A lot of stuff to share this week. 3 00:00:26,772 --> 00:00:31,052 S1: Let's see here. Um, okay, so I had a debate 4 00:00:31,052 --> 00:00:34,692 S1: with Marcus Hutchins about I, I don't know if you've 5 00:00:34,692 --> 00:00:37,652 S1: seen him talking about AI online, but he basically, uh, 6 00:00:38,212 --> 00:00:40,812 S1: him and a number of my friends, uh, Christopher Hoff 7 00:00:40,852 --> 00:00:44,452 S1: being one of them, they basically take opportunity whenever they 8 00:00:44,452 --> 00:00:47,892 S1: can to to basically point out how stupid it is 9 00:00:47,932 --> 00:00:50,772 S1: and how it makes tons of mistakes and how it's 10 00:00:50,772 --> 00:00:55,132 S1: basically all folly and, uh, how everyone's overestimating it. And 11 00:00:55,132 --> 00:00:56,932 S1: the whole the whole thing is going to come crashing 12 00:00:56,932 --> 00:00:59,292 S1: down and a bunch of scams, all for the purpose 13 00:00:59,292 --> 00:01:01,492 S1: of making money. And it's all just hype and marketing. 14 00:01:01,852 --> 00:01:05,212 S1: So that's kind of their position. Uh, Marcus is a 15 00:01:05,212 --> 00:01:08,532 S1: particularly good at making these points in a very humorous way. 16 00:01:08,732 --> 00:01:10,932 S1: And of course, he has a huge following because he 17 00:01:10,932 --> 00:01:16,051 S1: is Marcus Hutchins, probably the most famous malware researcher that 18 00:01:16,052 --> 00:01:20,782 S1: we've ever known. And, uh, because I have the opposite view, 19 00:01:20,782 --> 00:01:23,181 S1: I wanted to have this debate. So we had the debate. 20 00:01:23,182 --> 00:01:25,342 S1: It's on YouTube now. I highly suggest you go check 21 00:01:25,342 --> 00:01:28,062 S1: it out. And of course, the link is in the newsletter. 22 00:01:28,462 --> 00:01:31,662 S1: The conversation went extremely well. It was very cordial. It 23 00:01:31,661 --> 00:01:36,461 S1: was very, um, substantive. I think we both made points 24 00:01:36,462 --> 00:01:41,022 S1: that the other one received and appreciated. So I think 25 00:01:41,021 --> 00:01:43,222 S1: it was overall just great. Now, a lot of the 26 00:01:43,222 --> 00:01:47,782 S1: feedback has been that we weren't actually disagreeing and that, uh, 27 00:01:48,902 --> 00:01:52,742 S1: both things could be true. And this makes me think 28 00:01:52,742 --> 00:01:55,702 S1: we went a little bit too hard on being nice. And, uh, 29 00:01:55,702 --> 00:01:58,302 S1: maybe for part two, which we're already talking about, talking 30 00:01:58,302 --> 00:02:01,142 S1: about doing maybe for part two, we need to sort 31 00:02:01,142 --> 00:02:03,942 S1: of tighten it up, um, sharpen the pencils a little 32 00:02:03,942 --> 00:02:06,462 S1: bit and just be like, hey, look, you know, I 33 00:02:06,462 --> 00:02:11,501 S1: think you're wrong about this particular point. Let's let's debate that. And, um, 34 00:02:11,502 --> 00:02:13,142 S1: I think we got there. I think we did that. 35 00:02:13,182 --> 00:02:15,222 S1: I mean, it was over a two hour conversation, so 36 00:02:15,342 --> 00:02:18,082 S1: hopefully you could take that away in the conversation, but 37 00:02:18,082 --> 00:02:20,802 S1: I think in part two I at least myself, are 38 00:02:20,802 --> 00:02:22,322 S1: going to try to be a little bit tighter with 39 00:02:22,322 --> 00:02:26,322 S1: my arguments and a little bit more direct. And, um, 40 00:02:26,362 --> 00:02:28,602 S1: I don't want to say confrontational, but confrontational in the 41 00:02:28,602 --> 00:02:32,122 S1: most cordial way possible to basically try to pin down 42 00:02:32,121 --> 00:02:34,681 S1: these arguments and say, no, you know, this is where 43 00:02:34,681 --> 00:02:36,802 S1: I actually do disagree with you. I think you're mistaken 44 00:02:36,802 --> 00:02:40,202 S1: about that. And I think that will just be more 45 00:02:40,202 --> 00:02:42,762 S1: useful to people who are actually trying to make up 46 00:02:42,762 --> 00:02:45,681 S1: their minds. And most importantly, I think it'd be more 47 00:02:45,722 --> 00:02:49,242 S1: useful to Marcus and his points and to myself and 48 00:02:49,242 --> 00:02:52,362 S1: my points of like, if he thinks I'm wrong about 49 00:02:52,362 --> 00:02:55,082 S1: this and I think he's wrong about this, we're ultimately 50 00:02:55,082 --> 00:02:57,802 S1: trying to provide a service to our audiences and to 51 00:02:57,841 --> 00:03:00,721 S1: audiences we haven't talked to yet. And, you know, future audiences. 52 00:03:01,202 --> 00:03:04,722 S1: We're trying to make that point. We're trying to, you know, um, 53 00:03:05,562 --> 00:03:07,802 S1: not just have a discussion, but but also try to 54 00:03:07,962 --> 00:03:11,122 S1: show that we're right about this discussion. And it doesn't 55 00:03:11,121 --> 00:03:12,722 S1: mean we're right about everything. It doesn't mean we've got 56 00:03:12,722 --> 00:03:15,281 S1: it all figured out. We don't. It's the future. You 57 00:03:15,332 --> 00:03:18,092 S1: can't predict it. But I think you see my point. 58 00:03:18,091 --> 00:03:20,132 S1: So that's what I'm going to try to do in 59 00:03:20,132 --> 00:03:22,772 S1: part two. And after that, I might do a specific 60 00:03:22,812 --> 00:03:26,292 S1: sort of argument breakdowns or something even more tactical than that. 61 00:03:26,532 --> 00:03:28,732 S1: But in the meantime, I recommend you go check out 62 00:03:28,732 --> 00:03:33,412 S1: the episode. It's quite good. All right. Cloud code definitely 63 00:03:33,412 --> 00:03:35,972 S1: my most holy crap. Are you serious? I tool that 64 00:03:35,972 --> 00:03:39,412 S1: I've ever used. And again, I talked about this. I've 65 00:03:39,452 --> 00:03:41,132 S1: been talking about this for a couple of weeks now. 66 00:03:41,612 --> 00:03:45,412 S1: I'm telling you, like Klein and Cursor, those all still 67 00:03:45,612 --> 00:03:49,212 S1: feel powerful. And I still use cursor sometimes. But here's 68 00:03:49,212 --> 00:03:52,172 S1: the distinction I. I feel like cursor is a tool, 69 00:03:53,012 --> 00:03:56,732 S1: whereas Klein. Well, and Klein is very similar, although I'm 70 00:03:56,732 --> 00:03:59,692 S1: not using Klein as much. I'm using cursor now, and 71 00:04:00,172 --> 00:04:02,412 S1: I would say that those are tools as opposed to 72 00:04:02,572 --> 00:04:05,892 S1: cloud code is more like a coworker. It's more like 73 00:04:05,892 --> 00:04:08,532 S1: another engineer that I'm working with, and I'm just kind 74 00:04:08,532 --> 00:04:10,772 S1: of like sending work over there, like like I'm a 75 00:04:10,772 --> 00:04:14,212 S1: senior engineer and they're a junior engineer, and I'm sending 76 00:04:14,352 --> 00:04:16,231 S1: them work, and they're going and doing tests for me, 77 00:04:16,232 --> 00:04:19,352 S1: and they're coming back and giving me a status report saying, 78 00:04:19,352 --> 00:04:22,912 S1: here's what I got done, right. It's really, really that good. 79 00:04:23,432 --> 00:04:25,752 S1: And I'm just blown away by it. And I can't 80 00:04:25,752 --> 00:04:31,552 S1: even imagine, like, what gets better from here. So I 81 00:04:31,552 --> 00:04:34,712 S1: talked about this metric metric. I think I mentioned it 82 00:04:34,712 --> 00:04:37,832 S1: in this week's show or the one previous, but it's 83 00:04:37,832 --> 00:04:41,472 S1: a metric for how long term of tasks I can 84 00:04:41,472 --> 00:04:44,991 S1: actually accomplish. I think it's a really, really cool metric metric. 85 00:04:45,032 --> 00:04:49,152 S1: I thought they might have been mispronouncing or misstating miter miter, 86 00:04:49,392 --> 00:04:53,671 S1: but no, it's metric. And it basically talks about, you know, 87 00:04:53,712 --> 00:04:57,192 S1: the fact that I struggles with long context problems with 88 00:04:57,192 --> 00:04:59,872 S1: large scope problems where it just gets confused and just 89 00:04:59,872 --> 00:05:02,912 S1: can't remember its place, can't remember all the different subtasks, 90 00:05:02,912 --> 00:05:06,352 S1: and it just kind of spirals into like chaos. And 91 00:05:06,351 --> 00:05:08,112 S1: that's a huge problem for AI, and it's kind of 92 00:05:08,152 --> 00:05:11,552 S1: the big bottleneck right now, along with context, which is 93 00:05:11,552 --> 00:05:13,722 S1: very much related to this, and along with working memory, 94 00:05:13,722 --> 00:05:17,202 S1: which is very much related to this. So metric, this metric, 95 00:05:17,322 --> 00:05:21,762 S1: I believe it's a pretty good capture of this overall 96 00:05:21,762 --> 00:05:25,042 S1: collection of problems that I currently has, and how good 97 00:05:25,041 --> 00:05:28,522 S1: of a job we're doing at addressing those. So all 98 00:05:28,562 --> 00:05:31,602 S1: that to say, I think Claude, is Claude code is 99 00:05:31,601 --> 00:05:35,041 S1: the best right now at actually dealing with this. So 100 00:05:35,762 --> 00:05:40,002 S1: something to keep an eye on. Highly recommend this conversation 101 00:05:40,002 --> 00:05:44,642 S1: with Arthur. Arthur Kroeber and Dwarkesh Patel on China. So, 102 00:05:44,682 --> 00:05:49,442 S1: Arthur Kroeber or Kroeber? Sorry, Arthur Kroeber. That's a little 103 00:05:49,442 --> 00:05:53,202 S1: bit difficult to say. So he's a China expert. The 104 00:05:53,202 --> 00:05:55,842 S1: the China expert that everyone's talking about right now, or 105 00:05:55,842 --> 00:05:58,362 S1: at least in my circles, like there's a lot of 106 00:05:58,362 --> 00:06:02,722 S1: video of, um, Peter Zeihan and he's just talking about 107 00:06:02,722 --> 00:06:05,242 S1: how China is so screwed and they, you know, their 108 00:06:05,282 --> 00:06:09,202 S1: population collapse and, you know, they just they're in really 109 00:06:09,202 --> 00:06:11,522 S1: bad shape and they're going to be done any year now. 110 00:06:12,022 --> 00:06:16,062 S1: And I believe he's so popular because everyone wants to 111 00:06:16,062 --> 00:06:17,822 S1: hear that. I definitely wanted to hear that. I was 112 00:06:17,822 --> 00:06:20,022 S1: definitely happy to hear it. And what it did was 113 00:06:20,062 --> 00:06:24,062 S1: kind of expose my bias because it's confirmation bias. It's 114 00:06:24,062 --> 00:06:27,062 S1: also good news. And those are both very infectious. They 115 00:06:27,101 --> 00:06:29,942 S1: kind of burrow into the brain. And I'm a little 116 00:06:29,981 --> 00:06:31,861 S1: bothered that it worked so well on me for so 117 00:06:31,862 --> 00:06:34,142 S1: long for like a couple of years where I'm just like, oh, 118 00:06:34,142 --> 00:06:37,262 S1: this guy's super bright and China is actually super screwed. Well, 119 00:06:37,662 --> 00:06:39,902 S1: it turns out if you ask a lot of people 120 00:06:39,902 --> 00:06:42,622 S1: and if you sort of dig around, a lot of 121 00:06:42,622 --> 00:06:46,702 S1: China experts and a lot of economists don't think Zion 122 00:06:46,702 --> 00:06:50,102 S1: is all that smart. They don't think he's all that 123 00:06:50,142 --> 00:06:54,782 S1: educated and knowledgeable about China or China's economy. And some 124 00:06:54,782 --> 00:06:56,382 S1: of that could be hate. And this is why I 125 00:06:56,382 --> 00:06:58,662 S1: was a little bit confused. And maybe I am still 126 00:06:58,662 --> 00:07:01,022 S1: a little bit confused. I'm open to the fact that 127 00:07:01,062 --> 00:07:03,422 S1: Zion actually still does know a lot of stuff. And 128 00:07:03,422 --> 00:07:05,582 S1: I think that is the case. The problem is, I've 129 00:07:05,582 --> 00:07:09,421 S1: seen him talk about cyber, and he basically said that, um, 130 00:07:09,822 --> 00:07:14,112 S1: the Facebook app is key, logging every single keystroke that 131 00:07:14,112 --> 00:07:16,992 S1: you do at all times, even if you don't have 132 00:07:16,992 --> 00:07:21,672 S1: it open and it's uploading all of those keystrokes. Everything 133 00:07:21,672 --> 00:07:25,032 S1: you're doing on your computer, up to Facebook at all times. 134 00:07:26,072 --> 00:07:27,712 S1: And the problem is, he said that with the same 135 00:07:27,712 --> 00:07:31,472 S1: exact conviction that he's talking about with China. And I'm like, what? 136 00:07:31,752 --> 00:07:34,312 S1: Wait a minute. So I have to go and review 137 00:07:34,312 --> 00:07:36,592 S1: all the all the dumb shit that he said about China, 138 00:07:36,592 --> 00:07:39,592 S1: which I have been absorbing. And I'm like, this guy, 139 00:07:39,592 --> 00:07:43,312 S1: is this lost about this? And it's not just some esoteric, 140 00:07:43,352 --> 00:07:46,912 S1: like one philosophical view versus another. He's talking about a 141 00:07:46,912 --> 00:07:49,712 S1: very tangible thing. Like, to me, this is like lying 142 00:07:49,712 --> 00:07:52,072 S1: about a stat or something. This is like he's claiming 143 00:07:52,072 --> 00:07:54,832 S1: that this is technically true when I know it not 144 00:07:54,832 --> 00:07:57,512 S1: to be true. And just just as a simple way 145 00:07:57,512 --> 00:08:00,672 S1: of like giving an argument why I know this not 146 00:08:00,672 --> 00:08:04,512 S1: to be true or believe very strongly, at least if 147 00:08:04,752 --> 00:08:09,392 S1: the Facebook app was up stealing and uploading the keystrokes 148 00:08:09,592 --> 00:08:12,092 S1: of all its users all the time. Do you not 149 00:08:12,092 --> 00:08:14,732 S1: think that this would be a huge story? Do you 150 00:08:14,772 --> 00:08:18,092 S1: not think that all the security researchers out there in 151 00:08:18,092 --> 00:08:21,532 S1: the world would not have noticed this? Right? Do you 152 00:08:21,532 --> 00:08:25,212 S1: think Facebook would have meta would have okayed this and 153 00:08:25,212 --> 00:08:27,612 S1: actually thought it was okay and it would never be caught? 154 00:08:28,172 --> 00:08:30,452 S1: Like that is just ridiculous. So he's claiming to have 155 00:08:30,452 --> 00:08:32,972 S1: this knowledge that no one else has and he's spreading it, 156 00:08:33,012 --> 00:08:37,412 S1: you know, and I'm just like, wow. Right. This is 157 00:08:37,412 --> 00:08:40,212 S1: where you have to really question things. So anyway, this 158 00:08:40,212 --> 00:08:43,852 S1: other person who came on the Dwarkesh podcast, this guy 159 00:08:44,012 --> 00:08:48,452 S1: Arthur Kroeber, is a lot more serious. You can just 160 00:08:48,452 --> 00:08:50,252 S1: instantly tell, like, he'll be like, well, I don't know 161 00:08:50,252 --> 00:08:52,532 S1: about that. I wouldn't go that far. Or, you know, 162 00:08:52,572 --> 00:08:57,572 S1: like he's not talking like an influencer where it's just like, yeah, 163 00:08:57,652 --> 00:09:00,252 S1: which is what zeihan it talks like, you know, if 164 00:09:00,292 --> 00:09:04,332 S1: you listen to it, all his statements are absolutes. Um, 165 00:09:05,412 --> 00:09:09,012 S1: that's not completely true, right? He, um, he sometimes says, oh, 166 00:09:09,052 --> 00:09:11,462 S1: we don't know for sure or whatever, but a lot 167 00:09:11,462 --> 00:09:15,742 S1: of his statements are absolutes. Um, and it's just like, 168 00:09:16,182 --> 00:09:19,021 S1: first of all, we don't know that much about China. Like, 169 00:09:19,542 --> 00:09:23,182 S1: the data is sometimes hard to know, right? Uh, especially 170 00:09:23,182 --> 00:09:25,062 S1: because we're going off of their claims or we're going 171 00:09:25,062 --> 00:09:27,742 S1: off of inference. We're trying to figure these things out. Um, 172 00:09:27,742 --> 00:09:30,942 S1: bottom line is, he's just way too sure and way 173 00:09:30,942 --> 00:09:35,702 S1: too confident about opinions that I'm not sure that that 174 00:09:35,702 --> 00:09:38,862 S1: he should have at that level of confidence. And, um, 175 00:09:38,902 --> 00:09:41,702 S1: it's great to hear someone on the Dwarkesh podcast. Notice 176 00:09:41,702 --> 00:09:43,822 S1: he didn't have Peter Zeihan on. He went and found 177 00:09:43,822 --> 00:09:47,342 S1: this guy who's a well respected economist, uh, and who 178 00:09:47,622 --> 00:09:49,382 S1: he covered so much ground. I mean, it was just 179 00:09:49,381 --> 00:09:53,062 S1: a fantastic conversation. And, uh, I would say the short 180 00:09:53,062 --> 00:09:55,982 S1: version is I came away thinking China is in much 181 00:09:55,982 --> 00:09:59,862 S1: better shape than Zeihan believes. And then a lot of 182 00:09:59,862 --> 00:10:03,182 S1: people believe, and it's more in line with what my 183 00:10:03,182 --> 00:10:07,102 S1: concerns were of where China might be. And essentially, Zeihan 184 00:10:07,102 --> 00:10:10,002 S1: was giving me some confidence that I was wrong, that 185 00:10:10,002 --> 00:10:12,722 S1: China wasn't actually in that good a position, but it 186 00:10:12,722 --> 00:10:15,802 S1: seemed obvious to me from like other economic markers that 187 00:10:15,802 --> 00:10:18,562 S1: they actually are thriving and they actually are, you know, 188 00:10:18,562 --> 00:10:22,522 S1: in a position to kind of dominate the US, uh, in, 189 00:10:23,322 --> 00:10:27,641 S1: in short order. So I would say, go listen to this. 190 00:10:27,642 --> 00:10:30,002 S1: If you care about China, go, go listen to this. 191 00:10:30,002 --> 00:10:32,881 S1: If you care about national security, economic security, the future 192 00:10:32,881 --> 00:10:36,602 S1: of the US, the future of China, this conversation is spectacular. 193 00:10:36,962 --> 00:10:39,962 S1: And it's partially because of how good the guest was, 194 00:10:39,962 --> 00:10:43,482 S1: but it's largely because Dwarkesh is so good at interviewing. 195 00:10:43,482 --> 00:10:48,762 S1: He is just very young, very smart, very curious. That's 196 00:10:48,761 --> 00:10:53,521 S1: the biggest thing is like incandescent curiosity. That's the thing 197 00:10:53,522 --> 00:10:55,682 S1: with Dwarkesh. He's just like, he can't wait to jump 198 00:10:55,682 --> 00:10:58,602 S1: in with his next question. He's moving the conversation along, 199 00:10:58,881 --> 00:11:01,322 S1: like his conversations with Tyler Cowen are some of my 200 00:11:01,322 --> 00:11:08,412 S1: favorite podcasts to listen to. Anyway, it's a fantastic episode. Okay. 201 00:11:08,452 --> 00:11:11,812 S1: Very fascinated by this. Uh, what do you notice, Psyop 202 00:11:11,812 --> 00:11:14,131 S1: that I keep seeing, especially on X, but it's on 203 00:11:14,131 --> 00:11:17,332 S1: lots of different social media, I would say. And what 204 00:11:17,332 --> 00:11:19,812 S1: it basically does is it shows you like the picture of, 205 00:11:19,852 --> 00:11:22,652 S1: like a nine over 11, uh, building falling. Right. One 206 00:11:22,652 --> 00:11:25,492 S1: of the twin towers falling. And it's like, watch it 207 00:11:25,492 --> 00:11:29,212 S1: again in slow motion. What did you notice? Does this 208 00:11:29,212 --> 00:11:31,892 S1: look normal to you? And what it's doing is like 209 00:11:31,932 --> 00:11:36,732 S1: seeding this this concept of like, question everything. You've you've 210 00:11:36,732 --> 00:11:43,212 S1: been lied to and it's like really, really pernicious. It's nasty. Um, the, 211 00:11:43,492 --> 00:11:46,612 S1: the other cool thing that it does from a marketing standpoint, 212 00:11:46,612 --> 00:11:48,172 S1: it requires you to watch the video all the way 213 00:11:48,172 --> 00:11:50,492 S1: through in multiple times. And I think a lot of 214 00:11:50,492 --> 00:11:52,412 S1: people do. That's why a lot of these things have 215 00:11:52,452 --> 00:11:55,332 S1: like millions upon millions of views. And maybe that's just 216 00:11:55,332 --> 00:11:59,612 S1: the propaganda people pushing it, but it implicitly asks you 217 00:11:59,612 --> 00:12:01,492 S1: to use your own special analysis skills to figure out 218 00:12:01,492 --> 00:12:05,012 S1: what's wrong with it. And gullible people love to believe 219 00:12:05,011 --> 00:12:07,712 S1: that they're experts in lots of different stuff. So they're like, oh, 220 00:12:07,752 --> 00:12:09,992 S1: you know, I'll use my special skills to figure out 221 00:12:09,992 --> 00:12:11,952 S1: what's wrong with the building. He's like, oh, it shouldn't 222 00:12:11,952 --> 00:12:14,192 S1: have fallen that fast. And like when there's a comment, 223 00:12:14,192 --> 00:12:16,352 S1: it's like, did it fall faster than it should have? 224 00:12:17,952 --> 00:12:20,072 S1: They will suddenly pop into their mind. Yeah. It did 225 00:12:20,072 --> 00:12:22,672 S1: fall faster. That's how that's not how fast buildings are 226 00:12:22,672 --> 00:12:25,752 S1: supposed to fall. Somehow they're now an expert in building 227 00:12:25,792 --> 00:12:29,872 S1: falling like velocities. Right? And this is just like conspiracy thinking. 228 00:12:29,912 --> 00:12:32,631 S1: This is what a lot of people have. Um, and 229 00:12:32,712 --> 00:12:34,631 S1: I would say everyone has it. It's a question is like, 230 00:12:34,631 --> 00:12:36,752 S1: how much do you have it in check? Right? And 231 00:12:36,752 --> 00:12:38,952 S1: then they paint the perfect picture in your brain without 232 00:12:38,952 --> 00:12:41,072 S1: actually saying anything. Sometimes they say it in the comment, 233 00:12:41,072 --> 00:12:45,032 S1: sometimes they hint at it, but ultimately. Right. The brain 234 00:12:45,032 --> 00:12:47,152 S1: watching tries to like, square the circle, coming up with 235 00:12:47,152 --> 00:12:50,592 S1: an explanation when there isn't one. Or there might not 236 00:12:50,592 --> 00:12:53,592 S1: actually be anything wrong with the video, but because they 237 00:12:53,592 --> 00:12:56,232 S1: told you there was something wrong with it, you still 238 00:12:56,272 --> 00:12:59,832 S1: are these gullible types who watch it leave feeling like, ah, 239 00:13:00,072 --> 00:13:02,152 S1: you know, they have been lying to us. That video 240 00:13:02,192 --> 00:13:05,072 S1: was crazy. I can't believe I saw that video. It's like, no, 241 00:13:05,202 --> 00:13:07,682 S1: it was just a regular video, just regular iPhone took 242 00:13:07,682 --> 00:13:11,161 S1: a picture and sent it up there. It's like it 243 00:13:11,162 --> 00:13:15,402 S1: just opens the opportunity for gullibility and conspiracy thinking like 244 00:13:15,642 --> 00:13:19,922 S1: I've never seen before. It's a brilliant tactic. So basically, 245 00:13:19,922 --> 00:13:22,642 S1: the conspiracy brain goes on special forces mission to solve 246 00:13:22,642 --> 00:13:26,041 S1: the puzzle, and then it's followed up with like over 247 00:13:26,042 --> 00:13:27,162 S1: the course of a day or a week or a 248 00:13:27,162 --> 00:13:30,762 S1: month or whatever, like hundreds more videos hitting their feet. 249 00:13:30,761 --> 00:13:32,122 S1: 1 or 2 of them will give you like an 250 00:13:32,122 --> 00:13:34,802 S1: explanation for this thing. It's like, oh, that's because they're 251 00:13:34,802 --> 00:13:37,082 S1: doing this. That's because of, you know, it's all a conspiracy. 252 00:13:37,082 --> 00:13:40,442 S1: That's because of this or whatever. And in maybe like 253 00:13:40,482 --> 00:13:43,082 S1: one out of the ten or something, they'll give you 254 00:13:43,082 --> 00:13:45,281 S1: a reason. But in the other nine it doesn't give 255 00:13:45,282 --> 00:13:47,002 S1: you a reason. It just asks you to think about 256 00:13:47,002 --> 00:13:50,442 S1: it and what the brain does, what, in my opinion, 257 00:13:50,482 --> 00:13:55,842 S1: these people's brains. And I'm putting that in quotes. What 258 00:13:55,842 --> 00:13:58,402 S1: they're doing is they're filling in the blank for the 259 00:13:58,402 --> 00:14:02,082 S1: other nine videos, right? They're like, I solved it. I 260 00:14:02,082 --> 00:14:05,702 S1: know the answer. Um, Um, I figured this out. This 261 00:14:05,702 --> 00:14:08,862 S1: is a conclusion that I've come to myself not realizing 262 00:14:08,862 --> 00:14:10,822 S1: it was either fed to them or they jumped to it, 263 00:14:10,822 --> 00:14:14,381 S1: and they're now sticking with it, right? And it's just like, 264 00:14:14,381 --> 00:14:19,702 S1: this is so nasty. And notice that the videos get 265 00:14:19,742 --> 00:14:22,662 S1: millions of views. Like, how are they getting so many views? 266 00:14:23,182 --> 00:14:25,982 S1: And so basically, like here are the narratives that I'm seeing. 267 00:14:26,022 --> 00:14:28,142 S1: And I'm starting to bookmark these things. I don't like 268 00:14:28,142 --> 00:14:31,662 S1: them because I don't want to like, um, encourage them. Uh, 269 00:14:31,662 --> 00:14:33,862 S1: but I bookmark them. So I could actually, uh, put 270 00:14:33,862 --> 00:14:37,062 S1: together some research on these accounts. But, uh, it's things 271 00:14:37,062 --> 00:14:39,102 S1: like whether it's changing, food is changing. The world is 272 00:14:39,102 --> 00:14:42,742 S1: basically different now in some inexplicable way, which you should 273 00:14:42,742 --> 00:14:45,342 S1: try to figure out. Uh, women aren't women anymore. Men 274 00:14:45,342 --> 00:14:47,742 S1: aren't men anymore. They used to be us. Used to 275 00:14:47,742 --> 00:14:50,342 S1: be on top. Now we're just a bunch of pussies 276 00:14:50,342 --> 00:14:53,462 S1: or whatever. Um, black people are nasty and violent. Nobody 277 00:14:53,462 --> 00:14:56,142 S1: likes them for good reason. Anything that's wrong with you, 278 00:14:56,142 --> 00:14:59,222 S1: it's from vaccines. US is the most racist country ever 279 00:14:59,222 --> 00:15:03,792 S1: with constant crimes against non-whites. So the stuff is designed 280 00:15:03,792 --> 00:15:07,112 S1: to anger you. It's not pointed towards or against any 281 00:15:07,112 --> 00:15:10,192 S1: particular group. What it's doing is finding all the trigger 282 00:15:10,192 --> 00:15:14,272 S1: issues and just blasting them right, with millions of views, 283 00:15:14,272 --> 00:15:17,072 S1: which that might be paid promotions, that might be the 284 00:15:17,072 --> 00:15:20,472 S1: fact that it's just viral and people are resharing or whatever. Um, 285 00:15:20,512 --> 00:15:23,792 S1: liberals and black people and gay people and LGBTQ people 286 00:15:23,792 --> 00:15:26,632 S1: have basically ruined the country. Russia is way nicer than 287 00:15:26,632 --> 00:15:28,712 S1: the US. It's full of pretty white girls and no 288 00:15:28,712 --> 00:15:31,352 S1: black people. You've been sold a lie that Russia is bad, 289 00:15:31,352 --> 00:15:33,632 S1: but people love it there. They actually love it there. 290 00:15:33,832 --> 00:15:36,992 S1: And most importantly, you are being lied to and you 291 00:15:36,992 --> 00:15:39,592 S1: should be angry about it. And this is the message, right? 292 00:15:39,632 --> 00:15:42,432 S1: It's like everywhere, but especially on X. And there are 293 00:15:42,432 --> 00:15:44,432 S1: like hundreds of these accounts. And if you click on 294 00:15:44,432 --> 00:15:48,192 S1: any of the accounts and you just scroll down, it's nonstop. 295 00:15:48,312 --> 00:15:52,192 S1: Just these types of posts and they are formulated with 296 00:15:52,232 --> 00:15:54,192 S1: like the same sort of quality and the same sort 297 00:15:54,192 --> 00:15:57,472 S1: of narrative hooks. It is completely obvious to me that 298 00:15:57,472 --> 00:16:00,032 S1: these are all part of campaigns. So the question is, 299 00:16:00,032 --> 00:16:02,532 S1: who's paying the money for these campaigns? Who is creating 300 00:16:02,532 --> 00:16:07,452 S1: this content? And my answer is it's absolutely propaganda. It 301 00:16:07,452 --> 00:16:10,172 S1: is enemies of the United States who are doing this. 302 00:16:10,292 --> 00:16:13,852 S1: Probably mostly Russia, but there could be others. It might 303 00:16:13,852 --> 00:16:15,812 S1: even be like internal groups who are just trying to 304 00:16:15,852 --> 00:16:19,012 S1: stoke anger for particular internal candidates. I have no idea. 305 00:16:19,012 --> 00:16:21,532 S1: I haven't done the research on exactly who these groups are, 306 00:16:22,412 --> 00:16:25,052 S1: but I think we know from the past that it's 307 00:16:25,052 --> 00:16:28,052 S1: likely to involve Russia to a decent amount, but they're 308 00:16:28,052 --> 00:16:31,372 S1: not the only people doing propaganda. And the bigger point 309 00:16:31,372 --> 00:16:34,572 S1: is that AI is also going to make this so 310 00:16:34,572 --> 00:16:36,692 S1: much easier. Like what happens when you. I can write 311 00:16:36,692 --> 00:16:38,932 S1: this stuff. I have no idea if I actually did 312 00:16:38,932 --> 00:16:40,812 S1: write a bunch of this stuff. Maybe that's why I'm 313 00:16:40,812 --> 00:16:42,772 S1: seeing so much of it now, because it's more scalable. 314 00:16:43,012 --> 00:16:46,132 S1: But I would love for someone to go and take 315 00:16:46,132 --> 00:16:47,972 S1: a look at all these different accounts and actually figure 316 00:16:47,972 --> 00:16:50,052 S1: out who they are. See if there's any, you know, 317 00:16:50,092 --> 00:16:53,452 S1: Osint like, uh, clues in there that we can, we 318 00:16:53,452 --> 00:16:55,212 S1: can use to track them down. I want to know 319 00:16:55,212 --> 00:16:56,612 S1: which countries are doing this. I want to know how 320 00:16:56,612 --> 00:16:58,452 S1: much they're spending. I want to know how they're getting 321 00:16:58,452 --> 00:17:02,382 S1: so many views. I think it's fascinating. just fascinating to 322 00:17:02,382 --> 00:17:06,702 S1: see how propaganda works. And um, one other caveat I 323 00:17:06,702 --> 00:17:08,742 S1: want to mention here is that I'm saying this as 324 00:17:08,742 --> 00:17:12,342 S1: an ex-military person who was an Intel like a little bit. 325 00:17:12,382 --> 00:17:15,302 S1: I basically spent a long time studying influence operations in 326 00:17:15,302 --> 00:17:17,542 S1: the Army. When I was assigned to S-2 for my 327 00:17:17,542 --> 00:17:22,182 S1: battalion in the 101st airborne. And does that make me 328 00:17:22,222 --> 00:17:26,982 S1: like an absolute expert in influence operations or propaganda? Not really. 329 00:17:26,982 --> 00:17:28,542 S1: I've read a whole bunch of books on it. I 330 00:17:28,582 --> 00:17:31,902 S1: studied it when I was in the military. Um, I 331 00:17:31,902 --> 00:17:36,782 S1: was sort of assigned to doing counterterrorism type influence stuff 332 00:17:36,862 --> 00:17:39,182 S1: a little bit when I was actually in the military, 333 00:17:39,422 --> 00:17:42,341 S1: but I'm not like a trained Intel person from the military. 334 00:17:42,342 --> 00:17:46,182 S1: I was, uh, airborne infantry, so I was assigned to 335 00:17:46,222 --> 00:17:48,901 S1: it because I seemed to have a talent for it. 336 00:17:48,902 --> 00:17:51,341 S1: But it wasn't like I'm not formally trained in it 337 00:17:51,782 --> 00:17:54,022 S1: or whatever. So it just just keep in mind, I 338 00:17:54,022 --> 00:17:57,102 S1: read a whole bunch of books. Doesn't mean I'm an expert. Um, 339 00:17:57,142 --> 00:17:59,022 S1: what I'm trying to do is to just get you 340 00:17:59,022 --> 00:18:02,282 S1: to see my perspective and how I'm seeing this very 341 00:18:02,282 --> 00:18:05,161 S1: clearly as being part of campaigns. And everyone who I 342 00:18:05,162 --> 00:18:08,402 S1: know who actually is trained in this stuff. Um, as 343 00:18:08,402 --> 00:18:11,722 S1: much or more than I am. They also agree other 344 00:18:11,722 --> 00:18:15,002 S1: people trained in this way also agree that this is 345 00:18:15,162 --> 00:18:20,242 S1: propaganda being sent against the United States. From who? Who knows? Right. 346 00:18:20,242 --> 00:18:22,522 S1: You have to do that research. Uh, but but I 347 00:18:22,522 --> 00:18:24,681 S1: would say just watch for it. Um. Oh, and I've 348 00:18:24,682 --> 00:18:27,802 S1: got some questions here. Have you noticed any of your 349 00:18:27,802 --> 00:18:31,401 S1: friends feeling these feelings of, like, America is screwed? Oh, 350 00:18:31,402 --> 00:18:35,282 S1: I hate that group. That group ruined everything. Um, on 351 00:18:35,282 --> 00:18:39,322 S1: X particularly, but also just in, uh, social media in general. 352 00:18:39,362 --> 00:18:42,562 S1: Have you noticed yourself feeling more angry towards other groups 353 00:18:42,561 --> 00:18:45,962 S1: or towards, um, just like the past used to be 354 00:18:45,962 --> 00:18:48,881 S1: better and feeling like, yeah, Russia had it figured out 355 00:18:48,882 --> 00:18:51,482 S1: because Russia seemed super nice or whatever. And are you 356 00:18:51,522 --> 00:18:53,722 S1: consuming a whole bunch of social media? All this to 357 00:18:53,762 --> 00:18:59,052 S1: basically ask, are you feeling yourself being influenced in this way? right? 358 00:18:59,092 --> 00:19:01,091 S1: Because that's what I have to ask myself all the time, 359 00:19:01,132 --> 00:19:04,092 S1: because I'm constantly trying to clean myself of bias. Not 360 00:19:04,092 --> 00:19:06,812 S1: always effective. And I end up catching it like I've 361 00:19:06,852 --> 00:19:10,212 S1: talked about multiple times in the newsletter and on the podcast. 362 00:19:10,772 --> 00:19:14,012 S1: I'm constantly checking myself having bias. It's the other reason 363 00:19:14,012 --> 00:19:18,452 S1: I do the Telos thing is to constantly be able 364 00:19:18,452 --> 00:19:20,732 S1: to ask I and say, what mistakes am I making? 365 00:19:20,772 --> 00:19:24,052 S1: What biases do I have that I seem to be, uh, 366 00:19:24,492 --> 00:19:28,852 S1: you know, ignoring. So anyway, watch for it. Think about it. Uh, 367 00:19:28,852 --> 00:19:32,892 S1: something to keep in mind. Uh, traditional advertising is in 368 00:19:32,892 --> 00:19:36,612 S1: serious trouble. This is something, uh, just incredible. There's some 369 00:19:36,652 --> 00:19:41,652 S1: ads of, like, um, the ones I particularly like, these 370 00:19:41,692 --> 00:19:47,012 S1: I ads, uh, from Vo3 from Google. They're basically Bigfoot ads. Uh, Yeti, 371 00:19:47,052 --> 00:19:52,332 S1: Bigfoot and Yeti ads. And they're as funny as, like, 372 00:19:52,372 --> 00:19:55,252 S1: Super Bowl ads, which they pay millions of dollars for. 373 00:19:55,412 --> 00:19:57,252 S1: And a number of groups have made these things, and 374 00:19:57,252 --> 00:19:59,302 S1: it does actually take some talent to make it because 375 00:19:59,302 --> 00:20:01,022 S1: you have to write the script, you have to write 376 00:20:01,022 --> 00:20:03,821 S1: the prompt. So I think there is comedic human talent 377 00:20:03,821 --> 00:20:06,982 S1: going into these. The question is how much and how 378 00:20:06,982 --> 00:20:10,702 S1: long will that last? But I find it really interesting 379 00:20:10,702 --> 00:20:14,381 S1: that we're seeing these things have like hundreds of millions 380 00:20:14,382 --> 00:20:19,942 S1: of views. Okay, these Bigfoot little videos which are sometimes 381 00:20:19,982 --> 00:20:22,782 S1: are selling things and sometimes they're just comedy skits or whatever, 382 00:20:22,821 --> 00:20:25,902 S1: hundreds of millions of views. And I think a lot 383 00:20:25,902 --> 00:20:28,581 S1: of advertising is going to move in this direction. There's 384 00:20:28,622 --> 00:20:31,422 S1: one selling like a night vision goggle, and I'm like, 385 00:20:31,782 --> 00:20:33,822 S1: I think I kind of want it. Like it didn't 386 00:20:33,821 --> 00:20:36,102 S1: say anything about the night vision goggle. It was just funny. 387 00:20:36,182 --> 00:20:37,702 S1: And I'm like, where do I click? Where do I 388 00:20:37,742 --> 00:20:41,942 S1: buy it? Also, I need a night vision goggle. Um, 389 00:20:42,382 --> 00:20:44,982 S1: I don't actually. So hopefully I won't go and click 390 00:20:44,982 --> 00:20:49,222 S1: that and buy something unnecessary. All right. Cybersecurity A top 391 00:20:49,222 --> 00:20:52,222 S1: hacker on hacker one is a fully automated AI agent 392 00:20:52,222 --> 00:20:55,262 S1: called called XPO. And it's actually this is a slight correction. 393 00:20:55,302 --> 00:20:59,082 S1: Got a little errata here. live. It's actually the top 394 00:20:59,082 --> 00:21:02,002 S1: hacker in the US, so there are higher scores globally, 395 00:21:02,561 --> 00:21:06,802 S1: but the highest US score is from a fully automated AI. 396 00:21:07,682 --> 00:21:10,161 S1: And that's beating out all the top hackers in the US. 397 00:21:10,722 --> 00:21:15,162 S1: That is ridiculous. Absolutely ridiculous. And this should not be possible, 398 00:21:15,722 --> 00:21:20,402 S1: given what Marcus believes about AI, for example, or other 399 00:21:20,402 --> 00:21:23,002 S1: people who don't believe that AI is intelligent, to be 400 00:21:23,002 --> 00:21:27,362 S1: able to go and research and find and illustrate and 401 00:21:27,402 --> 00:21:32,482 S1: actually submit a report that gets you results on one 402 00:21:32,482 --> 00:21:34,402 S1: of these platforms, you have to you have to show 403 00:21:34,402 --> 00:21:36,601 S1: that it's actually a real vulnerability, and it has to 404 00:21:36,602 --> 00:21:38,962 S1: not be a dupe. You have to be first, and 405 00:21:38,962 --> 00:21:41,242 S1: it has to be a significant issue. Right? We already 406 00:21:41,242 --> 00:21:43,682 S1: have a high standard for what gets accepted. You know, 407 00:21:43,722 --> 00:21:47,282 S1: Bugcrowd has this, Hackerone has this. So this is hacker one, 408 00:21:47,282 --> 00:21:52,922 S1: accepting these actual vulnerabilities from actual AI. And it's number 409 00:21:52,922 --> 00:21:57,372 S1: one in the entire United States. So anyone who thinks 410 00:21:57,652 --> 00:22:00,532 S1: you can't automate this stuff, it's not possible or it's 411 00:22:00,532 --> 00:22:02,132 S1: going to take years, it's going to take three years 412 00:22:02,132 --> 00:22:05,372 S1: or five years or ten years. Nope. It already happened. 413 00:22:05,612 --> 00:22:09,012 S1: It already happened. Number one hacker in the US. Uh, 414 00:22:09,012 --> 00:22:11,811 S1: for for bug bounties, for mostly web vulnerabilities. Keep that 415 00:22:11,811 --> 00:22:15,972 S1: in mind. Anthropic says multiple models resort to blackmail to 416 00:22:16,012 --> 00:22:20,732 S1: avoid shutdown. Russian hackers beat Gmail toufar with app specific 417 00:22:20,732 --> 00:22:25,851 S1: password social engineering. So this was, uh, intelligence group. Was this. Yeah. 418 00:22:25,892 --> 00:22:30,732 S1: Russian group. Apt29 trick targets into generating Gmail app specific 419 00:22:30,732 --> 00:22:35,172 S1: passwords by sending fake State Department PDFs, giving them instructions. 420 00:22:35,172 --> 00:22:39,012 S1: And they filled it out. Stargazer's trick Minecraft players into 421 00:22:39,012 --> 00:22:43,852 S1: installing malware through fake mods. 500 fake GitHub repositories to 422 00:22:43,892 --> 00:22:49,892 S1: distribute malware disguised as Minecraft mods. 16 billion credentials leak. 423 00:22:50,012 --> 00:22:52,492 S1: You've probably seen a million stories about it. It's complete garbage. 424 00:22:52,492 --> 00:22:56,312 S1: It's just concatenation. Combinations of previous breaches. I don't know 425 00:22:56,311 --> 00:23:01,752 S1: how so many big names like security outfits, media groups 426 00:23:01,752 --> 00:23:05,191 S1: like totally just I guess it was so exciting they 427 00:23:05,192 --> 00:23:09,351 S1: couldn't resist the the headline. Anyway. Quite sad. It's a 428 00:23:09,592 --> 00:23:13,552 S1: nothing burger. Minnesota shooter allegedly used data brokers to find 429 00:23:13,552 --> 00:23:17,032 S1: victim addresses. Uh, good time for a reminder that data 430 00:23:17,071 --> 00:23:19,592 S1: brokers are way nastier, at least in my opinion, than 431 00:23:19,592 --> 00:23:21,832 S1: the dark web, because they have more data, they're way 432 00:23:21,832 --> 00:23:26,392 S1: more organized, and they're actually legal. They're collecting and selling 433 00:23:26,392 --> 00:23:31,552 S1: this data legally, and they have way more data on 434 00:23:31,552 --> 00:23:33,992 S1: all of us, and it's way higher quality and it's 435 00:23:34,032 --> 00:23:37,512 S1: way better organized, like the dark web. And criminals have 436 00:23:37,512 --> 00:23:40,552 S1: nothing on this, and it's completely legitimate. But we don't 437 00:23:40,592 --> 00:23:45,112 S1: talk about it because it's legal. That's ridiculous. Aflac gets 438 00:23:45,112 --> 00:23:49,912 S1: hit by a scattered spider in ongoing insurance company attack wave. So, uh, yeah, 439 00:23:49,952 --> 00:23:53,152 S1: this group scatters spider. They kind of focus on areas, 440 00:23:53,152 --> 00:23:55,282 S1: and they moved off their previous one. And they've been 441 00:23:55,282 --> 00:23:58,242 S1: going after a lot of insurance companies. Apple and Google 442 00:23:58,242 --> 00:24:00,962 S1: keep selling Chinese VPNs that could spy on you. So 443 00:24:00,962 --> 00:24:03,881 S1: someone found a whole bunch in the Apple App Store 444 00:24:03,882 --> 00:24:10,002 S1: and in Google that were Chinese VPNs and associated with, uh, 445 00:24:10,442 --> 00:24:15,441 S1: not good groups, I believe. Chinese government. Um, I'm not 446 00:24:15,442 --> 00:24:18,842 S1: sure about that. It was associated with something clearly malicious, though. Um, 447 00:24:18,842 --> 00:24:20,962 S1: although I would just say if it's a Chinese company, 448 00:24:21,002 --> 00:24:25,922 S1: just kind of assume that, uh, national security OpenAI gets 449 00:24:25,922 --> 00:24:32,322 S1: $200 million defense contract to build AI tools. Yeah, Pentagon 450 00:24:32,321 --> 00:24:35,202 S1: gave them $200 million to develop AI for national security 451 00:24:35,202 --> 00:24:38,361 S1: and war fighting. I've said a lot on this before, 452 00:24:38,362 --> 00:24:40,002 S1: so I'm not going to say it again right now. 453 00:24:40,242 --> 00:24:43,002 S1: I'll talk about it again later. Uh, this is good 454 00:24:43,002 --> 00:24:47,282 S1: and potentially bad. China's military is using AI for intelligence operations. 455 00:24:47,282 --> 00:24:51,162 S1: Recording future recorded future found that China's People's Liberation Army 456 00:24:51,162 --> 00:24:54,462 S1: has moved beyond just talking about AI for intelligence is 457 00:24:54,502 --> 00:24:58,742 S1: actually procuring and deploying generative AI tools for military intelligence work. 458 00:24:59,022 --> 00:25:03,902 S1: They're using it for like, uh, processing satellite imagery, generating 459 00:25:03,902 --> 00:25:08,141 S1: intelligence reports that just basically like augmenting their military using 460 00:25:08,142 --> 00:25:13,982 S1: these tools, including OpenAI. Iran's using hacked security cameras to 461 00:25:14,022 --> 00:25:19,022 S1: guide missile strikes on Israel. Russia's deadliest Kiev attack this 462 00:25:19,022 --> 00:25:22,742 S1: year kills 15 as peace talks have stalled. Kind of 463 00:25:22,782 --> 00:25:25,662 S1: a kind of a lull and an advantage for Russia. 464 00:25:25,702 --> 00:25:29,702 S1: I think, uh, with the stuff happening in Iran. Former 465 00:25:29,702 --> 00:25:32,182 S1: Army sergeant shows how not to be a spy. So 466 00:25:32,182 --> 00:25:34,942 S1: he used Google to search a, quote, countries that don't 467 00:25:34,942 --> 00:25:39,822 S1: extradite to us and, uh, emailed China directly, uh, to 468 00:25:39,862 --> 00:25:45,861 S1: share this information. Uh, he's sharing this, uh, classified information, uh, to, 469 00:25:45,902 --> 00:25:47,622 S1: to send it to China, but he's using his own 470 00:25:47,622 --> 00:25:52,672 S1: personal email accounts. Yeah. Uh. All right. I don't know 471 00:25:52,672 --> 00:25:55,712 S1: why that's as funny as it seems to be. Taiwan 472 00:25:55,712 --> 00:26:00,712 S1: gets Ukraine tested drone software to counter China. ex-CIA analyst 473 00:26:00,752 --> 00:26:05,032 S1: gets 37 months for leaking Israel's Iran attack plans. A 474 00:26:05,071 --> 00:26:10,071 S1: CF William Rodman photographed top secret documents about Israel's plans 475 00:26:10,071 --> 00:26:12,792 S1: to attack Iran, edited them to hide their source, and 476 00:26:12,792 --> 00:26:17,552 S1: then shared them with people before the documents went went 477 00:26:17,712 --> 00:26:23,631 S1: viral on telegram. Deep tech allegedly using shell companies to 478 00:26:23,672 --> 00:26:29,032 S1: support the Chinese military. US pressures Vietnam to remove Chinese 479 00:26:29,032 --> 00:26:35,111 S1: components from tech manufacturing. I meter okay, that's the one 480 00:26:35,112 --> 00:26:37,232 S1: I was talking about. Better metric for AI and agents. 481 00:26:37,232 --> 00:26:41,152 S1: It's basically yeah. Long term tasks I think is really smart. Uh, 482 00:26:41,152 --> 00:26:45,032 S1: Andrej Karpathy says prompts are software 3.0. So we did 483 00:26:45,032 --> 00:26:47,912 S1: a major talk two weeks ago and basically said that 484 00:26:47,912 --> 00:26:53,012 S1: software 3.0 is prompting, Prompting to an eye to have 485 00:26:53,012 --> 00:26:54,972 S1: it build the application for you. And he calls the 486 00:26:54,972 --> 00:27:00,532 S1: software 3.0. I love his really high level, you know, 487 00:27:01,172 --> 00:27:03,452 S1: thought leading type of talks, and this was definitely one 488 00:27:03,452 --> 00:27:05,532 S1: of them and absolutely loved it. I love his other 489 00:27:05,532 --> 00:27:09,331 S1: talks too, but I love talks that are about, you know, 490 00:27:09,412 --> 00:27:12,732 S1: thinking about things differently. He also said, we're in the 491 00:27:12,732 --> 00:27:15,052 S1: decade of agents and not the year of agents, which 492 00:27:15,052 --> 00:27:18,372 S1: I think is really smart. Sam Altman says GPT five 493 00:27:18,412 --> 00:27:22,412 S1: is coming in the summer. Google uses YouTube videos to 494 00:27:22,452 --> 00:27:27,012 S1: train Vo three without creator consent. Google also released a 495 00:27:27,012 --> 00:27:30,932 S1: stable version of Gemini 2.5, which includes new models with 496 00:27:30,932 --> 00:27:33,651 S1: different sizes and capabilities. This one is not in here now, 497 00:27:33,652 --> 00:27:36,811 S1: but they also just released a command line tool, which 498 00:27:36,811 --> 00:27:38,372 S1: I haven't messed with. I'm going to mess with later 499 00:27:38,372 --> 00:27:42,292 S1: today as soon as I'm done recording actually. And um, yeah, 500 00:27:42,292 --> 00:27:44,772 S1: it looks like maybe a competitor to Cloud Code, although 501 00:27:44,772 --> 00:27:47,252 S1: it looks like it might be more of a general agent. 502 00:27:47,252 --> 00:27:49,572 S1: So we're starting to move towards like the Da stuff 503 00:27:49,982 --> 00:27:52,622 S1: that I've been talking about forever. Wicks bought an AI 504 00:27:52,662 --> 00:27:54,902 S1: that lets anyone build software by chatting, so they bought 505 00:27:54,902 --> 00:27:57,821 S1: a company that is an AI tool for building software. 506 00:27:58,222 --> 00:28:01,502 S1: So this is starting to become a thing. Oh, the 507 00:28:01,502 --> 00:28:03,022 S1: other thing I want to mention right now is there 508 00:28:03,022 --> 00:28:05,581 S1: is a company called This is Major. There's a company 509 00:28:05,582 --> 00:28:09,141 S1: called Cluley, and it is doing very much like what 510 00:28:09,142 --> 00:28:12,022 S1: I wrote about in 2016, where it's like a digital assistant, 511 00:28:12,022 --> 00:28:15,141 S1: but it's not just taking tasks from you, it's actually 512 00:28:15,142 --> 00:28:18,462 S1: being more proactive. So it's actually recording your whole screen. 513 00:28:18,462 --> 00:28:21,342 S1: It's doing it's watching what you're doing. And you could 514 00:28:21,342 --> 00:28:24,022 S1: just actively ask it things. It could go and do 515 00:28:24,022 --> 00:28:27,502 S1: things on your behalf like more proactively. It is kind 516 00:28:27,502 --> 00:28:30,662 S1: of like you remember the Microsoft recall thing. It's kind 517 00:28:30,662 --> 00:28:33,542 S1: of like that. There's also Google has moved towards this. 518 00:28:33,582 --> 00:28:36,582 S1: Anthropic has moved towards this with like computer use. Everything 519 00:28:36,582 --> 00:28:39,222 S1: is moving all in the same direction, right? It's all 520 00:28:39,222 --> 00:28:43,421 S1: moving in the same direction of your Da is the 521 00:28:43,422 --> 00:28:45,862 S1: thing that understands everything about you. I just read my 522 00:28:45,862 --> 00:28:47,942 S1: book right before I came on this. I was talking 523 00:28:47,942 --> 00:28:54,322 S1: to Joseph Thacker, one of the best hackers in the world. And, um, yeah, 524 00:28:54,322 --> 00:28:55,562 S1: I was talking about I'm like, dude, you got to 525 00:28:55,562 --> 00:28:57,282 S1: finish reading this book. So I went and read it 526 00:28:57,282 --> 00:28:59,522 S1: real quick and it takes like 20 minutes to read 527 00:28:59,522 --> 00:29:02,202 S1: the whole thing. It's 17,000 words, but it goes really fast. 528 00:29:02,242 --> 00:29:04,722 S1: It's it's written pretty simply and it's online for free, right? 529 00:29:04,722 --> 00:29:08,802 S1: You don't have to go and buy anything because that's ridiculous. But, um, anyway, 530 00:29:08,842 --> 00:29:11,122 S1: it's all pushing towards what I was talking about in 531 00:29:11,162 --> 00:29:13,722 S1: that thing, and I've got multiple examples in there. It's 532 00:29:13,722 --> 00:29:17,242 S1: basically it knows everything about you and it's constantly advocating 533 00:29:17,242 --> 00:29:20,042 S1: for you on your behalf. It's constantly trying to adjust 534 00:29:20,042 --> 00:29:24,482 S1: the world around you using the API ification of everything. 535 00:29:24,482 --> 00:29:28,602 S1: Universal demonization. Right? Everything gets an API. This is what's 536 00:29:28,602 --> 00:29:32,082 S1: in the book. All companies get an API. I should 537 00:29:32,082 --> 00:29:33,962 S1: stop saying book. I should start saying it's in the 538 00:29:34,002 --> 00:29:38,362 S1: piece or it's in the content or whatever. Um, companies 539 00:29:38,362 --> 00:29:42,682 S1: get an API, people get APIs, objects get APIs, businesses 540 00:29:42,682 --> 00:29:47,282 S1: get APIs. Your Da speaks API, right? That was kind 541 00:29:47,282 --> 00:29:50,092 S1: of the core thing for the for this, for this concept, 542 00:29:50,092 --> 00:29:53,652 S1: for this book in 2016. Right. It's like your Da 543 00:29:53,812 --> 00:29:56,132 S1: speaks the language of these other things. Does that sound 544 00:29:56,132 --> 00:29:58,652 S1: like anything to you? Does that sound like agent to agent? 545 00:29:58,652 --> 00:30:02,772 S1: Does that sound like MCP? MCP is basically an implementation 546 00:30:03,052 --> 00:30:08,252 S1: of universal demonization. That's what it is. And I said 547 00:30:08,252 --> 00:30:11,212 S1: right there in the content, it's like, look, I have 548 00:30:11,212 --> 00:30:12,532 S1: no idea how this is going to work, but it's 549 00:30:12,532 --> 00:30:15,732 S1: probably going to be HTTP based because the predictions I'm 550 00:30:15,732 --> 00:30:18,892 S1: making there are about direction. It's stochastic. Right? It's not 551 00:30:18,892 --> 00:30:20,932 S1: about like who's going to win whatever. You can't predict 552 00:30:20,972 --> 00:30:26,332 S1: tech like that anyway. Cluley is heading directly in this direction. Um, 553 00:30:26,332 --> 00:30:28,931 S1: I'm wearing a pendant right now, which is recording everything 554 00:30:28,932 --> 00:30:31,532 S1: that would be input into my Da. My da's name 555 00:30:31,532 --> 00:30:34,332 S1: is Chi, by the way. Um, haven't fully spun him 556 00:30:34,332 --> 00:30:37,332 S1: up yet, but that's, uh, he's coming together. Starting to 557 00:30:37,332 --> 00:30:41,452 S1: bring these pieces together. My big challenge being in security. Um, 558 00:30:41,532 --> 00:30:43,292 S1: don't even need to be in security to figure this out. 559 00:30:43,292 --> 00:30:46,912 S1: But my big challenge is take this company Cluley, for example, 560 00:30:46,912 --> 00:30:48,872 S1: I don't know anything about them. I don't know anything 561 00:30:48,872 --> 00:30:51,072 S1: about the founder. I don't know anything about the team. 562 00:30:51,552 --> 00:30:54,792 S1: More importantly, I don't know anything about their security. So 563 00:30:54,792 --> 00:30:56,392 S1: do you think I'm going to have it recording all 564 00:30:56,392 --> 00:30:59,952 S1: of my screens with all of my microphones and uploading 565 00:30:59,952 --> 00:31:03,592 S1: all of that content to their servers so their eyes 566 00:31:03,632 --> 00:31:06,592 S1: can parse it? Right. What if I'm talking to somebody 567 00:31:06,592 --> 00:31:08,712 S1: over here about some government thing and someone else shares 568 00:31:08,712 --> 00:31:11,032 S1: with me? Hey, you know, we just got breached. Don't 569 00:31:11,032 --> 00:31:13,152 S1: mention it to anyone. Someone else is telling me about 570 00:31:13,152 --> 00:31:16,752 S1: their life, right? Um, I had a friend die recently. 571 00:31:16,792 --> 00:31:19,072 S1: What if I'm talking to my friend about, you know, 572 00:31:19,152 --> 00:31:21,392 S1: my other friend who's died? All of this is being 573 00:31:21,392 --> 00:31:24,952 S1: uploaded constantly to a third party. Not to Google, not 574 00:31:24,952 --> 00:31:27,472 S1: to Apple. Okay. Because. Because I could see doing that 575 00:31:27,472 --> 00:31:30,032 S1: with Google or Apple once they, you know, once they 576 00:31:30,032 --> 00:31:32,392 S1: sort it out, once Apple sorts out, you know, the 577 00:31:32,392 --> 00:31:36,232 S1: successor to Siri or whatever. Google they have rock solid security. 578 00:31:36,232 --> 00:31:38,752 S1: I'm not worried about them getting breached or them losing 579 00:31:38,752 --> 00:31:42,512 S1: the data or whatever. Other than small issues. But ultimately 580 00:31:42,512 --> 00:31:46,722 S1: those two, maybe even Amazon, maybe even Amazon. They also 581 00:31:46,722 --> 00:31:49,482 S1: have rock solid security in terms of like how many 582 00:31:49,482 --> 00:31:52,082 S1: billions of dollars they spend on it. The quality of 583 00:31:52,122 --> 00:31:54,882 S1: their team, the size of their team, etc.. Right. But 584 00:31:54,882 --> 00:31:59,161 S1: a third party, a new startup called Cluley and it's 585 00:31:59,322 --> 00:32:01,362 S1: all this AI hype is around it. How big is 586 00:32:01,362 --> 00:32:03,962 S1: their security team? How much are they thinking about this? 587 00:32:04,002 --> 00:32:05,642 S1: Not only that, but how good are they going to 588 00:32:05,642 --> 00:32:09,682 S1: be at defending against a halfway decent attacker, let alone 589 00:32:09,682 --> 00:32:11,242 S1: a state actor who wants to come and steal the 590 00:32:11,242 --> 00:32:13,922 S1: entire thing and see all this context that all these 591 00:32:13,922 --> 00:32:17,322 S1: people are spraying up into the cloud. Right? I'm very 592 00:32:17,562 --> 00:32:21,122 S1: forward leaning for tech and for AI, but do not 593 00:32:21,682 --> 00:32:24,482 S1: go in and record every single thing, every single text, 594 00:32:24,482 --> 00:32:29,162 S1: every single screenshot. This thing is screenshot Screenshotting your your 595 00:32:29,282 --> 00:32:32,842 S1: displays all the time, all the texture on there. You 596 00:32:32,842 --> 00:32:35,082 S1: see on there it could see the app that it's doing. 597 00:32:35,082 --> 00:32:37,802 S1: It could see your slack conversations. It sees all this right. 598 00:32:37,842 --> 00:32:43,442 S1: So you should be careful. You should be careful about 599 00:32:43,442 --> 00:32:46,142 S1: who you're sending this stuff to. I would say if 600 00:32:46,142 --> 00:32:48,142 S1: it's Google or Apple, which is not it's not out 601 00:32:48,142 --> 00:32:51,022 S1: yet for them, but with Microsoft. Microsoft is another one. 602 00:32:51,022 --> 00:32:53,542 S1: I wouldn't worry too much about sending that stuff to Microsoft. 603 00:32:53,862 --> 00:32:56,342 S1: I'm sure there'll be problems, but they are a legit 604 00:32:56,342 --> 00:33:00,022 S1: security team because they are a legit company. The startups 605 00:33:00,022 --> 00:33:01,462 S1: are what you have to worry about, and I'm not 606 00:33:01,462 --> 00:33:04,022 S1: saying there's anything wrong with Cluley security, I'm just saying 607 00:33:04,022 --> 00:33:08,902 S1: that they are a startup and they're probably not prioritizing security. Um, 608 00:33:08,902 --> 00:33:12,342 S1: even if they know how to write, um, show me 609 00:33:12,342 --> 00:33:14,582 S1: that they hire an awesome security team, maybe I'll feel 610 00:33:14,622 --> 00:33:18,382 S1: better about it, but there's still a startup. Um, so anyway, 611 00:33:19,382 --> 00:33:23,342 S1: all that is, um, bringing us towards where it's going, 612 00:33:23,382 --> 00:33:27,702 S1: which is? Your Da will be listening all the time. 613 00:33:27,702 --> 00:33:30,102 S1: It will have all of your context. My telos file. 614 00:33:30,102 --> 00:33:32,582 S1: That is part of what Chi has, right? Oh, by 615 00:33:32,582 --> 00:33:36,142 S1: the way, um, Network Truck just released a video about Telos. 616 00:33:36,142 --> 00:33:38,302 S1: Just came out, uh, a couple hours ago. You got 617 00:33:38,302 --> 00:33:40,462 S1: to go check that out. But that Telos file that 618 00:33:40,462 --> 00:33:42,542 S1: he talks about there, that I have, that he has 619 00:33:42,582 --> 00:33:45,272 S1: a whole bunch of my friends have. That telos file 620 00:33:45,272 --> 00:33:48,472 S1: is like my soul. This this thing is like it's traumas. 621 00:33:48,472 --> 00:33:51,752 S1: It's history. It's my journal. It's like my goals, my aspirations, 622 00:33:51,752 --> 00:33:54,352 S1: like my problems, my challenges, all this different stuff. Right? 623 00:33:55,192 --> 00:33:58,632 S1: All this is available to Chi. Chi knows this. Chi 624 00:33:58,672 --> 00:34:02,432 S1: also knows that I like, you know, prawn cocktail. It 625 00:34:02,432 --> 00:34:04,952 S1: also knows that I don't sit anywhere where there's like 626 00:34:04,952 --> 00:34:07,552 S1: smokers nearby. Right. So it knows all these different things, 627 00:34:07,592 --> 00:34:11,832 S1: knows preferences for food, you know, so it can help me, uh, 628 00:34:12,152 --> 00:34:14,392 S1: optimize my surroundings. It could order for me. It could 629 00:34:14,392 --> 00:34:16,552 S1: do all these things. How is it doing that? By 630 00:34:16,552 --> 00:34:21,192 S1: manipulating the APIs that it has access to for the 631 00:34:21,192 --> 00:34:23,672 S1: world around us. Right. If I go into a restaurant 632 00:34:23,672 --> 00:34:25,992 S1: and I can control the TV, well, I'm not going 633 00:34:25,992 --> 00:34:27,832 S1: to control the TV. I'm not going to ask them 634 00:34:27,832 --> 00:34:30,352 S1: to change the channel. No, Chi is going to change 635 00:34:30,352 --> 00:34:33,592 S1: the channel to table tennis. If I'm not overruled by 636 00:34:33,592 --> 00:34:37,072 S1: other people trying to have it change to something else, right? Why? 637 00:34:37,112 --> 00:34:39,872 S1: Because this is all available on media. This is part 638 00:34:39,872 --> 00:34:41,472 S1: of the API. This is part of the demon for 639 00:34:41,472 --> 00:34:45,012 S1: that restaurant. It's just an API key can submit to 640 00:34:45,052 --> 00:34:47,412 S1: it on my behalf. And this will be the case 641 00:34:47,412 --> 00:34:51,092 S1: for temperature, for sounds, for what's playing on the video, 642 00:34:51,212 --> 00:34:54,652 S1: for having doors open for you, you know, for having 643 00:34:54,692 --> 00:34:58,372 S1: materials ready for you and summaries created. Most importantly, Chi 644 00:34:58,412 --> 00:35:02,612 S1: will be reading a billion different stories and articles and 645 00:35:02,612 --> 00:35:07,412 S1: research papers and reports and raw data sets constantly. While 646 00:35:07,412 --> 00:35:10,012 S1: I'm in the middle of this sentence and while I'm sleeping, 647 00:35:10,012 --> 00:35:11,652 S1: and while I'm in the middle of another project and 648 00:35:11,652 --> 00:35:14,972 S1: doing some security work or whatever I'm doing, it's processing 649 00:35:14,972 --> 00:35:17,052 S1: all of that and getting it ready for me. The 650 00:35:17,052 --> 00:35:20,252 S1: other thing it's going to do is this dynamic content generation, 651 00:35:20,252 --> 00:35:22,452 S1: which is a blog post I put out who knows when. 652 00:35:22,452 --> 00:35:27,612 S1: Sometime last year Chi will be making me videos. Okay, 653 00:35:27,772 --> 00:35:32,692 S1: Chi will deepfake Chuck, who just made that full size video, 654 00:35:32,972 --> 00:35:35,132 S1: which is awesome, which you want to go watch the 655 00:35:35,372 --> 00:35:38,492 S1: the actual full thing, but Chi will make me a 656 00:35:38,492 --> 00:35:41,292 S1: 32nd version of that if I need it. And maybe 657 00:35:41,332 --> 00:35:44,062 S1: he just cuts up the video from from Chuck, but 658 00:35:44,062 --> 00:35:46,542 S1: maybe Kai makes me a whole new one with Chuck's 659 00:35:46,542 --> 00:35:50,462 S1: face or Richard Feynman's face on it, and he can 660 00:35:50,462 --> 00:35:52,622 S1: make a video for a blog post which didn't actually 661 00:35:52,622 --> 00:35:55,742 S1: have a video. Or he could turn a video into 662 00:35:55,742 --> 00:35:58,942 S1: a blog post because maybe somebody doesn't like to watch videos. 663 00:35:59,222 --> 00:36:01,342 S1: Or maybe it's audio only because I'm out on a 664 00:36:01,342 --> 00:36:02,902 S1: run and I don't want to look at anything and 665 00:36:02,902 --> 00:36:07,182 S1: I can't look at anything. Bottom line is, Kai, your 666 00:36:07,222 --> 00:36:10,982 S1: da becomes the medium to you. Um, and this this 667 00:36:10,982 --> 00:36:12,222 S1: is kind of a big thing. I don't know why 668 00:36:12,222 --> 00:36:15,182 S1: this is, uh, becoming a whole episode about this or whatever, 669 00:36:15,182 --> 00:36:21,262 S1: but SEO is dead because websites are dead. Advertising is 670 00:36:21,262 --> 00:36:24,502 S1: screwed because advertising no longer is going to be landing 671 00:36:24,662 --> 00:36:28,182 S1: on the human. The advertising is going to get blocked 672 00:36:28,182 --> 00:36:31,862 S1: by the Da. Largely, the Da is going to go 673 00:36:31,862 --> 00:36:35,462 S1: and collect things. It's going to be the inbox and the, uh, 674 00:36:35,902 --> 00:36:41,032 S1: the inbound filter to every principles life, to every user's life. 675 00:36:41,672 --> 00:36:44,552 S1: The Da will then determine what gets through right? And 676 00:36:44,552 --> 00:36:48,352 S1: so now the whole attack point for influence against a 677 00:36:48,352 --> 00:36:51,512 S1: person becomes their digital assistant. If you want to sell 678 00:36:51,552 --> 00:36:53,752 S1: to them, you sell to their digital assistant. If you 679 00:36:53,752 --> 00:36:55,912 S1: want to sell to them, you get into the models 680 00:36:55,912 --> 00:36:58,952 S1: so that when the digital assistant goes and does its 681 00:36:58,952 --> 00:37:02,512 S1: research for what the best you know, pillowcase cover is, 682 00:37:03,032 --> 00:37:05,472 S1: it's going to find your product, right. So this is 683 00:37:05,472 --> 00:37:08,792 S1: already starting. People are already trying to poison models or 684 00:37:08,832 --> 00:37:12,432 S1: influence models, however you want to put it. Um, because 685 00:37:12,712 --> 00:37:18,592 S1: they understand that marketing and influence and uh, website stats 686 00:37:18,632 --> 00:37:21,312 S1: and all of that, all of that is shifting to 687 00:37:21,352 --> 00:37:25,272 S1: go towards the Da. Um, all this content. Nobody's reading 688 00:37:25,272 --> 00:37:28,592 S1: this stuff, right? There's too many YouTube videos. There's too 689 00:37:28,592 --> 00:37:31,072 S1: many blogs. Like, we already figured that out a long 690 00:37:31,072 --> 00:37:34,192 S1: time ago. There's a limit to this, but your Da 691 00:37:34,392 --> 00:37:37,032 S1: can go read all that stuff. It can go process 692 00:37:37,032 --> 00:37:39,692 S1: all that stuff. And it could use other services which 693 00:37:39,692 --> 00:37:43,412 S1: are companies themselves, that are consuming all of that content 694 00:37:43,412 --> 00:37:48,012 S1: and producing rankings and directories, which kind of like an 695 00:37:48,012 --> 00:37:51,252 S1: old like Yahoo directory or whatever, but constantly updated. And 696 00:37:51,252 --> 00:37:53,411 S1: your Da will go review that and be like, oh, 697 00:37:53,492 --> 00:37:56,772 S1: turns out that's the best tactical flashlight for right now. Or, 698 00:37:56,932 --> 00:37:59,732 S1: you know, that's the best, um, bed cover or the 699 00:37:59,732 --> 00:38:03,252 S1: best eight sleep competitor or whatever it is, right? And 700 00:38:03,252 --> 00:38:07,452 S1: so what ends up happening is we just we in 701 00:38:07,452 --> 00:38:10,092 S1: our Das, our Da is going to be our most 702 00:38:10,092 --> 00:38:14,252 S1: important relationship, um, for a great many people, right? Because 703 00:38:14,252 --> 00:38:15,612 S1: it could be our coach. It could be our friend. 704 00:38:15,612 --> 00:38:18,412 S1: It'll be our therapist. And this is something we have 705 00:38:18,412 --> 00:38:21,172 S1: to watch out for because we want our best friend, 706 00:38:21,212 --> 00:38:24,532 S1: our best human friends to be our best friends. Right? 707 00:38:24,572 --> 00:38:27,492 S1: We want our our husband and our wife and our 708 00:38:27,492 --> 00:38:31,052 S1: kids to be our best friends. And we want that 709 00:38:31,052 --> 00:38:33,652 S1: to be the center. We want humanity to be the center, right? 710 00:38:34,212 --> 00:38:37,172 S1: So we really have to watch this, this gulf that 711 00:38:37,172 --> 00:38:40,382 S1: we're about to go into this chasm or whatever, where 712 00:38:40,382 --> 00:38:43,742 S1: it's like this thing being offered. This Da is so 713 00:38:43,742 --> 00:38:45,742 S1: powerful and it's so awesome, and mine's going to be 714 00:38:45,742 --> 00:38:48,222 S1: so awesome very soon. I'm going to have mine kick 715 00:38:48,222 --> 00:38:52,262 S1: in long before most people do, uh, because it's already starting. 716 00:38:52,742 --> 00:38:55,302 S1: But everyone's going to have this soon. And what we 717 00:38:55,302 --> 00:38:58,462 S1: have to guard against is like cutting out actual humans 718 00:38:59,022 --> 00:39:03,582 S1: in exchange for this thing that is more responsive. It's 719 00:39:03,582 --> 00:39:06,102 S1: nicer to us or whatever. And like, it's going to 720 00:39:06,102 --> 00:39:09,662 S1: have advantages over humans. That's fine. But we've got to 721 00:39:09,702 --> 00:39:13,262 S1: maintain this is my admonition to everyone. We've got to 722 00:39:13,262 --> 00:39:19,062 S1: maintain the mentality of all of this is to enhance humanity. Okay. 723 00:39:19,102 --> 00:39:24,022 S1: The goal of AI, um, relationships, the goal of AI therapy, 724 00:39:24,022 --> 00:39:27,542 S1: the goal of AI anything should be to take the 725 00:39:27,542 --> 00:39:30,582 S1: things that you love about human to human interaction and 726 00:39:30,582 --> 00:39:34,062 S1: human to human value exchange and make them better. Find 727 00:39:34,102 --> 00:39:36,622 S1: ways to enhance them, find ways to do them more often. 728 00:39:36,622 --> 00:39:39,522 S1: Find them ways to do to do them in a 729 00:39:39,522 --> 00:39:46,362 S1: more enriched and deeper way. Right. Augmentation versus replacement. Augmentation 730 00:39:46,362 --> 00:39:50,682 S1: versus replacement. This is the trick that we cannot ignore. 731 00:39:50,842 --> 00:39:55,962 S1: And another way that I really like maximize this is, um, 732 00:39:55,962 --> 00:39:58,642 S1: all this infrastructure that I'm building, which, by the way, 733 00:39:59,082 --> 00:40:03,322 S1: the name for mine is called Pi Pi personal AI infrastructure. 734 00:40:04,322 --> 00:40:09,282 S1: My pi that I'm building, my personal AI infrastructure stresses 735 00:40:09,642 --> 00:40:12,322 S1: when a piece of content comes through that I should 736 00:40:12,322 --> 00:40:16,682 S1: watch in a slow version. That's analog, that's old school, 737 00:40:16,682 --> 00:40:19,762 S1: that's human. Sit down with a moleskine and a space 738 00:40:19,762 --> 00:40:24,562 S1: pen and take notes. Slow it down or pause it. 739 00:40:24,922 --> 00:40:27,962 S1: Take the notes and think. Maybe call your friend. Maybe 740 00:40:28,002 --> 00:40:30,922 S1: have an actual, you know, voice conversation with a friend 741 00:40:30,922 --> 00:40:33,122 S1: about this cool idea. I do this all the time 742 00:40:33,122 --> 00:40:35,882 S1: with Joseph and with Jason and be like, hey, have 743 00:40:35,882 --> 00:40:37,452 S1: you seen this thing? Hey, let's talk about it, blah 744 00:40:37,452 --> 00:40:41,572 S1: blah blah. And you know, you don't want to be like, oh, well, 745 00:40:41,572 --> 00:40:44,612 S1: that's not good. You know, the the pen is not 746 00:40:44,612 --> 00:40:47,172 S1: as good as the typing. You know, the Moleskine is 747 00:40:47,172 --> 00:40:49,572 S1: not as good as, you know, putting it in my 748 00:40:49,572 --> 00:40:52,852 S1: telo's file. No, this is not a competition. The telos 749 00:40:52,852 --> 00:40:56,892 S1: file should not be replacing things. It should be augmenting. Right. 750 00:40:56,932 --> 00:41:00,132 S1: It's just a journal. It's just capturing your life. Right. 751 00:41:00,172 --> 00:41:01,772 S1: You could do that on a moschino. You could do 752 00:41:01,772 --> 00:41:05,212 S1: that in a telos file with vim and markdown. Right. 753 00:41:06,012 --> 00:41:08,852 S1: We've got to stay focused on what actually matters and 754 00:41:08,852 --> 00:41:12,172 S1: what we value and what we're trying to enhance versus 755 00:41:12,172 --> 00:41:15,652 S1: what is shiny and cool and awesome. Right? I'm a 756 00:41:15,652 --> 00:41:18,972 S1: security guy. I love security, I love hacking, I love 757 00:41:18,972 --> 00:41:21,332 S1: hacking tools. I love all those things, I love I, 758 00:41:21,332 --> 00:41:23,692 S1: I love AI agents, I love automation, I love all 759 00:41:23,732 --> 00:41:27,612 S1: that stuff. Those are tools. Those are secondary. They serve 760 00:41:27,612 --> 00:41:32,132 S1: the purpose of human right. So so I think this 761 00:41:32,132 --> 00:41:36,592 S1: is absolutely critical to to just keep in mind, for yourself. 762 00:41:36,592 --> 00:41:38,752 S1: But also when you're hearing me talk about how cool 763 00:41:38,752 --> 00:41:41,232 S1: these things are, I just realize that they are secondary. 764 00:41:41,952 --> 00:41:45,432 S1: And now I have to find my place. Because what, uh, 765 00:41:45,552 --> 00:41:53,712 S1: what a divergence there. Um, let's see here. All right. 766 00:41:53,752 --> 00:41:57,312 S1: Meta's AI can produce nearly half of Harry Potter one, 767 00:41:57,312 --> 00:41:59,672 S1: so we could actually give the actual quotes. I'm curious 768 00:41:59,672 --> 00:42:02,912 S1: how this is possible because it's not a database. Um, 769 00:42:03,112 --> 00:42:05,712 S1: I don't understand how it's producing exact quotes if it's 770 00:42:05,712 --> 00:42:09,672 S1: not a database. Um, yeah, I guess I'll have to 771 00:42:09,712 --> 00:42:12,672 S1: figure that out. It's kind of interesting. Um, and I'm 772 00:42:12,672 --> 00:42:15,632 S1: not sure I could do that deterministically either. I think 773 00:42:15,632 --> 00:42:17,672 S1: a lot of that might be just collapsing on the 774 00:42:17,672 --> 00:42:21,352 S1: most likely because it's seen so many samples. Um, but 775 00:42:21,352 --> 00:42:23,112 S1: I don't think it could do it, like over and 776 00:42:23,112 --> 00:42:26,352 S1: over and over, because again, it's not an actual lookup. 777 00:42:26,512 --> 00:42:29,432 S1: It is a generation, which is why you can have, uh, 778 00:42:29,512 --> 00:42:35,842 S1: hallucinations and, uh, non-deterministic results technology, Jassy told employees that 779 00:42:35,842 --> 00:42:38,882 S1: Amazon will need fewer people doing current jobs and more 780 00:42:38,882 --> 00:42:41,442 S1: doing other types of jobs as AI gets rolled out 781 00:42:41,442 --> 00:42:45,042 S1: across the company. We got to listen to these folks 782 00:42:45,042 --> 00:42:47,282 S1: when they tell us. For one, I'm telling you that 783 00:42:47,282 --> 00:42:50,522 S1: this is happening, that they're replacing jobs like whatever that matters. 784 00:42:50,522 --> 00:42:53,402 S1: I'm just a guy on the internet talking, but all 785 00:42:53,402 --> 00:42:56,002 S1: these CEOs, lots of them are basically saying, yeah, yeah, 786 00:42:56,002 --> 00:42:58,642 S1: we're getting rid of jobs. We're getting we're changing the 787 00:42:58,642 --> 00:43:01,282 S1: entire company. We're thinking AI first. And a lot of 788 00:43:01,282 --> 00:43:04,522 S1: people are like, no, no, they're not going to replace jobs. 789 00:43:04,562 --> 00:43:08,042 S1: I can't replace jobs and I won't replace jobs. It's 790 00:43:08,042 --> 00:43:11,042 S1: a big, uh, sham or whatever. And I'm like, well, 791 00:43:11,762 --> 00:43:14,482 S1: the people who actually control the jobs are telling us this, 792 00:43:14,882 --> 00:43:16,442 S1: and you got to watch out because some people are 793 00:43:16,442 --> 00:43:19,402 S1: also selling an AI product. So it's kind of like marketing, 794 00:43:19,482 --> 00:43:21,962 S1: like I felt that way with Salesforce. They're like, oh, 795 00:43:22,002 --> 00:43:24,482 S1: we sell a virtual, you know, AI employee. Oh, and 796 00:43:24,482 --> 00:43:26,642 S1: also we're getting rid of a bunch of jobs because 797 00:43:26,642 --> 00:43:28,962 S1: we could do it internally. Well, that sounds like marketing. 798 00:43:29,402 --> 00:43:33,902 S1: But when Amazon's actual CEO tells you this when? When 799 00:43:33,902 --> 00:43:37,661 S1: the head of AI companies like the head of anthropic. 800 00:43:37,702 --> 00:43:40,582 S1: I love that guy. Like he's way different than Sam. 801 00:43:40,582 --> 00:43:44,422 S1: And Sam's awesome too. But I love how human focused 802 00:43:44,462 --> 00:43:47,182 S1: that the head of anthropic is. I love how he's like, 803 00:43:47,422 --> 00:43:49,902 S1: people get ready, this is going to take jobs like 804 00:43:49,902 --> 00:43:52,262 S1: it's going to suck, but it's going to potentially be awesome. 805 00:43:52,262 --> 00:43:55,382 S1: But get ready. Why would you not listen to that? 806 00:43:55,422 --> 00:43:57,302 S1: What do you think he has access to? What do 807 00:43:57,302 --> 00:43:59,661 S1: you think he's looking at to make him think that? Right? 808 00:43:59,702 --> 00:44:02,862 S1: This is not a stupid person, right? So just when 809 00:44:02,862 --> 00:44:04,742 S1: you when you see these CEOs that have no reason 810 00:44:04,742 --> 00:44:06,782 S1: to lie to you. And I hate to use this analogy, 811 00:44:06,782 --> 00:44:09,462 S1: but when a terrorist tells you I hate everything about 812 00:44:09,462 --> 00:44:11,422 S1: you and your culture and I want to destroy you, 813 00:44:11,582 --> 00:44:14,262 S1: and if I were to blow myself up and you 814 00:44:14,262 --> 00:44:16,342 S1: were to die as a result, I would go to heaven. 815 00:44:17,542 --> 00:44:19,262 S1: The one thing you shouldn't do is be like, yeah, 816 00:44:19,302 --> 00:44:21,982 S1: I'm just not really sure he believes it. No, he 817 00:44:21,982 --> 00:44:24,862 S1: believes it. He believes it. And you should be careful 818 00:44:24,862 --> 00:44:28,542 S1: at the park as a result. Okay. And this is, like, 819 00:44:28,542 --> 00:44:32,102 S1: the worst analogy ever. It's like the worst analogy ever. 820 00:44:32,192 --> 00:44:35,392 S1: because I'm not saying right. The heads of these companies 821 00:44:35,992 --> 00:44:38,952 S1: or the people making these models are terrorists. I'm just 822 00:44:38,952 --> 00:44:43,592 S1: saying that when somebody is crazy about a belief and 823 00:44:43,592 --> 00:44:45,992 S1: they are like telling you from the depths of their 824 00:44:45,992 --> 00:44:47,872 S1: soul that they believe it and it is true and 825 00:44:47,872 --> 00:44:51,232 S1: it is going to happen, it doesn't mean it's going 826 00:44:51,232 --> 00:44:54,752 S1: to happen, but it does mean that you should believe them, right? 827 00:44:54,792 --> 00:44:58,472 S1: You should believe them that they believe it. Again, with 828 00:44:58,472 --> 00:45:03,152 S1: the caveat of of marketing. Meta's expanding beyond Ray-Ban with 829 00:45:03,192 --> 00:45:05,912 S1: Oakley glasses aimed at athletes. I'm getting a pair of these. 830 00:45:05,912 --> 00:45:08,032 S1: My first pair was fantastic. I can't wait for the 831 00:45:08,032 --> 00:45:11,312 S1: battery to get much better. Resolution is improving. Apple, where 832 00:45:11,352 --> 00:45:14,312 S1: are you? We're looking for you. Gartner says I will 833 00:45:14,312 --> 00:45:20,312 S1: handle half of all business decisions by 2027. Um, did 834 00:45:20,312 --> 00:45:23,552 S1: I miss something that is way too fast? Way too fast. 835 00:45:23,592 --> 00:45:27,032 S1: It doesn't matter if we get asi si, not agi 836 00:45:27,072 --> 00:45:32,172 S1: if we get ASI next month. Okay, so it's a 837 00:45:32,212 --> 00:45:36,252 S1: 400 IQ or a 4000 IQ. It's smarter than Albert 838 00:45:36,252 --> 00:45:38,852 S1: Einstein and it runs on your phone for $0.03. It 839 00:45:38,852 --> 00:45:42,772 S1: doesn't matter if everyone had access to that, it still 840 00:45:43,012 --> 00:45:46,812 S1: wouldn't have half of all business decisions in 2027. It 841 00:45:46,812 --> 00:45:50,252 S1: doesn't matter how smart the AI is. Companies move slow 842 00:45:50,252 --> 00:45:52,572 S1: for a reason. There is friction built into the system. 843 00:45:53,492 --> 00:45:56,852 S1: You can't reinvent yourself. You can't reinvent the leadership team. 844 00:45:56,852 --> 00:45:59,612 S1: You got regulations to go by. You have employees like 845 00:46:00,012 --> 00:46:03,172 S1: there's way more friction in these companies. And and I'm 846 00:46:03,172 --> 00:46:05,292 S1: saying this with the experience of like working in like 847 00:46:05,332 --> 00:46:08,652 S1: fortune ten companies for years at a time. You know, 848 00:46:08,852 --> 00:46:11,252 S1: being at Apple, being at, you know, anyway, I don't 849 00:46:11,252 --> 00:46:15,452 S1: need to describe all that. Uh, trust me, companies move slow. Uh, 850 00:46:15,492 --> 00:46:18,572 S1: there's nothing that's going to make something handle half of 851 00:46:18,572 --> 00:46:22,532 S1: all business decisions by 2027. And maybe that was just 852 00:46:22,572 --> 00:46:26,132 S1: a bad, uh, take on what was actually said, but 853 00:46:26,132 --> 00:46:29,612 S1: I don't think so. iOS 26 opens up AirDrop and 854 00:46:29,612 --> 00:46:32,342 S1: Airplay tech. Two third party apps this is all part of. 855 00:46:32,662 --> 00:46:35,342 S1: If you haven't been paying attention to why Apple is 856 00:46:35,342 --> 00:46:39,342 S1: opening everything up. It's because of antitrust. They are very 857 00:46:39,342 --> 00:46:41,902 S1: afraid of getting crushed by antitrust, so they are opening 858 00:46:41,902 --> 00:46:45,942 S1: up the entire ecosystem. That's why you're seeing this. And 859 00:46:46,062 --> 00:46:49,702 S1: iOS 26 finally lets you set custom ringtones the easy way. 860 00:46:49,702 --> 00:46:54,142 S1: So that's cool. Humans. It looks like we found all 861 00:46:54,142 --> 00:46:56,342 S1: the missing matter in the universe. Did you know that 862 00:46:56,342 --> 00:47:00,542 S1: we were missing, I think 80%. Was it 80%? I 863 00:47:00,542 --> 00:47:03,182 S1: can't remember the percentage. We were missing most of the 864 00:47:03,182 --> 00:47:07,142 S1: energy in the universe. Like the the calculations, the equations, 865 00:47:07,142 --> 00:47:11,222 S1: they do not add up because basically, I think this 866 00:47:11,222 --> 00:47:16,142 S1: is correct. Um, if we only had the amount of 867 00:47:16,142 --> 00:47:19,062 S1: matter that we have seen, then the universe would be 868 00:47:19,062 --> 00:47:22,942 S1: expanding a lot faster. It's not expanding faster, which means 869 00:47:22,942 --> 00:47:26,022 S1: there's a lot more gravity holding it in, which means 870 00:47:26,022 --> 00:47:28,742 S1: more matter holding it in. And we have been unable 871 00:47:28,742 --> 00:47:31,562 S1: to find that matter, but we just did. So they 872 00:47:31,562 --> 00:47:35,522 S1: use these pulses to find and essentially it is all 873 00:47:35,522 --> 00:47:40,602 S1: in this fog, this very light, very thin matter, fog 874 00:47:40,602 --> 00:47:45,922 S1: between galaxies. And that fog altogether makes up the other 875 00:47:45,922 --> 00:47:48,402 S1: large percentage that they were looking for. And this is 876 00:47:48,402 --> 00:47:51,362 S1: almost guaranteed to be a Nobel Prize in physics. This 877 00:47:51,362 --> 00:47:53,722 S1: is like one of the biggest findings ever in all 878 00:47:53,762 --> 00:47:59,002 S1: of astronomy. And huge, huge news. Also, same week Rubin 879 00:47:59,002 --> 00:48:02,442 S1: Observatory takes its first images. Absolutely stunning. I'm getting a 880 00:48:02,442 --> 00:48:05,682 S1: print of this thing. It's like the Hubble Deep field image, 881 00:48:05,682 --> 00:48:08,882 S1: if you remember that. Well, that just times a million. 882 00:48:08,922 --> 00:48:13,322 S1: This thing is insane. This Rubin telescope is different and 883 00:48:13,322 --> 00:48:16,282 S1: better than anything else that's ever been put out there. 884 00:48:16,322 --> 00:48:18,482 S1: I've got like, 20 links to this thing and like 885 00:48:18,482 --> 00:48:20,681 S1: all the images and the story of it. YouTube video. 886 00:48:20,722 --> 00:48:24,402 S1: You got to go check this out. Now, they seem 887 00:48:24,402 --> 00:48:28,042 S1: to have been brought online to investigate dark matter. I 888 00:48:28,132 --> 00:48:30,892 S1: wonder if they were trying to figure out the the, uh, 889 00:48:31,092 --> 00:48:34,012 S1: the secret of dark matter and actually find it. It 890 00:48:34,012 --> 00:48:35,932 S1: would really suck if that were the case, because they 891 00:48:35,972 --> 00:48:39,052 S1: are announcing their existence the same exact week that the 892 00:48:39,052 --> 00:48:43,692 S1: whole mystery and puzzle was solved, so. Holy crap. If 893 00:48:43,692 --> 00:48:45,612 S1: I'm right about that, the timing has to be like 894 00:48:45,652 --> 00:48:47,732 S1: a little bit of a letdown. But still, the images 895 00:48:47,732 --> 00:48:50,612 S1: are just insane. And they're doing it with like, multiple 896 00:48:50,612 --> 00:48:56,732 S1: mirrors and a super, super sensitive sensor. MIT researchers using 897 00:48:56,772 --> 00:49:00,692 S1: EEG monitoring to track brain activity in essay writers and 898 00:49:00,732 --> 00:49:04,412 S1: found ChatGPT users show the lowest brain engagement. This is 899 00:49:04,412 --> 00:49:11,452 S1: basically saying AI makes you way dumber. Um, I think 900 00:49:11,452 --> 00:49:13,172 S1: we talked about this last week. I don't think we 901 00:49:13,172 --> 00:49:15,252 S1: need to go into it too much more, but sample 902 00:49:15,252 --> 00:49:18,932 S1: size was fairly small, but it absolutely showed that if 903 00:49:18,932 --> 00:49:21,092 S1: you rely on AI too much, it will make you 904 00:49:21,132 --> 00:49:23,092 S1: way worse. The way the way I think about this, 905 00:49:23,092 --> 00:49:27,672 S1: the way I explain this is, is, I think, fairly elegant. 906 00:49:27,912 --> 00:49:30,152 S1: If you were the type of person like me who 907 00:49:30,152 --> 00:49:33,432 S1: is going to use Chi Chi's job, my DA's job 908 00:49:33,632 --> 00:49:36,632 S1: is going to be to constantly question me, to constantly 909 00:49:36,672 --> 00:49:41,032 S1: harass me, to constantly debate me, pull out my bias, 910 00:49:41,072 --> 00:49:45,312 S1: challenge me, force me to like, make a statement speech. 911 00:49:45,472 --> 00:49:48,712 S1: You know, um, like I'm on an NPR debate or 912 00:49:48,712 --> 00:49:52,312 S1: IAC squared debate. I've got to, like, perfectly defend my position. Okay, 913 00:49:52,352 --> 00:49:54,312 S1: now switch to the other side and make and make 914 00:49:54,312 --> 00:49:57,232 S1: their choice or make make their argument better than they 915 00:49:57,232 --> 00:50:02,592 S1: made it. Right. I'm teaching Chi to be an augmentation system, 916 00:50:03,272 --> 00:50:06,232 S1: a training system to make me better as a human, 917 00:50:06,232 --> 00:50:09,672 S1: to make me a better first principles thinker, and to 918 00:50:09,712 --> 00:50:14,272 S1: make me more independent of a thinker. Right. I is 919 00:50:14,272 --> 00:50:17,232 S1: going to enable that 100%. It's going to enable that 920 00:50:17,232 --> 00:50:20,472 S1: for millions upon millions of people, hopefully hundreds of millions 921 00:50:20,472 --> 00:50:24,232 S1: of people, hopefully billions of people. That's the good scenario. 922 00:50:24,272 --> 00:50:27,122 S1: The bad scenario is you just use it like a crutch. 923 00:50:27,122 --> 00:50:29,682 S1: And that's what I believe we're seeing with this study 924 00:50:29,962 --> 00:50:34,202 S1: where people got way stupider. Right? Oh, the most interesting 925 00:50:34,402 --> 00:50:39,002 S1: part about this. They actually saw brain activity decrease. Okay. 926 00:50:39,322 --> 00:50:43,322 S1: So they have physical evidence of, I would say, just 927 00:50:43,322 --> 00:50:46,002 S1: people just being dumber, like their brains are not as 928 00:50:46,042 --> 00:50:50,162 S1: active because they are offloading that stuff to an AI. 929 00:50:50,362 --> 00:50:53,442 S1: And that is super frightening to me. That is like really, 930 00:50:53,442 --> 00:50:56,482 S1: really a problem. Just keep in mind it doesn't mean 931 00:50:56,522 --> 00:51:00,122 S1: it's it's necessary. It doesn't mean it's guaranteed. Right? You 932 00:51:00,122 --> 00:51:03,162 S1: can actively defend against this and have it actually be 933 00:51:03,162 --> 00:51:08,342 S1: a net positive. It's a choice. Massive study analyzing 1.2 934 00:51:08,342 --> 00:51:11,762 S1: million dating app ratings found men rate women as attractive 935 00:51:11,802 --> 00:51:16,442 S1: 13 times more than women rate men. Calling it the 936 00:51:16,442 --> 00:51:22,762 S1: attractiveness gap. Women rate only 4.5% of men as attractive. 937 00:51:23,122 --> 00:51:28,662 S1: That's quite sad. What gets measured. I will automate. Uh, 938 00:51:28,702 --> 00:51:30,662 S1: you could check that one out. Microsoft data shows the 939 00:51:30,662 --> 00:51:33,902 S1: infinite workday has become the norm. Analyze trillions of data 940 00:51:33,902 --> 00:51:37,262 S1: points and found that work now stretches endlessly, with employees 941 00:51:37,262 --> 00:51:41,262 S1: getting interrupted every two minutes during business hours. Yeah. Yep. 942 00:51:41,302 --> 00:51:45,062 S1: That's why I got out of the game. 2024 Baby 943 00:51:45,102 --> 00:51:49,542 S1: name details. Extreme spelling is exploding. Ten of the top 944 00:51:49,542 --> 00:51:52,742 S1: 20 Rising Boys names have X's or Z's in them. 945 00:51:53,582 --> 00:51:58,782 S1: Finland is firing up the world's largest sand battery. Evidently 946 00:51:58,782 --> 00:52:01,142 S1: store a lot of heat and sand. This is like 947 00:52:01,142 --> 00:52:04,342 S1: tripping me out. Private equity has taken over and destroyed 948 00:52:04,342 --> 00:52:07,862 S1: much of America. Megan Greenwald's new book argues that private 949 00:52:07,862 --> 00:52:12,742 S1: equity firms have systematically bought up the entire industries and 950 00:52:12,782 --> 00:52:19,662 S1: optimized them into bankruptcy. New Covid variant, called NB 1.8.1, 951 00:52:19,702 --> 00:52:23,222 S1: jumped from 15% to 37% of US cases in just 952 00:52:23,222 --> 00:52:28,032 S1: two weeks. and people are worried about a potential summer surge. 953 00:52:28,232 --> 00:52:31,872 S1: What was interesting here is there's actually CDC data on this, 954 00:52:31,872 --> 00:52:34,512 S1: which I'm surprised and happy that there's still data coming 955 00:52:34,512 --> 00:52:38,192 S1: out of the CDC discovery. Again, I'm not going to 956 00:52:38,192 --> 00:52:41,431 S1: read all these. They can get quite long. And there 957 00:52:41,432 --> 00:52:44,552 S1: is the newsletter for you to go check out. Heartwarming. I, 958 00:52:45,432 --> 00:52:47,432 S1: I am going to read this one. So a teacher 959 00:52:47,432 --> 00:52:50,712 S1: shows her students eye images. This thing is insane. You 960 00:52:50,752 --> 00:52:53,712 S1: got to go watch this video. It shows them. So 961 00:52:53,752 --> 00:52:55,472 S1: somebody's like, I want to be a football star. I 962 00:52:55,472 --> 00:52:57,712 S1: want to be, you know, a marine biologist. I want 963 00:52:57,712 --> 00:53:01,632 S1: to be an astronaut. It shows them age accelerated versions 964 00:53:01,632 --> 00:53:04,552 S1: of them, like in an astronaut suit or whatever. And 965 00:53:04,552 --> 00:53:08,272 S1: just like the looks on these kids faces, it is so, like, 966 00:53:08,272 --> 00:53:13,272 S1: so wholesome, right? Just imagine seeing that it is possible, right? 967 00:53:13,312 --> 00:53:14,992 S1: Do you believe your parents when they say it? Okay, 968 00:53:15,032 --> 00:53:16,672 S1: a little bit. Do you believe your teacher when they 969 00:53:16,672 --> 00:53:18,672 S1: say it's possible? A little bit, but how does that 970 00:53:18,672 --> 00:53:21,752 S1: compare to actually seeing your face grown up in an 971 00:53:21,752 --> 00:53:24,412 S1: actual astronaut suit? It is extraordinary. You got to check 972 00:53:24,412 --> 00:53:28,452 S1: it out. Good workers are often bad at interviews. Hormozi 973 00:53:28,452 --> 00:53:32,052 S1: is highest ROI spins. There are tons of great discovery links. Again, 974 00:53:32,052 --> 00:53:33,812 S1: I don't want to just go through and read the title. 975 00:53:34,372 --> 00:53:37,532 S1: Go check it out. Okay, this is the end of 976 00:53:37,532 --> 00:53:39,692 S1: the standard edition of the podcast, which includes just the 977 00:53:39,692 --> 00:53:41,612 S1: news items for the week to get the rest of 978 00:53:41,612 --> 00:53:44,332 S1: the episode, which includes much more of my analysis, the 979 00:53:44,332 --> 00:53:47,732 S1: ideas section, and the weekly member essay. Please consider becoming 980 00:53:47,732 --> 00:53:50,052 S1: a member. As a member, you get access to all 981 00:53:50,052 --> 00:53:53,492 S1: sorts of stuff, most importantly, access to our extraordinary community 982 00:53:53,492 --> 00:53:56,612 S1: of over a thousand brilliant and kind people in industries 983 00:53:56,612 --> 00:54:00,292 S1: like cybersecurity, AI, and the humanities. You also get access 984 00:54:00,292 --> 00:54:03,612 S1: to the UL Book Club, dedicated member content and events, 985 00:54:03,612 --> 00:54:06,572 S1: and lots more. Plus, you'll get a dedicated podcast feed 986 00:54:06,572 --> 00:54:08,132 S1: that you can put into your client that gets you 987 00:54:08,132 --> 00:54:11,012 S1: the full member edition of the podcast that basically doesn't 988 00:54:11,012 --> 00:54:12,612 S1: have this in it, and just goes all the way 989 00:54:12,612 --> 00:54:15,332 S1: through with all the different sections. So to become a 990 00:54:15,332 --> 00:54:17,932 S1: member and get all that, just head over to Daniel Store.com. 991 00:54:18,892 --> 00:54:22,212 S1: That's Daniel Miessler, and we'll see you next time.