1 00:00:19,372 --> 00:00:23,492 S1: All right. Welcome to episode 494. Sorry, it's been a while. 2 00:00:23,852 --> 00:00:27,772 S1: Had some, uh, config issues been working on with the 3 00:00:27,772 --> 00:00:31,612 S1: home gear. I think I got them sorted out, and 4 00:00:31,612 --> 00:00:35,172 S1: we are ready to, uh, keep going. So my stress 5 00:00:35,172 --> 00:00:45,692 S1: and excitement levels continue to rise. Feeling basically unbelievably excited 6 00:00:45,692 --> 00:00:48,531 S1: about the future and like, what we could potentially do 7 00:00:50,052 --> 00:00:52,892 S1: and like all the different ways that I'm using AI 8 00:00:52,972 --> 00:00:55,132 S1: and I'm trying to get other people to use AI. 9 00:00:55,692 --> 00:00:59,212 S1: I'm doing probably, I don't know, five sessions a week 10 00:00:59,212 --> 00:01:03,732 S1: with various friends, showing them how I'm building my stack, 11 00:01:03,732 --> 00:01:07,932 S1: how I'm building out my dachi, and how they can 12 00:01:07,972 --> 00:01:11,691 S1: build their own for their own purposes. A lot of 13 00:01:11,692 --> 00:01:14,932 S1: the stuff that I talked about in the augmented course 14 00:01:14,972 --> 00:01:19,112 S1: really is basically starting to happen, um, instantiated inside of 15 00:01:19,112 --> 00:01:27,232 S1: cloud code. So it's like this extreme optimism and thought 16 00:01:27,232 --> 00:01:31,392 S1: about possibilities and like, everything I could possibly do and 17 00:01:31,472 --> 00:01:36,712 S1: help other people, do, you know, escaping from the drudgery of, like, 18 00:01:36,752 --> 00:01:40,472 S1: corporate work? A lot of corporate work, not all corporate work, 19 00:01:40,512 --> 00:01:45,392 S1: of course. And just like thinking about a better future 20 00:01:45,592 --> 00:01:48,352 S1: and trying to actively build that future, especially when I 21 00:01:48,352 --> 00:01:52,712 S1: start talking about some of the stuff from previous episodes, 22 00:01:54,672 --> 00:01:59,032 S1: podcast episodes of the newsletter. And when I want to 23 00:01:59,032 --> 00:02:03,192 S1: call out is, uh, Damon, which is this new service 24 00:02:03,192 --> 00:02:07,912 S1: that I've stood up. So it's damon.com. And what I'm 25 00:02:07,912 --> 00:02:12,632 S1: essentially doing is bringing back this idea that I've always wanted, 26 00:02:14,152 --> 00:02:17,252 S1: I think, even before the book stuff. I think this 27 00:02:17,252 --> 00:02:20,652 S1: is like early 2000. There was this one service in 28 00:02:20,652 --> 00:02:24,092 S1: the past called Friendfeed, and I had it kind of 29 00:02:24,132 --> 00:02:29,612 S1: like integrated in my website at the time. And what 30 00:02:29,612 --> 00:02:32,931 S1: it did was it it was able to follow what 31 00:02:32,931 --> 00:02:36,172 S1: you were doing on Reddit, on Facebook, whatever the current 32 00:02:36,172 --> 00:02:38,492 S1: services were at the time. I know Reddit was definitely 33 00:02:38,492 --> 00:02:43,691 S1: one of them. Uh, book feeds, RSS feeds, basically you 34 00:02:43,692 --> 00:02:47,212 S1: could subscribe to or you could publish like click the 35 00:02:47,212 --> 00:02:51,531 S1: options for all these different accounts and connect them. And 36 00:02:51,532 --> 00:02:54,452 S1: then when someone comes to your Friendfeed page, they would 37 00:02:54,452 --> 00:02:58,131 S1: see like the books that you're looking at, like the stuff. Oh, 38 00:02:58,172 --> 00:03:00,652 S1: it was upvotes as well. So we could see what 39 00:03:00,651 --> 00:03:03,452 S1: you upvoted on like Hacker News and what you upvoted 40 00:03:03,452 --> 00:03:06,892 S1: on Digg and what you upvoted on Reddit. So it 41 00:03:06,892 --> 00:03:11,812 S1: was like this live dashboard thing that was going on 42 00:03:11,812 --> 00:03:15,962 S1: with that person. It was just unbelievably awesome. It was 43 00:03:15,962 --> 00:03:19,922 S1: called Friendfeed, and I have always been trying to get 44 00:03:19,922 --> 00:03:25,121 S1: back to that thing. And so Damon is essentially just 45 00:03:25,121 --> 00:03:28,841 S1: like a more pure form of that because rather than 46 00:03:28,841 --> 00:03:31,522 S1: it being a third party service, the idea is that 47 00:03:31,522 --> 00:03:35,682 S1: each of us hosts our own just using a common protocol, 48 00:03:36,442 --> 00:03:40,242 S1: and we all just know how to call it, you know, 49 00:03:40,282 --> 00:03:43,242 S1: it's a common protocol. Therefore there's clients, there's servers, and 50 00:03:43,282 --> 00:03:48,842 S1: like it's very simple. But the idea is that world. 51 00:03:48,842 --> 00:03:50,882 S1: And it wouldn't just be humans that have this right. 52 00:03:50,922 --> 00:03:55,242 S1: It's it's also the businesses and the buildings and everything. Um, 53 00:03:55,242 --> 00:03:57,962 S1: so everything gets an API, but it's like this format. 54 00:03:58,482 --> 00:04:01,922 S1: But if I'm in a coffee shop and, you know, 55 00:04:01,962 --> 00:04:04,722 S1: I share people with things around me, they have their favorite. 56 00:04:04,762 --> 00:04:07,922 S1: Their top book is also their top fantasy book is 57 00:04:07,922 --> 00:04:10,402 S1: also name of the wind, which is my top fantasy book. 58 00:04:11,142 --> 00:04:14,742 S1: So it would show me that right. My digital assistant 59 00:04:14,742 --> 00:04:16,742 S1: would be able to say, hey, you know, check out 60 00:04:16,742 --> 00:04:19,302 S1: this person next to you. You should say hi. They 61 00:04:19,302 --> 00:04:22,502 S1: also love name of the wind. Boom. Human connection. So 62 00:04:22,502 --> 00:04:26,502 S1: everything comes back to the human connection facilitated by the tech. 63 00:04:27,262 --> 00:04:29,302 S1: So that was one of the things I talked about 64 00:04:29,302 --> 00:04:33,861 S1: a lot in a previous part of the newsletter. I've 65 00:04:33,862 --> 00:04:41,462 S1: been blogging like, absolutely crazy. Just absolutely crazy. Like I've 66 00:04:41,702 --> 00:04:47,902 S1: been blogging higher quality stuff. Um, more thorough stuff. More 67 00:04:48,062 --> 00:04:52,822 S1: well researched stuff. I've put out probably ten of these 68 00:04:52,822 --> 00:04:57,342 S1: in like the last month. It is unbelievable the amount 69 00:04:57,342 --> 00:04:59,742 S1: of output I've been able to do because of Chi. 70 00:05:00,142 --> 00:05:03,822 S1: All because of my AI stack. Because what I'm now doing, 71 00:05:03,822 --> 00:05:08,262 S1: my workflow now is to use Whisper Flow to record 72 00:05:08,262 --> 00:05:13,041 S1: into vim or record into Apple notes or record into whatever. 73 00:05:14,522 --> 00:05:16,642 S1: And then I just copy and paste that into Kai 74 00:05:16,642 --> 00:05:21,882 S1: and I say create blog and it knows to not 75 00:05:21,882 --> 00:05:25,282 S1: change my language. It doesn't write anything for me. It 76 00:05:25,282 --> 00:05:28,442 S1: just puts my language in there. Um, I actually do 77 00:05:28,442 --> 00:05:33,562 S1: have a rewrite command, and if I use that, it's 78 00:05:33,562 --> 00:05:37,042 S1: allowed to fix grammar and typos and like sentence structure, 79 00:05:37,041 --> 00:05:39,842 S1: but it's still not allowed to change the content. And 80 00:05:39,842 --> 00:05:44,402 S1: what happens is it still reads exactly like me because 81 00:05:44,402 --> 00:05:46,962 S1: all it did is really clean up. Like, should that 82 00:05:46,962 --> 00:05:51,121 S1: have been a semicolon or a or a dash or 83 00:05:51,162 --> 00:05:56,642 S1: a period or whatever? And then I could say, okay, enrich. Okay. 84 00:05:56,682 --> 00:05:59,322 S1: You know, in like the movies it says like enhance. Well, 85 00:05:59,362 --> 00:06:02,241 S1: now I can say enrich. And what that does is 86 00:06:02,242 --> 00:06:05,522 S1: it goes and gets links, right? Because when I'm narrating, 87 00:06:05,522 --> 00:06:09,542 S1: this is insane. When I am narrating, I can literally 88 00:06:09,862 --> 00:06:13,422 S1: in Whisper flow. I say things like, yeah, and um, 89 00:06:13,462 --> 00:06:16,822 S1: go get a quote from Viktor Frankl about this and 90 00:06:16,822 --> 00:06:21,502 S1: put it here. Right. So I'm just narrating. I'm like, okay, 91 00:06:21,502 --> 00:06:24,101 S1: we need some data to back this one up. Um, 92 00:06:24,101 --> 00:06:28,862 S1: I'm talking about, you know, tech layoffs. So find the 93 00:06:28,862 --> 00:06:33,142 S1: latest data on this from a reputable source and link 94 00:06:33,182 --> 00:06:36,981 S1: to that here. Also create a visualization using our standard 95 00:06:37,182 --> 00:06:42,061 S1: D3 visualization. And boom. So I press enter to that thing. 96 00:06:42,782 --> 00:06:49,421 S1: And this thing fires off. And like 10s later I 97 00:06:49,462 --> 00:06:55,782 S1: have a linked like graph with real data and the 98 00:06:55,782 --> 00:06:59,742 S1: data sources listed and the data sources listed inside the graph. 99 00:07:01,902 --> 00:07:06,102 S1: It's all annotated, all the notes are there, and all 100 00:07:06,101 --> 00:07:09,602 S1: I did was talk. All I did was talk. It 101 00:07:09,602 --> 00:07:14,042 S1: is unbelievable. And, you know, for the longer ones, I 102 00:07:14,042 --> 00:07:16,802 S1: end up going back in multiple phases. But I'm working 103 00:07:16,802 --> 00:07:19,562 S1: on that workflow. Ideally, I'd like to have it as 104 00:07:19,562 --> 00:07:22,922 S1: much of it as done as possible. What has not 105 00:07:22,922 --> 00:07:25,282 S1: worked for me is kind of like iterating back and forth. 106 00:07:25,322 --> 00:07:29,442 S1: It's no fun to do that. I prefer to either 107 00:07:29,442 --> 00:07:33,162 S1: just capture like violently and like this giant thing and 108 00:07:33,162 --> 00:07:37,202 S1: then just work from there, or try to have the 109 00:07:37,202 --> 00:07:46,202 S1: thought fully parsed out. Done inside of the note over to, 110 00:07:47,242 --> 00:07:49,722 S1: you know, Claude. And he just takes it and runs 111 00:07:49,722 --> 00:07:51,962 S1: with it and just knows I've got all these different 112 00:07:51,962 --> 00:07:56,082 S1: custom commands for adding links for cleanup. I have a 113 00:07:56,082 --> 00:08:00,242 S1: separate command for publishing, which does like editor type stuff. 114 00:08:00,682 --> 00:08:03,322 S1: So it makes sure none of the links are broken. 115 00:08:03,362 --> 00:08:07,502 S1: It it, uh. It'd make sure I don't have any 116 00:08:07,542 --> 00:08:10,822 S1: unsubstantiated claims that it thinks I should add data to 117 00:08:11,422 --> 00:08:13,582 S1: and stuff like that. It checks to make sure quotes 118 00:08:13,582 --> 00:08:16,582 S1: actually came from that person, right? So I'm checking basically 119 00:08:16,582 --> 00:08:22,102 S1: for like, hallucinations and mistakes that I've made. I've got 120 00:08:22,102 --> 00:08:24,782 S1: a piece of errata for the newsletter that just went out. 121 00:08:25,262 --> 00:08:28,862 S1: So the open source package that does the bug bounty stuff, 122 00:08:28,902 --> 00:08:32,662 S1: searching for vulnerabilities that was not released by XPO. And 123 00:08:32,662 --> 00:08:35,022 S1: I guess this is a good example of how I'm 124 00:08:35,022 --> 00:08:38,862 S1: not using AI for for a whole lot of stuff 125 00:08:39,022 --> 00:08:42,422 S1: on the newsletter. Um, I actually wrote that and it 126 00:08:42,422 --> 00:08:47,102 S1: was incorrect. Um, I had just read both articles and 127 00:08:47,102 --> 00:08:50,622 S1: somehow in my mind I must have like crossed. I 128 00:08:50,622 --> 00:08:54,062 S1: don't know, cross pollinated or something. But anyway, yeah, that'll 129 00:08:54,102 --> 00:08:56,302 S1: be in the errata in the next newsletter. But since 130 00:08:56,302 --> 00:08:59,222 S1: I'm doing a podcast and it's after it, I can 131 00:08:59,222 --> 00:09:03,291 S1: correct it now. Uh, okay. What was I talking about? Yeah, 132 00:09:03,492 --> 00:09:09,092 S1: basically just this workflow is completely insane and I'm just 133 00:09:09,131 --> 00:09:13,732 S1: ultra excited about it now. All that to say, I'm 134 00:09:13,772 --> 00:09:17,291 S1: still extremely worried. Like I talked about in that other post, 135 00:09:17,412 --> 00:09:23,012 S1: that other post was talking about the. How worried I 136 00:09:23,011 --> 00:09:28,771 S1: am about the, uh, the future and especially for a 137 00:09:28,772 --> 00:09:33,531 S1: whole lot of workers globally, basically knowledge workers. I'm really worried. 138 00:09:33,932 --> 00:09:37,212 S1: And I'm not just worried about them. I'm worried about 139 00:09:37,212 --> 00:09:41,852 S1: systemic effects of them losing so many jobs and there 140 00:09:41,852 --> 00:09:47,252 S1: being some sort of like threshold where if it crosses that, 141 00:09:47,252 --> 00:09:50,532 S1: then enough people lose their jobs, it creates enough panic, 142 00:09:50,932 --> 00:09:56,452 S1: it creates enough reduction of income into local businesses, so 143 00:09:56,492 --> 00:09:59,812 S1: enough businesses close. So it kind of turns into like 144 00:09:59,852 --> 00:10:04,632 S1: a like a mini pandemic effect, where it's like every 145 00:10:04,672 --> 00:10:07,112 S1: I mean, that one was like different because it's so 146 00:10:07,112 --> 00:10:09,832 S1: obvious why it's happening. But in this case, it would 147 00:10:09,832 --> 00:10:13,031 S1: be like, oh, we're having some sort of economic meltdown. 148 00:10:13,312 --> 00:10:15,392 S1: And then so the business is closed because no one's 149 00:10:15,392 --> 00:10:17,912 S1: coming to the businesses, because they're panicking, because they don't 150 00:10:17,912 --> 00:10:21,472 S1: have any money. And it's just like this massive cascade effect. 151 00:10:21,952 --> 00:10:24,832 S1: So I'm really, really worried about that. And I'm really 152 00:10:24,832 --> 00:10:27,792 S1: worried about I don't think a lot of these people 153 00:10:27,792 --> 00:10:30,712 S1: are going to be able to reskill. I think this 154 00:10:30,712 --> 00:10:34,272 S1: whole reskilling narrative is like something we tell ourselves to 155 00:10:34,312 --> 00:10:36,952 S1: make ourselves feel good. And I'm not sure it's like 156 00:10:36,952 --> 00:10:42,752 S1: how realistic it is. And it's it's like, really, I 157 00:10:42,752 --> 00:10:45,472 S1: don't know. It's really depressing to me, the whole thing. 158 00:10:45,832 --> 00:10:49,192 S1: So thinking about that side of it and how little 159 00:10:49,192 --> 00:10:54,271 S1: I can possibly do is, is just not good for 160 00:10:54,272 --> 00:10:56,992 S1: me to dwell on. So I literally dwell on it. 161 00:10:56,992 --> 00:11:01,932 S1: I sit inside of the the suck, you know? And 162 00:11:01,932 --> 00:11:03,772 S1: I sit inside there and I'm like, what can I do? 163 00:11:03,812 --> 00:11:07,531 S1: What can I do? Um, how bad is it going 164 00:11:07,532 --> 00:11:09,651 S1: to be? I have to, you know, share how bad 165 00:11:09,652 --> 00:11:12,012 S1: it's going to be or whatever. So then I write 166 00:11:12,011 --> 00:11:15,052 S1: a post like that which goes into all the detail, 167 00:11:15,052 --> 00:11:17,172 S1: and I get it all out, and I basically put 168 00:11:17,172 --> 00:11:21,052 S1: it there, and then I like, basically say, okay, that's it. 169 00:11:21,252 --> 00:11:24,452 S1: We're not thinking about this anymore right now that we're 170 00:11:24,452 --> 00:11:28,131 S1: setting that aside because the solution is always the same. 171 00:11:28,652 --> 00:11:34,771 S1: The solution is to expand the mind to, you know, 172 00:11:34,812 --> 00:11:37,692 S1: brush that off to to have the stoic mindset and 173 00:11:37,692 --> 00:11:44,412 S1: be like, okay, yep. That super sucks. Okay. Well. Anyway. Right. 174 00:11:44,452 --> 00:11:47,612 S1: I've been talking about this human 3.0 thing. I did 175 00:11:47,612 --> 00:11:49,771 S1: that previous post called The End of Work. I didn't 176 00:11:49,772 --> 00:11:53,651 S1: think this would be happening this fast when I wrote that, but. 177 00:11:56,772 --> 00:11:59,232 S1: In the end of work, I've already talked about it 178 00:11:59,272 --> 00:12:03,752 S1: in David Graeber's book Bullshit Jobs. Right. I've always seen 179 00:12:03,752 --> 00:12:06,952 S1: this current structure of like 9 to 5, and everyone, 180 00:12:07,032 --> 00:12:09,712 S1: or a vast majority of people can't stand to go 181 00:12:09,712 --> 00:12:12,472 S1: to work. It's cutting out 9 to 5 is the 182 00:12:12,472 --> 00:12:16,072 S1: most important hours of the day. And all they have 183 00:12:16,072 --> 00:12:18,232 S1: time to do is to get ready for the next 184 00:12:18,631 --> 00:12:21,192 S1: day of work, which they don't want to go to, 185 00:12:21,232 --> 00:12:24,992 S1: to have a bunch of stupid meetings. Like it? It's 186 00:12:24,992 --> 00:12:27,672 S1: a bad situation that we're in, right? It's also bad 187 00:12:27,672 --> 00:12:33,152 S1: that it's going to, you know, catastrophically come to an end. 188 00:12:33,192 --> 00:12:36,392 S1: I would say, or at least, you know, very badly 189 00:12:36,392 --> 00:12:40,472 S1: come to an end, I would say. Disruptively. Right. It's 190 00:12:40,472 --> 00:12:44,832 S1: going to be extremely disruptive to millions upon millions of people, 191 00:12:46,712 --> 00:12:49,072 S1: maybe billions. I don't know, it depends how you count, 192 00:12:49,072 --> 00:12:53,232 S1: but it's going to be massively disruptive to the global economy. 193 00:12:53,272 --> 00:12:58,772 S1: What is what is happening but going back to the positive, 194 00:12:58,772 --> 00:13:01,651 S1: I think that needs to happen. It needs to happen 195 00:13:01,652 --> 00:13:05,132 S1: that we get rid of this corporate structure. I mean, 196 00:13:05,172 --> 00:13:08,732 S1: corp okay, this all comes from the military, the hierarchy. 197 00:13:09,172 --> 00:13:14,692 S1: You work for Mr. Johnson. Mr. Johnson works for Mrs. Williams. Right. 198 00:13:14,692 --> 00:13:19,292 S1: And Mrs. Williams is trying to get her presentation together, uh, 199 00:13:19,292 --> 00:13:22,052 S1: to give to Mrs. Clark, because she's the one who 200 00:13:22,052 --> 00:13:24,212 S1: put all this together, and we're all just happy to 201 00:13:24,212 --> 00:13:31,772 S1: work for her. It's a very submissive, very hierarchical. Very 202 00:13:31,772 --> 00:13:37,292 S1: unhealthy structure to have. Most people in any company be 203 00:13:37,292 --> 00:13:42,532 S1: in this, like. Oh, yeah, I only contribute this little piece. Oh, 204 00:13:42,572 --> 00:13:45,652 S1: they don't ask me for ideas. I just help them 205 00:13:45,652 --> 00:13:48,132 S1: implement them. I'm just happy to be a part of it. 206 00:13:48,572 --> 00:13:54,372 S1: Like that's cool for 1950s or whatever. I mean, we 207 00:13:54,372 --> 00:13:56,872 S1: have to move through stages as humans. Well, it's time 208 00:13:56,872 --> 00:14:00,551 S1: to get out of this stage. We should all be founders. 209 00:14:00,552 --> 00:14:03,432 S1: We should all be CEOs, not not CEOs in the 210 00:14:03,432 --> 00:14:08,752 S1: old style. I'm using the CEO term for what it 211 00:14:08,752 --> 00:14:11,272 S1: is in the future, which won't be called a CEO. 212 00:14:11,312 --> 00:14:16,552 S1: We are all individual contributors to society and we should 213 00:14:16,552 --> 00:14:21,352 S1: be sharing with each other. We have companies. Yes, we have, 214 00:14:22,152 --> 00:14:27,512 S1: you know, products. Yes, we have offerings. Yes. And of course, 215 00:14:27,512 --> 00:14:32,112 S1: somebody gives us something in return for us giving them 216 00:14:32,112 --> 00:14:35,472 S1: that service. Right. So there's still a currency here, but 217 00:14:35,472 --> 00:14:38,792 S1: that currency won't be won't be based on this negative 218 00:14:38,792 --> 00:14:42,472 S1: infrastructure that we currently have, where it's all based on 219 00:14:42,472 --> 00:14:46,352 S1: your labor and it's all based on your, uh, you know, 220 00:14:46,392 --> 00:14:51,472 S1: your participation in this negative hierarchical structure that creates all 221 00:14:51,472 --> 00:14:54,422 S1: these bullshit jobs like David Graeber talks about in the book. 222 00:14:55,542 --> 00:14:59,462 S1: And the way I think about this is this is 223 00:14:59,462 --> 00:15:02,302 S1: just a natural progression. And this idea, I think I 224 00:15:02,342 --> 00:15:06,582 S1: got most from Harari and I can't remember which book 225 00:15:06,582 --> 00:15:08,262 S1: it was. I think it might have been the second 226 00:15:08,262 --> 00:15:15,902 S1: one after sapiens. So that basically what he talked about 227 00:15:15,902 --> 00:15:20,742 S1: is that, um, governments and systems of organization and systems 228 00:15:20,742 --> 00:15:25,302 S1: of government are basically they're time bound based on what 229 00:15:25,302 --> 00:15:29,102 S1: is needed for the society at that time. So it's like, 230 00:15:29,142 --> 00:15:34,022 S1: it's not that we shouldn't have had, uh, capitalism or 231 00:15:34,022 --> 00:15:39,302 S1: we shouldn't have communism or whatever. It's like those structures 232 00:15:39,702 --> 00:15:44,582 S1: are like best cases. They are the things that fit 233 00:15:44,742 --> 00:15:48,222 S1: where humanity is at the time. And it's a good 234 00:15:48,222 --> 00:15:52,482 S1: way of looking at like, okay, why didn't Iraq suddenly 235 00:15:52,482 --> 00:15:58,082 S1: become a democracy as soon as we got rid of Saddam? Well, 236 00:15:58,082 --> 00:16:02,202 S1: that's not where the population is ready for. Right. And 237 00:16:02,242 --> 00:16:07,322 S1: maybe there's an argument that, like, we are no longer, uh, 238 00:16:07,642 --> 00:16:11,402 S1: up to the task of maintaining a democracy because our 239 00:16:11,402 --> 00:16:18,802 S1: population has declined so much. Cognitively, philosophically, education wise, or 240 00:16:18,802 --> 00:16:23,562 S1: I would say knowledge wise. And so we're losing our 241 00:16:24,122 --> 00:16:28,682 S1: ticket or we're losing our ability to be worthy of democracy. 242 00:16:29,642 --> 00:16:33,122 S1: That's one argument. That's one argument. So I like this framing. 243 00:16:33,122 --> 00:16:36,082 S1: I like the structure. And so essentially what we would 244 00:16:36,082 --> 00:16:40,442 S1: be saying here is that the capitalist structure is super 245 00:16:40,482 --> 00:16:43,802 S1: useful at a certain phase of our maturity and development, 246 00:16:44,162 --> 00:16:46,962 S1: and that ideally we would move out of it and 247 00:16:46,962 --> 00:16:50,982 S1: move into something more like, not like a Stalin communism, 248 00:16:50,982 --> 00:16:53,902 S1: not like a mao communism. It wouldn't even really be 249 00:16:53,902 --> 00:16:56,862 S1: communism because it wouldn't be like state control of things. 250 00:16:56,862 --> 00:17:00,702 S1: It would be more like shared human control of things 251 00:17:01,142 --> 00:17:05,142 S1: and shared human voting. It would be actual democracy instead 252 00:17:05,142 --> 00:17:12,861 S1: of democracy. Republic. Of democracy or a republic. Because we 253 00:17:12,862 --> 00:17:17,022 S1: would have the tech, presumably, to do so. And you 254 00:17:17,022 --> 00:17:19,342 S1: could use other things for the whole federal versus state 255 00:17:19,342 --> 00:17:23,302 S1: or whatever. We could basically figure out these problems. Right. 256 00:17:23,502 --> 00:17:26,942 S1: And the more advanced we are as a population, the 257 00:17:26,942 --> 00:17:29,502 S1: easier that would be to do, because you just make 258 00:17:29,502 --> 00:17:31,822 S1: a good argument and people are like, oh, yeah, it's 259 00:17:31,821 --> 00:17:34,262 S1: a pretty good argument. Yeah, let's do it that way. Right. 260 00:17:34,302 --> 00:17:40,982 S1: Which we clearly don't have today. But so the point 261 00:17:40,982 --> 00:17:44,702 S1: of all of this is to say that the only 262 00:17:44,702 --> 00:17:48,882 S1: way I can function is to take all of my 263 00:17:50,282 --> 00:17:54,442 S1: concern and worry and sort of pain around. Holy crap, 264 00:17:54,442 --> 00:17:57,722 S1: this is going to get bad and use that exact 265 00:17:57,722 --> 00:18:01,482 S1: same force and that exact same energy and just be like, no, 266 00:18:01,642 --> 00:18:05,802 S1: you know what? I am not thinking that way. I 267 00:18:05,802 --> 00:18:08,362 S1: don't know that we have a chance to get out 268 00:18:08,362 --> 00:18:12,722 S1: of this clean. We might exterminate ourselves. We might go 269 00:18:12,722 --> 00:18:15,561 S1: into massive civil war. We might have, like, a huge 270 00:18:15,561 --> 00:18:19,002 S1: amount of problems. There's a million what ifs. There is 271 00:18:19,002 --> 00:18:21,202 S1: nothing I can do about it. There's nothing you could 272 00:18:21,202 --> 00:18:23,282 S1: do about it. This is a thing that is happening 273 00:18:23,321 --> 00:18:26,401 S1: to the planet right now, and it's just kind of 274 00:18:26,442 --> 00:18:30,802 S1: playing out. And it's really weird with this whole free 275 00:18:30,802 --> 00:18:33,802 S1: will thing. It's like, I know that we don't have 276 00:18:33,802 --> 00:18:38,122 S1: the free will to fix this, but guess what? There's 277 00:18:38,162 --> 00:18:41,122 S1: a guaranteed way it doesn't get fixed is to think 278 00:18:41,122 --> 00:18:43,122 S1: that you don't have the free will to fix it. 279 00:18:43,522 --> 00:18:46,421 S1: Which means I behave as if. And this is another 280 00:18:46,422 --> 00:18:49,462 S1: essay I'm working on, but you behave as if you 281 00:18:49,462 --> 00:18:55,381 S1: had the freedom, and ironically, that ends up bringing forth 282 00:18:55,622 --> 00:18:59,542 S1: a more positive thing. Now you could say, well, that 283 00:18:59,542 --> 00:19:02,821 S1: was what was going to happen anyway. Well, cool. Whatever. 284 00:19:03,222 --> 00:19:04,942 S1: I don't have time to think about that right now. 285 00:19:04,942 --> 00:19:07,782 S1: I need to think about positivity and trying to make 286 00:19:07,782 --> 00:19:11,022 S1: it happen because we are humans. Actually, the name of 287 00:19:11,022 --> 00:19:13,382 S1: this essay, and I know I'm sort of going on 288 00:19:13,382 --> 00:19:17,061 S1: a tangent here, but hopefully it's okay. The name of 289 00:19:17,061 --> 00:19:21,742 S1: this essay is Reality is Layer Dependent, which means at 290 00:19:21,742 --> 00:19:26,502 S1: my layer right now, when I am trying to fix 291 00:19:26,502 --> 00:19:30,382 S1: the world at my layer right now as a human, 292 00:19:31,302 --> 00:19:36,022 S1: free will does exist. Because if I go outside and 293 00:19:36,022 --> 00:19:38,382 S1: someone has stolen my trash cans and I see them 294 00:19:38,382 --> 00:19:41,102 S1: across the street in front of their house and they're 295 00:19:41,102 --> 00:19:46,161 S1: full of garbage. I'm going to walk across the street 296 00:19:46,162 --> 00:19:50,602 S1: and ask them to not do that again. Please. I'm 297 00:19:50,642 --> 00:19:53,561 S1: not going to say I'm not going to walk across 298 00:19:53,561 --> 00:19:55,362 S1: the street and say, hey, I just want to let 299 00:19:55,362 --> 00:19:58,642 S1: you know, I know you don't have free will. And, um, 300 00:19:59,242 --> 00:20:03,042 S1: you there was nothing you could have possibly done to 301 00:20:03,082 --> 00:20:06,482 S1: not steal my garbage cans and fill them up full 302 00:20:06,482 --> 00:20:10,722 S1: of garbage. That is not a conversation I'm going to have. 303 00:20:11,202 --> 00:20:14,242 S1: I'm going to have a conversation about them changing their behavior, 304 00:20:14,762 --> 00:20:19,282 S1: which means at this layer of of reality, which is 305 00:20:19,282 --> 00:20:24,321 S1: human interaction and the perception of free will. These are 306 00:20:24,321 --> 00:20:28,362 S1: the these are the types of conversations that we have. Right. 307 00:20:29,442 --> 00:20:33,722 S1: And layers below this, it's kind of lame and stupid. Uh, 308 00:20:33,922 --> 00:20:38,842 S1: the biological layer, the chemical layer, the physical layer. You 309 00:20:38,842 --> 00:20:42,641 S1: talk to atoms about the fact that somebody stole garbage cans. 310 00:20:42,752 --> 00:20:45,752 S1: They don't understand that. You know what I mean? So. 311 00:20:48,872 --> 00:20:54,832 S1: All this coming back to. I'm trying to focus. I 312 00:20:54,872 --> 00:20:58,272 S1: recommend other people do this as well. Take all that 313 00:20:58,952 --> 00:21:03,311 S1: hurt energy, all that negativity, all that everything, and just 314 00:21:03,311 --> 00:21:07,712 S1: move it over into this positivity category because everything could 315 00:21:07,712 --> 00:21:12,351 S1: go horribly wrong. Or this could be the origin story 316 00:21:12,712 --> 00:21:16,992 S1: for a few scrappy people got together and they planned 317 00:21:16,992 --> 00:21:21,672 S1: on building a better internet. That was human based and 318 00:21:21,952 --> 00:21:28,351 S1: they just, you know, steadfastly refused to give into the suck. 319 00:21:28,352 --> 00:21:31,311 S1: And they started building the better thing. And, oh, the 320 00:21:31,311 --> 00:21:35,272 S1: idea started catching on. And when things got bad enough, 321 00:21:35,272 --> 00:21:39,392 S1: people leaned over into the positive side that that's actually 322 00:21:39,392 --> 00:21:42,732 S1: what we ended up building. and it turns out we survived. 323 00:21:42,732 --> 00:21:44,812 S1: It was way better. We got rid of the corporate 324 00:21:44,811 --> 00:21:48,492 S1: work we're now teaching in schools that everyone is a creator. 325 00:21:48,492 --> 00:21:53,252 S1: Everyone is a builder. Everyone is a entrepreneur. Everyone has 326 00:21:53,692 --> 00:21:59,132 S1: value to give to the world, right? Not the special few. 327 00:21:59,292 --> 00:22:01,932 S1: Your job is not to get out of school and 328 00:22:01,932 --> 00:22:04,492 S1: find a really cool company, because that's a really smart 329 00:22:04,492 --> 00:22:09,172 S1: person and you're going to give your little meager skills, 330 00:22:09,332 --> 00:22:14,932 S1: you know, to, to help them succeed. Screw that. This 331 00:22:14,932 --> 00:22:18,812 S1: is new world. This is we all have that capability, right? 332 00:22:19,492 --> 00:22:23,612 S1: So if we have a tiny chance of this happening, 333 00:22:24,412 --> 00:22:26,972 S1: that chance is larger if if we all try and 334 00:22:26,972 --> 00:22:31,892 S1: push towards it. So that's that's why I'm very. Sort 335 00:22:31,892 --> 00:22:36,611 S1: of sad and very excited all at the same time. Okay. 336 00:22:36,652 --> 00:22:41,292 S1: That was the first line of the newsletter. 2020 two 337 00:22:41,292 --> 00:22:45,032 S1: minutes in. That was literally the first line of the newsletter. Okay, 338 00:22:45,032 --> 00:22:52,552 S1: we're scrolling down. Incredibly awesome conversation with my buddy Jason 339 00:22:52,552 --> 00:22:57,072 S1: Haddox over the weekend about automated recon pen testing type 340 00:22:57,112 --> 00:22:59,992 S1: stuff that we're building together. So mine is called Helios. 341 00:23:00,032 --> 00:23:03,111 S1: I've had it for like, whatever, 15 years, I don't know, 342 00:23:03,152 --> 00:23:06,352 S1: 12 years, something like that. And before it was just 343 00:23:06,352 --> 00:23:11,111 S1: like nasty Python and bash and stuff like that. Basic 344 00:23:11,112 --> 00:23:15,112 S1: automation stuff, linuxi type automation stuff, using a whole bunch 345 00:23:15,112 --> 00:23:18,872 S1: of custom tools, a whole bunch of project discovery tools 346 00:23:19,472 --> 00:23:21,552 S1: and like it does the job. But now I'm basically 347 00:23:21,552 --> 00:23:25,552 S1: doing a full AI upgrade on it, and Jason's been 348 00:23:25,552 --> 00:23:29,032 S1: working on his stack as well. And so, uh, we 349 00:23:29,032 --> 00:23:32,072 S1: talked a bunch about, uh, what I'm doing currently with 350 00:23:32,071 --> 00:23:36,272 S1: my stack with Kai, and he pings me like, I 351 00:23:36,272 --> 00:23:39,892 S1: don't know, like 15 minutes later, he's like, dude, um, 352 00:23:40,492 --> 00:23:44,812 S1: my thing just found a bug. Um, a P1 bug. 353 00:23:45,372 --> 00:23:47,892 S1: And it turns out that, um, he had this password 354 00:23:47,892 --> 00:23:52,012 S1: field for a thing that he was testing, and, uh, 355 00:23:53,172 --> 00:23:55,532 S1: it was just the password field, and it was for 356 00:23:55,532 --> 00:24:01,091 S1: some sort of admin site, and he was thinking, like, 357 00:24:01,132 --> 00:24:02,691 S1: do I? Okay, what do I want to try to 358 00:24:02,692 --> 00:24:06,732 S1: brute force this thing? It doesn't look extremely vulnerable. And 359 00:24:06,732 --> 00:24:10,572 S1: his I found something that he and I would not 360 00:24:10,571 --> 00:24:13,172 S1: have thought of. Uh, right. I mean, it's somewhere in 361 00:24:13,172 --> 00:24:16,891 S1: my methodology, but it's not part of my main methodology. Um, 362 00:24:17,372 --> 00:24:19,932 S1: which is to start messing with parameters. That's fine, but 363 00:24:19,972 --> 00:24:23,692 S1: it in one shot, it's first thing that it tried 364 00:24:23,692 --> 00:24:27,811 S1: was to add a parameter to post, which was id 365 00:24:27,852 --> 00:24:32,652 S1: equals one and instantly bypassed. So an actual live real 366 00:24:32,652 --> 00:24:37,552 S1: world P1. And he said he had to mess with 367 00:24:37,552 --> 00:24:39,952 S1: it a little bit to get up to that point. 368 00:24:40,472 --> 00:24:44,512 S1: But in terms of like doing an actual exploit, a P1, 369 00:24:45,232 --> 00:24:47,792 S1: doing something that he wasn't going to do, I wouldn't 370 00:24:47,792 --> 00:24:50,631 S1: have thought of, uh, at least in that immediately. I'm 371 00:24:50,632 --> 00:24:54,192 S1: sure somewhere deep in our methodology, we have to not 372 00:24:54,192 --> 00:24:57,191 S1: only modify parameters, but also to add parameters. I mean, 373 00:24:57,192 --> 00:25:00,992 S1: that's a pretty common thing, actually, but it didn't come 374 00:25:00,992 --> 00:25:04,072 S1: to mind, right? That's the point. It was like, that 375 00:25:04,071 --> 00:25:07,872 S1: was pretty cool. It was pretty cool. So I think 376 00:25:07,872 --> 00:25:11,072 S1: the takeaway that we have there is just like, this 377 00:25:11,071 --> 00:25:13,792 S1: stuff is not theoretical. This stuff is like getting very, 378 00:25:13,792 --> 00:25:17,552 S1: very tangible. And my point about this has always been 379 00:25:17,552 --> 00:25:23,071 S1: that we just have an orchestration problem, like the individual 380 00:25:23,071 --> 00:25:26,952 S1: steps don't require all that much intelligence. It's not super 381 00:25:26,952 --> 00:25:30,512 S1: hard to do this stuff. The problem is keeping the 382 00:25:30,512 --> 00:25:33,232 S1: plot when you're doing it and you're doing like, you know, 383 00:25:33,272 --> 00:25:37,292 S1: 175 different mini tasks, and you're gathering all this context 384 00:25:37,292 --> 00:25:40,532 S1: and there's too much context, you know, falling out of 385 00:25:40,532 --> 00:25:43,252 S1: different places. So like a big part of what I'm 386 00:25:43,252 --> 00:25:47,652 S1: doing with Kai is breaking context into separate subfolders with 387 00:25:47,652 --> 00:25:52,891 S1: their separate cloud MD files and their separate large pieces 388 00:25:52,892 --> 00:25:57,052 S1: of context, so that the subagents are consuming context and 389 00:25:57,052 --> 00:25:59,852 S1: they're not polluting the main agent with that context. Right. 390 00:25:59,892 --> 00:26:03,331 S1: Because you absolutely need this when you're, you know, parsing 391 00:26:03,372 --> 00:26:06,732 S1: like complete Doms and stuff like that. Um, there's just 392 00:26:06,732 --> 00:26:09,492 S1: so much content in there. Then you have, uh, haystack 393 00:26:09,492 --> 00:26:13,212 S1: performance issues, right? You're trying to find like endpoints inside 394 00:26:13,212 --> 00:26:16,852 S1: of a giant, you know, Dom full of JavaScript. It's 395 00:26:16,852 --> 00:26:19,972 S1: like it's pretty nasty. And it takes up a lot 396 00:26:19,972 --> 00:26:24,811 S1: of tokens, and it's traditionally not great for for AI. 397 00:26:24,972 --> 00:26:27,212 S1: And the other thing to mention about this is like, 398 00:26:27,212 --> 00:26:30,171 S1: you know, when you should be using AI and know 399 00:26:30,172 --> 00:26:33,841 S1: when you shouldn't be, right. Um, I'm going to talk 400 00:26:33,842 --> 00:26:37,602 S1: about this somewhere else. But one thing that's really cool is, 401 00:26:37,642 --> 00:26:42,561 S1: is to think about the fact that, um, you should 402 00:26:42,561 --> 00:26:44,962 S1: use as little AI as possible. I talk about this 403 00:26:44,962 --> 00:26:48,522 S1: a little bit later in the newsletter, but uses little 404 00:26:48,522 --> 00:26:54,682 S1: AI's possible. Ideally, the whole workflow would be legacy traditional 405 00:26:54,722 --> 00:27:00,841 S1: deterministic tech, just regular code, right. So like addition, subtraction, 406 00:27:00,842 --> 00:27:06,082 S1: multiplication like no reason to enter AI into this. Right. 407 00:27:06,842 --> 00:27:08,522 S1: And then there are certain types of problems that do 408 00:27:08,522 --> 00:27:12,562 S1: require that that do require intelligence. So if it requires 409 00:27:12,561 --> 00:27:15,922 S1: intelligence it's the type of thing like hey is this 410 00:27:15,922 --> 00:27:19,561 S1: a cat, for example. That's not an if this, if this, 411 00:27:19,561 --> 00:27:22,482 S1: then that statement that you can give to regular code, 412 00:27:22,762 --> 00:27:25,081 S1: which is why we needed machine learning, which is why 413 00:27:25,082 --> 00:27:29,442 S1: we use AI. Right. So just knowing when to use 414 00:27:29,442 --> 00:27:33,982 S1: which is super important. Um. All right. Scroll down. Peace. 415 00:27:34,222 --> 00:27:39,262 S1: About limitations to creativity. And I realized there's actually a 416 00:27:39,262 --> 00:27:41,622 S1: third limitation. So I'm writing another blog post that kind 417 00:27:41,622 --> 00:27:45,462 S1: of unifies them all. But the third one, the third 418 00:27:45,502 --> 00:27:48,982 S1: type of limitation to creativity, is not even thinking about 419 00:27:48,982 --> 00:27:51,462 S1: some options for creating a new solution or solving a 420 00:27:51,462 --> 00:27:55,061 S1: problem because it was previously impossible. It's like being stuck 421 00:27:55,061 --> 00:27:58,182 S1: in like old world, old mind, old tech or whatever. 422 00:27:59,061 --> 00:28:04,902 S1: And I got an example from this weekend, so I 423 00:28:04,902 --> 00:28:09,662 S1: was I'm not very happy with like my, my, um. 424 00:28:12,061 --> 00:28:17,142 S1: Web analytics stack. So my web analytics stack is currently fathom, 425 00:28:17,902 --> 00:28:21,342 S1: but I would much rather be using a chartbeat. But 426 00:28:21,342 --> 00:28:23,662 S1: even Chartbeat doesn't do everything. It doesn't have a menu 427 00:28:23,662 --> 00:28:29,062 S1: bar thing. Um, it also doesn't have historical metrics, right? 428 00:28:29,122 --> 00:28:31,922 S1: I can't use it like Google Analytics. So I for 429 00:28:32,122 --> 00:28:35,162 S1: like over a decade, I use Google Analytics and Chartbeat 430 00:28:35,162 --> 00:28:38,442 S1: Chartbeat to look at Google Analytics to go and get 431 00:28:38,442 --> 00:28:44,442 S1: reports from. And that's kind of lame, but I never 432 00:28:44,442 --> 00:28:46,282 S1: questioned it. I never thought it was a problem. I 433 00:28:46,282 --> 00:28:48,322 S1: was just like the way the world is, no big deal. 434 00:28:48,362 --> 00:28:53,722 S1: Like that's just reality. Well, I was messing with Full Fathom, 435 00:28:54,042 --> 00:28:56,242 S1: which is my current replacement for both of those, because 436 00:28:56,242 --> 00:28:59,642 S1: Chartbeat is too expensive now and Google Analytics is garbage now. 437 00:29:00,522 --> 00:29:03,202 S1: So I use fathom, and it's like it doesn't have 438 00:29:03,202 --> 00:29:08,522 S1: the Google Analytics, like, I guess like enterprise feel. Um, 439 00:29:08,522 --> 00:29:13,482 S1: it doesn't look nearly as good as Google Analytics or Chartbeat. Um, 440 00:29:13,522 --> 00:29:16,522 S1: and it kind of has like a, uh, I found 441 00:29:16,522 --> 00:29:19,322 S1: a way to give a menu bar icon, but I realized, 442 00:29:19,362 --> 00:29:22,362 S1: hold on, I actually know how Chartbeat works. It works 443 00:29:22,362 --> 00:29:24,242 S1: by having a piece of JavaScript on the page that 444 00:29:24,242 --> 00:29:28,622 S1: sends a beacon every few seconds. And that's how it 445 00:29:28,622 --> 00:29:35,022 S1: actually sends the difference between. It can tell the difference between, um, well, 446 00:29:35,182 --> 00:29:37,782 S1: its primary functionality is that it can tell you who's 447 00:29:37,782 --> 00:29:42,062 S1: reading the website. Right? Like right now there's 165 people 448 00:29:42,062 --> 00:29:45,382 S1: reading my website. And that's that's my current count. I'm 449 00:29:45,382 --> 00:29:48,822 S1: looking at my menu bar that's 165 people with a 450 00:29:48,822 --> 00:29:53,862 S1: tab open that are reading it. Right. So that that's 451 00:29:53,862 --> 00:29:56,382 S1: what I care about. I care about who's currently looking 452 00:29:56,382 --> 00:30:01,102 S1: at what pages. Right. Google analytics doesn't have this. Chartbeat 453 00:30:01,142 --> 00:30:04,142 S1: has this. It's the only one I've found that really 454 00:30:04,142 --> 00:30:07,942 S1: had this feature. And they also had a guy that 455 00:30:07,942 --> 00:30:09,942 S1: was like super pretty and it was just like a 456 00:30:09,982 --> 00:30:12,862 S1: really cool way to visualize it. And I'm like, this 457 00:30:12,862 --> 00:30:14,662 S1: is the best thing ever. So I paid for it 458 00:30:14,662 --> 00:30:18,902 S1: forever and ever and ever. And now I just suddenly 459 00:30:18,902 --> 00:30:22,022 S1: realized on Sunday while I was working on the newsletter, hey, 460 00:30:22,062 --> 00:30:25,342 S1: I wonder if I could just make chartbeat. So I'm like, hey, 461 00:30:25,342 --> 00:30:29,282 S1: I want a piece of JavaScript called analytics. I'm going 462 00:30:29,322 --> 00:30:31,602 S1: to serve it inside my page. Here's where you're going 463 00:30:31,642 --> 00:30:35,042 S1: to put it. Here's the functionality it's going to have. Um, 464 00:30:35,082 --> 00:30:37,762 S1: here's how often it's going to beacon. Set up an 465 00:30:37,762 --> 00:30:42,562 S1: endpoint on Cloudflare to receive the beacon. You're going to 466 00:30:42,562 --> 00:30:45,962 S1: parse referrers like this, blah, blah, blah, basically. And I 467 00:30:46,282 --> 00:30:48,922 S1: gave it a whole bunch of spec sort of stuff, 468 00:30:48,962 --> 00:30:53,402 S1: you know, um, basically a full spec of what I 469 00:30:53,402 --> 00:30:59,122 S1: want to build, uh, in the Google Analytics JavaScript, also 470 00:30:59,122 --> 00:31:02,442 S1: in the receiving endpoint, also in the front end showing it. 471 00:31:02,962 --> 00:31:04,842 S1: And I give it a bunch of examples and stuff 472 00:31:04,842 --> 00:31:09,162 S1: like that. And in about 18 minutes of me just 473 00:31:09,162 --> 00:31:12,282 S1: talking to it and giving it to it in that format, 474 00:31:13,602 --> 00:31:17,482 S1: it then produced the full spec and wrote the entire thing. 475 00:31:17,682 --> 00:31:21,882 S1: So within 18 minutes I literally had a chartbeat replacement. 476 00:31:22,842 --> 00:31:26,262 S1: And I then spent many hours after that trying to, 477 00:31:26,302 --> 00:31:30,142 S1: you know, tweaking and adding fields and, you know, tweaking 478 00:31:30,142 --> 00:31:32,742 S1: the color scheme slightly and display and stuff like that. 479 00:31:32,742 --> 00:31:35,662 S1: So it's not like I was done after 18 minutes, 480 00:31:35,662 --> 00:31:40,382 S1: but I had live visitors showing up and showing up 481 00:31:40,382 --> 00:31:43,462 S1: on my graph, and the site looked pretty much like 482 00:31:43,462 --> 00:31:48,102 S1: Chartbeat within 18 minutes. That is ridiculous. And the whole 483 00:31:48,102 --> 00:31:51,542 S1: point of talking about this is, yeah, it's cool. So 484 00:31:51,542 --> 00:31:53,702 S1: I want to talk about it. I think it's cool. 485 00:31:53,702 --> 00:31:58,542 S1: I love, uh, live telemetry from the world is just 486 00:31:58,582 --> 00:32:01,222 S1: a thing that, like, I'm obsessed with. You have a map. 487 00:32:01,222 --> 00:32:06,342 S1: You realize that there's some human who's reading some thought, uh, and, 488 00:32:06,382 --> 00:32:09,582 S1: you know, maybe maybe you've seen this person before. Maybe 489 00:32:09,582 --> 00:32:11,982 S1: you've walked by them in a coffee shop or something. 490 00:32:12,942 --> 00:32:15,702 S1: It's just like human connection, I think is really cool. 491 00:32:17,182 --> 00:32:21,502 S1: But the thing I want to focus on here is that. 492 00:32:23,372 --> 00:32:26,612 S1: Literally the second before I had this thought, I was 493 00:32:26,612 --> 00:32:31,692 S1: caged by this concept of like, well, that's too bad. 494 00:32:31,692 --> 00:32:35,212 S1: Short beats expensive because there's no way I could make 495 00:32:35,212 --> 00:32:38,931 S1: a chart beat. I wasn't actually even thinking there's no 496 00:32:38,932 --> 00:32:41,012 S1: way I could make a chart beat. That was an 497 00:32:41,012 --> 00:32:46,812 S1: automatic default barrier inside my brain. I didn't even have 498 00:32:46,812 --> 00:32:51,452 S1: to think that the barrier was there. The barrier is invisible. Now, 499 00:32:51,492 --> 00:32:54,012 S1: keep in mind, I've been able to do this, this 500 00:32:54,172 --> 00:32:57,812 S1: advanced level of like building for like the last couple 501 00:32:57,852 --> 00:33:01,292 S1: of years, I've been building tons of stuff. Why did 502 00:33:01,292 --> 00:33:04,852 S1: I never think to build a better analytics system? I've 503 00:33:04,892 --> 00:33:10,292 S1: paid for Fathom Analytics. I've paid them hundreds upon hundreds 504 00:33:10,292 --> 00:33:14,492 S1: of dollars for a thing that's worse than what I 505 00:33:14,492 --> 00:33:21,212 S1: made in 18 minutes. Let that sink in, okay? 18 506 00:33:21,212 --> 00:33:25,232 S1: minutes to be way better than my Google Analytics solution 507 00:33:25,912 --> 00:33:30,592 S1: and my Chartbeat solution and my fathom solution. The limitation, 508 00:33:30,592 --> 00:33:33,192 S1: the reason I hadn't done it, is because it hadn't 509 00:33:33,192 --> 00:33:35,912 S1: popped into my mind that I was able to do it. 510 00:33:37,072 --> 00:33:42,072 S1: And what really breaks my brain is I'm still currently 511 00:33:42,072 --> 00:33:46,712 S1: limited by an infinite number of those walls and those barriers. 512 00:33:46,832 --> 00:33:52,832 S1: So in this blog this is type three creativity limitation. 513 00:33:53,592 --> 00:33:57,952 S1: And we seek. My argument is we seek or should 514 00:33:57,992 --> 00:34:02,152 S1: seek type three freedom. In fact, all three type one 515 00:34:02,152 --> 00:34:09,192 S1: freedom is freedom from internal um not not being able 516 00:34:09,192 --> 00:34:14,952 S1: to reach our internal childlike creativity. So the freedom is 517 00:34:14,952 --> 00:34:17,392 S1: to be able to and the restraint is not being 518 00:34:17,392 --> 00:34:21,092 S1: able to. And then the second one is audience capture. 519 00:34:21,292 --> 00:34:23,411 S1: So it's like you're worried about what your parents and 520 00:34:23,412 --> 00:34:26,332 S1: your friends and your your surroundings. It's not really audience. 521 00:34:26,332 --> 00:34:32,332 S1: It's more like peer or societal capture and like limitation. 522 00:34:32,572 --> 00:34:36,931 S1: So you literally stop having good thoughts because you can 523 00:34:36,932 --> 00:34:39,572 S1: only think along certain lines of what you're expected to 524 00:34:39,572 --> 00:34:44,172 S1: produce and what is acceptable to produce. So that's layer two. 525 00:34:44,172 --> 00:34:49,052 S1: And level three is thinking in the past being limited 526 00:34:49,052 --> 00:34:54,852 S1: by the past. So it's like I grew up thinking 527 00:34:54,852 --> 00:34:58,372 S1: only certain things were possible. And I can't even know 528 00:34:58,372 --> 00:35:01,012 S1: all the ways that those limitations are still on me. 529 00:35:01,532 --> 00:35:06,532 S1: I mean, I was born in the 70s, right? So phone, 530 00:35:06,732 --> 00:35:10,212 S1: cell phones, right? I mean, all these things that we 531 00:35:10,212 --> 00:35:16,052 S1: have now, how many of them? How many of them 532 00:35:16,492 --> 00:35:19,832 S1: are like limiting me in me thinking that those things 533 00:35:19,832 --> 00:35:22,592 S1: are possible. I feel like I'm way ahead of the 534 00:35:22,592 --> 00:35:27,992 S1: game against most people in the world, honestly, including extremely 535 00:35:27,992 --> 00:35:30,152 S1: young kids who grew up with tech. I feel like 536 00:35:30,152 --> 00:35:33,352 S1: I'm more unbounded than most, but I still feel very 537 00:35:33,352 --> 00:35:38,512 S1: bound by this. I feel extremely sort of caged by this. 538 00:35:38,512 --> 00:35:41,712 S1: So I'm writing this post to kind of remind myself. 539 00:35:42,232 --> 00:35:45,592 S1: Do you have all three types of these freedoms? Oh, 540 00:35:45,592 --> 00:35:47,512 S1: and by the way, the reason I got on to 541 00:35:47,512 --> 00:35:53,592 S1: this at all is from reading, uh, Rilke and Rilke's 542 00:35:53,632 --> 00:35:56,432 S1: Letters to a Young Poet, um, which is a book 543 00:35:56,432 --> 00:35:59,312 S1: that I've been recommending. And that's where I first started 544 00:35:59,312 --> 00:36:01,152 S1: thinking about this. But then I started thinking about it 545 00:36:01,152 --> 00:36:04,232 S1: from this book, Mathematica, which I just finished. And oh 546 00:36:04,272 --> 00:36:09,512 S1: my God, it is extraordinary. So anyway, um, a bunch 547 00:36:09,512 --> 00:36:12,312 S1: of that stuff is in the previous two posts about creativity, 548 00:36:12,312 --> 00:36:13,992 S1: but I'm about to do this third one that kind 549 00:36:13,992 --> 00:36:16,372 S1: of unites all of them. Oh, and by the way, 550 00:36:16,372 --> 00:36:19,012 S1: I'm going to tell Kai to once I'm done with 551 00:36:19,012 --> 00:36:22,172 S1: the third post to go and update the other two 552 00:36:22,612 --> 00:36:25,932 S1: to point to the third one. And I can also 553 00:36:25,932 --> 00:36:29,052 S1: say in the third one, hey, I wrote these previous 554 00:36:29,052 --> 00:36:31,812 S1: two posts and it's going to Kai is going to 555 00:36:31,852 --> 00:36:34,932 S1: automatically go find those two posts and link to them. 556 00:36:34,932 --> 00:36:37,972 S1: So this is an example of a thing that you know, 557 00:36:38,012 --> 00:36:41,372 S1: in the whatever 26 years that I've been blogging, I've 558 00:36:41,372 --> 00:36:45,892 S1: had to do that manually. And it's just really, really 559 00:36:45,892 --> 00:36:48,532 S1: great to think that I could just say one thing 560 00:36:48,532 --> 00:36:53,092 S1: into Whisper Flow, and all of that gets dynamically updated 561 00:36:53,652 --> 00:36:58,532 S1: by Kai. Okay. So, uh, all right, so that's the, uh, 562 00:36:58,572 --> 00:37:07,532 S1: whole chart beat creativity thing. Um, oh. Had a really cool, 563 00:37:07,532 --> 00:37:11,332 S1: interesting conversation with Matthew Brown from Trail of Bits. So 564 00:37:11,332 --> 00:37:13,732 S1: we were talking about this AI system design thing that 565 00:37:13,732 --> 00:37:16,272 S1: I was talking about earlier, and he gave me a 566 00:37:16,312 --> 00:37:19,752 S1: couple really cool, um, ways to think about this and like, 567 00:37:19,792 --> 00:37:22,952 S1: design principles that I'm going to use for my meta 568 00:37:22,952 --> 00:37:27,112 S1: system designer. Um, and one of the things he said 569 00:37:27,112 --> 00:37:32,472 S1: that was really cool is problem design, problem understanding and 570 00:37:32,472 --> 00:37:36,992 S1: problem problem categorization is what it is really. So he's 571 00:37:36,992 --> 00:37:40,912 S1: talking about problems that could be solved or problems that 572 00:37:40,912 --> 00:37:44,272 S1: are like identifying a cat in a picture. That is 573 00:37:44,272 --> 00:37:46,912 S1: a certain type of problem that code doesn't work for. 574 00:37:47,032 --> 00:37:51,192 S1: You need machine learning, right? Or you need intelligence. Which 575 00:37:51,192 --> 00:37:53,512 S1: machine learning is like a proxy for that. And then 576 00:37:53,512 --> 00:37:55,912 S1: AI is a proxy for it as well or a 577 00:37:55,952 --> 00:37:59,592 S1: type of it. So but there's other problems where you 578 00:37:59,592 --> 00:38:02,232 S1: should not be using AI at all. Now, the reason 579 00:38:02,232 --> 00:38:05,112 S1: I was talking to him is because he's, uh, he 580 00:38:05,112 --> 00:38:08,352 S1: led the team on Trail of Bits team that won 581 00:38:08,352 --> 00:38:13,142 S1: second place in the AI competition, which is a fully 582 00:38:13,142 --> 00:38:19,942 S1: automated discovery and and patching system. Agent system. It involves 583 00:38:19,942 --> 00:38:24,102 S1: multiple agents broken into multiple components. And so my main 584 00:38:24,102 --> 00:38:26,782 S1: question for him was like, are you relying on the 585 00:38:26,782 --> 00:38:29,702 S1: model the most or are you relying on the system 586 00:38:29,702 --> 00:38:31,982 S1: the most? And he ended up saying system as well. 587 00:38:31,982 --> 00:38:34,782 S1: I've thought it was system. So I was so happy 588 00:38:34,782 --> 00:38:36,902 S1: to hear I wasn't wrong about that, at least in 589 00:38:36,902 --> 00:38:41,622 S1: his opinion. And um, yeah, he just had some really 590 00:38:41,622 --> 00:38:45,982 S1: cool ideas about like how you design the agent system, um, 591 00:38:46,142 --> 00:38:48,422 S1: to constrain it and make sure you give it, like, 592 00:38:48,462 --> 00:38:52,022 S1: only specific types of context. But, um, that episode should 593 00:38:52,022 --> 00:38:55,462 S1: come out soon. Um, that's going to be on YouTube 594 00:38:55,462 --> 00:38:59,302 S1: and also in the, in the audio and the regular podcast. 595 00:39:00,342 --> 00:39:04,062 S1: And so, uh, really cool conversation. That guy is sharp. 596 00:39:04,102 --> 00:39:06,382 S1: I mean, this guy's been doing AI forever. Like, he's 597 00:39:06,382 --> 00:39:11,802 S1: probably going to get hired by Google for $74 trillion 598 00:39:11,842 --> 00:39:16,042 S1: would not surprise me. Um, a few weeks back I 599 00:39:16,042 --> 00:39:19,242 S1: posted about feeling really guilty about eating meat, and then 600 00:39:19,242 --> 00:39:22,802 S1: I saw a Dwarkesh episode where he had this, uh, 601 00:39:22,842 --> 00:39:26,762 S1: meat guy on or animals, you know, kindness guy on, 602 00:39:27,202 --> 00:39:29,122 S1: and he's like, yeah, we've got this thing called Farm 603 00:39:29,122 --> 00:39:32,362 S1: kind and it's a charity. And like, you know, even 604 00:39:32,362 --> 00:39:36,242 S1: $10 can save a lot of animals from suffering. Um, 605 00:39:36,282 --> 00:39:39,602 S1: and just just to, like, clarify why I care about this, 606 00:39:40,242 --> 00:39:44,802 S1: it really upsets me how similar we are to some animals, 607 00:39:44,922 --> 00:39:48,842 S1: especially like pigs, even chickens. I would say, like, obviously, 608 00:39:48,842 --> 00:39:50,522 S1: I don't know what's in the mind of a chicken, 609 00:39:50,882 --> 00:39:57,322 S1: but a pig looks and feels very normal to me. 610 00:39:57,362 --> 00:40:00,402 S1: It looks and feels like a a more advanced animal. 611 00:40:01,002 --> 00:40:04,402 S1: If you hear a pig squeal, especially if it's like 612 00:40:04,402 --> 00:40:08,802 S1: in pain or something like that sounds like pain to 613 00:40:09,022 --> 00:40:13,462 S1: my brain. It sounds like pain. And, um, there are 614 00:40:13,462 --> 00:40:16,742 S1: people in, uh. I think it's groups in Africa. Or 615 00:40:16,742 --> 00:40:24,062 S1: maybe it's Australia. Um, they call humans long pig. Uh, 616 00:40:24,062 --> 00:40:28,382 S1: because the flesh is similar and in flesh means muscle. 617 00:40:29,382 --> 00:40:32,542 S1: I mean, they're also pink, and a lot of humans 618 00:40:32,542 --> 00:40:36,502 S1: are pink, right? I mean, there's so many similarities here, 619 00:40:36,502 --> 00:40:40,181 S1: and I'm just like, oh, evidently we taste the same. 620 00:40:40,182 --> 00:40:42,542 S1: That was that was the point of of being called 621 00:40:42,542 --> 00:40:46,542 S1: Long Pig because, uh, these groups eat humans. They also 622 00:40:46,542 --> 00:40:49,622 S1: eat pigs, uh, I assume like wild boar or whatever. 623 00:40:49,622 --> 00:40:51,102 S1: And they're like, yeah, they kind of taste the same. 624 00:40:51,102 --> 00:40:54,822 S1: And I'm like, okay, uh, good to know. Uh, how 625 00:40:54,822 --> 00:40:59,222 S1: do I get out of here? Um, but I just like. Oh, 626 00:40:59,262 --> 00:41:01,302 S1: and then you see the pictures of, like, this farm 627 00:41:01,302 --> 00:41:06,142 S1: kind thing and just rows and rows and rows like 628 00:41:06,142 --> 00:41:08,882 S1: matrix where you see all those, you know, human eggs 629 00:41:08,882 --> 00:41:13,562 S1: or whatever, but it's rows and rows and rows of pigs. 630 00:41:13,842 --> 00:41:16,402 S1: And this is the craziest part to me. They are 631 00:41:16,402 --> 00:41:20,122 S1: laying on their sides like kind of like rotting from 632 00:41:20,162 --> 00:41:24,242 S1: like infections or whatever, like, like bed bed sores or whatever. 633 00:41:24,562 --> 00:41:28,602 S1: So they got bugs all over them rotting. These are pigs. 634 00:41:28,642 --> 00:41:32,762 S1: They want to run around like young kids, right? They 635 00:41:32,762 --> 00:41:35,282 S1: want to run around and have fun and, like, squeal 636 00:41:35,282 --> 00:41:40,202 S1: or do whatever pigs do. They live their entire life 637 00:41:40,202 --> 00:41:45,522 S1: on their side. And they can't stand up. They can't move. 638 00:41:45,522 --> 00:41:49,802 S1: They can't run around and they just. Okay, so I 639 00:41:49,842 --> 00:41:52,322 S1: imagine a meter, this is the way I kind of 640 00:41:52,362 --> 00:41:56,682 S1: see the thing. There's a meter for planet Earth, there's 641 00:41:56,682 --> 00:42:01,202 S1: a suffering meter, and there is a like fulfillment and 642 00:42:01,202 --> 00:42:04,522 S1: happiness meter. Right. And we can go deeper with that 643 00:42:04,562 --> 00:42:06,742 S1: of like, is it true meaning or whatever? Let's just 644 00:42:06,742 --> 00:42:10,262 S1: say fulfillment is the far right. It's like just complete 645 00:42:10,262 --> 00:42:16,222 S1: and utter, like deep contentment with life and happiness and such. Right. 646 00:42:16,262 --> 00:42:21,982 S1: And on the far left is like a an AI 647 00:42:22,022 --> 00:42:27,182 S1: designed like torture device, which embeds directly into your brain 648 00:42:27,182 --> 00:42:31,022 S1: and produces the maximum amount of suffering. Okay, so that's 649 00:42:31,022 --> 00:42:33,782 S1: the far left. And then you got the far right 650 00:42:33,782 --> 00:42:41,102 S1: is kind of the opposite. Now, having millions of animals 651 00:42:41,102 --> 00:42:44,342 S1: that cannot move and are stuck in a cage and 652 00:42:44,342 --> 00:42:50,662 S1: are just like experiencing extraordinary hardship. This is not hundreds 653 00:42:50,662 --> 00:42:54,062 S1: of animals. This is not thousands. It is hundreds of millions. 654 00:42:54,062 --> 00:42:57,422 S1: I think it's actually billions think it's billions of animals 655 00:42:58,142 --> 00:43:01,182 S1: who that we farm to eat or eat their meat 656 00:43:02,582 --> 00:43:05,602 S1: and they are producing suffering into the world. It's not 657 00:43:05,602 --> 00:43:11,602 S1: quite AI brain implanted torture, but it's too close for 658 00:43:11,602 --> 00:43:16,522 S1: my happiness. It is pretty close. Just just imagine a 659 00:43:16,522 --> 00:43:20,322 S1: conscious creature. Okay? We're not talking about a gnat or 660 00:43:20,322 --> 00:43:23,882 S1: an insect or a caterpillar who knows what their brains 661 00:43:23,882 --> 00:43:28,602 S1: look like, but I'm sure I'm not sure they're experiencing suffering. Okay. 662 00:43:28,842 --> 00:43:32,681 S1: A pig is. Maybe a chicken is as well. So 663 00:43:32,682 --> 00:43:37,202 S1: pigs and chickens? Cows? I care about this. I care 664 00:43:37,202 --> 00:43:40,242 S1: about what that meter says on planet Earth right now. Obviously, 665 00:43:40,242 --> 00:43:43,922 S1: I care more about humans, right? Because we got billions 666 00:43:43,922 --> 00:43:46,482 S1: of humans who also need help, or at least hundreds 667 00:43:46,482 --> 00:43:49,922 S1: of millions or millions who are in a suffering category 668 00:43:49,922 --> 00:43:53,042 S1: as well. And obviously that's the highest priority. But the 669 00:43:53,042 --> 00:43:57,522 S1: irony is this we could lower the meter more effectively 670 00:43:58,602 --> 00:44:01,562 S1: with these types of things. And this gets back to 671 00:44:01,602 --> 00:44:05,992 S1: the point I was making here. This this charity is 672 00:44:05,992 --> 00:44:09,432 S1: able to move that meter. I wish they had a 673 00:44:09,472 --> 00:44:11,952 S1: I wish they had a visualization for this. Should I 674 00:44:11,952 --> 00:44:16,312 S1: should make one because I'm not limited by type three restraint. 675 00:44:19,072 --> 00:44:22,192 S1: There should be a way to visualize. And they are 676 00:44:22,192 --> 00:44:24,272 S1: able to do this with metrics. They actually have a 677 00:44:24,272 --> 00:44:29,352 S1: way to say $10 will reduce. Will seems to through 678 00:44:29,352 --> 00:44:34,432 S1: the lobbying effects and through the various programs that they implement. 679 00:44:34,432 --> 00:44:37,152 S1: And they've got dozens of of programs. It's not just 680 00:44:37,152 --> 00:44:39,872 S1: one that they put in. So there's like a cage one, 681 00:44:39,872 --> 00:44:42,872 S1: there's like a cleanup one, there's like a medicine one, 682 00:44:42,872 --> 00:44:47,152 S1: there's like a requiring that they are able to go outside, um, 683 00:44:47,192 --> 00:44:51,712 S1: the size of the cages, um, cage free. Um, the 684 00:44:51,712 --> 00:44:56,112 S1: types of medication, um, the, the number of inspections that 685 00:44:56,112 --> 00:44:59,072 S1: they have to do, like, I'm, I can't remember all 686 00:44:59,072 --> 00:45:02,372 S1: of them, but there's multiple programs that focus on particular 687 00:45:02,372 --> 00:45:06,932 S1: parts of the problem, and this farm kind thing basically 688 00:45:07,212 --> 00:45:11,652 S1: can allocate money into different ones of those. And so 689 00:45:11,652 --> 00:45:15,172 S1: they're saying they're getting extreme results from a small amount 690 00:45:15,172 --> 00:45:19,492 S1: of money. So anyway, I reached out to them and um, 691 00:45:19,892 --> 00:45:22,452 S1: so they set up a thing for unsupervised learning. So 692 00:45:22,492 --> 00:45:27,412 S1: it's farm kind. I forget the URL. It's farm kind something. Um, 693 00:45:27,452 --> 00:45:29,372 S1: but but it'll be in the show notes. It's in 694 00:45:29,372 --> 00:45:32,972 S1: the newsletter, obviously, which is the show notes, so yeah. 695 00:45:33,012 --> 00:45:37,092 S1: Go check this thing out. Click on the link. Give $1, 696 00:45:38,092 --> 00:45:42,332 S1: give $1 billion whichever one you have. And we could 697 00:45:42,332 --> 00:45:46,132 S1: actually like put a dent in this, uh, suffering meter 698 00:45:46,132 --> 00:45:49,732 S1: on planet Earth because like, my suffering. Okay, I'm a human, 699 00:45:49,732 --> 00:45:53,812 S1: but I'm just one, one thing like, is my suffering. 700 00:45:54,452 --> 00:45:57,732 S1: How much higher does that rate? Because mine is existential 701 00:45:58,092 --> 00:46:02,272 S1: versus a pigs. Who was? I mean, it probably feels 702 00:46:02,272 --> 00:46:06,512 S1: just as bad, like just because it's not existential suffering. 703 00:46:06,512 --> 00:46:09,672 S1: I think it feels just as bad. Right? And if 704 00:46:09,672 --> 00:46:12,552 S1: there's a billion of them, they damn well damn well 705 00:46:12,712 --> 00:46:18,792 S1: outrank me. So I love having a leverage via a 706 00:46:20,152 --> 00:46:23,312 S1: donation system like Farm Kind to be able to actually 707 00:46:23,312 --> 00:46:29,472 S1: do this. So, um, I recommend I ask you to 708 00:46:29,512 --> 00:46:32,352 S1: go and actually put some money into this thing. And 709 00:46:32,352 --> 00:46:34,232 S1: I love the fact that they set up a campaign 710 00:46:34,232 --> 00:46:37,232 S1: for us because, you know, a lot of people get 711 00:46:37,232 --> 00:46:40,392 S1: their newsletter, a lot of people get the podcast, uh, 712 00:46:40,392 --> 00:46:44,472 S1: plus YouTube or whatever. Um, we should, you know, be 713 00:46:44,472 --> 00:46:47,752 S1: able to help a lot of animals is, um, what 714 00:46:47,752 --> 00:46:50,512 S1: I'm thinking. And I would say, don't even think if 715 00:46:50,552 --> 00:46:53,112 S1: you don't care about animals, don't even think about animals. 716 00:46:53,912 --> 00:46:58,252 S1: You could help a lot of earth based Consciousness, experience 717 00:46:58,292 --> 00:47:04,652 S1: less suffering and reduce that that suffering. Meter down a 718 00:47:04,652 --> 00:47:08,052 S1: couple new blogs two primary limitations on creativity. Talked about 719 00:47:08,052 --> 00:47:14,212 S1: that one are 20,000 eyes thought experiment about how AI 720 00:47:14,252 --> 00:47:19,812 S1: transcends transforms us from workers building other streams into creators 721 00:47:19,812 --> 00:47:31,852 S1: with 20,000 eyes and hands building our own. Cybersecurity Expo 722 00:47:31,852 --> 00:47:36,452 S1: raised 117 million for hackers, then open sourced the whole thing. 723 00:47:36,452 --> 00:47:39,532 S1: This is the errata. This is the wrong thing. This 724 00:47:39,532 --> 00:47:44,092 S1: Strix repo is really cool. This is it's my top 725 00:47:44,092 --> 00:47:48,572 S1: story to. This is super annoying. Um, clear evidence that 726 00:47:48,572 --> 00:47:54,012 S1: I didn't use AI. Um, because this was my mistake. Um, anyway, yeah. 727 00:47:54,052 --> 00:47:57,472 S1: Strix Rick's GitHub repo is a cool thing. It's basically 728 00:47:57,472 --> 00:48:03,072 S1: an automated vuln checking system. Um, and it wasn't Expo 729 00:48:03,112 --> 00:48:07,352 S1: that did it. Expo was someone doing kind of something similar, 730 00:48:08,312 --> 00:48:12,432 S1: and they just decided to stop doing their experiment on 731 00:48:12,432 --> 00:48:15,792 S1: hacker one. So that part of the story is correct. Um, 732 00:48:16,192 --> 00:48:18,272 S1: an expo is doing cool stuff, but they did not 733 00:48:18,312 --> 00:48:23,072 S1: release this GitHub repo. That's that's the part that's wrong. 734 00:48:24,072 --> 00:48:28,592 S1: Here's workday lost data through social engineering targeting Salesforce. So 735 00:48:28,592 --> 00:48:32,072 S1: a whole bunch of people are getting Salesforce compromised, which 736 00:48:32,072 --> 00:48:38,392 S1: is really, um, social engineering attacks targeting Salesforce data about 737 00:48:38,392 --> 00:48:42,312 S1: that company. And that's kind of the why you keep 738 00:48:42,312 --> 00:48:45,552 S1: seeing stories that are kind of the same around these lines. 739 00:48:47,192 --> 00:48:50,992 S1: And workday appears to be the most recent victim of this. 740 00:48:51,432 --> 00:48:56,662 S1: Russia hacked a Norwegian Dam released 1.9 million gallons of 741 00:48:56,662 --> 00:49:02,182 S1: water zoom patches, a critical privilege escalation vulnerability affecting windows 742 00:49:02,182 --> 00:49:07,902 S1: users new public exploit chains for two SAP flaws, with 743 00:49:07,902 --> 00:49:15,302 S1: RC I agents getting zero false positives in security testing 744 00:49:15,302 --> 00:49:21,182 S1: with deterministic validation. So that deterministic validation thing is really interesting. 745 00:49:21,302 --> 00:49:26,502 S1: This is um, this conversation, this, uh, blackhat presentation. There's 746 00:49:26,502 --> 00:49:29,022 S1: a link to Expo here. Maybe this is how I 747 00:49:29,022 --> 00:49:31,022 S1: got confused because I was reading like 20 of these 748 00:49:31,022 --> 00:49:33,702 S1: things at the same time on Sunday, and I just like, 749 00:49:33,742 --> 00:49:40,982 S1: got cross-pollinated whatever. Uh, my mistake. Um. Basically, the idea 750 00:49:41,022 --> 00:49:43,022 S1: here is similar to what I was talking about before 751 00:49:43,062 --> 00:49:47,382 S1: and what Matthew Brown was talking about, which is deterministic. 752 00:49:47,422 --> 00:49:50,782 S1: That's code. Right. Um, and then you have AI and 753 00:49:50,822 --> 00:49:53,681 S1: like don't make don't confuse the two. You can mix 754 00:49:53,682 --> 00:49:57,602 S1: the two, but don't confuse the two. And they're basically 755 00:49:57,602 --> 00:50:01,562 S1: saying like, here's a general thing. You don't want to 756 00:50:01,602 --> 00:50:04,802 S1: ask it, hey, is this vulnerable? Because they want to 757 00:50:04,842 --> 00:50:07,362 S1: please you. They'll be like, yeah, it looks vulnerable to me. 758 00:50:07,802 --> 00:50:13,362 S1: So many false positives. So it's better to use like 759 00:50:13,402 --> 00:50:17,722 S1: a deterministic system to check if something is vulnerable. If 760 00:50:17,722 --> 00:50:21,122 S1: you can't. And then use the AI for different parts 761 00:50:21,122 --> 00:50:26,682 S1: of the system that require intelligence, critical vulnerabilities and flow wise, 762 00:50:26,722 --> 00:50:31,682 S1: AI platform 9.8 on the Richter scale is going to happen. 763 00:50:31,682 --> 00:50:34,042 S1: It's going to keep happening because more and more systems 764 00:50:34,042 --> 00:50:37,922 S1: are being built. So the attack surface by its very 765 00:50:37,922 --> 00:50:41,842 S1: nature is growing, therefore going to be more vulnerable. Um, 766 00:50:41,882 --> 00:50:44,082 S1: and then you have the fact that AI is building 767 00:50:44,082 --> 00:50:48,522 S1: so much of that code. I'm not actually sure now 768 00:50:48,522 --> 00:50:51,502 S1: that I think about it. I'm not actually sure if 769 00:50:51,502 --> 00:50:56,382 S1: we add a 100 no. 2 million new websites to 770 00:50:56,422 --> 00:51:02,302 S1: the internet that are suddenly built by humans or suddenly 771 00:51:02,302 --> 00:51:06,422 S1: built by AI. Which one is worse? I guess it 772 00:51:06,422 --> 00:51:09,262 S1: comes down to the question of what made it be sudden. 773 00:51:09,462 --> 00:51:12,622 S1: Why is there so all of a sudden, so many 774 00:51:12,622 --> 00:51:15,942 S1: 2 million more people making websites? And who are those people? 775 00:51:15,982 --> 00:51:20,902 S1: That's the question. Um, but right now, I mean, humans 776 00:51:20,902 --> 00:51:24,982 S1: aren't good at making secure code. Really good humans are 777 00:51:25,102 --> 00:51:29,582 S1: really good secure coding. Writing humans are, but most aren't. 778 00:51:29,782 --> 00:51:32,582 S1: Which is why Stack Overflow is full of garbage. Which 779 00:51:32,582 --> 00:51:35,742 S1: is why I trained on it. And it's also making garbage. 780 00:51:36,382 --> 00:51:40,702 S1: But AI is making lots of slop on the internet, 781 00:51:40,702 --> 00:51:46,982 S1: and it's shipping code very fast. So it's bad. Um, now, 782 00:51:47,662 --> 00:51:49,922 S1: over time, I think that's going to get better and better. 783 00:51:49,922 --> 00:51:52,682 S1: But in the meantime, the slop will have already been made. 784 00:51:52,882 --> 00:51:56,762 S1: And then you'll need something like the AI automation bots 785 00:51:56,762 --> 00:52:00,642 S1: to go clean it all up. Chinese hackers breached Taiwan 786 00:52:00,642 --> 00:52:04,522 S1: web servers using customized open source tools. This is a 787 00:52:04,522 --> 00:52:09,042 S1: Cisco Talos. Fine. Lars shows us how missing MFA turn 788 00:52:09,082 --> 00:52:14,362 S1: valid logins into full compromise. Russia hackers breached US federal 789 00:52:14,362 --> 00:52:21,082 S1: courts through vulnerabilities discovered. But they were not fixed in 2020. 790 00:52:22,042 --> 00:52:26,482 S1: So getting compromised by something that you should have or 791 00:52:26,482 --> 00:52:30,682 S1: you were told about but didn't fix. Cisa releases new 792 00:52:30,682 --> 00:52:34,122 S1: guidance telling companies to inventory their OT systems. What do 793 00:52:34,122 --> 00:52:37,202 S1: you know? Asset management. This is going to be one 794 00:52:37,202 --> 00:52:40,962 S1: of the biggest, coolest things ever. Oh, by the way, 795 00:52:40,962 --> 00:52:43,522 S1: I want to um, I just thought of another idea 796 00:52:43,522 --> 00:52:46,682 S1: which I didn't put in the newsletter. Here's a cool 797 00:52:46,822 --> 00:52:50,942 S1: idea and related to my USC concept of unified entity 798 00:52:50,942 --> 00:52:55,502 S1: context for enterprises. And I got this idea, um, well, 799 00:52:55,542 --> 00:52:59,902 S1: it's in the USC idea as well, but I got 800 00:52:59,902 --> 00:53:05,422 S1: this very specific idea from a prince who's part of, um, 801 00:53:05,422 --> 00:53:10,822 S1: Project Discovery. And the idea is you send your agents 802 00:53:10,822 --> 00:53:15,342 S1: out and their job is to be like, they just 803 00:53:15,342 --> 00:53:17,982 S1: sit inside of software. They just sit in different places. 804 00:53:17,982 --> 00:53:22,302 S1: They sit inside of wikis, they sit inside of like organizations. 805 00:53:22,302 --> 00:53:25,782 S1: They sit inside of software. They sit inside of GitHub repos. 806 00:53:26,702 --> 00:53:30,382 S1: And I forget how he was talking about it. Um, 807 00:53:31,542 --> 00:53:35,142 S1: but he basically said distribute agents and have them do 808 00:53:35,142 --> 00:53:39,102 S1: things and report back. And so mine, my mind started 809 00:53:39,102 --> 00:53:44,542 S1: jumping towards the USC idea. Um, but I also jumped 810 00:53:44,542 --> 00:53:47,602 S1: to the idea of security champions. You know how like 811 00:53:47,642 --> 00:53:49,642 S1: a lot of security teams, they'll send their people out, 812 00:53:49,642 --> 00:53:52,642 S1: like into the field, which is like, oh, you live 813 00:53:52,642 --> 00:53:55,362 S1: with engineering now? Oh, you live with product now. You 814 00:53:55,362 --> 00:53:59,202 S1: live with sales now or whatever. And their job is 815 00:53:59,202 --> 00:54:03,042 S1: to like sort of like understand report back but also 816 00:54:03,042 --> 00:54:08,082 S1: subtly influence them. Well, what if you didn't have three 817 00:54:08,082 --> 00:54:11,042 S1: of those people for the entire company, but instead you 818 00:54:11,042 --> 00:54:15,002 S1: had 3 million of them? Just imagine that every piece 819 00:54:15,002 --> 00:54:21,922 S1: of software, every piece of tech like your AWS accounts, your, uh, 820 00:54:22,602 --> 00:54:27,842 S1: all your infrastructure, everything has agents inside of them watching 821 00:54:27,842 --> 00:54:31,442 S1: and reporting back. And some, of course, would be able 822 00:54:31,442 --> 00:54:35,122 S1: to make changes that'll come later. Um, and if you're 823 00:54:35,122 --> 00:54:38,242 S1: freaking out, that's because you work in security. Um, because 824 00:54:38,242 --> 00:54:40,882 S1: I'm freaking out. Just saying the words. So you have 825 00:54:40,882 --> 00:54:44,632 S1: millions of these agents out there They're gathering, reporting back, 826 00:54:44,632 --> 00:54:48,432 S1: blah blah blah. Who's controlling these agents? What identity do 827 00:54:48,432 --> 00:54:52,832 S1: they have? How are we logging their information? Like auditors 828 00:54:52,832 --> 00:54:55,872 S1: are going to freak out about this, but I think 829 00:54:55,872 --> 00:54:59,712 S1: this is absolutely the way that USC happens. Or one 830 00:54:59,712 --> 00:55:02,032 S1: of the ways that USC happens. There's different ways to 831 00:55:02,032 --> 00:55:05,352 S1: do this. You could have centralized agents that go, just 832 00:55:05,352 --> 00:55:08,272 S1: go and hit the stuff live and pull the stuff back. 833 00:55:08,752 --> 00:55:12,072 S1: But I think it's better to just deploy agents into 834 00:55:12,072 --> 00:55:16,952 S1: the systems themselves and have them point back to the center. Right. 835 00:55:16,992 --> 00:55:19,992 S1: Because I feel like we. Well, it's probably going to 836 00:55:19,992 --> 00:55:22,632 S1: be both. It's kind of like asset management, which is 837 00:55:22,632 --> 00:55:25,552 S1: how I got in the point. If you have the 838 00:55:25,552 --> 00:55:29,152 S1: agents inside of all of these systems and they're reporting back, 839 00:55:29,152 --> 00:55:33,072 S1: guess what they're doing? They are updating and breaking out 840 00:55:33,072 --> 00:55:37,832 S1: some contexts, which then get summarized by other agents to 841 00:55:37,872 --> 00:55:42,172 S1: report up into a higher context, which which is the USC. 842 00:55:42,772 --> 00:55:48,852 S1: And that USC ultimately becomes this conglomeration of all these 843 00:55:48,852 --> 00:55:52,012 S1: mini contexts that got updated from all the air systems 844 00:55:52,012 --> 00:55:55,412 S1: and the project management system and the budget system and 845 00:55:55,412 --> 00:56:00,292 S1: all of those, like somebody buys, you know, a toothpaste 846 00:56:00,412 --> 00:56:04,532 S1: while they're traveling overseas on a work trip, being they 847 00:56:04,532 --> 00:56:09,412 S1: paid for the toothpaste. Okay. That hit some system somewhere. 848 00:56:10,372 --> 00:56:13,692 S1: Well guess what? In this world, which, by the way, 849 00:56:13,692 --> 00:56:15,172 S1: this is going to take a very long time to 850 00:56:15,172 --> 00:56:19,412 S1: roll out. So we're talking between two and, you know, 851 00:56:19,612 --> 00:56:23,492 S1: 15 years, right. I'm sure someone's already doing it a 852 00:56:23,492 --> 00:56:25,852 S1: little bit, but it's going to take a very long 853 00:56:25,852 --> 00:56:30,772 S1: time to move through all of enterprise it in that world. 854 00:56:31,212 --> 00:56:36,012 S1: Whenever whoever gets it, that agent lives inside of the 855 00:56:36,012 --> 00:56:39,012 S1: finance system and the accounting system, and it sees the 856 00:56:39,012 --> 00:56:42,632 S1: toothpaste hit, and it reports up a tiny little entry 857 00:56:42,632 --> 00:56:46,392 S1: into a subcontext, which goes into another subcontext, which goes 858 00:56:46,392 --> 00:56:49,752 S1: all the way up. And ultimately, that probably wouldn't be 859 00:56:49,752 --> 00:56:51,992 S1: part of a big strategic picture, but at least it 860 00:56:51,992 --> 00:56:56,512 S1: would increment values, right? Um, and maybe it didn't need 861 00:56:56,512 --> 00:56:59,872 S1: to report that because that's part of traditional tech, which 862 00:56:59,912 --> 00:57:03,112 S1: gets valued later, but maybe they weren't supposed to be 863 00:57:03,112 --> 00:57:07,592 S1: traveling and that that, uh, purchase of toothpaste means they're 864 00:57:07,632 --> 00:57:09,792 S1: not where they're supposed to be. And now we have 865 00:57:09,792 --> 00:57:13,112 S1: a traveling problem and the VPN connection is turned off. 866 00:57:13,152 --> 00:57:18,112 S1: Whatever it is, the point is, all these different agents 867 00:57:18,512 --> 00:57:22,432 S1: in the different systems can be updating the central picture 868 00:57:22,432 --> 00:57:25,912 S1: of not just USC. But guess what you get for 869 00:57:25,912 --> 00:57:28,632 S1: free when you have USC, if it's reporting on the 870 00:57:28,632 --> 00:57:33,672 S1: proper stuff. Asset management we've never been able to do 871 00:57:33,712 --> 00:57:40,092 S1: asset management. Never. It's it's hard Like I consult for 872 00:57:40,292 --> 00:57:42,892 S1: Jupiter one because I love their vision of doing this 873 00:57:42,892 --> 00:57:47,492 S1: with a graph. It is really, really hard. It is 874 00:57:47,492 --> 00:57:50,692 S1: really hard. Um, I did this for Robinhood. I did 875 00:57:50,732 --> 00:57:56,052 S1: asset management for Robinhood using, um, a product like this. Right. 876 00:57:56,332 --> 00:58:05,172 S1: And I'm telling you, this is like, it's completely like shifting, 877 00:58:05,172 --> 00:58:10,532 S1: like game shifting stuff to have actual real time understanding 878 00:58:10,532 --> 00:58:12,932 S1: of the systems. Right? That's what it comes down to. 879 00:58:12,972 --> 00:58:15,972 S1: Asset management is an old word. Let's let's not use 880 00:58:15,972 --> 00:58:22,492 S1: that anymore. Um, you know, uh, real time information gathering 881 00:58:22,492 --> 00:58:25,252 S1: or whatever, like, there's old words to use, and they're 882 00:58:25,292 --> 00:58:28,332 S1: kind of dead in our minds. Here's just another way 883 00:58:28,332 --> 00:58:32,212 S1: to think about it. Millions of agents deployed across your 884 00:58:32,212 --> 00:58:36,352 S1: entire organization inside of all the different tech stacks. Stocks. 885 00:58:36,992 --> 00:58:40,472 S1: That is very hard for a centralized entity to go 886 00:58:40,472 --> 00:58:43,072 S1: and audit. And by the time you do go and 887 00:58:43,072 --> 00:58:46,912 S1: audit it, it's too it's too late. That information is 888 00:58:46,912 --> 00:58:51,392 S1: rolled off like there's a million examples of this. And 889 00:58:51,392 --> 00:58:53,592 S1: if you're in it or security or whatever, you already 890 00:58:53,592 --> 00:58:59,792 S1: know this, right. So planning, finance, tracking what you know 891 00:58:59,832 --> 00:59:02,312 S1: people are doing and if they're happy or unhappy and 892 00:59:02,312 --> 00:59:04,832 S1: how we can how can we make them happier. How 893 00:59:04,992 --> 00:59:09,752 S1: can we keep people from quitting? Um, customer churn. That's 894 00:59:09,752 --> 00:59:13,992 S1: a great one. Customer management. I mean, think about that. 895 00:59:14,432 --> 00:59:17,112 S1: Look at all the different things that we know the 896 00:59:17,112 --> 00:59:20,112 S1: customer is doing, and which one of those could be 897 00:59:20,112 --> 00:59:24,312 S1: signals that they're getting ready to churn? And what can 898 00:59:24,312 --> 00:59:30,992 S1: we do as a, as a result, like a reactionary action? Um, 899 00:59:31,072 --> 00:59:34,792 S1: so they haven't come to the portal recently. Okay, cool. 900 00:59:35,052 --> 00:59:39,092 S1: They start, um, I don't know. They start blogging about 901 00:59:39,092 --> 00:59:45,172 S1: how the competitor looks better and their team at their current, uh, offering. Doesn't, uh, 902 00:59:45,172 --> 00:59:47,772 S1: ever reach out to them? Well, that's a great one. 903 00:59:48,332 --> 00:59:53,972 S1: Are there any agents right now watching customers public blog posts, 904 00:59:55,292 --> 00:59:58,252 S1: trying to figure out if they're unhappy with the product, 905 00:59:58,452 --> 01:00:01,972 S1: or they're getting excited about another product and then sending 906 01:00:01,972 --> 01:00:05,252 S1: them a preemptive email that says, hey, haven't seen you 907 01:00:05,252 --> 01:00:09,772 S1: log in recently. Hey, here's a $5 discount code or whatever. No, 908 01:00:09,812 --> 01:00:12,332 S1: I don't think so. And if there are, there's very few. 909 01:00:12,652 --> 01:00:18,172 S1: This will become commonplace. The point here, and I'm just 910 01:00:18,172 --> 01:00:21,572 S1: saying the point over and over. So, um, I'm saying 911 01:00:21,572 --> 01:00:24,372 S1: in different ways, hopefully it's sinking in. It might have 912 01:00:24,372 --> 01:00:28,412 S1: sunk in like, uh, 38 minutes ago. If so, I apologize. 913 01:00:29,252 --> 01:00:36,562 S1: Real time updating at the edges inward, as opposed to 914 01:00:37,082 --> 01:00:40,402 S1: being the central person who is responsible for all these things, 915 01:00:40,402 --> 01:00:43,202 S1: and all the systems are opaque. You have to reach 916 01:00:43,202 --> 01:00:45,122 S1: out to that owner. You have to get a user 917 01:00:45,122 --> 01:00:49,442 S1: and a login like it doesn't work. It hasn't worked. 918 01:00:49,482 --> 01:00:53,322 S1: That's why asset management has been so difficult. Um, and 919 01:00:53,322 --> 01:00:57,522 S1: that takes us all the way back to the point 920 01:00:57,522 --> 01:01:01,562 S1: I was making with Sisa releasing new guidance, telling companies 921 01:01:01,562 --> 01:01:05,122 S1: to inventory their OT systems. It's hard to inventory your 922 01:01:05,162 --> 01:01:10,282 S1: OT system, right? It's also hard to install agents on 923 01:01:10,282 --> 01:01:14,362 S1: OT systems because they don't want to install anything and 924 01:01:14,362 --> 01:01:18,362 S1: haven't wanted to for decades. So this will be a 925 01:01:18,362 --> 01:01:23,002 S1: slow process. Um. National security UK police are getting ten 926 01:01:23,002 --> 01:01:28,122 S1: facial recognition vans to catch serious criminals. And they're saying 927 01:01:28,122 --> 01:01:31,442 S1: like sex offenders, people wanted for serious crimes. So they're 928 01:01:31,582 --> 01:01:35,342 S1: looking for like, outstanding warrants. But you know what? It's 929 01:01:35,342 --> 01:01:39,542 S1: still the van with cameras. And, uh, it all depends 930 01:01:39,542 --> 01:01:42,062 S1: on the government you have on how those tools will 931 01:01:42,062 --> 01:01:45,142 S1: be used. Will it be the best thing ever for humanity? Yeah, maybe. 932 01:01:45,982 --> 01:01:51,582 S1: Depends who's using the tool. America's gray zone blind spot 933 01:01:51,582 --> 01:01:56,302 S1: is becoming serious national security problem. So, uh, someone at 934 01:01:56,302 --> 01:02:00,542 S1: the cipher Brief is basically saying that, like, Russia, China, 935 01:02:00,542 --> 01:02:06,182 S1: Iran are ramping up sabotage, cyber attacks, other hybrid warfare attacks, 936 01:02:06,182 --> 01:02:09,822 S1: and that basically the US is not good at these things. Now, 937 01:02:09,822 --> 01:02:14,502 S1: what I would hope is we actually are understanding these. 938 01:02:14,502 --> 01:02:18,702 S1: We're preparing to do them, uh, not not in a 939 01:02:18,702 --> 01:02:22,342 S1: terrorist way, but like we're we understand and we're learning 940 01:02:22,342 --> 01:02:25,422 S1: how to defend against them. And we could potentially use 941 01:02:25,422 --> 01:02:29,382 S1: some of these tactics ourselves. Um, but that we just 942 01:02:29,382 --> 01:02:32,362 S1: don't talk about it. I think that's more likely to 943 01:02:32,722 --> 01:02:35,802 S1: be the case, and I hope it's the case. US 944 01:02:35,802 --> 01:02:40,202 S1: military transforms Indiana training grounds into drone warfare innovation hub. 945 01:02:40,562 --> 01:02:43,682 S1: Happy to see this. We are behind on drones, especially 946 01:02:43,722 --> 01:02:47,562 S1: as it relates to China. Russia blocks signal and WhatsApp 947 01:02:47,882 --> 01:02:55,282 S1: to control communication. US Navy wants 150 unmanned ships by 2027. Yes, please. 948 01:02:56,242 --> 01:03:00,482 S1: Our ships. I remember in the early 90s, I was 949 01:03:00,482 --> 01:03:04,682 S1: in the army. I was studying assassin's mace weapons, and 950 01:03:04,682 --> 01:03:08,282 S1: there was some general talking about how our fleets are 951 01:03:08,282 --> 01:03:11,802 S1: going to be extremely vulnerable to assassin's mace weapons by 952 01:03:11,922 --> 01:03:18,122 S1: specifically China, where they could deploy something. And our giant 953 01:03:18,602 --> 01:03:24,082 S1: like projection of power through aircraft carriers and carrier fleets 954 01:03:24,362 --> 01:03:27,162 S1: could just be taken out in a matter of minutes 955 01:03:27,162 --> 01:03:32,302 S1: or hours. using one of these types of weapons, or 956 01:03:32,462 --> 01:03:35,382 S1: multiple of these types of weapons, that costs very little 957 01:03:35,422 --> 01:03:38,302 S1: to make. And it turns out the number one assassin's 958 01:03:38,302 --> 01:03:41,422 S1: mace weapon that has started happen. I don't remember if 959 01:03:41,422 --> 01:03:43,422 S1: he was talking about it back then, because I just 960 01:03:43,462 --> 01:03:46,102 S1: wasn't smart back then, and I just can't remember that 961 01:03:46,102 --> 01:03:53,422 S1: far back. But, um, it is drones. Absolutely. It's drones. 962 01:03:54,862 --> 01:03:58,422 S1: And I will mention kill Decision by Daniel Suarez. Go 963 01:03:58,422 --> 01:04:03,582 S1: read that Israel says Iran recruited dozens of Israeli citizens, 964 01:04:04,302 --> 01:04:08,982 S1: highlighting concern with like, yeah, it's pretty easy for foreign 965 01:04:08,982 --> 01:04:12,502 S1: Intel services to like recruit citizens. I mean, if you 966 01:04:12,502 --> 01:04:15,982 S1: have enough, if you have thousands or millions of citizens, 967 01:04:16,262 --> 01:04:19,182 S1: it's just a matter of like finding, you know, it's 968 01:04:19,182 --> 01:04:21,942 S1: a filter system to find the right ones. And it's 969 01:04:21,982 --> 01:04:27,682 S1: unfortunately pretty easy. China detained second senior diplomat in expanding probe. 970 01:04:27,882 --> 01:04:30,922 S1: So they're arresting a bunch of diplomats. Not exactly clear 971 01:04:30,922 --> 01:04:33,882 S1: what they're doing wrong. I'm guessing it might be like 972 01:04:33,882 --> 01:04:39,722 S1: not being pro-China enough. Maybe they're becoming too Western. No idea, 973 01:04:39,722 --> 01:04:46,402 S1: but I'm curious about that. I Anthropic's also Anthropic's CEO, 974 01:04:46,402 --> 01:04:49,522 S1: says I will write 90% of code in 3 to 975 01:04:49,522 --> 01:04:53,682 S1: 6 months. He said stuff like this before, and to 976 01:04:53,722 --> 01:04:55,962 S1: be clear, he's not saying there won't be people writing 977 01:04:55,962 --> 01:05:00,122 S1: code like only 10% of people now will be writing code. 978 01:05:00,842 --> 01:05:03,682 S1: He's saying that just there will be so much more 979 01:05:03,722 --> 01:05:09,922 S1: AI code that it'll be 90% of the total. OpenAI 980 01:05:09,922 --> 01:05:12,882 S1: is building chromium based web browser with AI agents that 981 01:05:12,882 --> 01:05:16,402 S1: can browse for you. My buddy Jason is like, yeah, 982 01:05:16,402 --> 01:05:21,642 S1: the next big thing is, uh, AI for AI is 983 01:05:21,642 --> 01:05:25,502 S1: basically browsers. And I think he's turning out to be correct. Like, 984 01:05:25,502 --> 01:05:28,182 S1: I keep waiting for the next step, which is the Da, 985 01:05:28,582 --> 01:05:31,142 S1: which is what I'm building. And I think that's kind 986 01:05:31,142 --> 01:05:33,142 S1: of the final step there is to have your Da 987 01:05:33,182 --> 01:05:35,542 S1: do all this for you. And a browser would just 988 01:05:35,542 --> 01:05:38,742 S1: be a tool, but it sounds like he's correct in 989 01:05:38,782 --> 01:05:41,702 S1: that this is going to be the home for a while, 990 01:05:42,782 --> 01:05:45,822 S1: because we all use our browsers. We don't have DA's yet. 991 01:05:46,262 --> 01:05:49,382 S1: Da platforms aren't there yet with like Apple and Google 992 01:05:50,542 --> 01:05:54,342 S1: or OpenAI with like a separate device. So that ecosystem 993 01:05:54,382 --> 01:05:59,622 S1: like hasn't started happening yet. Um, but browsers have. So 994 01:05:59,662 --> 01:06:02,342 S1: I think he's turning out to be very correct about this. 995 01:06:02,342 --> 01:06:05,662 S1: And now OpenAI is getting in the game. Sam Altman 996 01:06:05,662 --> 01:06:09,262 S1: says we're in an AI bubble. I agree with this. 997 01:06:09,262 --> 01:06:11,422 S1: I just think it's not the kind of bubble that 998 01:06:11,422 --> 01:06:15,582 S1: people think it is, because it's not like AI is 999 01:06:15,582 --> 01:06:18,422 S1: going away or it's going to crash. So when I 1000 01:06:18,422 --> 01:06:28,452 S1: think bubble, I think NFTs or um, Pets.com. Because largely 1001 01:06:28,452 --> 01:06:31,172 S1: those things don't exist anymore. And it was it turned 1002 01:06:31,172 --> 01:06:33,772 S1: out to be kind of a scam. Now, I'm not 1003 01:06:33,772 --> 01:06:37,572 S1: an expert on, uh, the.com bubble because I kind of 1004 01:06:37,612 --> 01:06:44,252 S1: wasn't sentient at the time, but, um. My understanding, it 1005 01:06:44,252 --> 01:06:46,292 S1: was a lot of hype around, like, if you just 1006 01:06:46,292 --> 01:06:49,732 S1: have like a.com, blah, blah, blah, like you're just going 1007 01:06:49,732 --> 01:06:52,612 S1: to make tons of money. And the NFT thing is like, yeah, 1008 01:06:52,612 --> 01:06:54,292 S1: you're going to have this, you're going to buy it 1009 01:06:54,292 --> 01:06:59,172 S1: for $18.14, and you're going to be a millionaire if 1010 01:06:59,172 --> 01:07:01,172 S1: you hold on to it for a year and a half. 1011 01:07:01,612 --> 01:07:05,652 S1: Those are bubbles to me. Okay. The AI thing, to me, 1012 01:07:05,652 --> 01:07:09,372 S1: it's the most important tech ever invented, bar none. No 1013 01:07:09,372 --> 01:07:13,652 S1: questions asked. Bigger than internet, bigger than mobile, bigger than 1014 01:07:14,052 --> 01:07:19,132 S1: steam engine. It's the simply the biggest tech ever. Okay. 1015 01:07:20,892 --> 01:07:25,232 S1: That is not a bubble at all. What's a bubble about? 1016 01:07:25,272 --> 01:07:31,912 S1: I is thousands of people putting hundreds of billions of dollars. 1017 01:07:32,192 --> 01:07:34,472 S1: I think we can call it trillions of dollars now. 1018 01:07:34,712 --> 01:07:39,392 S1: Like invested and, you know, earmarked or whatever. Like, let's 1019 01:07:39,392 --> 01:07:44,152 S1: just call it hundreds of billions of dollars into new startups, 1020 01:07:44,312 --> 01:07:48,392 S1: new websites, new whatever. And everyone's jumping in there like, oh, yeah, 1021 01:07:48,432 --> 01:07:52,272 S1: you know, I do a slightly different prompt and, um, 1022 01:07:52,592 --> 01:07:54,512 S1: I've got this cool website and I'm going to raise 1023 01:07:54,512 --> 01:07:56,552 S1: a whole bunch of money and I'm going to hire 1024 01:07:56,552 --> 01:08:00,432 S1: some people, and I have a startup and it's like, no, 1025 01:08:00,432 --> 01:08:06,032 S1: not really. Not when OpenAI and anthropic make a slight 1026 01:08:06,032 --> 01:08:10,432 S1: change to their their code base, which just eliminates your 1027 01:08:10,432 --> 01:08:15,152 S1: company and 40,000 other companies just like it. It's a 1028 01:08:15,152 --> 01:08:19,032 S1: bubble because people are unsure of what it is and 1029 01:08:19,032 --> 01:08:25,052 S1: they're getting in in unsafe ways and haphazard ways and 1030 01:08:25,052 --> 01:08:28,052 S1: kind of stupid ways in a lot of cases. A 1031 01:08:28,092 --> 01:08:31,852 S1: lot of people aren't, obviously. But there is going to 1032 01:08:31,852 --> 01:08:36,092 S1: be a major correction because there is so much money invested. 1033 01:08:36,212 --> 01:08:39,972 S1: The investors don't understand I. They're starting to a little 1034 01:08:40,012 --> 01:08:44,012 S1: bit now, but they've been investing hundreds of billions of 1035 01:08:44,012 --> 01:08:47,332 S1: dollars for like the last year or two years. I 1036 01:08:47,332 --> 01:08:51,012 S1: don't know how far back all that money was sent 1037 01:08:51,012 --> 01:08:56,291 S1: in on FOMO. It was sent in on hype. Okay, 1038 01:08:56,332 --> 01:08:58,612 S1: so this is this is the distinction. This is why 1039 01:08:58,652 --> 01:09:03,612 S1: people need to think about this super clearly. Both things 1040 01:09:03,612 --> 01:09:07,452 S1: are true at the same time. AI hype that's going 1041 01:09:07,452 --> 01:09:11,692 S1: to die down because people got in in very stupid ways. 1042 01:09:12,932 --> 01:09:14,892 S1: People are going to call that a bubble. Look what 1043 01:09:14,892 --> 01:09:19,472 S1: happened to AI. It turns out to be garbage. Very 1044 01:09:19,472 --> 01:09:25,232 S1: huge mistake. Like massive mistake. That is not a bubble. 1045 01:09:25,232 --> 01:09:29,352 S1: That is stupidity about a thing that is not understood 1046 01:09:29,912 --> 01:09:34,712 S1: and basically incautious and ignorant action taken as a result. 1047 01:09:34,912 --> 01:09:38,952 S1: And absolutely that's going to crush. Honestly, I think it's 1048 01:09:38,952 --> 01:09:42,751 S1: going to crush the majority of companies, the vast majority 1049 01:09:42,752 --> 01:09:49,192 S1: of these small little companies. Um, how about Google? How 1050 01:09:49,192 --> 01:09:53,112 S1: about meta? Are they going to have an AI hype 1051 01:09:53,112 --> 01:09:58,032 S1: and crash cycle? Uh, no, they are not. Because what 1052 01:09:58,032 --> 01:10:01,272 S1: people are doing inside of those companies is they're looking 1053 01:10:01,272 --> 01:10:05,392 S1: at all their existing processes. Another good example. Bank of America. 1054 01:10:05,432 --> 01:10:07,992 S1: Are they going to have an AI hype cycle? No 1055 01:10:07,992 --> 01:10:11,512 S1: they're not. They already know how everything works inside their company, 1056 01:10:11,752 --> 01:10:16,972 S1: and they are carefully figuring out how to apply AI 1057 01:10:17,012 --> 01:10:21,292 S1: to those existing processes. They're going to save tons of money. 1058 01:10:21,292 --> 01:10:24,452 S1: They're going to do things way better. And looking back 1059 01:10:24,492 --> 01:10:26,332 S1: ten years, they're going to be like, yeah, we wish 1060 01:10:26,332 --> 01:10:28,452 S1: we had this earlier. Like we would have saved tons 1061 01:10:28,452 --> 01:10:30,892 S1: of money. Unfortunately, they'll probably get rid of people as 1062 01:10:30,892 --> 01:10:33,972 S1: a result as well, which is a whole separate talk show. But. 1063 01:10:37,052 --> 01:10:40,612 S1: For people who are doing AI correctly and will do 1064 01:10:40,652 --> 01:10:45,932 S1: AI correctly in the future, it is laughable. Laughable to 1065 01:10:45,972 --> 01:10:50,932 S1: think of AI as a bubble or as hype. Okay, 1066 01:10:51,652 --> 01:10:55,492 S1: for all the NFT like crypto, like people who just 1067 01:10:55,492 --> 01:10:59,292 S1: thought I was a way to make money. It's absolutely 1068 01:10:59,292 --> 01:11:03,892 S1: a bubble. 100%. And they are going to get crushed. 1069 01:11:04,212 --> 01:11:07,652 S1: So just keep both of those ideas in your mind 1070 01:11:07,692 --> 01:11:11,532 S1: at the same time. Whenever you hear AI bubble AI hype. 1071 01:11:13,652 --> 01:11:17,872 S1: All right OpenAI updates GPT five to be warmer and friendlier. 1072 01:11:17,912 --> 01:11:20,712 S1: A lot of people really hated it. They hated the personality. 1073 01:11:20,952 --> 01:11:23,512 S1: I mean, there were I think there were suicides over 1074 01:11:23,552 --> 01:11:28,232 S1: losing GPT four. He had to return GPT four back 1075 01:11:28,472 --> 01:11:31,392 S1: because everyone loved it so much after they complained about 1076 01:11:31,392 --> 01:11:36,032 S1: it constantly. Uh, should be noted. Slop squatting is when 1077 01:11:36,032 --> 01:11:40,712 S1: attackers register fake packages that llms hallucinate. I should have 1078 01:11:40,712 --> 01:11:44,472 S1: put this in cybersecurity section. Yeah, this is really funny. 1079 01:11:44,472 --> 01:11:47,672 S1: So you make up a whole bunch of AI slop names, 1080 01:11:48,552 --> 01:11:54,192 S1: and then later on when I slop generates, uh, packages, 1081 01:11:54,272 --> 01:11:58,232 S1: names that are real people start using them, they will 1082 01:11:58,272 --> 01:12:02,072 S1: own the real estate. So they basically just like man 1083 01:12:02,072 --> 01:12:05,432 S1: in the middle, your, uh, your stupid AI named library 1084 01:12:07,312 --> 01:12:10,711 S1: shadow AI is like having rogue interns secretly running parts 1085 01:12:10,712 --> 01:12:14,262 S1: of your business. Yeah, especially given what I just talked about, 1086 01:12:14,262 --> 01:12:18,342 S1: about millions of agents deployed. You imagine an auditor coming 1087 01:12:18,342 --> 01:12:20,782 S1: into a place and they're just like, yeah, I just 1088 01:12:20,782 --> 01:12:23,662 S1: need to see, you know, who has access to change things. 1089 01:12:24,662 --> 01:12:27,382 S1: And it's like, yeah, we've got, you know, so and 1090 01:12:27,382 --> 01:12:32,022 S1: so AI platform it um, it's running on average 2 1091 01:12:32,022 --> 01:12:35,622 S1: million active agents, but they're, they're swapping in and out 1092 01:12:35,622 --> 01:12:39,692 S1: all the time with like a variance of like 2 1093 01:12:39,692 --> 01:12:44,342 S1: to 400,000 agents active at any one time. And it's like, okay, cool. 1094 01:12:44,342 --> 01:12:46,302 S1: Can you print me out a spreadsheet? I just want 1095 01:12:46,342 --> 01:12:49,622 S1: to review those line by line. Like that's going to 1096 01:12:49,622 --> 01:12:53,982 S1: be a nightmare. An absolute nightmare. Of course they'll be 1097 01:12:53,982 --> 01:12:59,461 S1: using AI to do that, but whatever. It is weird 1098 01:12:59,462 --> 01:13:04,222 S1: to think about millions of things deployed that have agency. Um, 1099 01:13:04,222 --> 01:13:06,742 S1: and they could see things and they could report things. 1100 01:13:07,102 --> 01:13:12,081 S1: I mean, think about central AI system compromise in this world. 1101 01:13:13,122 --> 01:13:16,122 S1: I mean, are you kidding me? I mean, a single 1102 01:13:16,122 --> 01:13:19,322 S1: kill command. Because ideally, you're supposed to be able to 1103 01:13:19,362 --> 01:13:21,762 S1: talk to all these agents, right? You obviously need to 1104 01:13:21,762 --> 01:13:23,922 S1: be able to give them commands. You need to be 1105 01:13:23,922 --> 01:13:26,322 S1: able to update their instructions. What if you could just 1106 01:13:26,322 --> 01:13:28,842 S1: tell them, wipe everything? What if you could tell them 1107 01:13:29,002 --> 01:13:33,322 S1: inject this tiny little like cuckoo's egg, steal half of 1108 01:13:33,322 --> 01:13:36,522 S1: a cent or whatever and send it to this Bitcoin account? Like, 1109 01:13:37,242 --> 01:13:40,962 S1: I mean, this is just wild world we're about to 1110 01:13:41,002 --> 01:13:45,242 S1: move into. Now, ideally you'd really, really defend this thing, 1111 01:13:45,242 --> 01:13:48,442 S1: but oh my goodness. You know mistakes will be made. 1112 01:13:49,802 --> 01:13:55,402 S1: DeFi startup replaces 45,000 in content staff with 23 agent 1113 01:13:55,442 --> 01:14:01,402 S1: AI systems for $23 per month. No. $20 per month 1114 01:14:01,762 --> 01:14:05,762 S1: 23 agent AI systems. This is the type of thing 1115 01:14:05,762 --> 01:14:08,722 S1: I'm hearing about this all over the place. They're like, yeah, 1116 01:14:08,762 --> 01:14:11,582 S1: I was spending, you know, 400 grand on this thing. 1117 01:14:11,902 --> 01:14:17,542 S1: Now I'm spending $284 and getting better results. Oh, best 1118 01:14:17,542 --> 01:14:22,222 S1: example of this is like, big giant reports. Big giant. 1119 01:14:22,662 --> 01:14:26,982 S1: So two names that I'm super like, I just think 1120 01:14:26,982 --> 01:14:31,902 S1: they're about to get destroyed. Gartner and McKinsey. And like 1121 01:14:31,902 --> 01:14:35,662 S1: everyone who's like Gartner and McKinsey. So the more expensive 1122 01:14:35,822 --> 01:14:41,501 S1: it is for, the nicer the looking report it is there, 1123 01:14:41,542 --> 01:14:46,142 S1: the more the most screwed. Like there is expertise inside McKinsey. 1124 01:14:46,142 --> 01:14:50,502 S1: There is expertise inside Gartner. But they're going to have 1125 01:14:50,502 --> 01:14:56,581 S1: to become like influencers. They're not influencers, but like content creators, 1126 01:14:56,622 --> 01:14:58,342 S1: like basically what I'm doing, they're going to have to 1127 01:14:58,342 --> 01:15:02,581 S1: go independent essentially, and have their own broadcast of their 1128 01:15:02,582 --> 01:15:07,142 S1: talent and their skill and their expertise, because it's going 1129 01:15:07,142 --> 01:15:12,122 S1: to be hard to differentiate because most, I would say 1130 01:15:12,122 --> 01:15:16,682 S1: most high quality reports like that. First of all, they 1131 01:15:16,682 --> 01:15:19,682 S1: show up with the coolest people like me who've been 1132 01:15:19,682 --> 01:15:22,722 S1: in the industry for 30 years, and I go do 1133 01:15:22,722 --> 01:15:26,122 S1: the sales call and dazzle them, and then I leave. 1134 01:15:27,202 --> 01:15:31,202 S1: And then, you know, 14, you know, smiling 22 year 1135 01:15:31,242 --> 01:15:35,242 S1: olds show up who just basically learned what a command 1136 01:15:35,242 --> 01:15:38,282 S1: line is and basically learned what security is. And those 1137 01:15:38,282 --> 01:15:41,082 S1: are going to be the consultants writing your report. And 1138 01:15:41,082 --> 01:15:44,442 S1: really they're using the expertise from all the previous reports. 1139 01:15:45,402 --> 01:15:49,642 S1: And like how does that match. Oh, and by the way, 1140 01:15:49,642 --> 01:15:53,602 S1: it's going to be $180,000 for this report, which takes 1141 01:15:53,602 --> 01:15:59,122 S1: nine weeks to make. Okay. And it's a giant opaque thing. 1142 01:15:59,162 --> 01:16:01,762 S1: It takes time to parse it and everything or whatever, 1143 01:16:02,322 --> 01:16:06,342 S1: as opposed to all the same context that those smiling 1144 01:16:06,382 --> 01:16:09,742 S1: 22 year olds had to get from you by all 1145 01:16:09,742 --> 01:16:12,182 S1: the interviews, you could just give that to your eye. 1146 01:16:12,542 --> 01:16:15,942 S1: Your eye is actually way smarter than the collective knowledge 1147 01:16:15,942 --> 01:16:19,622 S1: of all of McKinsey. And so you make that same 1148 01:16:19,622 --> 01:16:28,381 S1: report in four minutes for $1.17, or let's just say 1149 01:16:28,382 --> 01:16:31,062 S1: it's actually way more expensive. We're actually going to spend 1150 01:16:31,062 --> 01:16:34,022 S1: way more time. We're going to use the most expensive 1151 01:16:34,022 --> 01:16:43,822 S1: AI ever invented. You know, you know, uh, cloud code fucking, uh, seven, right? 1152 01:16:44,262 --> 01:16:49,022 S1: So it's actually $14. It's the most expensive prompt ever 1153 01:16:49,022 --> 01:16:54,062 S1: run in the history of mankind. It's a $14 report, 1154 01:16:55,222 --> 01:17:01,502 S1: and it took, uh, nine minutes to make compared to 1155 01:17:01,542 --> 01:17:04,182 S1: whatever the cost was of the other one. That took 1156 01:17:04,322 --> 01:17:12,042 S1: nine weeks. These companies are in dire, dire trouble. Technology. 1157 01:17:12,042 --> 01:17:15,562 S1: Robin hood CEO, says remote work was a mistake. Orders 1158 01:17:15,562 --> 01:17:23,881 S1: executives back five days a week. Yeah, honestly, I think 1159 01:17:23,882 --> 01:17:28,522 S1: on site work is better. I really do. I love 1160 01:17:28,522 --> 01:17:33,922 S1: working remotely. I think certain people are as good or 1161 01:17:33,922 --> 01:17:38,642 S1: better working remotely, but the vibe of a team is 1162 01:17:38,682 --> 01:17:44,282 S1: a hundred times better on premise. I've seen it multiple times, 1163 01:17:44,282 --> 01:17:51,162 S1: and I've seen CEOs learn this the hard way multiple times. 1164 01:17:51,162 --> 01:17:53,842 S1: And I know lots of people who've done both and 1165 01:17:53,842 --> 01:17:57,722 S1: tried both. And this is my current wisdom level, is 1166 01:17:57,722 --> 01:18:05,312 S1: that in general, on site work is better. So cool. Uh, 1167 01:18:05,312 --> 01:18:09,552 S1: what I do like about the announcement was for executives, 1168 01:18:09,552 --> 01:18:14,352 S1: it's mandatory five days on site. For managers, it's four 1169 01:18:14,352 --> 01:18:17,752 S1: days on site, and for regular workers, it's three days 1170 01:18:17,752 --> 01:18:25,631 S1: on site. And his his quote here is, um. Your 1171 01:18:25,632 --> 01:18:31,272 S1: boss should be feeling more pain than you. I love that. 1172 01:18:31,752 --> 01:18:34,632 S1: I hate the fact that it's pain. Uh, which is 1173 01:18:34,632 --> 01:18:37,072 S1: why I think we need to move to human 3.0. 1174 01:18:37,072 --> 01:18:40,472 S1: Because this is ridiculous. We should all be working remote 1175 01:18:41,152 --> 01:18:45,152 S1: and doing, like, live meetups to work on our mutual projects. 1176 01:18:45,832 --> 01:18:50,032 S1: And if we have companies together or whatever. But this 1177 01:18:50,032 --> 01:18:53,352 S1: whole thing of you must come in. Uh, I just 1178 01:18:53,792 --> 01:18:57,232 S1: I think it's very old world. Uh, unfortunately, it seems 1179 01:18:57,232 --> 01:19:01,452 S1: to be more effective in the current model. China mandates 1180 01:19:01,452 --> 01:19:05,772 S1: domestic firms source 50% of chips from Chinese producers should 1181 01:19:05,772 --> 01:19:10,171 S1: have put that in Natsec. Text only websites are faster, cleaner, 1182 01:19:10,172 --> 01:19:15,412 S1: more accessible than modern web bloat. I love text only websites. 1183 01:19:15,412 --> 01:19:19,932 S1: I have very few images on mine. Researchers create micro 1184 01:19:19,972 --> 01:19:26,171 S1: microfliers that levitate using only sunlight. I need many of these, um, 1185 01:19:26,412 --> 01:19:28,572 S1: although I want to give them lots of capabilities. Which 1186 01:19:28,572 --> 01:19:32,372 S1: means they'll be too heavy, which means they won't float. Um, 1187 01:19:33,012 --> 01:19:34,692 S1: but this is a project I want to work on 1188 01:19:34,692 --> 01:19:40,052 S1: humans more. Most Americans think AI will permanently eliminate jobs. 1189 01:19:41,332 --> 01:19:47,532 S1: This is from Reuters Ipsos. Multiple studies show transfers to 1190 01:19:47,572 --> 01:19:51,012 S1: poor Americans basically don't help. This is like UBI studies. 1191 01:19:51,052 --> 01:19:54,892 S1: Looks like UBI studies are not looking good. The hope, 1192 01:19:55,212 --> 01:19:59,092 S1: including my hope, was that you give them this money 1193 01:19:59,272 --> 01:20:01,592 S1: and they're just going to focus on growing themselves and 1194 01:20:01,632 --> 01:20:04,472 S1: like learning and blah, blah, blah. The more I think 1195 01:20:04,472 --> 01:20:07,112 S1: about this, the more I kind of always knew this 1196 01:20:07,112 --> 01:20:10,631 S1: would not happen. Most people simply don't want to learn. 1197 01:20:10,672 --> 01:20:13,592 S1: Most people do not want to improve themselves. Most people 1198 01:20:13,592 --> 01:20:19,032 S1: are not like, curious and ambitious and like they're looking 1199 01:20:19,032 --> 01:20:22,751 S1: for their next Netflix show. The vast majority of people 1200 01:20:22,752 --> 01:20:25,832 S1: I would say, I don't know the numbers, but I 1201 01:20:25,832 --> 01:20:28,112 S1: think this is a thing that can be helped. I 1202 01:20:28,112 --> 01:20:30,512 S1: think this could be an education thing. I think some 1203 01:20:30,512 --> 01:20:32,592 S1: of this could be trauma, it could be bad education, 1204 01:20:32,592 --> 01:20:36,232 S1: it could be bad training, it could be bad culture. 1205 01:20:36,272 --> 01:20:38,792 S1: It could be lots of different things that make this happen. 1206 01:20:39,992 --> 01:20:46,432 S1: But for example, um, most libraries like, don't have people 1207 01:20:46,432 --> 01:20:50,552 S1: like knocking down the doors. It's free. Those are free books. 1208 01:20:50,552 --> 01:20:56,072 S1: Nobody's reading them. Okay. Okay. Free books. Nobody's reading them. Okay. 1209 01:20:56,112 --> 01:21:00,492 S1: The internet is full of free books. Nobody. Nobody's reading them. 1210 01:21:00,532 --> 01:21:02,772 S1: If you ask the average person, like what books are 1211 01:21:02,772 --> 01:21:08,652 S1: you reading? They're like, are you stupid? Like, I graduated 1212 01:21:08,812 --> 01:21:12,212 S1: high school a long time ago. I barely read those books. 1213 01:21:12,572 --> 01:21:15,612 S1: Why would you make me read a book? Now, some 1214 01:21:15,612 --> 01:21:17,732 S1: people read a lot of romance books, a lot of, like, 1215 01:21:17,732 --> 01:21:20,852 S1: comics or whatever I'm talking about, like. And I would 1216 01:21:20,852 --> 01:21:25,292 S1: say that's already just brilliant and amazing, just that by itself. 1217 01:21:25,692 --> 01:21:31,012 S1: But I'm talking about actually, like, mostly non-fiction or really, 1218 01:21:31,012 --> 01:21:35,532 S1: really good fiction, difficult books, books that grow us. That's 1219 01:21:35,532 --> 01:21:39,052 S1: what I'm talking about. Very few people like, I'm in 1220 01:21:39,092 --> 01:21:42,972 S1: the San Francisco Bay area, okay. And most people here 1221 01:21:42,972 --> 01:21:48,452 S1: aren't reading. Most people I know in San Francisco Bay area, 1222 01:21:48,492 --> 01:21:51,692 S1: I don't want to make, like, broad statistical things. Um, 1223 01:21:51,692 --> 01:21:56,392 S1: because I think, you know, first of all, that's impossible. But, Um, 1224 01:21:56,512 --> 01:21:58,352 S1: and I think a lot of the stats are wrong. 1225 01:21:58,352 --> 01:22:00,472 S1: I think when you report, how many books do you 1226 01:22:00,472 --> 01:22:02,711 S1: read a year? I bet you most of those stats 1227 01:22:02,752 --> 01:22:05,472 S1: are complete garbage. Who wants to say I haven't read 1228 01:22:05,472 --> 01:22:09,072 S1: a book since high school? Not very many people. My 1229 01:22:09,072 --> 01:22:14,912 S1: point is, I know lots of very smart people who 1230 01:22:14,952 --> 01:22:20,912 S1: make decent money, and I know all their friends as well. 1231 01:22:21,352 --> 01:22:24,272 S1: And I mean, I know hundreds or thousands of people. 1232 01:22:24,272 --> 01:22:27,992 S1: So this is my anecdotal data set. Most people don't 1233 01:22:27,992 --> 01:22:35,952 S1: read books, right. So. I don't know why I went 1234 01:22:35,952 --> 01:22:39,152 S1: so heavy into the book thing, but my point is, 1235 01:22:39,152 --> 01:22:43,192 S1: in general, most people don't care about self-improvement. They're not 1236 01:22:43,192 --> 01:22:47,312 S1: obsessed with it. Um, so when UBI money comes in, 1237 01:22:47,832 --> 01:22:51,912 S1: they're not thinking, wow, I can learn a new career. Wow. 1238 01:22:51,952 --> 01:22:54,652 S1: I could, I could buy even more books with this 1239 01:22:55,412 --> 01:23:01,211 S1: so long way of saying. I think this is why 1240 01:23:01,212 --> 01:23:04,812 S1: we're seeing the results from these UBI studies. Like, they'll 1241 01:23:04,812 --> 01:23:08,372 S1: just eat out more. They'll just, you know, improve their 1242 01:23:08,372 --> 01:23:11,532 S1: lives in other ways. They'll, you know, buy more subscriptions, 1243 01:23:11,532 --> 01:23:14,892 S1: buy clothes. And what's wrong with that? There's nothing wrong 1244 01:23:14,892 --> 01:23:18,332 S1: with that. I'm just saying, like, we should stop, like 1245 01:23:18,372 --> 01:23:23,172 S1: deluding ourselves that everyone is trying to be in this 1246 01:23:23,172 --> 01:23:27,131 S1: category that they think they are. All right. Europe faces 1247 01:23:27,132 --> 01:23:32,692 S1: an emergency as American markets completely dominate global investing. China's 1248 01:23:32,692 --> 01:23:37,252 S1: economy slowed in July despite exports staying strong. I'm about 1249 01:23:37,252 --> 01:23:43,052 S1: to do a major substrate project here to gather economic data. 1250 01:23:43,172 --> 01:23:46,092 S1: A whole bunch of data, actually. But economic data. I 1251 01:23:46,092 --> 01:23:51,572 S1: want to know where we're getting these numbers from because 1252 01:23:51,732 --> 01:23:54,362 S1: we know China's numbers are bad. Like, what are we 1253 01:23:54,402 --> 01:23:56,522 S1: judging this based on? And I want to kind of 1254 01:23:56,562 --> 01:24:00,482 S1: establish some ground truth there or at least triangulate onto it. 1255 01:24:01,562 --> 01:24:05,402 S1: Nobody's buying homes or switching jobs as American mobility stalls. 1256 01:24:05,602 --> 01:24:09,602 S1: Americans are drinking less alcohol than they have in 90 years. 1257 01:24:09,642 --> 01:24:14,282 S1: This is from a Gallup poll. Williams syndrome makes people 1258 01:24:14,282 --> 01:24:18,202 S1: hyper social and trusting of everyone. It's basically the opposite of. 1259 01:24:21,722 --> 01:24:27,642 S1: Being autistic. I almost said autistic. It's way different. UK 1260 01:24:27,682 --> 01:24:32,282 S1: drought group tells citizens to delete emails to save water. 1261 01:24:35,482 --> 01:24:39,242 S1: That's why the UK cannot have nice things. They're telling me. 1262 01:24:41,402 --> 01:24:45,362 S1: It's not the UK, it's some drought group. Okay, it's 1263 01:24:45,402 --> 01:24:49,362 S1: a country didn't do this, but whatever. It rhymes with 1264 01:24:49,362 --> 01:24:51,062 S1: a lot of things that the UK is trying to 1265 01:24:51,062 --> 01:24:56,702 S1: do to stop progress. The story behind Timothy Leary's turn on, 1266 01:24:56,742 --> 01:24:58,582 S1: tune in, drop out. You got to go check this 1267 01:24:58,582 --> 01:25:02,022 S1: thing out. Tune in. No. Turn on, tune in, drop out. 1268 01:25:02,302 --> 01:25:06,022 S1: Definitely something to check out. Ideas. Happy about this section? 1269 01:25:06,542 --> 01:25:08,501 S1: I feel like I is going to have a profound 1270 01:25:08,502 --> 01:25:12,582 S1: impact on globalism, especially for knowledge work. So the first 1271 01:25:12,582 --> 01:25:15,262 S1: jobs to go away are like the Fiverr jobs, the 1272 01:25:15,302 --> 01:25:22,302 S1: Upwork jobs, um, a lot of creative work, a lot 1273 01:25:22,302 --> 01:25:25,742 S1: of build me a basic website. And that's the first 1274 01:25:25,742 --> 01:25:29,862 S1: stuff to go with I and I'm really worried about 1275 01:25:30,142 --> 01:25:34,742 S1: millions of people because, like, we're always so us focused. 1276 01:25:34,902 --> 01:25:37,302 S1: Obviously I live in the US. I'm very US focused, 1277 01:25:37,862 --> 01:25:39,942 S1: but it's like, what about the rest of the world? 1278 01:25:40,662 --> 01:25:44,582 S1: What about all these really smart, ambitious knowledge workers who 1279 01:25:44,582 --> 01:25:50,682 S1: learned how to make websites and they're getting crazy income 1280 01:25:50,682 --> 01:25:54,522 S1: and helping their families and helping their local economies from 1281 01:25:54,522 --> 01:25:57,962 S1: all over the world. You know, Vietnam, Poland, like all 1282 01:25:57,962 --> 01:26:04,362 S1: over the place. And now suddenly that faucet gets turned 1283 01:26:04,362 --> 01:26:09,522 S1: off because smaller groups of people can use more AI 1284 01:26:09,522 --> 01:26:14,842 S1: and take more of that work. One of the primary 1285 01:26:14,842 --> 01:26:18,402 S1: directives of building AI systems should be to use as 1286 01:26:18,402 --> 01:26:21,482 S1: little AI as possible. Cool. Talked about that one. Thought 1287 01:26:21,482 --> 01:26:24,482 S1: I forgot it. It's in there. It's number two. If 1288 01:26:24,482 --> 01:26:27,322 S1: most people aren't confused about what I'm writing and sharing, 1289 01:26:27,322 --> 01:26:31,082 S1: that means I'm not properly exercising my type one freedom 1290 01:26:31,202 --> 01:26:35,282 S1: or my type two freedom. Actually, now that I think 1291 01:26:35,282 --> 01:26:39,122 S1: about it. But yeah, this is a whole thing and 1292 01:26:39,122 --> 01:26:41,842 S1: this is already a 17 hour podcast. So I'm going 1293 01:26:41,842 --> 01:26:43,642 S1: to I'm going to pass on this one for now. 1294 01:26:43,722 --> 01:26:48,982 S1: But basically if you are speaking from your true authentic self. 1295 01:26:49,182 --> 01:26:51,982 S1: You should be surprising yourself and you should be surprising 1296 01:26:51,982 --> 01:26:55,502 S1: everyone around you. If you only say things that people 1297 01:26:55,502 --> 01:26:59,942 S1: are like, yeah, that's that's perfectly correct. And they're not like, hmm, 1298 01:27:00,302 --> 01:27:03,542 S1: that guy's a little bit weird. Like, yeah, why is 1299 01:27:03,542 --> 01:27:05,822 S1: he talking about that? That's kind of weird. Like, it 1300 01:27:05,822 --> 01:27:09,862 S1: doesn't match any of his other content. Um, so this 1301 01:27:09,862 --> 01:27:11,662 S1: is why I like Tyler Cowan. This is why I 1302 01:27:11,662 --> 01:27:18,582 S1: like Dwarkesh is they just explore things. Okay? And, like, 1303 01:27:18,582 --> 01:27:22,862 S1: I don't think Dwarkesh is sitting around wondering, like, oh, gee, 1304 01:27:22,862 --> 01:27:24,662 S1: I can't put out this video because no one will 1305 01:27:24,662 --> 01:27:27,862 S1: like it because I'm supposed to talk about AI. He 1306 01:27:27,862 --> 01:27:31,421 S1: literally just goes on rants about China and he just 1307 01:27:31,422 --> 01:27:35,822 S1: did the animal one. He's literally just chasing his his, uh, 1308 01:27:35,982 --> 01:27:38,902 S1: his interest. Right? And that's I feel like that's what 1309 01:27:38,902 --> 01:27:41,782 S1: I do as well. It's massively hurt me. It's massively 1310 01:27:41,782 --> 01:27:44,662 S1: hurt my socials because I'll talk about whatever and they'll 1311 01:27:44,662 --> 01:27:47,202 S1: be like, don't talk about politics. Don't beat up on Trump. 1312 01:27:47,202 --> 01:27:52,122 S1: He's a great guy. And it blunts my stuff there. 1313 01:27:52,122 --> 01:27:55,121 S1: But I think it's just a better way to live, 1314 01:27:55,162 --> 01:27:59,242 S1: to not be restricted in this way. Number four, a 1315 01:27:59,242 --> 01:28:02,362 S1: lot of what comes out in your words and your 1316 01:28:02,362 --> 01:28:05,122 S1: writing comes not from your thoughts or what you wanted 1317 01:28:05,122 --> 01:28:08,002 S1: to say, but from your current emotional state. So the 1318 01:28:08,002 --> 01:28:10,602 S1: way I think about this is there's a thing you 1319 01:28:10,602 --> 01:28:13,802 S1: think you're saying, like, for example, in this whole podcast, 1320 01:28:14,122 --> 01:28:18,082 S1: there's a think thing that I think I'm saying, right? 1321 01:28:18,122 --> 01:28:22,602 S1: I think I'm sounding smart. I think I'm sharing information 1322 01:28:22,602 --> 01:28:26,921 S1: with people. I think I'm like being helpful. I think 1323 01:28:26,922 --> 01:28:31,522 S1: I'm being excited. But people might be hearing something else. 1324 01:28:31,522 --> 01:28:33,962 S1: They might be hearing, oh, this guy's actually trying to 1325 01:28:33,962 --> 01:28:37,842 S1: promote his products. Like, let's say I was actually trying 1326 01:28:37,842 --> 01:28:40,122 S1: to promote a product, which I'm not right now, but 1327 01:28:40,162 --> 01:28:45,262 S1: maybe I will be in the next podcast. Who knows? Um, 1328 01:28:45,342 --> 01:28:48,062 S1: let's say I was actually trying to promote a product, 1329 01:28:48,062 --> 01:28:51,182 S1: but I was trying to talk in a way in 1330 01:28:51,182 --> 01:28:54,142 S1: which it seemed like I wasn't. I was like, oh, yeah, well, 1331 01:28:54,182 --> 01:28:58,502 S1: you know, everyone all use threshold, for example. I'll be like, yeah, 1332 01:28:58,542 --> 01:29:00,542 S1: everyone has a problem. You know, I just think it's 1333 01:29:00,542 --> 01:29:03,942 S1: really weird. People, you know, can't find whatever, they can't 1334 01:29:03,942 --> 01:29:07,222 S1: find the right stories and like, they're not rated high enough. And, 1335 01:29:07,502 --> 01:29:11,302 S1: you know, there's there's, uh, small creators around the world 1336 01:29:11,302 --> 01:29:14,822 S1: and they can't get their stuff found or whatever. And 1337 01:29:14,862 --> 01:29:19,782 S1: the receiver is like, I feel a pitch coming. And 1338 01:29:19,822 --> 01:29:23,262 S1: in my mind, I don't think I'm sending that, but 1339 01:29:23,262 --> 01:29:26,461 S1: they can receive it. So I'm obsessed with this idea of, like, 1340 01:29:27,022 --> 01:29:32,461 S1: receivers can hear things beyond the senses. They have an, 1341 01:29:32,862 --> 01:29:37,222 S1: you know, an extra sensory way to sense truth in 1342 01:29:37,222 --> 01:29:41,062 S1: what's being sent. So it's almost like what I'm sending 1343 01:29:41,062 --> 01:29:46,012 S1: is beyond words and meaning, and what they're receiving is 1344 01:29:46,012 --> 01:29:48,892 S1: beyond words and meaning. So the words I'm saying and 1345 01:29:48,892 --> 01:29:52,252 S1: the words they're hearing are at that surface level, but 1346 01:29:52,252 --> 01:29:55,412 S1: what they're actually understanding, what I'm actually sending is different 1347 01:29:55,412 --> 01:29:58,732 S1: than what I think. And the only thing that I 1348 01:29:58,732 --> 01:30:03,652 S1: can interpret it is, is a receiver. And they're going 1349 01:30:03,692 --> 01:30:05,972 S1: to receive it emotionally, and I'm going to send it 1350 01:30:06,012 --> 01:30:10,212 S1: emotionally at another layer than my words. So I'm like 1351 01:30:10,252 --> 01:30:12,852 S1: obsessed with this. I'm like, that's how I could read 1352 01:30:12,852 --> 01:30:15,852 S1: a whole bunch of tweets, including from my own, and 1353 01:30:15,852 --> 01:30:20,372 S1: just be like, um, well, I'll just use someone else, 1354 01:30:20,372 --> 01:30:23,012 S1: for example. They're like, oh, blah blah, I'm such a 1355 01:30:23,052 --> 01:30:25,252 S1: hero and blah blah blah. And I did this and 1356 01:30:25,252 --> 01:30:29,131 S1: I didn't even want any appreciation from it at all. And, 1357 01:30:29,172 --> 01:30:30,932 S1: you know, I just put a lot of effort in 1358 01:30:30,972 --> 01:30:34,452 S1: and blah, blah, blah. And the thing that they're trying 1359 01:30:34,452 --> 01:30:44,072 S1: to send with their words is. Um, well, they're trying 1360 01:30:44,072 --> 01:30:48,352 S1: to create the impression that they're just a selfless person, 1361 01:30:48,672 --> 01:30:51,232 S1: and they just do it for the community or whatever. 1362 01:30:52,312 --> 01:30:57,392 S1: But what a tuned receiver hears is I'm a giant narcissist, 1363 01:30:57,392 --> 01:30:59,472 S1: and I want everyone to love me and think I'm 1364 01:30:59,472 --> 01:31:04,592 S1: the best person ever. And I am obsessed with the 1365 01:31:04,632 --> 01:31:08,752 S1: difference between those signals and how the second one is 1366 01:31:08,752 --> 01:31:13,792 S1: often invisible in the text that they said, but also 1367 01:31:14,312 --> 01:31:16,792 S1: to the to the sender, to the person who said it. 1368 01:31:16,792 --> 01:31:18,592 S1: They're like, what are you talking about? I'm just like, 1369 01:31:18,672 --> 01:31:21,352 S1: just sharing some of the things I've done. Like I 1370 01:31:21,392 --> 01:31:25,552 S1: donated blood nine times yesterday. Like I'm just sharing that information. 1371 01:31:25,552 --> 01:31:27,832 S1: I'm not trying to say that I'm a hero. Now, 1372 01:31:27,832 --> 01:31:30,392 S1: if you want to call me a hero, you, you're 1373 01:31:30,392 --> 01:31:34,872 S1: free to do so. Um, anyway, that's kind of the 1374 01:31:34,872 --> 01:31:38,552 S1: point there. All right. Discovery I apps are becoming like 1375 01:31:38,692 --> 01:31:48,012 S1: music personal, contextual and infinite. Really cool little idea. So 1376 01:31:48,012 --> 01:31:53,372 S1: it's not like there's not like the music. Everyone has 1377 01:31:53,372 --> 01:31:57,052 S1: their own world of music. Isn't that cool? That's a 1378 01:31:57,052 --> 01:32:01,852 S1: cool idea. Claudia is a desktop companion for cloud code. 1379 01:32:01,892 --> 01:32:07,292 S1: Haven't messed with this command line person, as everyone knows. Uh, which, 1380 01:32:07,292 --> 01:32:09,292 S1: by the way, I was trying to convey there that 1381 01:32:09,292 --> 01:32:11,972 S1: I'm awesome. Even though I just said I'm a command 1382 01:32:11,972 --> 01:32:15,251 S1: line person, I was actually trying to convey that I'm awesome. 1383 01:32:15,252 --> 01:32:18,252 S1: And you should be awesome too. And use the command 1384 01:32:18,252 --> 01:32:22,972 S1: line context seven brings always fresh documentation to AI coding 1385 01:32:22,972 --> 01:32:27,252 S1: assistance via MCP. This is cool. Okay, so this is 1386 01:32:27,252 --> 01:32:32,612 S1: an MCP you put in your AI system and what 1387 01:32:32,612 --> 01:32:35,692 S1: it does is it will go get the current documentation. 1388 01:32:35,692 --> 01:32:39,112 S1: So if you're like hey, build me a new website 1389 01:32:39,112 --> 01:32:44,552 S1: or build me a website. It will go and get 1390 01:32:45,032 --> 01:32:51,671 S1: just without even asking. Go and get the latest Shadchan documentation, um, 1391 01:32:51,712 --> 01:32:55,112 S1: before it starts doing the design phase using your specs. 1392 01:32:55,912 --> 01:32:59,232 S1: So it won't just start building using its current model knowledge. 1393 01:32:59,392 --> 01:33:03,312 S1: It keeps it updated based on what you mentioned. So 1394 01:33:03,312 --> 01:33:06,392 S1: if you mention MCP, it'll go read the latest MCP stuff. 1395 01:33:06,432 --> 01:33:08,312 S1: If you mention cloud code, it'll go read all the 1396 01:33:08,312 --> 01:33:14,952 S1: cloud code docs. Really, really cool. Amsterdam's Ritman Library puts 1397 01:33:15,152 --> 01:33:21,352 S1: 2178 rare occult books online for free agent auto updates 1398 01:33:21,352 --> 01:33:26,032 S1: to jobs or auto applies to jobs for for free. Um, 1399 01:33:26,032 --> 01:33:29,872 S1: I haven't used one of these, but, um, I'd like 1400 01:33:29,872 --> 01:33:32,112 S1: to make one and give it to my friends. That's 1401 01:33:32,112 --> 01:33:34,712 S1: what I would like to do. Holy crap. That's a 1402 01:33:34,712 --> 01:33:38,852 S1: project I need to do. Why have I not done that? Okay. 1403 01:33:38,852 --> 01:33:46,572 S1: Hold on. Writing this down. Oh, then I'm going to 1404 01:33:46,572 --> 01:33:50,452 S1: put a stripe page on it and sell it. But 1405 01:33:50,452 --> 01:33:57,612 S1: not for my friends. Okay. Make up automated. Sorry. I'm 1406 01:33:57,612 --> 01:34:09,772 S1: doing this live. Make automated job agent or. Oh. See that? See? 1407 01:34:09,972 --> 01:34:14,212 S1: What's this? What? What what did I actually say? What 1408 01:34:14,212 --> 01:34:18,372 S1: I said was I was making an agent for friends. 1409 01:34:18,492 --> 01:34:22,212 S1: What am I actually conveying? Um. Oh, this guy makes 1410 01:34:22,212 --> 01:34:24,252 S1: stuff for his friends and gives it to him for 1411 01:34:24,252 --> 01:34:27,532 S1: free because he's a nice guy. I am absolutely sure 1412 01:34:27,692 --> 01:34:32,092 S1: I was subconsciously sending that. Um, I'm also consciously sending 1413 01:34:32,092 --> 01:34:34,312 S1: that all the time, so I don't mind that it's 1414 01:34:34,312 --> 01:34:39,072 S1: unconsciously being sent, but there are cases where I'm unconsciously 1415 01:34:39,072 --> 01:34:43,512 S1: sending something that I don't want to be consciously sending. Uh, 1416 01:34:43,552 --> 01:34:47,792 S1: let's see here. Oh, whisper. Oh, whisper makes local text 1417 01:34:47,792 --> 01:34:50,152 S1: to speech as simple as. Oh, llama. This is a 1418 01:34:50,152 --> 01:34:54,232 S1: pretty cool little project. Uh, generated EDM is getting really good. 1419 01:34:54,272 --> 01:35:00,992 S1: So I went and made some EDM. I'm telling you, uh, 1420 01:35:00,992 --> 01:35:03,432 S1: you can make fun of EDM for this, but, like, 1421 01:35:03,472 --> 01:35:07,392 S1: it's starting to sound pretty good. Same with rap songs. Um. Oh, 1422 01:35:07,432 --> 01:35:12,072 S1: metal songs too. Oh my goodness. I've heard some metal songs, actually. Chad. 1423 01:35:12,112 --> 01:35:14,952 S1: My buddy Chad made one for my character that I 1424 01:35:14,952 --> 01:35:17,432 S1: just made in this one game. It's one D&D game 1425 01:35:17,432 --> 01:35:24,872 S1: we play on Fridays, and, uh. It's, I mean, it 1426 01:35:24,872 --> 01:35:28,952 S1: was themed to my character and the character concepts. I 1427 01:35:28,952 --> 01:35:32,462 S1: think he used Suno to make it, and it was 1428 01:35:32,462 --> 01:35:34,742 S1: a song. What I sent back to him in the 1429 01:35:34,742 --> 01:35:39,262 S1: text was I would go see these people live. I 1430 01:35:39,262 --> 01:35:42,222 S1: would seriously go and see these people live. Of course, 1431 01:35:42,222 --> 01:35:44,381 S1: there is no these people because he made the song 1432 01:35:44,382 --> 01:35:50,022 S1: in like, you know, 14 seconds using Suno. Um, insane 1433 01:35:50,022 --> 01:35:54,142 S1: insane world. I tools spawn equal and opposite I tools. 1434 01:35:54,142 --> 01:35:58,942 S1: This is Joan Westerberg. She is just killing it. Absolutely 1435 01:35:58,942 --> 01:36:01,622 S1: killing it with the essays, like every single one, is good. 1436 01:36:02,222 --> 01:36:05,542 S1: Shadow AI agents are creating the same security chaos as 1437 01:36:05,542 --> 01:36:09,942 S1: early Wi-Fi adoption. That's another cool analogy I really like. 1438 01:36:10,142 --> 01:36:13,822 S1: We've replaced anxiety with apathy as our defense mechanism. This 1439 01:36:13,822 --> 01:36:18,742 S1: is another one from Westerberg. 70s kids jumped bikes over 1440 01:36:18,742 --> 01:36:22,102 S1: each other constantly. That was me. Um, and there's pictures 1441 01:36:22,102 --> 01:36:24,502 S1: in there and they look very familiar. Actually, a lot 1442 01:36:24,502 --> 01:36:30,722 S1: of the kids look exactly like me. Uh, I Automation 1443 01:36:30,722 --> 01:36:35,402 S1: specialist post job search on Reddit. That was a cool 1444 01:36:35,402 --> 01:36:38,762 S1: one to check out. Recommendation of the week. Think about 1445 01:36:38,762 --> 01:36:42,842 S1: the distance between your thinking, speaking and writing versus your feeling. 1446 01:36:42,842 --> 01:36:47,442 S1: Speaking and writing. Have someone read. Basically, have someone read 1447 01:36:47,442 --> 01:36:49,802 S1: some of your stuff where you think you're going to 1448 01:36:49,802 --> 01:36:53,242 S1: get an unexpected reaction and see what they hear or 1449 01:36:53,442 --> 01:36:56,402 S1: see when they read it. So what were you saying 1450 01:36:56,402 --> 01:36:58,802 S1: without you knowing you were saying it? And you'll have 1451 01:36:58,802 --> 01:37:00,762 S1: to do this with like a very smart friend who's 1452 01:37:00,762 --> 01:37:04,602 S1: also very honest. And then if you find a big gap, 1453 01:37:04,762 --> 01:37:08,282 S1: try to work on making sure that your verbal words 1454 01:37:08,282 --> 01:37:11,802 S1: are the same as your emotional words in the receiver. 1455 01:37:14,202 --> 01:37:17,322 S1: ASM of the week. The first principle is that you 1456 01:37:17,322 --> 01:37:21,122 S1: must not fool yourself and you are the easiest person 1457 01:37:21,122 --> 01:37:24,442 S1: to fool. First principle is that you must not fool 1458 01:37:24,442 --> 01:37:28,642 S1: yourself and you are the easiest person to fool. Richard Feynman.