1 00:00:00,740 --> 00:00:05,000 S1: Unsupervised Learning is a podcast about trends and ideas in cybersecurity, 2 00:00:05,030 --> 00:00:10,010 S1: national security, AI, technology and society, and how best to 3 00:00:10,039 --> 00:00:17,600 S1: upgrade ourselves to be ready for what's coming. All right, 4 00:00:17,600 --> 00:00:20,090 S1: welcome to Unsupervised Learning. This is Daniel Miessler, and we're 5 00:00:20,090 --> 00:00:22,849 S1: going to start off with this really cool story from 6 00:00:22,850 --> 00:00:27,470 S1: Reversing Labs. So basically what their team found was that 7 00:00:27,470 --> 00:00:34,130 S1: it's possible to exploit local models on hugging face if 8 00:00:34,130 --> 00:00:38,420 S1: they use the pickle format, which is used by TensorFlow models. 9 00:00:38,420 --> 00:00:41,210 S1: So it's used by a lot of different models. And 10 00:00:41,210 --> 00:00:45,650 S1: basically what pickle is, is it's a Python library for 11 00:00:45,650 --> 00:00:49,010 S1: parsing model files. Essentially it's a parser, right? It's a 12 00:00:49,010 --> 00:00:52,580 S1: parsing library. And everyone knows in security that parsing is 13 00:00:52,580 --> 00:00:56,060 S1: the root of all evil. So what they figured out 14 00:00:56,060 --> 00:01:01,480 S1: is that the scanning system, the security software for hugging 15 00:01:01,510 --> 00:01:06,160 S1: face would basically break. If you get over a certain 16 00:01:06,160 --> 00:01:09,369 S1: size or if you have like a broken file. So 17 00:01:09,400 --> 00:01:12,760 S1: what they figured out is that somebody had actually embedded 18 00:01:12,760 --> 00:01:17,920 S1: malware in a couple of these models. And this kind 19 00:01:17,920 --> 00:01:22,090 S1: of breaks people's brains because you think essentially that if 20 00:01:22,090 --> 00:01:25,360 S1: you're running a local model, it's safe, right? That that's 21 00:01:25,360 --> 00:01:27,940 S1: what everyone has been saying. If you run a local model, 22 00:01:27,940 --> 00:01:31,810 S1: it's completely safe. It's the services that are dangerous, right. 23 00:01:31,840 --> 00:01:35,919 S1: Like the deep seek service obviously that's in China, so 24 00:01:35,920 --> 00:01:38,560 S1: it's dangerous. But if it's a local model, it's fine. 25 00:01:38,650 --> 00:01:43,030 S1: What they found was look, it's still software, right? Running 26 00:01:43,030 --> 00:01:45,940 S1: a model is still software. You have a parsing library 27 00:01:45,940 --> 00:01:51,550 S1: involved if it's a pickle file based model. And yeah, 28 00:01:51,580 --> 00:01:56,169 S1: they were basically able to find malware inside of this 29 00:01:56,170 --> 00:01:59,380 S1: thing that was trying to dial home. It was trying 30 00:01:59,380 --> 00:02:02,590 S1: to dial home to an IP address inside of China. 31 00:02:02,590 --> 00:02:06,700 S1: So really, really cool finding and it's going to be 32 00:02:06,700 --> 00:02:10,780 S1: in the newsletter next week. But in the meantime, you 33 00:02:10,780 --> 00:02:14,410 S1: can get it from this URL right here. They they're 34 00:02:14,410 --> 00:02:20,230 S1: calling the attack nullify which is pretty good name. Nullify nullify. 35 00:02:20,230 --> 00:02:24,639 S1: And yeah it nullifies I guess hugging face security. So 36 00:02:24,669 --> 00:02:28,329 S1: yeah that was that was cool. All right. Continuing down here. 37 00:02:28,360 --> 00:02:31,450 S1: Oh yeah. We did the augmented version three. We talked 38 00:02:31,450 --> 00:02:35,169 S1: about Telos Files. This was last Friday. It was a 39 00:02:35,500 --> 00:02:38,680 S1: great class. Best of the three. It was definitely the 40 00:02:38,680 --> 00:02:41,650 S1: best of the three courses in my opinion went a 41 00:02:41,650 --> 00:02:45,010 S1: full four hours. It was quite good. The next one 42 00:02:45,010 --> 00:02:47,020 S1: I think is going to be online, although I said 43 00:02:47,020 --> 00:02:49,180 S1: that for the last one. Um, so it might be 44 00:02:49,180 --> 00:02:51,670 S1: another live class but haven't announced that one yet, so 45 00:02:51,669 --> 00:02:54,100 S1: don't worry about that. Um, going back to roots on 46 00:02:54,100 --> 00:02:56,950 S1: the story format here. So I'm basically if a story 47 00:02:56,950 --> 00:03:00,300 S1: is only one sentence, It's going to be one sentence. 48 00:03:00,480 --> 00:03:02,100 S1: I went back and looked at a whole bunch of 49 00:03:02,100 --> 00:03:05,640 S1: my old episodes where I just felt really comfortable with 50 00:03:05,639 --> 00:03:08,910 S1: doing that writing, and I was surprised at how short 51 00:03:08,940 --> 00:03:11,250 S1: a lot of the things were. I would just be like, hey, yeah, 52 00:03:11,250 --> 00:03:13,860 S1: this thing happened. I think it's kind of cool because 53 00:03:13,860 --> 00:03:18,150 S1: of this reason. Boom, whatever. That's like 12 words or whatever. 54 00:03:18,180 --> 00:03:20,730 S1: And I had a link in boom, that was it. 55 00:03:20,730 --> 00:03:23,580 S1: And I'm just going to kind of go back to that. 56 00:03:23,580 --> 00:03:26,970 S1: I've been so I focused, I want to write more 57 00:03:26,970 --> 00:03:30,179 S1: because I could use AI to help with summarizing, but 58 00:03:30,210 --> 00:03:33,180 S1: I want to focus more on manual writing of that 59 00:03:33,180 --> 00:03:36,330 S1: first sentence and use the AI more to like, gather 60 00:03:36,330 --> 00:03:40,800 S1: data and gather support and sources for the sentence that 61 00:03:40,800 --> 00:03:43,500 S1: I write myself, which I'm already writing a whole lot 62 00:03:43,530 --> 00:03:45,240 S1: of the sentence, but I just want to start from 63 00:03:45,240 --> 00:03:49,080 S1: scratch and write the sentence completely myself, which is how 64 00:03:49,080 --> 00:03:52,320 S1: I was doing it, obviously going back to 2015. So 65 00:03:52,320 --> 00:03:54,900 S1: that's a little bit of a change there. Uh, put 66 00:03:54,900 --> 00:03:58,140 S1: a bunch of new fabric patterns focusing on analyzing your 67 00:03:58,170 --> 00:04:01,290 S1: Telos file. And this is part of the the the 68 00:04:01,290 --> 00:04:04,590 S1: course just put out. And they all start with t underscore. 69 00:04:04,590 --> 00:04:07,830 S1: So if you look for those patterns those are telos oriented. 70 00:04:07,920 --> 00:04:11,250 S1: Um yeah. Had some loud music in my raycast video. 71 00:04:11,280 --> 00:04:14,970 S1: I'm re-uploading that I've been talking to the team about it. 72 00:04:14,970 --> 00:04:19,050 S1: Sometimes you might hear in the podcast edit, you're never 73 00:04:19,050 --> 00:04:21,450 S1: supposed to hear that the team is supposed to take 74 00:04:21,450 --> 00:04:24,900 S1: that and edit that out, which is why I say 75 00:04:24,900 --> 00:04:27,870 S1: the word. So I've had some talks with the team 76 00:04:27,870 --> 00:04:31,770 S1: and they are working on fixing that. But apologies in 77 00:04:31,770 --> 00:04:34,620 S1: the meantime. And yeah, I had some shenanigans with an 78 00:04:34,620 --> 00:04:37,890 S1: automated post to my X account a few days ago, 79 00:04:37,890 --> 00:04:40,440 S1: and it turns out it was like an automated like 80 00:04:40,470 --> 00:04:43,950 S1: Twitter application that looks like it might have been compromised 81 00:04:43,950 --> 00:04:47,669 S1: on the back end. Maybe it not sure exactly, but yeah, 82 00:04:47,700 --> 00:04:50,010 S1: it was really kind of weird. And I got an 83 00:04:50,040 --> 00:04:51,989 S1: email out to the security team to see if they 84 00:04:51,990 --> 00:04:54,450 S1: want to fix it. See, what have we got here? Um, 85 00:04:54,450 --> 00:04:59,789 S1: someone's using 2.8 million. This is under security. 2.8 million IPS. 86 00:04:59,790 --> 00:05:02,370 S1: To brute force the passwords of basically every type of 87 00:05:02,370 --> 00:05:05,520 S1: VPN and firewall device out there, and most of the 88 00:05:05,520 --> 00:05:08,219 S1: IPS are in Brazil, but lots of other countries as well. 89 00:05:08,250 --> 00:05:11,250 S1: Apple just patched another zero day that Citizens Lab says 90 00:05:11,250 --> 00:05:16,589 S1: was being used in extremely sophisticated, targeted attacks against specific 91 00:05:16,589 --> 00:05:20,640 S1: journalists and dissidents. Ken Wong from CSA released a detailed 92 00:05:20,640 --> 00:05:25,080 S1: framework called maestro. And the thing that this does, it's 93 00:05:25,080 --> 00:05:28,979 S1: a threat modeling framework, but it's specifically focused on AI agents, 94 00:05:28,980 --> 00:05:32,010 S1: and it's addressing gaps in things like stride and pasta 95 00:05:32,010 --> 00:05:35,429 S1: and other frameworks that just were pre AI. So they 96 00:05:35,430 --> 00:05:39,480 S1: don't really handle AI that well. Uh, also uh, deals 97 00:05:39,480 --> 00:05:43,830 S1: with emerging threats like goal misalignment, model extraction, adversarial attacks, 98 00:05:43,830 --> 00:05:48,750 S1: stuff like that. And researchers at Watchtower discovered 150 abandoned 99 00:05:48,779 --> 00:05:53,159 S1: around 150 abandoned S3 buckets used by a whole bunch 100 00:05:53,160 --> 00:05:58,150 S1: of software companies like governments. Romance, various types of infrastructure pipelines. 101 00:05:58,150 --> 00:06:03,880 S1: And the concern is if those S3 buckets and the artifacts, 102 00:06:03,880 --> 00:06:07,120 S1: the assets inside of those S3 buckets are being consumed 103 00:06:07,120 --> 00:06:09,940 S1: as part of a pipeline, well, that's ripe for a 104 00:06:09,940 --> 00:06:13,330 S1: supply chain attack. Cloudflare had a significant outage because someone 105 00:06:13,330 --> 00:06:16,630 S1: tried to block a phishing URL, accidentally turned off their 106 00:06:16,630 --> 00:06:21,820 S1: entire R2 storage, which R2 is basically S3. It's cloudflare's 107 00:06:21,820 --> 00:06:27,580 S1: answer to S3. And yeah, Cloudflare always has really good postmortems, 108 00:06:27,580 --> 00:06:30,580 S1: so you definitely want to check that one out. While 109 00:06:30,580 --> 00:06:34,420 S1: ARM released a report on API security that includes a 110 00:06:34,420 --> 00:06:38,500 S1: bunch of stats on AI services that use APIs, and 111 00:06:38,500 --> 00:06:40,720 S1: the big takeaway for me, which I kind of already 112 00:06:40,720 --> 00:06:45,640 S1: believed is AI security right now is largely API security, 113 00:06:45,790 --> 00:06:51,130 S1: and with the caveat being that it's AI API security, 114 00:06:51,130 --> 00:06:53,440 S1: so that your API has to be strong because that's 115 00:06:53,440 --> 00:06:57,570 S1: like that's the parsing system for going into your application. 116 00:06:57,570 --> 00:07:00,090 S1: But the added twist with AI is it's not only 117 00:07:00,120 --> 00:07:04,680 S1: API security, it's also whatever is being consumed through that API. 118 00:07:04,710 --> 00:07:08,010 S1: You also have the problem of prompt injection. And prompt 119 00:07:08,010 --> 00:07:12,810 S1: injection can like travel and like detonate multiple times or 120 00:07:12,810 --> 00:07:15,390 S1: get stored and detonate later. It's kind of like a 121 00:07:15,390 --> 00:07:19,230 S1: stored cross-site scripting type of vibe, except for it's even 122 00:07:19,230 --> 00:07:22,080 S1: more nuanced because it's English and it can detonate at 123 00:07:22,080 --> 00:07:28,890 S1: multiple levels, and different parsers can detonate the content differently, right? 124 00:07:28,920 --> 00:07:31,410 S1: It could be that you have like a five layer 125 00:07:31,440 --> 00:07:36,900 S1: nested prompt injection attack, and maybe some systems peel off 126 00:07:36,900 --> 00:07:39,510 S1: two of the layers, or some systems peel off only 127 00:07:39,510 --> 00:07:42,450 S1: one of the layers. Some systems peel off four of 128 00:07:42,450 --> 00:07:46,020 S1: the layers, but the fifth one gets through. So there's 129 00:07:46,020 --> 00:07:49,050 S1: a lot of extra complexity because you're using English to 130 00:07:49,080 --> 00:07:55,380 S1: try to trick systems. But big, big focus, however, on 131 00:07:55,380 --> 00:07:59,940 S1: AI security or API security with AI security. Ex-google engineers 132 00:07:59,940 --> 00:08:05,610 S1: facing espionage charges for passing confidential IP to China, Estonia, 133 00:08:05,640 --> 00:08:09,060 S1: Latvia and Lithuania just cut their last major tie to 134 00:08:09,090 --> 00:08:13,470 S1: Russia by switching their power grids from Soviet era systems 135 00:08:13,560 --> 00:08:17,010 S1: to the European continental network. That is great to see 136 00:08:17,010 --> 00:08:21,690 S1: that independence happening. And a bunch of Russian drone operators 137 00:08:21,690 --> 00:08:25,740 S1: were attacked with kind of like that, um, Hezbollah attack 138 00:08:25,740 --> 00:08:29,730 S1: where they attacked the pagers, but they were receiving booby 139 00:08:29,760 --> 00:08:32,880 S1: trapped headsets. But the problem was when they opened up 140 00:08:32,880 --> 00:08:36,420 S1: the packages which they were going to use, the packaging 141 00:08:36,420 --> 00:08:40,140 S1: was not perfect. And because the packaging was not perfect, 142 00:08:40,350 --> 00:08:43,949 S1: they they detected tampering and they were like, hey, something 143 00:08:43,980 --> 00:08:46,470 S1: is weird here. And of course, they probably thought of 144 00:08:46,470 --> 00:08:49,770 S1: the Hezbollah situation and they did not use them, but 145 00:08:49,770 --> 00:08:52,079 S1: if they did, it could have been similar to like 146 00:08:52,080 --> 00:08:55,280 S1: the Hezbollah attack where a bunch of Russian drone operators 147 00:08:55,280 --> 00:09:01,370 S1: get killed or injured from modified headsets. I in tech, 148 00:09:01,400 --> 00:09:05,839 S1: Gumroad says they're no longer hiring junior or mid-level engineers 149 00:09:05,840 --> 00:09:08,810 S1: because AI is handling most of that work. Now, this 150 00:09:08,809 --> 00:09:11,720 S1: is colossal. And also this is not an AI company. 151 00:09:11,720 --> 00:09:17,089 S1: So Salesforce said something like this. But Salesforce also makes 152 00:09:17,120 --> 00:09:21,170 S1: agents for or what is that thing called agent Force 153 00:09:21,170 --> 00:09:23,630 S1: or something like that. It's they have their own product. 154 00:09:23,630 --> 00:09:26,000 S1: So when the head of Salesforce says, hey, you know, 155 00:09:26,030 --> 00:09:28,280 S1: we don't even need engineers because our product is so 156 00:09:28,280 --> 00:09:33,020 S1: awesome that maybe it's true, maybe it's marketing. In this case, 157 00:09:33,020 --> 00:09:36,260 S1: this is just a regular company. Gumroad they're saying they 158 00:09:36,290 --> 00:09:40,670 S1: are not hiring junior or mid-level engineers anymore. The only 159 00:09:40,670 --> 00:09:44,990 S1: people they're going to hire are senior people, senior devs, 160 00:09:44,990 --> 00:09:48,439 S1: or senior architects who are really good with AI. I've 161 00:09:48,440 --> 00:09:52,160 S1: been saying this since very early 23, like February of 162 00:09:52,179 --> 00:09:55,059 S1: 23 and so have a lot of other people, right? 163 00:09:55,090 --> 00:09:57,370 S1: I'm not the only one. It's it's kind of obvious 164 00:09:57,370 --> 00:09:59,890 S1: where this is heading, right. And you get into a 165 00:09:59,890 --> 00:10:02,530 S1: real big problem with like the pipeline. But at the 166 00:10:02,530 --> 00:10:05,950 S1: same time this seems a little bit extreme. It seems 167 00:10:05,950 --> 00:10:10,209 S1: like they might be overstating their case, because I do 168 00:10:10,240 --> 00:10:13,960 S1: know the current state of AI agents, and it's not 169 00:10:13,960 --> 00:10:18,790 S1: quite at the level of replacing fully replacing junior engineers. 170 00:10:19,150 --> 00:10:23,199 S1: I don't know, in some tasks you can mid-level engineers? 171 00:10:23,200 --> 00:10:26,080 S1: I don't think so. I don't I don't think so. 172 00:10:26,080 --> 00:10:31,450 S1: Not right now. Not in February of 25. Maybe in 173 00:10:31,450 --> 00:10:34,390 S1: two months, maybe in six months, maybe in a year, 174 00:10:34,390 --> 00:10:39,670 S1: maybe inside of 2025. Yes. And of course, the details 175 00:10:39,670 --> 00:10:42,490 S1: are nuanced based on the use case, but I would 176 00:10:42,490 --> 00:10:49,360 S1: say junior engineers maybe, but not quite. Mid-level engineers? No, 177 00:10:49,600 --> 00:10:53,200 S1: and definitely not senior engineers. So I think this might 178 00:10:53,200 --> 00:10:56,890 S1: be a little bit padded, but, um, yeah, I would 179 00:10:56,890 --> 00:10:59,920 S1: say there's a lot of daylight between these tools and 180 00:10:59,920 --> 00:11:04,930 S1: a human developer with 105 IQ. There really is there's 181 00:11:04,929 --> 00:11:07,000 S1: a there's a lot of daylight between those two things 182 00:11:07,000 --> 00:11:11,350 S1: still in February of 25. So one of Chatgpt's main architects, 183 00:11:11,380 --> 00:11:15,609 S1: John Schulman, left anthropic after only five months. And the 184 00:11:15,610 --> 00:11:20,020 S1: rumor is he's going to join Mira at her AI startup. 185 00:11:20,020 --> 00:11:23,260 S1: I'm not sure if that actually happened, but seems like 186 00:11:23,260 --> 00:11:26,620 S1: a lot of people are leaving both anthropic and OpenAI 187 00:11:26,650 --> 00:11:30,069 S1: to go work with Mira. Uh, new demo shows OpenAI 188 00:11:30,100 --> 00:11:37,060 S1: assistant giving remarkably natural sales calls, handling objections and questions. Yeah, 189 00:11:37,059 --> 00:11:40,510 S1: if you're a mediocre salesperson, you should be afraid. You 190 00:11:40,510 --> 00:11:45,010 S1: should be afraid and either do something else or get not. Mediocre. 191 00:11:45,010 --> 00:11:49,360 S1: Anthropic released an anthropic economic index report on how AI 192 00:11:49,390 --> 00:11:52,589 S1: is being used in the workforce. They see 36% of 193 00:11:52,590 --> 00:11:56,850 S1: people use AI for at least a quarter of their tasks. Interesting. 194 00:11:56,880 --> 00:12:00,120 S1: Good report. I like the content coming out of anthropic. 195 00:12:00,150 --> 00:12:03,479 S1: I feel like their their paper about like how we 196 00:12:03,480 --> 00:12:07,260 S1: view agents really good essays. I like the essay that 197 00:12:07,260 --> 00:12:12,420 S1: came out from the CEO Dario. Yeah, I like the 198 00:12:12,420 --> 00:12:16,110 S1: thought leadership that they're putting out in addition to their models, 199 00:12:16,110 --> 00:12:18,480 S1: by the way. A new model is supposed to be 200 00:12:18,480 --> 00:12:22,350 S1: coming out from anthropic. I would have bet money that 201 00:12:22,350 --> 00:12:24,900 S1: it would have been in January. Not a lot of money. 202 00:12:24,900 --> 00:12:29,640 S1: I would have put it at 70 or 80% chance 203 00:12:29,640 --> 00:12:32,640 S1: of it being in January. So I am quite surprised 204 00:12:32,640 --> 00:12:37,349 S1: that they have not released a new model family, or 205 00:12:37,350 --> 00:12:41,820 S1: just an update to sonnet or opus or haiku or whatever. 206 00:12:41,820 --> 00:12:44,670 S1: I think they're probably going to do a big family upgrade, 207 00:12:44,670 --> 00:12:47,969 S1: maybe even a new family. I think they're going to 208 00:12:47,970 --> 00:12:51,350 S1: do something big, but I thought they would have wanted 209 00:12:51,350 --> 00:12:53,870 S1: to do it in January, so I'm surprised that it's 210 00:12:53,870 --> 00:12:57,020 S1: not out yet, but the rumor is basically that it 211 00:12:57,020 --> 00:12:59,300 S1: is coming soon, maybe in the next couple of weeks. 212 00:12:59,330 --> 00:13:03,800 S1: Lee Robinson says AI is finally enabling truly personal software, 213 00:13:03,800 --> 00:13:08,090 S1: where people can build exactly what they need without extra bloat. 100% 214 00:13:08,090 --> 00:13:10,939 S1: agreed their LinkedIn is testing an AI tool where you 215 00:13:10,940 --> 00:13:13,010 S1: just talk to the interface about what you're looking for, 216 00:13:13,010 --> 00:13:15,890 S1: and it returns results. Another way to say this is 217 00:13:15,890 --> 00:13:19,730 S1: like LinkedIn is working on an interface that everyone's going 218 00:13:19,760 --> 00:13:22,910 S1: to have soon. Like so many people, especially developers are 219 00:13:22,910 --> 00:13:27,860 S1: just talking to like cursor or to Visual Studio or 220 00:13:27,860 --> 00:13:30,740 S1: to whatever AI tool. You just talk to it and 221 00:13:30,740 --> 00:13:32,929 S1: it's like it figures out what you mean and it 222 00:13:32,929 --> 00:13:37,010 S1: just starts working. All software is going to be like that, right? Um, 223 00:13:37,040 --> 00:13:40,160 S1: in my opinion, you're going to be talking to your, uh, 224 00:13:40,160 --> 00:13:42,740 S1: your own AI model and it's going to be the 225 00:13:42,740 --> 00:13:45,200 S1: interface to this tool. But you could also talk to 226 00:13:45,230 --> 00:13:47,570 S1: the tool directly in the meantime, because we don't have 227 00:13:47,570 --> 00:13:51,650 S1: full Das yet. but it seems obvious that voice is 228 00:13:51,650 --> 00:13:54,500 S1: going to become more and more of our interface to things. 229 00:13:54,770 --> 00:13:57,650 S1: Because as nice as a keyboard is, and we'll still 230 00:13:57,650 --> 00:14:00,200 S1: use the keyboard for some things, and some people will 231 00:14:00,200 --> 00:14:05,000 S1: prefer the keyboard, but it is friction compared to just talking, 232 00:14:05,030 --> 00:14:06,980 S1: especially when there's lots of back and forth and stuff 233 00:14:06,980 --> 00:14:09,470 S1: like that. Chick fil A is using drones to fly over, 234 00:14:09,470 --> 00:14:12,290 S1: study and optimize their drive thrus, helping them achieve the 235 00:14:12,290 --> 00:14:16,580 S1: highest per restaurant revenue in the US. So they got 236 00:14:16,580 --> 00:14:21,230 S1: an aerial film studies unit, and it helped one location 237 00:14:21,230 --> 00:14:25,940 S1: boost drive thru sales by 50% in 2022. So they've 238 00:14:25,940 --> 00:14:30,770 S1: got a new Atlanta location serving 700 cars per hour. 239 00:14:30,800 --> 00:14:33,380 S1: If you ever go there during a rush, it's like 240 00:14:33,380 --> 00:14:37,010 S1: it disrupts traffic on the street outside. So now they 241 00:14:37,010 --> 00:14:40,850 S1: got drones flying over like optimizing the thing. Mad respect 242 00:14:40,850 --> 00:14:42,980 S1: to chick fil A. Mad respect to chick fil A. 243 00:14:43,010 --> 00:14:46,490 S1: They're just like, yeah, they know how to run a business. 244 00:14:46,520 --> 00:14:50,920 S1: Same with In-N-Out, right? Same kind of vibe. Like really 245 00:14:51,100 --> 00:14:55,240 S1: focused on optimization. Apple's making a smart home display called 246 00:14:55,240 --> 00:14:59,950 S1: the home Pad. Basically a seven inch square display. Um, also, 247 00:14:59,950 --> 00:15:03,250 S1: there's a big announcement coming out. When is it? The 19th. 248 00:15:03,280 --> 00:15:06,610 S1: Apple is teasing some kind of big product. So it 249 00:15:06,610 --> 00:15:09,760 S1: might actually be this might actually be this. It's like 250 00:15:09,760 --> 00:15:13,360 S1: a seven inch square display that you put different places 251 00:15:13,360 --> 00:15:16,600 S1: in the house. So it's like a touch display for 252 00:15:16,600 --> 00:15:19,060 S1: managing your, your home. So it could be that or 253 00:15:19,060 --> 00:15:21,070 S1: it could be like a new surprise product. Not sure 254 00:15:21,070 --> 00:15:23,260 S1: what that's going to be. Uber is in a weird 255 00:15:23,260 --> 00:15:27,790 S1: spot because they're just the middleman between users and a 256 00:15:27,790 --> 00:15:31,210 S1: service like Waymo. So if you're anywhere near San Francisco, 257 00:15:31,240 --> 00:15:33,280 S1: you're coming to visit or you live there, you should 258 00:15:33,280 --> 00:15:38,110 S1: be using Waymo. Waymo is way better. Uber is just 259 00:15:38,140 --> 00:15:42,040 S1: it's amazing. It's absolutely great. Car comes and picks you up. 260 00:15:42,040 --> 00:15:45,970 S1: There's no driver. It's extremely safe. It's like the best 261 00:15:45,970 --> 00:15:50,170 S1: driver you've ever seen. It's way better than a human 262 00:15:50,170 --> 00:15:53,710 S1: Uber driver. Probably better than you as a driver or 263 00:15:53,710 --> 00:15:57,580 S1: me as a driver. It's just extraordinarily high quality. And 264 00:15:57,580 --> 00:16:01,750 S1: the question is, if Waymo gets big and becomes the 265 00:16:01,750 --> 00:16:04,540 S1: big thing, why wouldn't you just have the Waymo app 266 00:16:04,540 --> 00:16:07,780 S1: like I do? I don't I why would I use 267 00:16:07,810 --> 00:16:11,440 S1: the Uber app to call Waymo? Now one answer is 268 00:16:11,620 --> 00:16:13,960 S1: maybe I just want one app. And if it's the 269 00:16:13,960 --> 00:16:16,990 S1: same user experience or better, then I would use the 270 00:16:16,990 --> 00:16:19,239 S1: one app. But the problem is if you have a 271 00:16:19,240 --> 00:16:21,640 S1: DoorDash app and you have a Waymo app, those are 272 00:16:21,640 --> 00:16:25,360 S1: direct lines into the thing. I will tell you the 273 00:16:25,360 --> 00:16:28,570 S1: answer though. I'll tell you the answer, which is going 274 00:16:28,600 --> 00:16:31,690 S1: to be one or 2 or 3 years out, which 275 00:16:31,690 --> 00:16:37,300 S1: is you won't need any of those because your single interface, 276 00:16:37,300 --> 00:16:40,450 S1: going back to the previous point, your single interface will 277 00:16:40,450 --> 00:16:44,500 S1: be your Da, your digital assistant will be the single interface. 278 00:16:44,500 --> 00:16:48,360 S1: So you won't be calling Waymo. This is what people 279 00:16:48,360 --> 00:16:51,450 S1: don't realize. This is the book that I wrote in 2016. 280 00:16:51,480 --> 00:16:54,120 S1: I don't talk about the book enough. I should talk 281 00:16:54,150 --> 00:16:58,110 S1: about it. I literally predicted this stuff in the book 282 00:16:58,140 --> 00:17:00,870 S1: in 2016, which is why I wrote the book. It's 283 00:17:00,870 --> 00:17:03,840 S1: kind of a crappy book, honestly. It's more like a 284 00:17:03,840 --> 00:17:07,830 S1: structured blog post. It's very short, but I literally wrote 285 00:17:07,830 --> 00:17:10,919 S1: it down because I saw it so clearly, including the 286 00:17:10,920 --> 00:17:14,220 S1: AI stuff, including the Da stuff, and it just seems 287 00:17:14,220 --> 00:17:16,530 S1: logical to me. It seemed logical to me then. It 288 00:17:16,530 --> 00:17:20,880 S1: seems logical to me now. There won't be a Waymo app, 289 00:17:20,910 --> 00:17:23,850 S1: or they there will be, but it will be a 290 00:17:23,850 --> 00:17:31,740 S1: Waymo API. Companies become APIs, okay? Companies become APIs. That 291 00:17:31,740 --> 00:17:35,730 S1: is what's going to happen. Waymo is just an API. 292 00:17:35,940 --> 00:17:39,719 S1: Uber is just an API. If I go to the 293 00:17:39,720 --> 00:17:45,450 S1: corner and I have had a few Moscow mules with Tito's, 294 00:17:45,590 --> 00:17:49,250 S1: Which is my go to drink, and, um, I'm not driving. 295 00:17:49,250 --> 00:17:50,659 S1: My car is right there, but I'm not going to 296 00:17:50,660 --> 00:17:53,450 S1: drive because I've had too much to drink, and, um, 297 00:17:53,570 --> 00:17:56,479 S1: I need to go home. I'm going to say to 298 00:17:56,510 --> 00:17:59,750 S1: my da, give me a ride home. It is then 299 00:17:59,750 --> 00:18:03,290 S1: going to look at all the different services it's going. 300 00:18:03,320 --> 00:18:05,570 S1: First of all, it might know that I have a preference, 301 00:18:05,570 --> 00:18:09,109 S1: but anyway, it has all the different things available. All 302 00:18:09,109 --> 00:18:13,250 S1: the cab companies have an API, Uber has an API, 303 00:18:13,280 --> 00:18:17,090 S1: Waymo has an API. Maybe there's a competitor to Waymo, 304 00:18:17,119 --> 00:18:20,270 S1: maybe it's BYD. Hopefully it isn't. But let's say there's 305 00:18:20,270 --> 00:18:24,169 S1: a Chinese competitor. Oh, Tesla. Tesla will have an API 306 00:18:24,200 --> 00:18:26,690 S1: because there will be Tesla auto driving cars as well. 307 00:18:26,690 --> 00:18:30,080 S1: So that's a competitor to Waymo. All these things have 308 00:18:30,080 --> 00:18:34,220 S1: an API. My Da knows my preferences. It knows what 309 00:18:34,220 --> 00:18:36,320 S1: I like inside the car. It knows the cars that 310 00:18:36,320 --> 00:18:39,800 S1: I like. It knows when I got angry about an 311 00:18:39,800 --> 00:18:43,670 S1: experience before. It's judging all those things based on my 312 00:18:43,670 --> 00:18:47,859 S1: current AI I state it's selecting from which one. Also, 313 00:18:47,859 --> 00:18:50,709 S1: it could be which one can get to here the fastest. 314 00:18:50,710 --> 00:18:54,490 S1: Maybe it's raining. Maybe I need a car immediately. And 315 00:18:54,520 --> 00:18:58,090 S1: a car is literally like 100ft away from me and 316 00:18:58,090 --> 00:19:01,450 S1: it can pick me up in 30s. Maybe it selects that, 317 00:19:01,450 --> 00:19:05,440 S1: but it makes the choice for me. I literally say 318 00:19:05,440 --> 00:19:07,840 S1: I literally walk out of the thing and I say, 319 00:19:07,900 --> 00:19:11,470 S1: give me a ride home. And 28 seconds later, a 320 00:19:11,470 --> 00:19:15,100 S1: car stops and opens up and says, hey, Daniel, um, 321 00:19:15,100 --> 00:19:17,890 S1: I'm here to take you home. And it's an automated car. 322 00:19:17,890 --> 00:19:21,730 S1: It's autonomous and it takes me home. I did not 323 00:19:21,730 --> 00:19:26,050 S1: pick up my phone. I did not open an application. 324 00:19:26,050 --> 00:19:30,520 S1: That is ridiculous. That is ridiculous. I will not be 325 00:19:30,520 --> 00:19:34,780 S1: opening apps. I will be talking to my da. My 326 00:19:34,810 --> 00:19:38,860 S1: da is the proxy. It is my interface to the world. 327 00:19:38,859 --> 00:19:43,540 S1: My da is looking around at San Francisco at the 328 00:19:43,540 --> 00:19:49,270 S1: millions and millions of demons which are APIs for objects, 329 00:19:49,450 --> 00:19:53,530 S1: including the objects which are businesses. It has the ability 330 00:19:53,530 --> 00:19:59,920 S1: to see and read and interpret and manipulate and interface 331 00:19:59,920 --> 00:20:04,510 S1: with all of those millions of APIs to create the 332 00:20:04,510 --> 00:20:07,869 S1: best possible experience for me, which in this case involves 333 00:20:07,869 --> 00:20:11,889 S1: getting me home quickly while not driving drunk. That's it. 334 00:20:12,369 --> 00:20:15,010 S1: If you think about the future of AI and the 335 00:20:15,010 --> 00:20:18,370 S1: future of tech, this is the biggest thing that nobody's 336 00:20:18,369 --> 00:20:20,620 S1: talking about because they haven't figured it out yet. They 337 00:20:20,619 --> 00:20:23,110 S1: just haven't figured this out yet. Like, it's I don't 338 00:20:23,109 --> 00:20:25,750 S1: see anybody talking about it yet. I keep talking about 339 00:20:25,750 --> 00:20:27,640 S1: it every once in a while, like I am now, 340 00:20:27,700 --> 00:20:30,100 S1: but in a couple of years, people are going to 341 00:20:30,100 --> 00:20:32,590 S1: be like, you know what? I tell you what? You know, 342 00:20:32,590 --> 00:20:35,169 S1: it's going to be cool. Everything's going to have some 343 00:20:35,170 --> 00:20:40,210 S1: sort of interface. And then the AI on the individual's 344 00:20:40,210 --> 00:20:44,700 S1: device is going to interact with those services on behalf 345 00:20:44,700 --> 00:20:48,270 S1: of the user. And I'm going to be like, yeah, yeah, 346 00:20:48,690 --> 00:20:51,359 S1: that's exactly what's going to happen. And the next piece 347 00:20:51,359 --> 00:20:53,280 S1: on top of that, okay, I'm just going to pull 348 00:20:53,280 --> 00:20:55,440 S1: it up. The next piece on top of that is 349 00:20:55,440 --> 00:20:59,100 S1: going to be the interface. Okay. What's this thing called 350 00:20:59,190 --> 00:21:03,270 S1: invoking raycast. Going to do this real quick. Digital assistants. 351 00:21:03,810 --> 00:21:08,760 S1: Step one digital assistants are personal AIS know everything about 352 00:21:08,760 --> 00:21:12,090 S1: us and they are everything to us. That is our 353 00:21:12,090 --> 00:21:16,439 S1: single interface. Then two everything gets an API, which is 354 00:21:16,440 --> 00:21:19,619 S1: a demon or an aura or an API. Whatever we 355 00:21:19,619 --> 00:21:21,990 S1: want to call that, I will call it a demon 356 00:21:21,990 --> 00:21:26,460 S1: for a person, which is Greek for soul. Three our 357 00:21:26,460 --> 00:21:29,760 S1: Das will mediate everything. This is the point that I 358 00:21:29,760 --> 00:21:32,220 S1: was just making. We'll say what we want in our 359 00:21:32,250 --> 00:21:36,420 S1: Das will manipulate external demons APIs to make it happen, 360 00:21:36,420 --> 00:21:42,090 S1: for our Das will become our active advocates and defenders. 361 00:21:42,090 --> 00:21:46,730 S1: They will actively monitor, filter and manipulate the world to 362 00:21:46,760 --> 00:21:51,920 S1: protect us, to advocate on our behalf. Okay. That's four. 363 00:21:51,950 --> 00:21:54,859 S1: And millions of examples in here. You should definitely go 364 00:21:54,859 --> 00:21:56,990 S1: read this one. I'll try to link to it somewhere. 365 00:21:56,990 --> 00:22:02,359 S1: And five A thriving ecosystem of Da modules. These Da 366 00:22:02,359 --> 00:22:07,070 S1: modules are the APIs. The companies become the APIs. The 367 00:22:07,070 --> 00:22:10,490 S1: companies are the APIs. The companies put themselves into the 368 00:22:10,490 --> 00:22:16,250 S1: world via the APIs. Six AR interfaces show us the 369 00:22:16,250 --> 00:22:21,379 S1: demons around us through an interface that our Da is 370 00:22:21,380 --> 00:22:26,030 S1: providing us. So some companies will provide the filters through 371 00:22:26,060 --> 00:22:31,639 S1: which we see the demons. So for example, this filter 372 00:22:31,760 --> 00:22:35,540 S1: that my Da is using, let's say it's from a 373 00:22:35,540 --> 00:22:41,900 S1: company called chroma. Okay. This company called chroma displays people's 374 00:22:41,900 --> 00:22:46,160 S1: demons in really cool, colorful ways. That's the only thing 375 00:22:46,160 --> 00:22:50,899 S1: this company does. So my Da knows that I like 376 00:22:50,900 --> 00:22:54,530 S1: to use the chroma interface. So when I walk around 377 00:22:54,530 --> 00:22:58,159 S1: and I see people's demons, this girl, for example, this 378 00:22:58,160 --> 00:23:03,440 S1: girl is super creative. She's hyper creative. And the chroma 379 00:23:03,470 --> 00:23:07,700 S1: interface when it reads a demon. When my da reads 380 00:23:07,730 --> 00:23:11,270 S1: a demon, it reads her demon. We're in a Starbucks. 381 00:23:11,270 --> 00:23:13,520 S1: I look over at this girl. This is what I 382 00:23:13,520 --> 00:23:15,590 S1: see in my visor. This is what I see in 383 00:23:15,590 --> 00:23:18,830 S1: my apple glasses, or my meta glasses or my Google glasses. 384 00:23:18,830 --> 00:23:22,580 S1: Whatever the company is, this is what I see. I 385 00:23:22,580 --> 00:23:26,300 S1: see this glowing fire around this girl and I'm like, 386 00:23:26,300 --> 00:23:29,480 S1: Holy crap, she must be an artist. She must be 387 00:23:29,480 --> 00:23:35,240 S1: like a musician. She's like on fire with like, creativity. 388 00:23:35,300 --> 00:23:38,389 S1: My Da knew that because it read the thing, but 389 00:23:38,390 --> 00:23:42,910 S1: most importantly, it went through the filter of the chroma interface, 390 00:23:42,910 --> 00:23:46,120 S1: and that's what made the display inside of my visual. 391 00:23:46,119 --> 00:23:50,770 S1: Inside of my. Ah, right. So some people don't want 392 00:23:50,800 --> 00:23:52,960 S1: to see this. They want to see a more tech version. 393 00:23:52,960 --> 00:23:54,940 S1: They want to see a more minimalist version. They want 394 00:23:54,970 --> 00:24:00,580 S1: to see a green outline for creative people or whatever. Right. 395 00:24:00,580 --> 00:24:07,960 S1: So those filters are themselves separate companies with separate APIs 396 00:24:07,960 --> 00:24:11,950 S1: and software that run on the, ah, interface, etc.. Right. 397 00:24:11,950 --> 00:24:17,170 S1: So it's all this giant marketplace of filters and demons 398 00:24:17,170 --> 00:24:20,860 S1: and companies that are available to me. Right. And if 399 00:24:20,859 --> 00:24:22,990 S1: I go back up here, look, this is where the 400 00:24:22,990 --> 00:24:26,139 S1: the this is all the demons. I can't read that. 401 00:24:26,140 --> 00:24:29,200 S1: That is millions of demons literally on the screen right now. 402 00:24:29,230 --> 00:24:31,659 S1: I as a human cannot read this. My Da is 403 00:24:31,660 --> 00:24:35,800 S1: reading it for me and advocating on my behalf. Right? 404 00:24:36,010 --> 00:24:39,040 S1: So you have this giant ecosystem of these things. You 405 00:24:39,040 --> 00:24:41,800 S1: have the ah, R interface. Here's another way of viewing 406 00:24:41,800 --> 00:24:43,899 S1: a person. This is another way of viewing a person. 407 00:24:43,900 --> 00:24:48,220 S1: This one I think it yeah, this is very human. 408 00:24:48,220 --> 00:24:51,340 S1: This is somebody going through hard times. So if I 409 00:24:51,340 --> 00:24:54,340 S1: look over and I see this on somebody, I'm going 410 00:24:54,369 --> 00:24:57,010 S1: to be like, hey, can I buy this person a coffee? Hey, 411 00:24:57,010 --> 00:24:59,740 S1: how are you doing? Like you're going through some shit. 412 00:24:59,770 --> 00:25:03,250 S1: What's going on? Do you need somebody to talk to? Right. 413 00:25:03,280 --> 00:25:08,290 S1: But this is why I. And to me, and tech 414 00:25:08,320 --> 00:25:10,959 S1: is a human thing. I want to be able to 415 00:25:10,960 --> 00:25:14,050 S1: connect with this person. And maybe I should read it 416 00:25:14,050 --> 00:25:18,520 S1: on their face. Right. But this is a public demon 417 00:25:18,520 --> 00:25:21,790 S1: that she's broadcasting, so she's allowing this to go out 418 00:25:21,790 --> 00:25:26,320 S1: into the world. That life's messed up right now. So 419 00:25:26,320 --> 00:25:28,179 S1: I'm going to go say hi. Right. I'm going to 420 00:25:28,180 --> 00:25:29,919 S1: go see if they need a friend. So these are 421 00:25:29,920 --> 00:25:32,770 S1: the different demon options. Look there's going to be a 422 00:25:32,770 --> 00:25:37,750 S1: whole ecosystem around this. But anyway, I digress. Okay. Ted's 423 00:25:37,750 --> 00:25:40,500 S1: Chris Anderson is looking for somebody to take over the 424 00:25:40,500 --> 00:25:44,040 S1: entire Ted organization, and he's running the search like a 425 00:25:44,070 --> 00:25:49,080 S1: Willy Wonka contest where anybody can apply. Interesting. Christie's is 426 00:25:49,080 --> 00:25:52,109 S1: doing their first AI only art auction, and it's a 427 00:25:52,109 --> 00:25:56,460 S1: lot of traditional artists getting really mad about this. Yeah. Understandably, 428 00:25:56,460 --> 00:25:59,550 S1: Google says they are getting rid of their diversity hiring 429 00:25:59,580 --> 00:26:03,480 S1: targets for 24, saying it was positive discrimination, which yeah, 430 00:26:03,510 --> 00:26:07,649 S1: it might technically be true. That is the purpose of 431 00:26:07,650 --> 00:26:12,510 S1: affirmative action is because that particular group needs extra help. 432 00:26:12,510 --> 00:26:15,840 S1: So I you got to keep in mind what Dei 433 00:26:15,840 --> 00:26:20,220 S1: is for. Like I agree, Dei has caused damage. It's 434 00:26:20,220 --> 00:26:23,489 S1: it's been poorly implemented. The concept is not broken. The 435 00:26:23,490 --> 00:26:27,869 S1: concept is great and we should still continue doing it. 436 00:26:28,200 --> 00:26:31,020 S1: I think what they are saying, I'm giving you a 437 00:26:31,020 --> 00:26:34,470 S1: charitable reading here. What they're saying is that we're still 438 00:26:34,470 --> 00:26:37,650 S1: doing Dei the correct way, which is we still want 439 00:26:37,650 --> 00:26:42,440 S1: to hire everyone who's capable and who's best qualified for 440 00:26:42,440 --> 00:26:46,610 S1: the job, and we are absolutely looking for underrepresented groups 441 00:26:46,609 --> 00:26:51,260 S1: to do that. What they are saying, charitable interpretation, again, 442 00:26:51,260 --> 00:26:54,740 S1: is that we are not specifically going to say we 443 00:26:54,740 --> 00:27:00,230 S1: must hire this many of this particular group, and we 444 00:27:00,230 --> 00:27:05,090 S1: exclude other qualified candidates because they are not those groups. 445 00:27:05,090 --> 00:27:07,580 S1: They are saying they are discarding that because they're getting 446 00:27:07,580 --> 00:27:09,800 S1: sued over it. And it's not a good idea and 447 00:27:09,800 --> 00:27:13,460 S1: it's reverse discrimination and all of that. So the question 448 00:27:13,460 --> 00:27:16,070 S1: is how does this look in a few years? Is 449 00:27:16,070 --> 00:27:21,380 S1: it that they're actually getting rid of all Dei, all 450 00:27:21,410 --> 00:27:25,070 S1: affirmative action and they're just using this as cover? Yeah. 451 00:27:25,070 --> 00:27:28,219 S1: So here's here's my comment on this surprising not surprising 452 00:27:28,220 --> 00:27:32,780 S1: that all these programs disappeared overnight on January 21st. What 453 00:27:32,780 --> 00:27:35,179 S1: does that tell you? And like I say here, it 454 00:27:35,180 --> 00:27:37,490 S1: tells me that they couldn't wait for a reason or 455 00:27:37,490 --> 00:27:40,489 S1: opportunity to do this, so they've been wishing they could 456 00:27:40,490 --> 00:27:43,220 S1: do it already. So what I'm hoping is it doesn't 457 00:27:43,220 --> 00:27:46,790 S1: just get rid of all affirmative action, which like I said, 458 00:27:46,790 --> 00:27:49,850 S1: the core concept is correct and we need to continue 459 00:27:49,880 --> 00:27:54,110 S1: doing it until we actually have equal opportunity, which we 460 00:27:54,109 --> 00:27:57,590 S1: do not right now. Doctors are now a major client 461 00:27:57,590 --> 00:28:01,220 S1: base for weight loss drugs like Ozempic. This link said 462 00:28:01,220 --> 00:28:04,790 S1: something like 40% of doctors are on Ozempic. I like that, 463 00:28:04,790 --> 00:28:09,770 S1: although few decades ago all doctors were smoking camels, so 464 00:28:09,800 --> 00:28:14,780 S1: doesn't necessarily mean it's perfect advice. New York City is. 465 00:28:14,780 --> 00:28:18,560 S1: Subway crime dropped by 36% in January because they added 466 00:28:18,590 --> 00:28:24,140 S1: 1200 more police? Yeah, only 147 subway crimes in January 25th, 467 00:28:24,140 --> 00:28:28,370 S1: compared to 231 in January. And they got 300 specifically 468 00:28:28,369 --> 00:28:33,230 S1: for overnight trains. Love it. Yeah. More cops like it. 469 00:28:33,500 --> 00:28:37,240 S1: Measles outbreak is hitting the least vaccinated parts of Texas, 470 00:28:37,240 --> 00:28:41,020 S1: with nine cases in an area where only 82% of 471 00:28:41,020 --> 00:28:45,910 S1: kids are vaccinated. You actually need 95% for herd immunity. 472 00:28:46,030 --> 00:28:50,320 S1: We're also in the worst flu season the last 15 years. But, uh, 473 00:28:50,320 --> 00:28:53,050 S1: but Covid was actually super annoying. So let's not talk 474 00:28:53,050 --> 00:28:57,190 S1: about let's not talk about, uh, flu or pandemics or 475 00:28:57,190 --> 00:29:02,110 S1: anything like that that could actually benefit humanity because Covid 476 00:29:02,110 --> 00:29:04,840 S1: was annoying. I think that's a good policy. One of 477 00:29:04,840 --> 00:29:08,080 S1: my favorite thinkers, Robin Hanson, breaks down how different social 478 00:29:08,080 --> 00:29:12,910 S1: circles value different status markers. And specifically, he's looking at 479 00:29:12,910 --> 00:29:17,200 S1: his own friend group and how people that he hangs 480 00:29:17,200 --> 00:29:21,940 S1: out with pursue and signal value. And he says most 481 00:29:21,940 --> 00:29:25,060 S1: intellectuals that he knows are actually chasing fame and prestige. 482 00:29:25,060 --> 00:29:29,170 S1: He's not talking about his core friends, his core friends. 483 00:29:29,380 --> 00:29:33,880 S1: The highest sort of social status is two polymaths who 484 00:29:33,880 --> 00:29:38,790 S1: follow evidence across disciplines. That's who I like to hang 485 00:29:38,790 --> 00:29:41,220 S1: out with, too. I think we should all like that. 486 00:29:41,310 --> 00:29:45,540 S1: After 12 years of Walmart domination. Amazon just jumped ahead 487 00:29:45,540 --> 00:29:52,560 S1: with 187.8 billion in quarterly revenue. Compared to Walmart's expected 488 00:29:52,590 --> 00:29:57,000 S1: 180 billion. So that's revenue. I'm curious about profit, but 489 00:29:57,000 --> 00:29:59,370 S1: not curious enough to look it up because why do 490 00:29:59,370 --> 00:30:03,120 S1: I care? Ideas paralyzed by crisis I'm a bit paralyzed 491 00:30:03,120 --> 00:30:06,930 S1: by what's going on right now in politics and specifically 492 00:30:06,930 --> 00:30:11,280 S1: with the government. I basically cycle between depressed, apathetic and 493 00:30:11,280 --> 00:30:14,040 S1: very angry. Did the government need to be audited and 494 00:30:14,040 --> 00:30:16,920 S1: cleaned up? Yes. Is the best way to start from 495 00:30:16,920 --> 00:30:20,250 S1: scratch and be aggressive with it? Possibly. Yeah, I'm kind 496 00:30:20,280 --> 00:30:23,220 S1: of excited about that. But you lose me when I 497 00:30:23,220 --> 00:30:27,000 S1: don't see you being careful about the programs that matter. 498 00:30:27,030 --> 00:30:32,100 S1: Like this whole. What was it? NCIS, NCIS. I didn't 499 00:30:32,100 --> 00:30:34,470 S1: look into it. I get depressed about the stuff, so 500 00:30:34,470 --> 00:30:38,580 S1: I just like I tune out. But there's no question 501 00:30:38,580 --> 00:30:41,730 S1: to me that a bunch of that stuff was bad 502 00:30:41,730 --> 00:30:44,550 S1: and needed to be removed. And there's also no question 503 00:30:44,550 --> 00:30:47,790 S1: to me or for me that a bunch of stuff 504 00:30:47,790 --> 00:30:51,030 S1: in there was actually really good and it wasn't very expensive, 505 00:30:51,030 --> 00:30:55,350 S1: and you should not have turned it off. Right? So 506 00:30:55,350 --> 00:30:57,810 S1: you lose me when you just throw everything out. And 507 00:30:57,810 --> 00:31:00,480 S1: maybe they didn't do that. They're claiming that they're putting 508 00:31:00,480 --> 00:31:04,020 S1: everything on the Doge website and it's all public and 509 00:31:04,020 --> 00:31:06,870 S1: it's all visible. I went and looked yesterday for the 510 00:31:06,870 --> 00:31:09,360 S1: first time, because I couldn't bring myself to pull up 511 00:31:09,360 --> 00:31:14,190 S1: a government website called Doge, having lived through Dogecoin. So 512 00:31:14,190 --> 00:31:15,840 S1: I'm just like, I don't want to go to that website. 513 00:31:15,870 --> 00:31:18,450 S1: But anyway, I go there, it's dot gov. I don't know, 514 00:31:18,480 --> 00:31:20,640 S1: it looked like there might have been some data published. 515 00:31:20,640 --> 00:31:23,280 S1: I'll probably use AI to figure out if it's like 516 00:31:23,310 --> 00:31:27,719 S1: actually publish if it's like legitimate transparency. I saw Elon 517 00:31:27,720 --> 00:31:30,300 S1: talk from the white House the other day on YouTube 518 00:31:30,300 --> 00:31:32,100 S1: in the middle of the night, which I should have 519 00:31:32,100 --> 00:31:35,100 S1: been going to sleep instead. But anyway, he looked honest 520 00:31:35,100 --> 00:31:38,190 S1: about it. He looked like he was actually trying to 521 00:31:38,220 --> 00:31:40,440 S1: uncover things and do the right thing, which is what 522 00:31:40,440 --> 00:31:42,959 S1: they were claiming they were doing. When I read the 523 00:31:42,960 --> 00:31:46,110 S1: media about it, it's like, oh, this is the Antichrist. 524 00:31:46,110 --> 00:31:49,260 S1: I don't think Elon is good at lying. When I 525 00:31:49,260 --> 00:31:52,140 S1: see him saying that he's trying to do the right 526 00:31:52,140 --> 00:31:55,050 S1: thing and he's trying to uncover this stuff, and he's 527 00:31:55,050 --> 00:31:58,650 S1: making it public. And this is exactly what they were 528 00:31:58,650 --> 00:32:02,760 S1: elected to do. I agree with him. However, you add 529 00:32:02,760 --> 00:32:08,100 S1: in Nazi salutes, you add in his personality on X 530 00:32:08,100 --> 00:32:12,780 S1: and how it's like slam dunk and like mean and rude. 531 00:32:12,780 --> 00:32:17,280 S1: It doesn't rhyme with the authenticity. I just saw him 532 00:32:17,280 --> 00:32:22,710 S1: giving about trying to do this doge thing correctly. So 533 00:32:23,010 --> 00:32:25,620 S1: I'm very torn on the whole thing. I'm very like, 534 00:32:25,650 --> 00:32:29,550 S1: you know, how how much can I actually believe what 535 00:32:29,550 --> 00:32:32,910 S1: I'm hearing right now from from him? Um, I do 536 00:32:32,950 --> 00:32:37,300 S1: know that the media portrayal of it, I believe it's wrong. 537 00:32:37,330 --> 00:32:41,140 S1: I believe he's they and a bunch of them and 538 00:32:41,140 --> 00:32:44,320 S1: a bunch of them are pure garbage. It just outright 539 00:32:44,320 --> 00:32:48,219 S1: pure garbage. I actually don't think Trump is nearly as 540 00:32:48,220 --> 00:32:49,959 S1: bad as a lot of people around him, and I 541 00:32:49,960 --> 00:32:52,330 S1: don't think Elon is nearly as bad as a lot 542 00:32:52,330 --> 00:32:54,910 S1: of the people around him. But at the same time, 543 00:32:54,910 --> 00:32:59,710 S1: I can't fully trust Elon. And after the the Nazi 544 00:32:59,710 --> 00:33:03,160 S1: salute stuff, honestly, I'm looking for a new car because 545 00:33:03,430 --> 00:33:05,560 S1: I don't want to go political here. I already did, 546 00:33:05,590 --> 00:33:09,070 S1: but I can't dig on Nazi salutes. I don't think 547 00:33:09,070 --> 00:33:12,580 S1: he's anti-Semitic, which is why so many Jewish people still 548 00:33:12,580 --> 00:33:16,690 S1: like him. He's very pro-Israel. He's very pro Jewish, very 549 00:33:16,690 --> 00:33:20,470 S1: pro Jewish hostages. So I don't think he's anti-Semitic. Why 550 00:33:20,500 --> 00:33:23,500 S1: are you doing Nazi salutes if you're not anti-Semitic? The 551 00:33:23,500 --> 00:33:28,690 S1: answer is the answer is he's appealing to AfD about 552 00:33:28,690 --> 00:33:33,250 S1: a culture war in Europe, and he's signaling to them 553 00:33:33,250 --> 00:33:37,210 S1: that he's on their side. Problem is, it's a fucking 554 00:33:37,210 --> 00:33:43,630 S1: Nazi salute that has implications that motivates neo-Nazis in the US, 555 00:33:43,660 --> 00:33:47,290 S1: in Europe, all over the world. And it's not nuanced 556 00:33:47,290 --> 00:33:50,770 S1: along the lines of like, oh, he's trolling and he's 557 00:33:50,800 --> 00:33:55,720 S1: actually not anti-Semitic. neo-Nazis are anti-Semitic. Hate to tell you. 558 00:33:56,170 --> 00:33:59,350 S1: So very upset with him looking for a new car 559 00:33:59,350 --> 00:34:01,840 S1: because of this. At the same time, I think the 560 00:34:01,840 --> 00:34:05,709 S1: narratives around what they're doing is pure evil. They're actually 561 00:34:05,710 --> 00:34:08,260 S1: trying to ruin the country. They're going to go off 562 00:34:08,260 --> 00:34:11,890 S1: their authoritarian. I don't think that's actually accurate. I think 563 00:34:11,890 --> 00:34:14,620 S1: they're actually trying to do good inside of the government. 564 00:34:14,620 --> 00:34:17,830 S1: This it's actually not the title of this little piece 565 00:34:17,830 --> 00:34:21,160 S1: I'm doing right now. Um, it's actually one I'm must 566 00:34:21,160 --> 00:34:23,380 S1: not have published that one yet. I'm still working on it. 567 00:34:23,410 --> 00:34:28,540 S1: This complexity, the ability to hold multiple conflicting things in 568 00:34:28,540 --> 00:34:31,649 S1: your mind is a thing that I believe that we 569 00:34:31,650 --> 00:34:35,730 S1: actually need my North Star. I guess I'm going here. 570 00:34:35,730 --> 00:34:39,180 S1: I didn't plan on going here. My North Star is 571 00:34:39,180 --> 00:34:44,940 S1: Star Trek the Next Generation. I want a multiracial, multicultural 572 00:34:44,940 --> 00:34:50,820 S1: future in which everyone has equal opportunity to thrive. I 573 00:34:50,820 --> 00:34:55,530 S1: believe in government programs. I believe in taxation. I believe 574 00:34:55,530 --> 00:35:00,000 S1: it is society's purpose and responsibility to make sure that 575 00:35:00,000 --> 00:35:03,660 S1: everyone growing up and even into adulthood, if they didn't 576 00:35:03,660 --> 00:35:07,230 S1: get this, they get the nurturing that they need to be, 577 00:35:07,230 --> 00:35:10,350 S1: even with everyone else. Then they can launch. They can 578 00:35:10,350 --> 00:35:14,280 S1: go to their the limits of their discipline, the limits 579 00:35:14,280 --> 00:35:19,020 S1: of their talent, the limits of their creativity, whatever. But 580 00:35:19,020 --> 00:35:22,500 S1: you can't have certain groups get left behind and then 581 00:35:22,500 --> 00:35:26,219 S1: just be like, oh, well, whatever. No, not whatever. If 582 00:35:26,250 --> 00:35:29,520 S1: we are successful, it is our job to lift them. 583 00:35:29,520 --> 00:35:31,970 S1: That is what a society is. That is what a 584 00:35:31,969 --> 00:35:36,080 S1: government is. That's what taxation is for. And if you 585 00:35:36,080 --> 00:35:39,980 S1: listen to somebody like Rogan, he was on with someone 586 00:35:39,980 --> 00:35:44,509 S1: who I really disagree with right now about vaccines. Weinstein, 587 00:35:44,510 --> 00:35:48,560 S1: one of the one of the Weinstein brothers. The other one, not, uh, 588 00:35:48,560 --> 00:35:52,430 S1: not Eric, his brother. Anyway, very much disagree. Very angry 589 00:35:52,430 --> 00:35:56,480 S1: with him about vaccines, very much disagree with him about vaccines. 590 00:35:56,480 --> 00:36:00,980 S1: And they, him and Joe, who are considered like extremely 591 00:36:00,980 --> 00:36:04,489 S1: right at this point by all my very left friends, 592 00:36:04,489 --> 00:36:08,540 S1: they are both like, we need government programs. We need 593 00:36:08,540 --> 00:36:12,230 S1: those government programs to help people who need them. We 594 00:36:12,230 --> 00:36:16,490 S1: cannot get rid of taxation. Taxation is great because it 595 00:36:16,489 --> 00:36:19,610 S1: helps people lift up to the bottom. And I'm like, 596 00:36:19,640 --> 00:36:24,379 S1: see these people? Look, these people all used to be liberals, 597 00:36:24,380 --> 00:36:28,219 S1: including Elon, including Joe. This is why I disagree with 598 00:36:28,219 --> 00:36:31,130 S1: Harris and I might be wrong about this. I also 599 00:36:31,130 --> 00:36:35,060 S1: disagree with my friend Marcus. Marcus Hutchins about this. He's 600 00:36:35,060 --> 00:36:37,850 S1: very much in the camp of Sam Harris about this. 601 00:36:37,850 --> 00:36:41,630 S1: And sorry, I ended up going very, very politics on 602 00:36:41,630 --> 00:36:44,149 S1: this one. Feel free to skip ahead to the discovery 603 00:36:44,150 --> 00:36:46,879 S1: section if you don't want to hear me, close this out. 604 00:36:46,880 --> 00:36:50,060 S1: But it's top of mind for me and I want 605 00:36:50,060 --> 00:36:56,180 S1: to mention it. I want to rescue center people. I 606 00:36:56,210 --> 00:37:00,230 S1: consider Elon a center person. I consider Joe a center person. 607 00:37:00,230 --> 00:37:03,980 S1: I consider, well, Sam Harris is definitely a center person, 608 00:37:03,980 --> 00:37:08,270 S1: but these people who I feel like have strayed and 609 00:37:08,270 --> 00:37:12,560 S1: gone crazy with vaccines and like gone crazy with conspiracies. 610 00:37:12,710 --> 00:37:16,520 S1: If they are still core liberal like I am, who 611 00:37:16,520 --> 00:37:20,810 S1: believe society's responsibility is to help everyone and lift them up, 612 00:37:20,810 --> 00:37:27,980 S1: and they want to build a multiracial, multicultural future which 613 00:37:28,010 --> 00:37:31,810 S1: also lifts people up and is humanistic in nature and 614 00:37:31,810 --> 00:37:37,600 S1: is based on maximizing benefit for all of humanity. That 615 00:37:37,600 --> 00:37:41,080 S1: means there are liberal. That's why I am a liberal. 616 00:37:41,110 --> 00:37:45,640 S1: If they believe that, then I will stick with them 617 00:37:45,640 --> 00:37:49,540 S1: through the nastiness of all this other crap. The moment 618 00:37:49,540 --> 00:37:53,140 S1: I see, the moment I see that they are actually 619 00:37:53,140 --> 00:37:56,770 S1: not going for that. And what they're actually going for 620 00:37:56,770 --> 00:38:03,160 S1: is like a white race is better thing where it's like, 621 00:38:03,160 --> 00:38:05,200 S1: does it matter where you came from? It doesn't matter 622 00:38:05,200 --> 00:38:09,130 S1: what you get. Um, your identity is what matters more. 623 00:38:09,190 --> 00:38:12,280 S1: Guess what? Those are the same on the far left 624 00:38:12,280 --> 00:38:15,040 S1: and the far right. It's most known on the far right. Right. 625 00:38:15,070 --> 00:38:17,859 S1: You must be light skinned. You must be white. You 626 00:38:17,860 --> 00:38:20,890 S1: must be whatever. You must be Jewish. You must be 627 00:38:20,920 --> 00:38:24,610 S1: whatever it is that they're obsessed with. They think that's 628 00:38:24,610 --> 00:38:27,130 S1: the best one. And they think everything else should fall 629 00:38:27,130 --> 00:38:31,650 S1: under that. That is that's a zero for me. That 630 00:38:31,650 --> 00:38:33,990 S1: is a do not pass go if that is the 631 00:38:33,989 --> 00:38:36,330 S1: world you are trying to build, I do not support 632 00:38:36,330 --> 00:38:39,210 S1: you and I will not support you. So the reason 633 00:38:39,210 --> 00:38:44,160 S1: I continue to support to some limited degree people like Joe, 634 00:38:44,190 --> 00:38:47,219 S1: people like Elon, is because I believe they're going for 635 00:38:47,219 --> 00:38:49,410 S1: the other one. They're going for the one that I'm 636 00:38:49,410 --> 00:38:53,009 S1: going for, which is this mutual lifting up of everyone 637 00:38:53,010 --> 00:38:57,719 S1: into a multiracial, multicultural future. And the moment they stop 638 00:38:57,719 --> 00:39:00,420 S1: supporting that, the more in anything they do that doesn't 639 00:39:00,420 --> 00:39:04,140 S1: support that, I fight them on because I'm not confused 640 00:39:04,140 --> 00:39:07,469 S1: about what my morals are. I'm not confused about my politics. 641 00:39:07,469 --> 00:39:10,739 S1: They have not changed. They will not change. This is 642 00:39:10,739 --> 00:39:15,600 S1: why I voted for Kamala, because somebody tried to change 643 00:39:15,600 --> 00:39:19,650 S1: the vote on vote verification day, and they're claiming that 644 00:39:19,650 --> 00:39:21,510 S1: it's not a crime. I'm not cool with that. I 645 00:39:21,510 --> 00:39:25,170 S1: can't vote for him. End of story. So all of 646 00:39:25,170 --> 00:39:30,060 S1: this to say that complexity is required to make sense 647 00:39:30,060 --> 00:39:35,730 S1: out of a complex world, right? There is criticism of 648 00:39:35,730 --> 00:39:39,780 S1: people like myself who try to manage this complexity of like, 649 00:39:39,810 --> 00:39:45,630 S1: looking at all sources and all claims and everything in 650 00:39:45,630 --> 00:39:48,990 S1: like multiple levels and multiple layers and trying to find 651 00:39:48,989 --> 00:39:50,820 S1: the good in it and trying to find the bad 652 00:39:50,820 --> 00:39:54,210 S1: in it. People, there are people who say, and these 653 00:39:54,210 --> 00:39:57,029 S1: are like my far left friends and my far right 654 00:39:57,030 --> 00:40:01,650 S1: friends who are like, um, you're both sides. No, no, 655 00:40:01,650 --> 00:40:05,670 S1: the world is both sides. Like, if you judged the 656 00:40:05,670 --> 00:40:09,960 S1: actual deep moral character of everyone that you hang out 657 00:40:09,960 --> 00:40:12,930 S1: with based on the worst thing that they've done, you 658 00:40:12,930 --> 00:40:15,540 S1: would not talk to anybody. You understand that? You would 659 00:40:15,540 --> 00:40:18,570 S1: not talk to anybody. Like a lot of the people 660 00:40:18,570 --> 00:40:22,110 S1: that we hang out with, they would become horrible people 661 00:40:22,110 --> 00:40:26,009 S1: if they had the power of the people who have power. Right? 662 00:40:26,030 --> 00:40:29,299 S1: And that's just reality. That's just like how the world works. 663 00:40:29,300 --> 00:40:33,710 S1: So I think we should focus on and I talked 664 00:40:33,710 --> 00:40:36,710 S1: about this somewhere else. I think we should focus on 665 00:40:36,710 --> 00:40:40,549 S1: figuring out this person that you're talking about, this person 666 00:40:40,550 --> 00:40:44,450 S1: that you are supporting, this person that you are criticizing, 667 00:40:44,450 --> 00:40:48,770 S1: this person that you think is the Antichrist, whatever. Figure 668 00:40:48,800 --> 00:40:52,430 S1: out the ideal world that they are trying to build, 669 00:40:52,430 --> 00:40:55,550 S1: then think about your world that you are trying to 670 00:40:55,550 --> 00:40:59,660 S1: build and see how similar they are. I would argue 671 00:40:59,810 --> 00:41:04,279 S1: if those two worlds are similar, then the work that 672 00:41:04,280 --> 00:41:07,790 S1: we need to do is to try to help them 673 00:41:07,790 --> 00:41:12,440 S1: and help yourself muddle through the garbage, muddle through the 674 00:41:12,440 --> 00:41:17,360 S1: the chaos, muddle through the noise, muddle through the emotion, 675 00:41:17,360 --> 00:41:21,380 S1: whatever to get to this mutual place. And this is 676 00:41:21,380 --> 00:41:25,930 S1: why I still support people who my friends on the 677 00:41:25,930 --> 00:41:30,310 S1: far left have abandoned and completely abandoned a long time ago. 678 00:41:30,340 --> 00:41:34,030 S1: I support them because I think, I think, and I 679 00:41:34,030 --> 00:41:36,609 S1: could be wrong about this. And the moment I'm wrong 680 00:41:36,610 --> 00:41:40,060 S1: about this, I switch to this fast. Okay. I think 681 00:41:40,060 --> 00:41:43,060 S1: that the thing they are trying to build is very 682 00:41:43,060 --> 00:41:47,320 S1: similar to the liberal thing that I want to build, 683 00:41:47,320 --> 00:41:51,339 S1: and I want to see happen. That's it. It's that simple. 684 00:41:51,400 --> 00:41:54,340 S1: If I think the world they are trying to build 685 00:41:54,340 --> 00:41:58,630 S1: is this liberal world and that they are wrong about it, 686 00:41:58,630 --> 00:42:02,350 S1: they are misguided about it. They don't know something. They're 687 00:42:02,350 --> 00:42:05,350 S1: super dumb about a thing. I don't care about any 688 00:42:05,350 --> 00:42:07,690 S1: of that. I care about what's inside of them. I 689 00:42:07,719 --> 00:42:09,759 S1: care about their soul. I care about what they're trying 690 00:42:09,760 --> 00:42:12,220 S1: to actually build. Right. So that's what I'm trying to 691 00:42:12,219 --> 00:42:17,020 S1: figure out. This is why I am less worried about 692 00:42:17,050 --> 00:42:21,010 S1: Trump now, because I see much more clear the thing 693 00:42:21,010 --> 00:42:23,920 S1: that he's trying to build. It's actually not that bad. 694 00:42:23,920 --> 00:42:28,120 S1: I don't like the guy because he's an egomaniac and like, 695 00:42:28,120 --> 00:42:31,150 S1: half the things he says like, drive me absolutely crazy. 696 00:42:31,270 --> 00:42:34,600 S1: And I think he's being influenced by Russia. I don't 697 00:42:34,600 --> 00:42:37,600 S1: think he's I don't think he's a Russian asset at all. 698 00:42:37,600 --> 00:42:43,210 S1: I don't think he's being, uh, directly manipulated by anyone 699 00:42:43,210 --> 00:42:47,860 S1: at all. I think he's being side manipulated by narratives 700 00:42:47,860 --> 00:42:50,860 S1: that are in his brain that could actually help people, 701 00:42:50,860 --> 00:42:54,520 S1: that we don't want to help. But the point is, 702 00:42:54,520 --> 00:42:58,029 S1: I think he wants a world, first of all, and 703 00:42:58,030 --> 00:43:02,290 S1: this is really a tangent, he actually wants less nuclear weapons. 704 00:43:02,290 --> 00:43:05,560 S1: He wants peace time. The downside for this guy is 705 00:43:05,560 --> 00:43:09,610 S1: he wants a statue, and he wants people to stand 706 00:43:09,610 --> 00:43:11,650 S1: up and salute when he walks in the room. The 707 00:43:11,650 --> 00:43:15,580 S1: guy's he's a narcissist. Okay, but half the world is 708 00:43:15,610 --> 00:43:19,779 S1: created by narcissists. If he's a narcissist who wants to 709 00:43:19,780 --> 00:43:26,190 S1: build an amazing world where Literally the whole planet loves him. 710 00:43:26,219 --> 00:43:29,730 S1: Does not fear him. He doesn't put everyone in jail. 711 00:43:29,730 --> 00:43:32,730 S1: He doesn't kill everyone. He's not stalling. He's not Hitler. 712 00:43:32,730 --> 00:43:35,700 S1: If that's the world he's trying to build, I'll take 713 00:43:35,700 --> 00:43:40,590 S1: those floors. I will take those floors. He wants to denuclearize. 714 00:43:40,950 --> 00:43:43,590 S1: He wants. He wants to, like, have all these countries 715 00:43:43,620 --> 00:43:46,980 S1: get along. If that's the world he's actually trying to build. 716 00:43:46,980 --> 00:43:50,430 S1: And I'm not sure of that. But if that is, 717 00:43:50,430 --> 00:43:53,490 S1: then I put him in a much lower threat level. 718 00:43:53,489 --> 00:43:57,120 S1: There are people around him in this administration who are 719 00:43:57,120 --> 00:43:59,549 S1: not trying to build that world. They are trying to 720 00:43:59,550 --> 00:44:05,280 S1: build something nasty, nasty, and they're using him and his, uh, 721 00:44:05,280 --> 00:44:09,450 S1: naive naivete to try to get what they want. And 722 00:44:09,450 --> 00:44:12,089 S1: so when I find those people, I'm like, yeah, that's garbage. 723 00:44:12,120 --> 00:44:16,560 S1: Total garbage. Elon, I think he's going for a better version. 724 00:44:16,560 --> 00:44:18,630 S1: I think he's going for a good world. I think 725 00:44:18,630 --> 00:44:22,050 S1: it rhymes in is very similar to mine. Very angry 726 00:44:22,050 --> 00:44:26,540 S1: with what he's currently doing, his actual actions. Right. But 727 00:44:26,540 --> 00:44:29,689 S1: if I find out that he's not trying to build 728 00:44:29,690 --> 00:44:32,960 S1: that world, he's trying to build this other world, this 729 00:44:32,960 --> 00:44:38,390 S1: race centered one, or this inequality built one or whatever, 730 00:44:38,810 --> 00:44:41,690 S1: that's it. I'm done with them. That's it. Right. That 731 00:44:41,690 --> 00:44:44,719 S1: is that is how I operate in this world. And 732 00:44:44,719 --> 00:44:48,290 S1: I would recommend I would appeal to you, especially if 733 00:44:48,320 --> 00:44:51,020 S1: you're one of my friends on the left and also 734 00:44:51,020 --> 00:44:53,420 S1: one of my friends on the right. Think about what 735 00:44:53,420 --> 00:44:57,230 S1: they're trying to build and let everything else fade away. 736 00:44:57,260 --> 00:45:00,020 S1: Let that be the center that you focus on in 737 00:45:00,020 --> 00:45:05,180 S1: your political conversations. And don't let media or noise or distraction, 738 00:45:05,210 --> 00:45:08,330 S1: you know, pull you away from that central conversation. What 739 00:45:08,330 --> 00:45:10,310 S1: are we trying to build? How can we get there? 740 00:45:10,340 --> 00:45:14,420 S1: And again, apologies for going into it is in the 741 00:45:14,420 --> 00:45:17,779 S1: ideas section. So it's not in a news section, but uh, 742 00:45:17,870 --> 00:45:20,660 S1: hopefully you didn't have to sit through that if you 743 00:45:20,660 --> 00:45:23,810 S1: didn't want to hear here politics, but hopefully it's a 744 00:45:23,810 --> 00:45:26,810 S1: good message that you got at the end. Moving on 745 00:45:26,840 --> 00:45:31,279 S1: to discovery Lmics. This is probably the coolest AI library 746 00:45:31,280 --> 00:45:34,850 S1: that you've never heard of. This thing abstracts your LLM 747 00:45:34,850 --> 00:45:38,480 S1: calls to a universal config. So basically you have a 748 00:45:38,480 --> 00:45:43,310 S1: little piece of like JavaScript or JSON like config. And 749 00:45:43,310 --> 00:45:47,839 S1: you can send this to any back end. It speaks 750 00:45:47,840 --> 00:45:50,690 S1: all the different AI back ends. Right. So you use 751 00:45:50,690 --> 00:45:52,910 S1: this in like a proxy and like a lambda or 752 00:45:52,940 --> 00:45:56,120 S1: like some sort of like AI function. So this is 753 00:45:56,120 --> 00:46:00,140 S1: for people like myself who are building AI. But you 754 00:46:00,140 --> 00:46:05,150 S1: don't have to write that code yourself. You have that 755 00:46:05,150 --> 00:46:11,150 S1: thing implemented in this LLM X library, and then every 756 00:46:11,150 --> 00:46:14,660 S1: single app that you build, every single component that you build, 757 00:46:14,660 --> 00:46:19,520 S1: all these different multi-level, uh, agent systems and everything, they 758 00:46:19,520 --> 00:46:23,259 S1: all have prompts. They all have models. They all have 759 00:46:23,260 --> 00:46:27,279 S1: configs being sent to the models. You do that in 760 00:46:27,280 --> 00:46:31,390 S1: this tiny little piece of config. That's what gets sent 761 00:46:31,390 --> 00:46:36,910 S1: to Lmax and it handles everything for you. And the 762 00:46:36,910 --> 00:46:40,180 S1: return comes back in a super clean way. It's the 763 00:46:40,180 --> 00:46:43,450 S1: coolest thing ever. It's from a friend of mine who 764 00:46:43,450 --> 00:46:46,720 S1: put this out, and it's like not getting near the 765 00:46:46,719 --> 00:46:50,200 S1: coverage that it should get. You should absolutely check it out. 766 00:46:50,560 --> 00:46:53,950 S1: MTR this thing combines traceroute and ping. I put this 767 00:46:53,980 --> 00:46:56,740 S1: in here every year or so. It's like a really 768 00:46:56,739 --> 00:47:00,490 S1: cool utility. It combines traceroute and ping and shows you 769 00:47:00,520 --> 00:47:03,399 S1: connection quality between hops in real time. So it's just 770 00:47:03,400 --> 00:47:07,870 S1: a cool geeky tool. RPG Map Bundle collection of print 771 00:47:07,870 --> 00:47:12,130 S1: and play RPG maps for your next role playing game. 772 00:47:12,160 --> 00:47:14,080 S1: Especially if you're doing like a one off and you 773 00:47:14,080 --> 00:47:17,440 S1: just need maps real quick. A blog and pure text files. 774 00:47:17,440 --> 00:47:19,930 S1: I think it's a cool idea. Just write however you 775 00:47:19,930 --> 00:47:24,089 S1: do it. Science is a strong link problem. Very cool 776 00:47:24,090 --> 00:47:28,380 S1: essay and a frustrated Redditor asks what career options exist 777 00:47:28,380 --> 00:47:32,250 S1: for those who consider themselves less intelligent. I thought this 778 00:47:32,250 --> 00:47:36,600 S1: was a really fascinating discussion on Reddit. Recommendation of the week. 779 00:47:36,600 --> 00:47:40,410 S1: Remember that there's only so much that one person can do. 780 00:47:40,530 --> 00:47:44,400 S1: Good books are always there for you. Go to the classics. 781 00:47:44,550 --> 00:47:46,680 S1: If you're wondering what to read because you don't want 782 00:47:46,680 --> 00:47:49,830 S1: to like current events or politics or anything to seep in. 783 00:47:49,860 --> 00:47:53,370 S1: Go to the classics. Read the classics. Maybe not political ones, 784 00:47:53,370 --> 00:47:57,719 S1: but tons of books. Books that are there to cradle 785 00:47:57,750 --> 00:48:02,280 S1: you and save you and enrich you while you get 786 00:48:02,280 --> 00:48:07,500 S1: some needed separation from the shit show that is reality. 787 00:48:07,739 --> 00:48:12,570 S1: And supplement this with journaling. Um. My friend Clint Gibler 788 00:48:12,570 --> 00:48:17,970 S1: really has continually emphasized how important journaling is to him, 789 00:48:17,969 --> 00:48:22,480 S1: and it's got me writing more journal entries, so I 790 00:48:22,510 --> 00:48:27,009 S1: highly suggest that as well. Especially combined with separating and 791 00:48:27,010 --> 00:48:29,799 S1: possibly reading good books, the good books will give you 792 00:48:29,800 --> 00:48:33,610 S1: ideas which you could also combine with your life experience 793 00:48:33,610 --> 00:48:36,520 S1: and turn that into journaling. And the combination of the 794 00:48:36,520 --> 00:48:40,810 S1: two is really beneficial. And the aphorism of the week. 795 00:48:40,900 --> 00:48:44,710 S1: It is one of the blessings of old friends that 796 00:48:44,710 --> 00:48:47,500 S1: you can afford to be stupid with them. It is 797 00:48:47,500 --> 00:48:50,710 S1: one of the blessings of old friends that you can 798 00:48:50,710 --> 00:48:56,259 S1: afford to be stupid with them. Ralph Waldo Emerson. Unsupervised 799 00:48:56,260 --> 00:48:59,500 S1: learning is produced on Hindenburg Pro using an SM seven 800 00:48:59,530 --> 00:49:03,280 S1: B microphone. A video version of the podcast is available 801 00:49:03,310 --> 00:49:06,910 S1: on the Unsupervised Learning YouTube channel, and the text version 802 00:49:06,910 --> 00:49:10,150 S1: with full links and notes is available at Daniel Metzler 803 00:49:10,150 --> 00:49:13,750 S1: Comm Slash newsletter. We'll see you next time.