WEBVTT - UL NO. 447: Sam Curry on Bug Bounty Careers, Slack Data Exfil, The Work Lie 0:00:01.129 --> 0:00:04.460 Welcome to Unsupervised Learning, a security, AI and meaning focused 0:00:04.460 --> 0:00:07.310 podcast that looks at how best to thrive as humans 0:00:07.310 --> 0:00:11.540 in a post AI world. It combines original ideas, analysis, 0:00:11.570 --> 0:00:14.780 and mental models to bring not just the news, but 0:00:14.780 --> 0:00:22.340 why it matters and how to respond. All right, welcome 0:00:22.340 --> 0:00:26.060 to unsupervised learning. This is Daniel Miessler. All right. What 0:00:26.060 --> 0:00:29.480 have we got here? Lots of content this week. Super 0:00:29.480 --> 0:00:32.840 excited for this episode. Uh, did an all text episode 0:00:32.840 --> 0:00:34.940 this time, which is kind of a callback to how 0:00:34.940 --> 0:00:37.520 I used to do it. Got upcoming speaking at the 0:00:37.520 --> 0:00:43.400 Snyk conference in October, Cyberstorm in Switzerland and Blackhat in Riyadh. 0:00:43.430 --> 0:00:47.000 One tool in AI that you should be trying out, 0:00:47.000 --> 0:00:50.870 that everyone is talking about, including Karpathy and a bunch 0:00:50.870 --> 0:00:54.800 of other people, is called cursor AI. So it's cursor.com 0:00:54.800 --> 0:00:57.230 is the domain. I thought it was cursor AI, but 0:00:57.230 --> 0:01:00.730 it is not. But the big feature appears to be 0:01:00.730 --> 0:01:04.630 that it basically it looks and feels like VSCode, but 0:01:04.630 --> 0:01:08.259 what you do is you upload your entire repository into it. 0:01:08.290 --> 0:01:11.710 I guess I guess technically VSCode could also see your 0:01:11.740 --> 0:01:14.259 all your code if it's all in there as well. 0:01:14.260 --> 0:01:17.110 But I think what cursor is supposedly doing well is 0:01:17.110 --> 0:01:19.900 it's taking all of that content and kind of using 0:01:19.900 --> 0:01:22.900 it in context to understand it better, not just the 0:01:22.900 --> 0:01:25.240 current file that you're in. So I believe that's the 0:01:25.240 --> 0:01:28.360 big feature. If I'm wrong about that, somebody correct me. Okay. 0:01:28.390 --> 0:01:33.280 My work, a couple of massive episodes or essays that 0:01:33.280 --> 0:01:35.620 I put out this week, and I actually sent them 0:01:35.620 --> 0:01:38.470 out directly, which I only do for things that I 0:01:38.470 --> 0:01:43.000 think are really decent and also evergreen. So one of 0:01:43.000 --> 0:01:46.600 them is called We Were Lied To About Work or 0:01:46.600 --> 0:01:48.610 the Real Problem with the job market. I had two 0:01:48.610 --> 0:01:52.000 different titles actually, but it's basically why layoffs, hiring the 0:01:52.000 --> 0:01:54.850 job market, and work in general just really sucks right now. 0:01:54.850 --> 0:01:56.559 And I would say it's probably one of my top 0:01:56.590 --> 0:02:00.970 20 essays ever. So highly, highly recommend that one. And 0:02:00.970 --> 0:02:03.850 I've got a new way to explain AI and specifically 0:02:03.880 --> 0:02:06.880 Llms to people, and I think this one is short enough. 0:02:06.880 --> 0:02:08.920 In fact, I'm just going to give the highlights of it. 0:02:08.919 --> 0:02:11.950 So let me take you into this one. So here 0:02:11.980 --> 0:02:16.510 is the basic concept five levels of LLM understanding. So 0:02:16.510 --> 0:02:19.630 the first level at the bottom is it's just predicting 0:02:19.660 --> 0:02:23.440 text like the next text token in a in a 0:02:23.440 --> 0:02:27.429 text sequence. That's all it's doing. It's not magical. Right. 0:02:27.430 --> 0:02:30.010 And this is kind of the most common argument for 0:02:30.010 --> 0:02:33.580 why llms and AI is not all that special or 0:02:33.580 --> 0:02:37.180 specifically AI based on Llms. People are just like, look, 0:02:37.210 --> 0:02:40.030 it's just next token prediction. No big deal. If you 0:02:40.030 --> 0:02:43.090 pull away from that one level, it's like, look, it's 0:02:43.090 --> 0:02:46.360 just predicting the next item in a sequence, okay? So 0:02:46.360 --> 0:02:50.919 it's just next token prediction okay. One abstraction away from there. 0:02:50.919 --> 0:02:54.880 And this all comes from this thing from Eliezer Yudkowsky 0:02:54.910 --> 0:02:59.429 who had this, this great kind of little statement on 0:02:59.430 --> 0:03:02.340 X that made me think about this, but essentially what 0:03:02.340 --> 0:03:06.090 it is, is for this level. Level three, it's predicting 0:03:06.090 --> 0:03:09.839 the next token in the description of an answer. Okay. 0:03:09.870 --> 0:03:16.110 So what Yudkowsky said was any predicting the next token 0:03:16.110 --> 0:03:20.850 is isomorphic with predicting the next token of an answer. Okay. 0:03:20.880 --> 0:03:23.220 And that is really, really powerful. And it wasn't the 0:03:23.220 --> 0:03:25.740 exact quote, but that's essentially it. If you go one 0:03:25.740 --> 0:03:31.350 level above that okay, it's predicting answers to insanely difficult questions. 0:03:31.350 --> 0:03:34.590 So so here are the four levels so far. It's 0:03:34.590 --> 0:03:37.920 predicting the next token in a piece of text. It's 0:03:37.920 --> 0:03:41.670 predicting the next token third level. It's predicting the next 0:03:41.670 --> 0:03:46.080 token of an answer. Second level it's predicting answers. And 0:03:46.080 --> 0:03:49.770 the top level it seems to know everything okay. And 0:03:49.770 --> 0:03:53.100 this all comes from this. Here it is right here 0:03:53.100 --> 0:03:57.390 it just predicts the next token literally any well posed 0:03:57.410 --> 0:04:01.340 problem is isomorphic to predict the next token of the answer. 0:04:01.340 --> 0:04:05.030 And literally anyone with a grasp of undergraduate comp sci 0:04:05.060 --> 0:04:07.820 is supposed to see that without being told. I don't 0:04:07.850 --> 0:04:10.160 agree with that last part. This part here, I don't 0:04:10.190 --> 0:04:13.070 agree with that. I don't think that's accurate. In fact, 0:04:13.070 --> 0:04:16.400 I think the vast majority of people that's very much 0:04:16.400 --> 0:04:18.979 not the case. Very much not the case. I would 0:04:18.980 --> 0:04:23.300 say 95%. So I would say I disagree on multiple 0:04:23.300 --> 0:04:26.360 levels for that second piece. The part that I like 0:04:26.390 --> 0:04:32.060 is this here, which I highlighted and separated on purpose, 0:04:32.060 --> 0:04:36.950 literally any well posed problem is isomorphic to predict the 0:04:36.950 --> 0:04:42.530 next token of the answer. That is extraordinary. That is brilliant. 0:04:42.529 --> 0:04:46.310 That is the answer to the it's just next token 0:04:46.310 --> 0:04:50.240 prediction argument. That is, the immediate counter of that well 0:04:50.270 --> 0:04:54.500 posed problem equals prediction of next token of answer. So 0:04:54.500 --> 0:04:58.010 that's that one. And basically this essay is a full 0:04:58.040 --> 0:05:02.630 breakdown of that. So it's a standalone evergreen essay about 0:05:02.630 --> 0:05:05.660 why this is a powerful concept and why you shouldn't 0:05:05.660 --> 0:05:09.800 trust that next token argument. Okay. Security. CrowdStrike did their 0:05:09.980 --> 0:05:15.680 2024 report talking about how North Koreans have infiltrated over 0:05:15.680 --> 0:05:19.339 100 US based companies in all sorts of important places 0:05:19.339 --> 0:05:22.250 like aerospace, defense, retail and tech. They didn't mention much 0:05:22.250 --> 0:05:25.490 about Blue Friday. Not sure why that was. State linked 0:05:25.490 --> 0:05:29.930 Chinese entities are using cloud services from Amazon and other 0:05:29.930 --> 0:05:34.940 competitors of Amazon to access advanced US chips and AI capabilities, 0:05:34.940 --> 0:05:38.000 so basically they can't get the chips themselves. But you 0:05:38.000 --> 0:05:41.239 can go to AWS. Anybody can, or most people can. 0:05:41.270 --> 0:05:44.029 You go to AWS, you sign up and you just 0:05:44.029 --> 0:05:48.740 attach Nvidia chips to that instance and you start doing workloads. 0:05:48.740 --> 0:05:51.890 And so this is what China is doing. They're like, well, 0:05:51.920 --> 0:05:54.080 we can't get the chips, but we can use a 0:05:54.080 --> 0:05:57.679 service that leverages the chips. So I'm sure Amazon and 0:05:57.680 --> 0:06:00.590 the government are in conversations about that trying to get 0:06:00.589 --> 0:06:04.190 that fixed. Cisco has patched multiple vulnerabilities, including a high 0:06:04.190 --> 0:06:08.600 severity bug in its Unified Communications Manager product. Thanks to 0:06:08.630 --> 0:06:12.860 Threatlocker for sponsoring. Two US lawmakers are urging the Commerce 0:06:12.860 --> 0:06:18.260 Department to investigate cybersecurity risks associated with TP-Link routers, citing 0:06:18.260 --> 0:06:22.219 vulnerabilities and potential data sharing with the Chinese government. So 0:06:22.220 --> 0:06:25.820 kind of a Huawei situation. A mini Huawei Quarks lab 0:06:25.820 --> 0:06:28.970 found a major back door in RFID cards made by 0:06:28.970 --> 0:06:34.310 Shanghai Fudan Microelectronics, one of China's top chip manufacturers. I 0:06:34.310 --> 0:06:37.339 don't think this is a China influence story. I think 0:06:37.339 --> 0:06:40.489 this is just vulnerabilities and chips, and the impact here 0:06:40.490 --> 0:06:43.730 is the fact that it's smart cards for like office 0:06:43.730 --> 0:06:46.729 stores and hotel rooms and whatever. And it's a big 0:06:46.730 --> 0:06:50.780 company that does this. And lots of RFID chips and 0:06:50.779 --> 0:06:54.020 cards come from China. So just an impact due to 0:06:54.050 --> 0:06:56.440 the size of the market type situation. What is this 0:06:56.440 --> 0:07:03.370 bouncing now? Now you're seeing my Diablo four chat messages. 0:07:03.370 --> 0:07:06.039 It's really important stuff I'm working on in Diablo four. 0:07:06.070 --> 0:07:08.440 All right. What are we talking about here? Yeah. Thanks 0:07:08.440 --> 0:07:12.460 to defender five for sponsoring. Next one. Researchers found a 0:07:12.460 --> 0:07:15.670 way to exfiltrate data from Slack's AI by using indirect 0:07:15.670 --> 0:07:18.910 prompt injection. US Navy is rolling out Starlink on its 0:07:18.940 --> 0:07:23.380 warships to provide high speed, reliable internet connections, improving operations 0:07:23.380 --> 0:07:27.790 and crew morale. AI and Tech Anthropic has published the 0:07:27.790 --> 0:07:31.720 system prompts for its latest AI models, including opus, Sonnet 0:07:31.720 --> 0:07:34.840 and Haiku. AGI bot is a Chinese company and they 0:07:34.840 --> 0:07:38.920 just unveiled unveiled a fleet of advanced humanoid robots to 0:07:38.950 --> 0:07:43.150 compete directly with Optimus, which is the one from Tesla, 0:07:43.150 --> 0:07:46.900 and they're designed for tasks ranging from household chores to 0:07:46.930 --> 0:07:50.410 industrial operations. And they're going to start shipping supposedly by 0:07:50.410 --> 0:07:53.590 the end of this year. So like immediately. And Optimus 0:07:53.620 --> 0:07:57.960 is nowhere near ready for that kind of timeline. So 0:07:57.960 --> 0:08:02.910 I'm basically anti-Chinese imports for both robotaxis and humanoid robots, 0:08:02.910 --> 0:08:07.740 because China is too far ahead and they're too cheap 0:08:07.740 --> 0:08:10.800 and I would say just too good. So I don't 0:08:10.800 --> 0:08:12.960 want to give them a head start. And I don't 0:08:12.960 --> 0:08:16.740 like being anti-competitive against any sort of country. I don't 0:08:16.740 --> 0:08:19.530 like slowing pressure from the outside, but if this were 0:08:19.530 --> 0:08:22.560 India or Ireland, I would actually be okay with them 0:08:22.560 --> 0:08:25.350 applying pressure to the US but not China because they're 0:08:25.350 --> 0:08:28.830 too obviously a malicious actor who actually just like, wants 0:08:28.830 --> 0:08:31.950 to crush the United States in like all aspects, including 0:08:31.950 --> 0:08:34.920 the United States going away as an economic power. So 0:08:34.920 --> 0:08:37.859 it's not like friendly competition. So I think we should 0:08:37.890 --> 0:08:41.280 actually just either tax the hell out of them or 0:08:41.280 --> 0:08:44.010 not allow them to function until until we have proper 0:08:44.010 --> 0:08:47.280 footing and like, can compete properly. And speaking of that, 0:08:47.280 --> 0:08:51.030 Tesla is hiring people to train its Optimus humanoid robot 0:08:51.030 --> 0:08:55.100 by wearing a motion capture suit and mimicking the different actions. 0:08:55.100 --> 0:08:57.679 You get like $48 an hour, but you got to 0:08:57.710 --> 0:09:01.160 walk over seven hours a day carrying £30 while wearing 0:09:01.160 --> 0:09:04.340 a VR headset. That's a tough job. Waymo is looking 0:09:04.340 --> 0:09:08.150 to launch a subscription service called Waymo Teen, so this 0:09:08.150 --> 0:09:11.900 is basically to help parents not have to shuffle kids around. 0:09:11.929 --> 0:09:15.020 Although I'm not sure, depending on the age of the teen, 0:09:15.020 --> 0:09:17.840 should they have their own car, I'm not sure. But anyway, 0:09:17.900 --> 0:09:23.780 cool idea. I scientists developed by the University of British Columbia, 0:09:23.809 --> 0:09:30.110 Oxford and Sakana AI is creating its own machine learning experiments. Okay, 0:09:30.140 --> 0:09:34.309 let's back up. An AI scientist is creating its own 0:09:34.309 --> 0:09:37.849 machine learning experiments and running them autonomously. I think this 0:09:37.850 --> 0:09:40.880 is where most innovation will come from. AI not just 0:09:40.880 --> 0:09:44.179 implementing tasks, but in doing new research. And I talked 0:09:44.179 --> 0:09:48.260 about it in a post there. Victor Miller, a mayoral 0:09:48.260 --> 0:09:51.350 candidate in Wyoming's capital city, has vowed to let its 0:09:51.350 --> 0:09:56.679 customized ChatGPT named Vic Virtual Integrated Citizen help run the 0:09:56.679 --> 0:09:59.950 local government. I'm actually working on how to articulate a 0:09:59.950 --> 0:10:03.910 political platform for any level of office using substrate. Basically, 0:10:03.910 --> 0:10:06.400 define exactly what you want to do, how it branches 0:10:06.400 --> 0:10:10.510 out with problem strategies, most importantly KPIs and promises. So 0:10:10.510 --> 0:10:12.670 you could literally say, like I talked about it in 0:10:12.670 --> 0:10:15.610 the substrate video, you could say, look, here's how I'm 0:10:15.610 --> 0:10:18.850 measuring myself. Here are my assumptions. Here's what I believe 0:10:18.850 --> 0:10:23.080 the problems are. Here are my strategies for attacking those problems. 0:10:23.080 --> 0:10:27.070 And here are my specific projects that I'm going to 0:10:27.070 --> 0:10:30.400 do to implement those strategies. And here's how much it's 0:10:30.400 --> 0:10:34.330 going to cost. Here's how I'm measuring myself. And here's 0:10:34.330 --> 0:10:37.840 the promise if I don't move these numbers by X amount, 0:10:37.840 --> 0:10:39.970 you should fire me in two years or in four 0:10:39.970 --> 0:10:42.790 years or whatever the term is. So I think this 0:10:42.790 --> 0:10:48.130 is where leadership is heading. Transparent descriptions of vision, strategy, KPIs, 0:10:48.130 --> 0:10:54.280 and promises. Sean Ammirati Arathi, a professor at Carnegie Mellon, 0:10:54.309 --> 0:10:58.210 noticed a massive up leveling of progress in his entrepreneurship 0:10:58.210 --> 0:11:01.570 class this year, thanks to generative AI tools like ChatGPT, 0:11:01.900 --> 0:11:06.520 GitHub Copilot, and Flow wise, AI. So they basically the 0:11:06.520 --> 0:11:09.970 students use these tools for marketing, coding, product development, and 0:11:09.970 --> 0:11:14.079 recruiting early customers. This is what I've been talking about 0:11:14.080 --> 0:11:17.500 with AI automation. If you were competing with a 95 0:11:17.530 --> 0:11:20.530 out of 100 person before and they were a 95 0:11:20.559 --> 0:11:23.410 because they went to CMU, well, now you're competing with 0:11:23.410 --> 0:11:26.350 a 130 out of 100 because they went to CMU 0:11:26.350 --> 0:11:29.230 and they're using AI for everything. And for me, in 0:11:29.230 --> 0:11:32.050 my own life, I read better articles because of AI. 0:11:32.080 --> 0:11:35.110 I get better ideas because of AI. Therefore I build 0:11:35.110 --> 0:11:37.990 better stuff because of AI. And this is all feeding 0:11:37.990 --> 0:11:41.200 on itself, just okay. And this all feeds itself and 0:11:41.200 --> 0:11:44.020 just makes that whole top of funnel even better. And 0:11:44.020 --> 0:11:47.980 I believe that your options are to upgrade or lose. Basically, 0:11:48.010 --> 0:11:51.540 GM is cutting over 1000 software engineers to streamline its 0:11:51.540 --> 0:11:57.300 software services organization, streamlining by cutting out 1000 devs. The 0:11:57.300 --> 0:12:00.630 way I see this is they're actually just taking everyone 0:12:00.630 --> 0:12:03.210 out because they're like, this is a huge waste of time. 0:12:03.210 --> 0:12:06.270 We hired a bunch of duds. Let's start from scratch 0:12:06.270 --> 0:12:09.179 and only hire the best possible people who are probably 0:12:09.179 --> 0:12:12.300 also massively augmented with AI. Yeah, so I got a 0:12:12.300 --> 0:12:15.540 whole bunch of other content killer cult members. This is 0:12:15.540 --> 0:12:19.650 my new, um, way of framing. This whole thing is like, 0:12:19.650 --> 0:12:23.070 killer cult members is what companies are looking for. It's 0:12:23.070 --> 0:12:25.980 kind of obvious that this is what startups are looking for, 0:12:25.980 --> 0:12:28.140 but I think corporations are going to look for this 0:12:28.140 --> 0:12:31.380 as well. So people talk about, oh, toxic work culture. 0:12:31.410 --> 0:12:35.280 Guess what? Companies only want employees who have a toxic 0:12:35.280 --> 0:12:38.610 work culture. They want people who show up and are like, 0:12:38.610 --> 0:12:42.180 I am religiously dedicated to this. I will sleep under 0:12:42.179 --> 0:12:44.550 my desk. I will work as much time as I 0:12:44.550 --> 0:12:48.030 need to. I am fully dedicated to this mission. I 0:12:48.059 --> 0:12:51.140 wear all the swag in public. I talk about this. 0:12:51.140 --> 0:12:53.630 I think about this all day long. That is what 0:12:53.630 --> 0:12:56.689 people want. That is what hiring managers want. That is 0:12:56.690 --> 0:13:02.449 what corporations want. Because that behavior, especially if you're on site, 0:13:02.450 --> 0:13:05.750 which everyone is pulling everyone back to be on prem now, 0:13:05.750 --> 0:13:08.780 if you are surrounded by other cult members with that 0:13:08.780 --> 0:13:11.900 sort of energy and keep in mind, this also has 0:13:11.900 --> 0:13:16.100 toxic aspects to it. There are massive downsides to this 0:13:16.100 --> 0:13:20.420 type of culture, but the downsides are not to the company. Really. 0:13:20.420 --> 0:13:24.860 The downsides are mostly to the employees themselves as a 0:13:24.860 --> 0:13:28.160 trade off for doing this much risk to get much, 0:13:28.309 --> 0:13:32.090 a lot of a reward, potentially in equity or pay 0:13:32.120 --> 0:13:36.620 or whatever. So this culture is good for companies and 0:13:36.620 --> 0:13:39.140 that's why they are fostering it. That's why they're saying 0:13:39.140 --> 0:13:42.590 you must be on site. So for example, OpenAI OpenAI 0:13:42.620 --> 0:13:44.630 requires you to be on site. You have to work 0:13:44.630 --> 0:13:48.290 on site. There are some exceptions, very few. But in 0:13:48.290 --> 0:13:52.339 general they make a cool office. In a cool city, 0:13:52.340 --> 0:13:55.160 they require the best. They hire the best talent. They 0:13:55.160 --> 0:13:59.449 basically tell you, or it's naturally implied that you must 0:13:59.450 --> 0:14:02.180 be one of these killer cult members. And then that 0:14:02.179 --> 0:14:06.710 is that energy which is synergistic with itself, builds and 0:14:06.710 --> 0:14:09.950 builds and builds and it produces the best products that 0:14:09.950 --> 0:14:12.860 people want to buy, and they get massive valuations. And 0:14:12.890 --> 0:14:18.110 that's capitalism and that's what works. What doesn't work is, okay, 0:14:18.110 --> 0:14:21.980 let's not do that. Let's have a company culture that's 0:14:21.980 --> 0:14:25.880 good for employees, and let's make sure there's work life balance. 0:14:25.880 --> 0:14:29.180 And let's make sure, like all these different policies and 0:14:29.180 --> 0:14:32.720 it ends up with a lot of people not working, 0:14:32.720 --> 0:14:36.140 not doing good work. It ends up with A-players hiring 0:14:36.170 --> 0:14:40.070 B players. B players hire C players. You end up 0:14:40.070 --> 0:14:43.130 with a bunch of C players, a few B's and 0:14:43.280 --> 0:14:46.790 and some D players. And after a while, the corporation 0:14:46.790 --> 0:14:52.090 looks at their entire headcount, spend their entire human resources spend, 0:14:52.090 --> 0:14:54.370 and they're like, I'm not getting value from this. This 0:14:54.370 --> 0:14:59.140 is not worth it. And they just fire everyone. And 0:14:59.140 --> 0:15:02.110 so when I see somebody as firing a thousand software 0:15:02.110 --> 0:15:05.170 engineers to streamline the software, I don't know exactly what's 0:15:05.170 --> 0:15:08.200 happening here. I'm arguing that what I'm saying might be 0:15:08.230 --> 0:15:11.890 happening here is definitely happening all over the place. And 0:15:11.890 --> 0:15:14.170 what they'll do is they'll go to zero. Then they'll 0:15:14.170 --> 0:15:18.220 find a players who are killer cult members and only 0:15:18.220 --> 0:15:20.860 hire those from now on. And here's the crazy part. 0:15:20.890 --> 0:15:25.060 100 of those people might be worth 1000 or 10,000 0:15:25.090 --> 0:15:28.630 C players who are interested in work life balance. Again, 0:15:28.630 --> 0:15:32.050 I'm not saying anything about like the total value of 0:15:32.050 --> 0:15:35.350 on society of work life balance or all of that. 0:15:35.350 --> 0:15:39.040 I've got completely separate ideas about that. Actually, in the 0:15:39.040 --> 0:15:41.500 essay We were Lied To About Work, you can see 0:15:41.500 --> 0:15:44.200 what I really feel about that whole thing. The point is, 0:15:44.200 --> 0:15:46.690 this is what companies are looking for. If you want 0:15:46.720 --> 0:15:49.330 to get hired, you have the answer. Meta is using 0:15:49.360 --> 0:15:52.870 AI to streamline system reliability investigations with a new root 0:15:52.870 --> 0:15:57.010 cause analysis system. System combines heuristic based retrieval and large 0:15:57.010 --> 0:16:02.230 language model based ranking, achieving 42% accuracy in identifying root 0:16:02.230 --> 0:16:06.310 causes at the investigation start. I didn't look to see 0:16:06.340 --> 0:16:09.250 or it wasn't in there, how that compares to humans 0:16:09.250 --> 0:16:12.760 and how fast, because that's the trick. Accuracy and speed 0:16:12.760 --> 0:16:15.940 and of course, cost. AI companies are shifting focus from 0:16:15.940 --> 0:16:20.410 creating godlike AI to building practical products. Who knew? So 0:16:20.410 --> 0:16:22.150 I don't think this is a bubble pop. I think 0:16:22.150 --> 0:16:25.240 it's a natural maturity of brand new tech that just 0:16:25.240 --> 0:16:28.120 came out. And because people are still figuring this stuff 0:16:28.120 --> 0:16:31.120 out and it's basically day one, like AI hasn't even 0:16:31.120 --> 0:16:34.330 gotten good yet, hasn't even started to get good. Canada 0:16:34.330 --> 0:16:40.690 is slapping 100% import tariffs on Chinese electric vehicles starting 0:16:40.720 --> 0:16:44.290 October 1st. We were just talking about that, former Google 0:16:44.320 --> 0:16:48.330 CEO Eric Schmidt predicts rapid advancements in AI, the potential 0:16:48.330 --> 0:16:52.530 to create significant apps like TikTok competitors in minutes within 0:16:52.530 --> 0:16:55.140 the next few years. I know what he's saying, but 0:16:55.140 --> 0:16:57.840 there's a difference between being able to run an app 0:16:57.840 --> 0:17:01.590 at scale versus creating it. Of course, he knows that 0:17:01.590 --> 0:17:04.950 better than I do or most people. But important point 0:17:04.950 --> 0:17:09.149 to make. Claude 3.5 can now create icalendar files from images. 0:17:09.150 --> 0:17:12.750 And this guy Greg's ramblings shows how you can use 0:17:12.750 --> 0:17:16.440 this feature to generate calendar entries by snapping a photo 0:17:16.440 --> 0:17:20.940 of a schedule or event flyer. AWS CEO Adam Selipsky 0:17:20.970 --> 0:17:25.110 predicts that within the next 24 months, most developers might 0:17:25.109 --> 0:17:28.890 not be coding anymore due to AI advancements. He says 0:17:28.890 --> 0:17:33.300 the real skill shift will be towards innovation and understanding 0:17:33.300 --> 0:17:39.030 customer needs, rather than writing code. 100% agree. Although most 0:17:39.030 --> 0:17:42.480 developers in 24 months. This is the type of thing 0:17:42.480 --> 0:17:45.920 where it happens way slower than you think and then 0:17:45.950 --> 0:17:48.800 way faster than you think. At the same time, most 0:17:48.800 --> 0:17:52.520 in 24 months will not be coding anymore. That is 0:17:52.520 --> 0:17:56.750 wildly off, wildly off. But people who are like, oh, 0:17:56.780 --> 0:17:58.730 it's going to take forever. It's going to take, you know, 0:17:58.760 --> 0:18:01.939 three years, five years, seven years, ten years, 15 years. 0:18:01.940 --> 0:18:05.659 They are also wildly off. Chinese companies have ramped up 0:18:05.660 --> 0:18:11.659 imports of chip production equipment. $26 billion in the first 0:18:11.660 --> 0:18:15.530 seven months of 24 on chip production equipment. They need 0:18:15.530 --> 0:18:20.240 to equip 18 new fabs expected to start operations in 24. 0:18:20.570 --> 0:18:23.120 There they are all in on this because they see 0:18:23.119 --> 0:18:26.359 the world turning against them and shutting them down. So 0:18:26.359 --> 0:18:29.570 they are all in on this. Question is, can they actually, 0:18:29.570 --> 0:18:33.560 even with all the fabs and the equipment and you know, 0:18:33.590 --> 0:18:36.500 the factories, do they have the know how to actually 0:18:36.500 --> 0:18:40.580 do what TSMC is doing? My current understanding from current 0:18:40.609 --> 0:18:44.300 knowledge combined with like the book chip. Wars know they 0:18:44.300 --> 0:18:47.230 don't have the knowledge. So even with all that stuff, 0:18:47.230 --> 0:18:50.170 it's not going to be enough humans. Cisco's laying off 7% 0:18:50.170 --> 0:18:54.040 of its workforce, around 6000 employees, as it pivots towards 0:18:54.070 --> 0:18:58.209 AI and cybersecurity. McKinsey's new study reveals that business leaders 0:18:58.210 --> 0:19:01.810 are missing the mark on why employees are quitting. They 0:19:01.840 --> 0:19:06.970 say companies are focusing on transactional perks like compensation and flexibility, 0:19:06.970 --> 0:19:12.909 but employees are actually seeking meaning. Belonging, holistic care and 0:19:12.910 --> 0:19:17.230 appreciation at work couldn't have been better timed with this 0:19:17.230 --> 0:19:22.660 week's essay. 24 brain samples collected in early 2424 brain 0:19:22.660 --> 0:19:29.440 samples in 2024 measured an average of 0.5% plastic by weight. 0:19:29.440 --> 0:19:33.490 So a brain is multiple pounds. Is it £3 or £2? 0:19:33.490 --> 0:19:37.030 I can't remember 3 or £4, something like that. It's heavy. 0:19:37.060 --> 0:19:39.430 You can feel this thing right. You can feel your head. 0:19:39.430 --> 0:19:43.930 It's heavy. Half a percent. Okay. Let's just type in here. 0:19:43.960 --> 0:19:46.120 What's the half a percent of the weight of an 0:19:46.119 --> 0:19:49.990 average brain? 1400 grams. Half a percent of that would 0:19:49.990 --> 0:19:56.230 be seven grams. Okay, I put 25g of coffee to 0:19:56.260 --> 0:20:00.070 brew when I brew coffee, so I roughly feel like 0:20:00.070 --> 0:20:02.950 I know that's about a third, about a third of 0:20:02.950 --> 0:20:05.020 the amount of coffee that I drink in the morning, 0:20:05.020 --> 0:20:08.260 which when you brew that thing, I mean, it's significant. 0:20:08.260 --> 0:20:13.750 This is a extremely non-trivial amount of plastic sitting inside 0:20:13.750 --> 0:20:16.480 the brain. And there's a lot of speculation right now 0:20:16.480 --> 0:20:20.080 which I don't put too much weight into it because, 0:20:20.109 --> 0:20:22.420 you know, how we have these health scares and like, oh, 0:20:22.450 --> 0:20:25.690 this is dangerous and that's dangerous or whatever, but seven 0:20:25.690 --> 0:20:31.510 grams of plastic, seven grams of plastic sitting inside of 0:20:31.510 --> 0:20:34.869 a brain. I feel like it can't be good unless 0:20:34.869 --> 0:20:38.199 it's like alien plastic that is like nanobots or something. 0:20:38.200 --> 0:20:41.260 But no, this is regular plastic. The question is, where 0:20:41.290 --> 0:20:43.210 is it coming from? How is it getting in there? 0:20:43.210 --> 0:20:45.660 Is it in all our foods? Is it in? Is 0:20:45.660 --> 0:20:48.600 it from drinking bottles? I'm actually super concerned about this 0:20:48.600 --> 0:20:51.510 because I drink energy drinks. That's why I switched to 0:20:51.540 --> 0:20:53.609 these because I think they have less. But I saw 0:20:53.609 --> 0:20:58.140 a crazy report about my favorite not energy drinks, protein drinks, 0:20:58.140 --> 0:21:01.740 my favorite protein drink. It was the core power one, 0:21:01.740 --> 0:21:04.560 and it was like off the charts in the amount 0:21:04.590 --> 0:21:07.350 of plastic, according to this one lab that ran it. 0:21:07.380 --> 0:21:10.050 Now who knows, maybe that lab was run by the 0:21:10.050 --> 0:21:12.930 counter product, which I went out and bought, by the way, 0:21:12.930 --> 0:21:16.800 which is Muscle Milk. Anyway, I'm concerned about this, but 0:21:16.800 --> 0:21:19.170 I'm also cautious because you don't know if it's a 0:21:19.170 --> 0:21:23.130 scare turn. That plastic could be completely inert and doesn't 0:21:23.130 --> 0:21:26.580 even matter. And the thing that's causing all this cancer 0:21:26.580 --> 0:21:30.630 and drop in testosterone and all this stuff is actually 0:21:30.630 --> 0:21:32.880 something else could be our sun exposure, which I think 0:21:32.880 --> 0:21:36.120 is a later. Yeah, it's two stories down. So anyway, 0:21:36.119 --> 0:21:39.450 lots of plastic in our brains. Not sure what that 0:21:39.450 --> 0:21:43.820 means yet. Gallup has released its 2023 Global Emotions Report, 0:21:43.820 --> 0:21:48.139 which measures the world's emotional temperature through Positive Experience Index. 0:21:48.140 --> 0:21:50.959 I'm opening this one because it is cool. Look at this. 0:21:50.990 --> 0:21:55.159 Experienced anger. Look at this. You got these country breakdowns. 0:21:55.160 --> 0:21:57.650 This is really cool. Then you got map views. And 0:21:57.650 --> 0:22:00.050 look at this. The map view. You can click on 0:22:00.050 --> 0:22:04.340 sadness stress worry pain. Look at that enjoyment. Okay. So 0:22:04.369 --> 0:22:10.460 dark is better right? Um. China. Yes. Yeah. Dark. Yep. Okay, 0:22:10.490 --> 0:22:14.720 so let's look at a light one. Yes, 100% dark 0:22:14.720 --> 0:22:19.369 is better because that's Afghanistan. Two thirds. No. Oh, man, 0:22:19.400 --> 0:22:24.379 that's so depressing. Afghanistan has two thirds. No enjoyment. However, 0:22:24.380 --> 0:22:28.040 the question was asked another one with that similar color turkey. 0:22:28.070 --> 0:22:31.310 I'm pretty sure that's turkey. Yeah that's Turkey. Uh, Ukraine. 0:22:31.310 --> 0:22:35.570 About half, for obvious reasons. What's this one? Morocco half. 0:22:35.600 --> 0:22:40.340 What's this one? Tunisia half. Roughly super happy over here. 0:22:40.340 --> 0:22:45.950 Somalia Why is Somalia so happy? See, I love these visualizations. Uzbekistan. 0:22:45.950 --> 0:22:48.889 Very happy. What have we got over here? Norway is happy. 0:22:48.920 --> 0:22:51.439 Keep forgetting. Norway is the far left one. What is 0:22:51.440 --> 0:22:54.980 this one? Over here. Estonia. Very happy. Iceland could have 0:22:54.980 --> 0:23:00.050 guessed that one. Mexico! Kicking total ass over here. Ireland seems. Yeah. 0:23:00.080 --> 0:23:04.850 Really happy. UK? Not so much. What are these? Over here. Indonesia. Indonesia. 0:23:04.880 --> 0:23:10.850 Very happy. Malaysia. Very happy anyway. Really cool stuff. Russian Federation. 0:23:10.880 --> 0:23:14.000 Pretty low score, actually. I think this is the actual 0:23:14.000 --> 0:23:18.199 worst one. Afghanistan, which is totally explainable. I mean, so 0:23:18.200 --> 0:23:20.780 this is one experience I've been having is I take 0:23:20.780 --> 0:23:23.149 a decent number of Ubers and I always talk to 0:23:23.180 --> 0:23:26.750 the person in the Bay area. My chances of getting 0:23:26.750 --> 0:23:31.190 someone from Afghanistan are like 80, 90%, and I talk 0:23:31.190 --> 0:23:33.800 to every single one of them for the entire duration 0:23:33.800 --> 0:23:38.270 of the trip. Usually they are interpreters, don't say translator, 0:23:38.270 --> 0:23:41.800 they are interpreters. And oftentimes they worked for the US government, 0:23:41.800 --> 0:23:43.900 which is how they got their visa to come over 0:23:43.900 --> 0:23:47.710 here and what they have back home if they brought 0:23:47.710 --> 0:23:51.100 their family. Most likely they didn't bring most of their family. 0:23:51.100 --> 0:23:54.340 Their family is in danger because they are here and 0:23:54.340 --> 0:23:58.000 it is going from bad to worse every single week. 0:23:58.030 --> 0:24:01.960 It just is so, so depressing. Anyway, I'm on a 0:24:01.960 --> 0:24:05.770 tangent here, but talk to your Uber drivers. Experience life 0:24:05.770 --> 0:24:11.080 through other people's eyes. Okay. Um. Data from surveys conducted 0:24:11.080 --> 0:24:14.530 in 142 countries. Mix of telephone, face to face, and 0:24:14.530 --> 0:24:18.400 some web stuff. About a thousand respondents per country. So 0:24:18.400 --> 0:24:20.950 that's not great, but I'm sure they're doing it scientifically, 0:24:20.950 --> 0:24:25.270 so it's a decent sample. Non-smokers who avoided the sun 0:24:25.300 --> 0:24:28.870 had a life expectancy similar to smokers who got the 0:24:28.869 --> 0:24:33.639 most sun. And this is nearly 30,000 Swedish women over 0:24:33.640 --> 0:24:37.960 20 years, by the way, Scandinavia likes to smoke. And Germany, 0:24:37.960 --> 0:24:40.169 this is the worst thing about Europe and there are 0:24:40.170 --> 0:24:42.720 many things competing for that right now. But one of 0:24:42.720 --> 0:24:45.240 the worst things about Europe is smoking. I go there 0:24:45.240 --> 0:24:48.960 and I'm just like, what is going on? Yeah, Switzerland. 0:24:48.990 --> 0:24:50.820 I'm about to go back to Switzerland and everyone's going 0:24:50.850 --> 0:24:54.630 to be smoking inside the restaurant. They're smoking inside the restaurant. 0:24:54.630 --> 0:24:56.820 I'm just like, what is? I thought you guys were 0:24:56.820 --> 0:25:00.300 the advanced group. Okay, the research suggests that avoiding sun 0:25:00.300 --> 0:25:04.919 is as risky as smoking. So this needs more research, obviously. 0:25:04.920 --> 0:25:07.860 But like I said, damn. For me, I get sun 0:25:07.859 --> 0:25:10.680 in the morning. I got decent amount of sun this morning. 0:25:10.680 --> 0:25:14.550 Massive boost for me. I put on the Waking Up app, 0:25:14.580 --> 0:25:17.550 listen to Sam do a ten minute thing. I do 0:25:17.550 --> 0:25:19.860 a ten minute walk out there, I do some breathing. 0:25:19.859 --> 0:25:22.170 That's how I start my day. When I'm on a routine, 0:25:22.170 --> 0:25:27.270 which I should be and often not enough. But I 0:25:27.270 --> 0:25:29.850 did today and yesterday and the day before. All right. 0:25:29.880 --> 0:25:34.080 Stanford researchers have found that blocking the whatever blocking this 0:25:34.080 --> 0:25:38.129 pathway in the brain can reverse the metabolic disruptions caused 0:25:38.130 --> 0:25:43.770 by Alzheimer's disease, providing cognitive improving cognitive functions in mice. 0:25:43.770 --> 0:25:46.920 I'm starting to feel like we're about to make massive 0:25:46.920 --> 0:25:51.090 progress on both Alzheimer's and cancer. And honestly, it's making 0:25:51.090 --> 0:25:52.980 me want to invest in like, the top three drug 0:25:52.980 --> 0:25:56.250 companies I think I'm already in, which one I'm in, 0:25:56.250 --> 0:25:58.620 the one that does wegovy. I can't remember the name 0:25:58.619 --> 0:26:01.710 of that one. I'll think of it. It's right there. 0:26:01.710 --> 0:26:04.440 It's right there. Anyway, I'm in that one, but I'm 0:26:04.440 --> 0:26:07.139 not in the Lilly one. I think they're called Eli Lilly, 0:26:07.140 --> 0:26:10.080 but they're rebranding to Lilly. So I figure if I 0:26:10.080 --> 0:26:13.470 get into the two top competitors, one or both of 0:26:13.470 --> 0:26:16.859 them is going to do something with Alzheimer's and cancer, 0:26:16.859 --> 0:26:21.899 and we're already doing it with obesity. Like, that's a trifecta. 0:26:21.930 --> 0:26:25.560 All we need now is like balding and aging. Aging 0:26:25.560 --> 0:26:27.600 is the big one. Aging and cancer I would say, 0:26:27.630 --> 0:26:31.800 are the big ones. And then like obesity and hair loss, 0:26:31.800 --> 0:26:36.300 that's amazing. So not investment advice, but I'm damn sure 0:26:36.330 --> 0:26:40.760 getting in all right. Air purifiers and two Helsinki daycare 0:26:40.760 --> 0:26:45.440 centers reduced sick kids days by 30%. And I don't 0:26:45.440 --> 0:26:48.590 think this is Covid or flu or anything. I think 0:26:48.590 --> 0:26:51.290 they're just talking about like, all cause based on the 0:26:51.290 --> 0:26:55.129 the study parts that I saw, University of Missouri scientists 0:26:55.130 --> 0:26:59.240 have developed a liquid based solution that removes over 98% 0:26:59.240 --> 0:27:02.930 of nanoplastics from water. It uses water repelling solvents to 0:27:02.960 --> 0:27:07.189 absorb the particles, which are easily separated and removed. I 0:27:07.190 --> 0:27:09.110 assume you just run it through a filter and those 0:27:09.109 --> 0:27:12.500 things will be big globs, and they get stopped behind 0:27:12.500 --> 0:27:14.990 by the filter. I expect to see lots more of this. 0:27:15.020 --> 0:27:17.930 Can't wait for the Huberman episode on this, because I 0:27:17.930 --> 0:27:20.930 wanted to be able to do this like cheaply at home. 0:27:20.930 --> 0:27:25.070 Eli Lilly's weight loss drug Tripeptide, found in Zepp bound 0:27:25.100 --> 0:27:29.359 and Manjaro, reduced the risk of developing type two diabetes 0:27:29.359 --> 0:27:36.649 by 94% in obese or overweight adults with pre-diabetes, 94%. 0:27:36.650 --> 0:27:39.790 And Apple Podcasts is losing ground to YouTube and Spotify, 0:27:39.790 --> 0:27:46.600 so recent study put YouTube at 31%, Spotify at 21%, 0:27:46.600 --> 0:27:50.050 and Apple Podcasts at 12. I don't do Spotify, I 0:27:50.050 --> 0:27:54.699 do YouTube and Apple Podcasts. But all right. Ideas thought 0:27:54.700 --> 0:27:58.360 of a cool idea for fabric, Telos and substrate. Maintain 0:27:58.359 --> 0:28:01.210 a list of everything I've been really, really wrong about, 0:28:01.210 --> 0:28:03.429 which I'm already building that list, and then write a 0:28:03.430 --> 0:28:06.070 fabric pattern that looks at that list. By the way, 0:28:06.070 --> 0:28:07.600 I'm just going to have it all in my same 0:28:07.630 --> 0:28:10.600 Telos file as I talked about in the augmented course, 0:28:10.600 --> 0:28:15.700 but basically just have this list listed as all my 0:28:15.700 --> 0:28:20.200 biggest mistakes, cognitive errors or whatever listed in there. Then 0:28:20.200 --> 0:28:23.290 I also have another section inside that same telos file, 0:28:23.290 --> 0:28:25.869 which is like all my current beliefs, my model of 0:28:25.869 --> 0:28:29.470 the world. And then the fabric pattern basically says evaluate 0:28:29.470 --> 0:28:33.160 all my current beliefs, look at all the previous mistakes 0:28:33.160 --> 0:28:36.670 that I've made and look for patterns. Look for what 0:28:36.670 --> 0:28:40.390 in my current beliefs might be broken in a similar 0:28:40.390 --> 0:28:43.030 way as the way I was wrong about the other things. 0:28:43.030 --> 0:28:45.670 First of all, it's going to help me diagnose why 0:28:45.670 --> 0:28:47.979 I was wrong. Is it because I have a bias 0:28:47.980 --> 0:28:50.590 towards this one thing? It's like, oh, you're so pro 0:28:50.590 --> 0:28:54.070 I that you are wrong about this thing because you 0:28:54.070 --> 0:28:57.790 thought technology was a solution. So turns out your bias 0:28:57.820 --> 0:29:00.670 is thinking that tech can solve too many things in 0:29:00.670 --> 0:29:03.970 the world. Good to know. Like I'm already actively defending 0:29:03.970 --> 0:29:07.270 against that bias because I know it's there doesn't mean 0:29:07.270 --> 0:29:11.470 I'm properly or adequately defending against that bias. Point is, 0:29:11.470 --> 0:29:13.150 I want to see the bias. I want to see 0:29:13.150 --> 0:29:15.310 it call me out on it, and look for other 0:29:15.310 --> 0:29:18.370 evidence that my current beliefs are broken because of that problem. 0:29:18.370 --> 0:29:22.450 Discovery Foof I uses fluff and I yeah, that's why 0:29:22.450 --> 0:29:25.720 it's called that. To find more web hacking targets by 0:29:25.720 --> 0:29:30.700 Joseph Thacker, go fuzzy recursively looks at JavaScript files and 0:29:30.700 --> 0:29:34.510 finds endpoints that can be tested. Analyze interviewer techniques is 0:29:34.510 --> 0:29:37.660 a new fabric pattern that will capture the je NE 0:29:37.660 --> 0:29:40.030 sais quoi. By the way, this is spelled wrong. I 0:29:40.030 --> 0:29:42.550 don't know why I didn't just use AI and spell 0:29:42.550 --> 0:29:46.480 this correctly. I always spell this wrong. N e is 0:29:46.510 --> 0:29:49.690 n a I s and I think the other one 0:29:49.690 --> 0:29:52.090 might even be wrong too, but I think the quote 0:29:52.120 --> 0:29:55.660 is right. But anyway, it's n I s I believe 0:29:55.690 --> 0:29:58.600 is how you spell that. Anyway, someone French gave me 0:29:58.600 --> 0:30:02.290 the proper spelling again, more I should be part of this. Um, 0:30:02.680 --> 0:30:05.590 back to my bias of I being the problem, but 0:30:05.590 --> 0:30:10.780 I've been using this on Dwarkesh and Tyler Cowen content 0:30:10.780 --> 0:30:14.680 analyze interviewer techniques. It basically finds out, it figures out 0:30:14.680 --> 0:30:17.470 and tells you why they're such a good interviewer. Harness 0:30:17.500 --> 0:30:19.420 is a quick tool I put together to test the 0:30:19.420 --> 0:30:22.480 efficacy of one prompt versus another. It runs both against 0:30:22.480 --> 0:30:25.450 an input and then scores the output according to a 0:30:25.450 --> 0:30:28.989 third objective prompt that rates how well they followed the 0:30:28.990 --> 0:30:33.880 plot and actually executed the instructions of what the prompts 0:30:33.880 --> 0:30:37.340 were trying to do in the first place. So super useful. 0:30:37.370 --> 0:30:40.700 Stayed in time are the same thing by Hillel Wayne. 0:30:40.700 --> 0:30:43.190 Don't Force yourself to become a bug bounty Hunter by 0:30:43.190 --> 0:30:47.540 Sam Curry 67 years of RadioShack catalogs have been scanned 0:30:47.540 --> 0:30:51.290 and are now available online. MD RSS is a go 0:30:51.320 --> 0:30:54.470 based tool that converts markdown files to RSS feeds. You 0:30:54.470 --> 0:30:57.380 can write articles in a local folder, and it automatically 0:30:57.380 --> 0:31:01.340 formats them into an RSS compliant XML file. Super cool. 0:31:01.340 --> 0:31:05.480 No hello, no quick call, no meetings without an agenda. 0:31:05.510 --> 0:31:08.870 You already know that's good. Roger Penrose's book, The Emperor's 0:31:08.900 --> 0:31:12.560 New Mind explores the relationship between the human mind and computers, 0:31:12.560 --> 0:31:16.880 arguing that human consciousness cannot be replicated by machines. I 0:31:16.880 --> 0:31:18.740 have the opposite view, which is why I'm going to 0:31:18.740 --> 0:31:21.650 read this book collection of free public APIs that are 0:31:21.650 --> 0:31:24.020 tested daily and the recommendation of the week. Take the 0:31:24.020 --> 0:31:26.930 time to read this week's main essay. We've been lied 0:31:26.960 --> 0:31:28.970 to about work, but more than just read it. Think 0:31:28.970 --> 0:31:31.040 about what it means. If I am right, think about 0:31:31.040 --> 0:31:33.290 what that means for you and your career, but also 0:31:33.290 --> 0:31:35.630 all the young people you know and care about. So 0:31:35.630 --> 0:31:37.670 I didn't talk about the solution in the piece, but 0:31:37.670 --> 0:31:40.460 it's essentially human 3.0, and I'm going to be talking 0:31:40.460 --> 0:31:43.310 a lot more about that, but start thinking about it now. 0:31:43.340 --> 0:31:46.640 Definitely recommend that you read this piece. And the aphorism 0:31:46.640 --> 0:31:49.070 of the week to fear love is to fear life, 0:31:49.070 --> 0:31:52.970 and those who fear life are already three parts dead 0:31:53.000 --> 0:31:55.430 to fear. Love is to fear life, and those who 0:31:55.430 --> 0:32:01.280 fear life are already three parts dead. Bertrand Russell. Unsupervised 0:32:01.280 --> 0:32:03.800 learning is produced and edited by Daniel Miessler on a 0:32:03.800 --> 0:32:08.660 Neumann U87 AI microphone using Hindenburg. Intro and outro music 0:32:08.660 --> 0:32:11.750 is by zombie with a Y. And to get the 0:32:11.750 --> 0:32:13.820 text and links from this episode, sign up for the 0:32:13.820 --> 0:32:19.459 newsletter version of the show at Daniel miessler.com/newsletter. We'll see 0:32:19.460 --> 0:32:20.240 you next time.