WEBVTT - UL NO. 444: Pizza Meter Intelligence, China Bypasses Bans, Securing AWS Secrets… 0:00:00.110 --> 0:00:03.680 Whether you're starting or scaling your company's security program, demonstrating 0:00:03.710 --> 0:00:07.220 top notch security practices and establishing trust is more important 0:00:07.220 --> 0:00:14.900 than ever. Vanta automates compliance for Soc2, ISO 27,001 and more, 0:00:14.900 --> 0:00:19.520 saving you time and money while helping you build customer trust. Plus, 0:00:19.520 --> 0:00:23.780 you can streamline security reviews by automating questionnaires and demonstrating 0:00:23.780 --> 0:00:27.500 your security posture with a customer facing trust center, all 0:00:27.500 --> 0:00:32.970 powered by AI. Over 7000 global companies like Atlassian, Flow 0:00:32.970 --> 0:00:36.090 Health and Quora use Vanta to manage risk and prove 0:00:36.090 --> 0:00:40.560 security in real time. Get $1,000 off Vanta when you 0:00:40.560 --> 0:00:48.180 go to Vanta comm slash unsupervised. That's vanta.com/supervised for $1,000 off. 0:00:49.979 --> 0:00:53.339 Welcome to Unsupervised Learning, a security, AI and meaning focused 0:00:53.350 --> 0:00:56.230 podcast that looks at how best to thrive as humans 0:00:56.230 --> 0:01:00.430 in a post AI world. It combines original ideas, analysis, 0:01:00.430 --> 0:01:03.670 and mental models to bring not just the news, but 0:01:03.670 --> 0:01:11.319 why it matters and how to respond. All right, welcome 0:01:11.319 --> 0:01:14.230 to unsupervised Learning. This is Daniel Miessler. I am in 0:01:14.230 --> 0:01:17.960 a hotel room in Vegas recording because I have to. 0:01:17.959 --> 0:01:22.250 But it is Hacker Week and I'm here for Defcon 0:01:22.250 --> 0:01:24.830 and Black Hat and all that good stuff, but I 0:01:24.830 --> 0:01:28.459 wanted to get the episode out. So first note here 0:01:28.459 --> 0:01:31.220 is that Osint is one of my favorite hobbies, and 0:01:31.220 --> 0:01:33.950 there's something called a pizza index. That's one of my 0:01:33.950 --> 0:01:37.970 favorite examples of this, which is how much pizza essentially 0:01:37.970 --> 0:01:41.390 the neighborhood around the Pentagon is ordering, which really means 0:01:41.390 --> 0:01:44.940 the Pentagon. And there's another index related to that, which 0:01:44.940 --> 0:01:48.480 is how many people are in the bars and this 0:01:48.480 --> 0:01:53.610 person real. Ben Geller posted a tweet about this. And 0:01:53.610 --> 0:01:57.210 essentially it says it shows that the people in the 0:01:57.210 --> 0:02:01.440 bars is like extremely low and the pizza meter is 0:02:01.440 --> 0:02:04.560 off the charts. And I just love this so much 0:02:04.560 --> 0:02:08.920 because it indicates pretty strongly that something is about to 0:02:08.919 --> 0:02:10.870 go down. And I've got a friend who used to 0:02:10.870 --> 0:02:13.870 be an analyst at the Pentagon, and he says, this 0:02:13.870 --> 0:02:16.990 is absolutely true. When people are ordering in pizza and 0:02:16.990 --> 0:02:20.619 nobody's going home, it's obviously because something is going down. 0:02:20.620 --> 0:02:22.840 And in this case, we kind of know what's going down, 0:02:22.840 --> 0:02:27.100 which is Iran is preparing to attack Israel and or 0:02:27.100 --> 0:02:31.130 whoever else. So that's what that is. But definitely check 0:02:31.130 --> 0:02:34.100 out this tweet. It's pretty interesting. So this is also 0:02:34.100 --> 0:02:37.220 why I can't wait to fully build out my agent framework, 0:02:37.220 --> 0:02:41.090 and for agent framework to become more tightly integrated with 0:02:41.090 --> 0:02:44.030 models and platforms, because it's going to allow a lot 0:02:44.030 --> 0:02:46.550 more people to do things like this. What I love 0:02:46.550 --> 0:02:50.480 about it is you could track all the different experts, right? 0:02:50.480 --> 0:02:52.250 I'm going to use a whole bunch of different stuff 0:02:52.250 --> 0:02:55.200 for this, but there will be some agent functionality in 0:02:55.200 --> 0:02:58.920 the middle to sort of handle, like orchestration and summarization 0:02:58.919 --> 0:03:01.650 and creating like an Intel report. But I love the 0:03:01.650 --> 0:03:08.130 idea of like gathering all these individual, hopefully standalone intelligence sources, 0:03:08.130 --> 0:03:12.780 aggregating them together, but also keeping them separate and then 0:03:12.780 --> 0:03:16.920 triangulating on truth based on that. And I heard some 0:03:16.919 --> 0:03:20.320 pretty cool ideas from, um, it was actually a friend 0:03:20.320 --> 0:03:23.709 of mine named John, uh, who was talking about how 0:03:23.710 --> 0:03:26.350 you want to rate those different sources in different ways. 0:03:26.350 --> 0:03:29.260 One way to rate them is to rate them based 0:03:29.260 --> 0:03:33.280 on their difference and their uniqueness of ideas relative to 0:03:33.280 --> 0:03:35.980 other people, because you don't know if they're actually just 0:03:35.980 --> 0:03:39.670 reading other people's Intel and following along, and you don't 0:03:39.670 --> 0:03:42.310 want to use eight of those people who are all 0:03:42.310 --> 0:03:46.820 following the same thing as eight different ends, right? Eight 0:03:46.820 --> 0:03:51.170 different sources of of data or sources of signal. So 0:03:51.170 --> 0:03:53.300 there's a whole bunch of cool stuff that you could 0:03:53.300 --> 0:03:56.480 do once you have the discrete signal coming in from 0:03:56.480 --> 0:03:59.270 all these different places. And then you could factor in 0:03:59.270 --> 0:04:02.960 things like prediction markets and stuff like that and just, uh, 0:04:03.350 --> 0:04:05.840 lots of different stuff you could do. But ultimately what 0:04:05.840 --> 0:04:09.570 I want is a daily Intel report, which is as 0:04:09.570 --> 0:04:12.030 good or better than what you would get from, like 0:04:12.030 --> 0:04:15.360 Stratfor back in the day. Or you know what a 0:04:15.360 --> 0:04:19.230 lot of these paid platforms would do. Um, or even like, 0:04:19.230 --> 0:04:21.869 you know, a high level government thing, I think we 0:04:21.870 --> 0:04:26.760 could build something really, really good that leverages the intelligence 0:04:26.760 --> 0:04:29.610 of all these different, really smart people who are posting 0:04:29.610 --> 0:04:31.920 their stuff online. We're not talking about private stuff. We're 0:04:31.930 --> 0:04:35.560 talking about people on Twitter. We're talking about people on 0:04:35.560 --> 0:04:39.640 different platforms, blogs. They're writing their stuff out there. And 0:04:39.640 --> 0:04:42.039 a lot of times nobody's reading it. But if you 0:04:42.040 --> 0:04:44.200 put the effort in, you could find all those signals 0:04:44.200 --> 0:04:48.490 and start triangulating. So really excited about that. Okay. The 0:04:48.490 --> 0:04:52.330 state of things. Yeah. I wrote a long piece about, uh, 0:04:52.330 --> 0:04:55.359 this is I posted it on X, a fairly long piece. 0:04:55.360 --> 0:04:57.470 I should probably turn it into a full blog, but 0:04:57.470 --> 0:05:01.310 it's a little bit, uh, long winded and it's got 0:05:01.310 --> 0:05:02.900 some politics in it, so I think I'm going to 0:05:02.900 --> 0:05:06.770 skip it, but I recommend going and checking it out 0:05:06.770 --> 0:05:09.260 if you're into that kind of stuff. And I spoke 0:05:09.260 --> 0:05:14.180 with Christine Gadsby, the head of product security operations at BlackBerry, 0:05:14.180 --> 0:05:16.970 and we talked about the role of AI in cybersecurity 0:05:16.970 --> 0:05:20.240 and a whole bunch of different topics. The topic list 0:05:20.240 --> 0:05:23.970 for this episode is quite large and you should absolutely 0:05:23.970 --> 0:05:26.640 check it out. Um, so so go check that out 0:05:26.640 --> 0:05:28.679 on the YouTube or you can click it in the 0:05:28.680 --> 0:05:32.609 newsletter as well. So for security, two critical ServiceNow vulnerabilities 0:05:32.610 --> 0:05:36.870 were reported by Asset Note company has reportedly paid a 0:05:36.870 --> 0:05:40.620 new record high 75 million to a ransomware group. And 0:05:40.620 --> 0:05:42.839 that seems like a lot of money. But it's not 0:05:42.839 --> 0:05:44.979 a lot compared to not being able to do business 0:05:44.980 --> 0:05:47.110 at all. So a lot of people kind of beat 0:05:47.110 --> 0:05:50.440 up people for paying ransoms. And it really is kind 0:05:50.440 --> 0:05:52.990 of similar to like, your kid gets stolen. It's like 0:05:52.990 --> 0:05:56.800 all the philosophy goes away when somebody has your kid 0:05:56.800 --> 0:06:00.040 and it's the same as a CEO or whoever. When 0:06:00.040 --> 0:06:01.960 you have the ability to pay some money and get 0:06:01.960 --> 0:06:05.680 business back online, sure, they might just ransom you again. Sure, 0:06:05.680 --> 0:06:08.470 they might do whatever. Sure, it might be bad for 0:06:08.470 --> 0:06:12.830 other people. But when business has stopped, things become quite 0:06:12.830 --> 0:06:15.830 clear to you in terms of what you need to do. 0:06:15.830 --> 0:06:19.460 So I'm not saying people should pay or anything like that. 0:06:19.460 --> 0:06:22.310 I'm not making any judgments. I'm just saying, well, essentially 0:06:22.310 --> 0:06:26.030 that don't make judgments. Try to avoid making judgments because 0:06:26.029 --> 0:06:28.580 it's really hard to be in that position. Digicert is 0:06:28.580 --> 0:06:34.529 revoking 83,000 TLS certificates due to a domain validation bug. 0:06:34.560 --> 0:06:37.830 China is getting around US bans on advanced AI chips 0:06:37.830 --> 0:06:42.900 through smuggling front companies and loopholes, basically finding ways to 0:06:42.900 --> 0:06:45.720 get the chips that are not supposed to be getting. 0:06:45.750 --> 0:06:49.170 Ransomware attacks are rising, with an 18% year on year 0:06:49.170 --> 0:06:54.089 increase reported by Zscaler, and I've always considered ransomware attacks 0:06:54.089 --> 0:06:57.180 to be something that we'd have to invent as a government. 0:06:57.180 --> 0:06:59.290 It would have to be like a government service if 0:06:59.290 --> 0:07:01.720 if it didn't exist in the marketplace, like as a 0:07:01.720 --> 0:07:04.690 way to test for bad security. And maybe you give 0:07:04.690 --> 0:07:07.510 like a fine or something if people keep having the mistake. 0:07:07.510 --> 0:07:10.390 But my intuition was that after a number of years 0:07:10.390 --> 0:07:13.780 that we'd get harder and harder because security would increase. 0:07:13.780 --> 0:07:16.989 So if these attacks are still increasing, I wonder what 0:07:16.990 --> 0:07:20.380 the reason is. Is it because attackers are moving to 0:07:20.380 --> 0:07:23.180 like the more vulnerable targets, or are they just getting 0:07:23.180 --> 0:07:26.150 better at finding the holes or something else? Or all 0:07:26.150 --> 0:07:28.429 of the above? Probably all of the above. But if 0:07:28.430 --> 0:07:32.180 somebody has more insight on why things aren't getting tighter 0:07:32.180 --> 0:07:34.790 or see, that's the trick is it doesn't mean things 0:07:34.790 --> 0:07:37.310 aren't getting harder. Just because the number of attacks are 0:07:37.310 --> 0:07:40.610 going up doesn't mean things aren't getting harder. They might 0:07:40.610 --> 0:07:43.310 just be getting better faster. Got a great analysis here 0:07:43.310 --> 0:07:47.730 of securing secrets in AWS. So the blog post discussing 0:07:47.730 --> 0:07:51.420 creating custom implants for evasion by building them in C 0:07:51.450 --> 0:07:55.770 and this thing details server setup, client functionality, and testing 0:07:55.770 --> 0:07:59.160 against security tools. The average cost of a data breach 0:07:59.160 --> 0:08:04.680 jumped 10% to 4.88 million in 23. China is tightening 0:08:04.680 --> 0:08:09.270 its civilian drone export rules starting September 1st to prevent 0:08:09.270 --> 0:08:13.040 use in military or terrorist activities. Yeah, I'm trying to 0:08:13.040 --> 0:08:17.060 figure out if this is CCP trying to keep it 0:08:17.060 --> 0:08:20.450 their stuff from being used against them, or if they're 0:08:20.450 --> 0:08:25.220 trying to make it easier to sell their products because 0:08:25.220 --> 0:08:29.330 they're playing nice and they're appearing to be good guys. 0:08:29.360 --> 0:08:33.050 AI and tech open AI has started rolling out its 0:08:33.050 --> 0:08:37.820 new ChatGPT voice feature for ChatGPT plus users, and it's 0:08:37.820 --> 0:08:40.189 quite good. It's it's quite a bit different. You can 0:08:40.190 --> 0:08:43.190 basically interrupt it. It sounds a lot more natural. I 0:08:43.190 --> 0:08:45.860 am getting a lot of voice artifacts though, like it'll 0:08:45.860 --> 0:08:49.310 sound like choppy and broken and a lot of weird pauses. 0:08:49.309 --> 0:08:51.350 Not like in a human way, but I think the 0:08:51.350 --> 0:08:54.260 platform might be overwhelmed. Or maybe, I don't know, maybe 0:08:54.260 --> 0:08:56.510 I need to restart the app. Maybe it was buggy. 0:08:56.510 --> 0:09:00.270 Not sure. Lots of AI talk at Blackhat, which, uh, yeah, 0:09:00.270 --> 0:09:02.910 already here and it's already happening. Um, another thing to 0:09:02.910 --> 0:09:06.330 mention about the ChatGPT stuff is, uh, Greg Brockman is 0:09:06.330 --> 0:09:11.340 taking a sabbatical. Uh, John Schulman, I think, is leaving 0:09:11.340 --> 0:09:14.640 the company. Is he the one that went to to anthropic? 0:09:14.640 --> 0:09:17.969 I can't remember. Another leader went to anthropic and another 0:09:17.970 --> 0:09:20.550 one left as well. So the three people left all 0:09:20.550 --> 0:09:24.880 at once. But it's not like a mass exodus all 0:09:24.880 --> 0:09:28.420 to one place they're not mad at. OpenAI seems to 0:09:28.420 --> 0:09:31.450 be fairly benign. Um, but it does look kind of 0:09:31.450 --> 0:09:33.640 weird to have an announcement where three people leave at 0:09:33.640 --> 0:09:36.160 the same time. A funniest joke I saw about this 0:09:36.160 --> 0:09:39.130 was that Sam Altman predicted that soon there would be 0:09:39.130 --> 0:09:43.390 a one person unicorn company, and the joke was, yeah, 0:09:43.390 --> 0:09:45.820 it might be your company. You might be the only 0:09:45.820 --> 0:09:48.680 one left. Um, I thought that was kind of clever. 0:09:48.679 --> 0:09:54.380 California's SB 1047 safe and secure innovation for Frontier Artificial 0:09:54.380 --> 0:09:57.740 Intelligence Models Act. That's a long name. It's looking to 0:09:57.740 --> 0:10:01.970 regulate large AI models by mandating safety features to prevent 0:10:01.970 --> 0:10:08.660 catastrophic incidents. Use risk based AI regulation began on August 1st, 0:10:08.660 --> 0:10:13.140 and it's got staggered deadlines based on low or no 0:10:13.140 --> 0:10:17.160 risk versus high risk and limited risk tiers. So that's 0:10:17.160 --> 0:10:20.370 starting to roll out. And OpenAI has launched the GPT 0:10:20.370 --> 0:10:24.569 four long output model. I've already switched all or at 0:10:24.570 --> 0:10:27.270 least a lot of my stuff. I switched my fabric 0:10:27.270 --> 0:10:31.740 prompt over to this. So it's got 64 output tokens, 0:10:31.740 --> 0:10:36.370 64,000 output tokens, which is 16 times more than the 0:10:36.370 --> 0:10:41.290 previous one, and it's 50% cheaper for most things. And 0:10:41.290 --> 0:10:43.479 a lot of people are saying that the benchmarks, it's 0:10:43.480 --> 0:10:46.420 actually much better than the previous one. So I consider 0:10:46.420 --> 0:10:50.110 it just a straight across upgrade plus being cheaper. So yeah, 0:10:50.110 --> 0:10:54.670 I already made that change. Google's experimental Gemini 1.5 Pro 0:10:54.670 --> 0:10:58.780 has claimed top spot on a bunch of leaderboards, surpassing 0:10:58.780 --> 0:11:03.740 GPT four and, uh, sonnet 35 with a score of 1300. 0:11:03.770 --> 0:11:06.290 I've not used it yet, because every time I try 0:11:06.290 --> 0:11:08.990 to use a Google product, I have to vomit. But 0:11:08.990 --> 0:11:11.810 I am going to try again soon to see, uh, 0:11:11.990 --> 0:11:14.810 if it's usable. Meta says it'll need ten times more 0:11:14.809 --> 0:11:18.260 computing power to train llama four compared to llama three. 0:11:18.290 --> 0:11:22.220 Elliott Management is calling Nvidia a bubble and says AI 0:11:22.220 --> 0:11:25.650 is overhyped. They mark. They argue that the market is 0:11:25.650 --> 0:11:29.880 overly optimistic about AI's potential and Nvidia's role in it. 0:11:29.880 --> 0:11:31.980 I think it's a bubble, but it's a bubble like 0:11:31.980 --> 0:11:36.059 the internet in 1995. In other words, there absolutely will 0:11:36.059 --> 0:11:39.030 be a burst of lots and lots of companies, right? 0:11:39.030 --> 0:11:44.160 Pets.com and companies like that, the AI equivalents, thousands of 0:11:44.160 --> 0:11:47.280 those companies are going to fail. Lots of investors are 0:11:47.280 --> 0:11:49.630 going to be very sad about this, but that's completely 0:11:49.630 --> 0:11:52.930 unrelated to what AI is up is about to do 0:11:52.929 --> 0:11:55.270 to the world. Right? So I think people shouldn't be 0:11:55.270 --> 0:11:57.970 confused about those two things. One happening doesn't mean that 0:11:57.970 --> 0:12:00.490 the other one is not going to happen. Bellingcat has 0:12:00.490 --> 0:12:05.380 put together a guide on identifying explosive ordnance in social 0:12:05.380 --> 0:12:09.850 media imagery. CrowdStrike is facing a massive lawsuit after Blue 0:12:09.850 --> 0:12:13.610 Friday crashed over 8 million computers globally. Intel is laying 0:12:13.610 --> 0:12:16.220 off over 15% of its workforce as part of a 0:12:16.220 --> 0:12:20.059 $10 billion cost reduction plan. Apple just posted a record 0:12:20.059 --> 0:12:25.219 breaking Q3 2024, $86 billion in revenue. And one thing 0:12:25.220 --> 0:12:28.729 that's interesting about this is Berkshire Hathaway just sold a 0:12:28.730 --> 0:12:31.130 whole bunch of stuff, uh, a whole bunch of Apple. 0:12:31.130 --> 0:12:34.370 And they sold it right before this crash happened. The 0:12:34.370 --> 0:12:37.220 crash happened. There was a giant recession that hit the 0:12:37.220 --> 0:12:39.380 United States, and then it went away the next day. 0:12:39.410 --> 0:12:45.050 Today was like a lot of that money came back. But, um. Yeah, strange. 0:12:45.050 --> 0:12:48.740 Who knows? It could happen again tomorrow. But very volatile, 0:12:48.740 --> 0:12:52.130 very emotional sort of time. I feel like in lots 0:12:52.130 --> 0:12:54.440 of different ways, and I feel like the stock market 0:12:54.440 --> 0:12:57.500 is matching that. But, uh, the other thing to mention 0:12:57.500 --> 0:13:02.250 about Apple is that their services money is now almost 0:13:02.250 --> 0:13:05.910 equal to their devices money, which is a huge tipping 0:13:05.910 --> 0:13:09.959 point or a milestone in terms of their growth. Apple 0:13:09.960 --> 0:13:13.199 is ramping up spending to get Apple intelligence ready for 0:13:13.200 --> 0:13:16.830 launch in the fall. I'm already using the beta, and 0:13:16.830 --> 0:13:19.080 it's pretty impressive, even though a lot of the features 0:13:19.080 --> 0:13:22.170 aren't rolled out yet. All right, human news. A lot 0:13:22.170 --> 0:13:25.140 of the world tried to push Huawei out of their infrastructure, 0:13:25.140 --> 0:13:30.250 but they're actually getting more successful, not less. Software company 0:13:30.250 --> 0:13:35.140 increased user engagement by eight times by drastically shortening their emails. 0:13:35.140 --> 0:13:40.059 Netlify fees. Is that it? Yeah. Netlify fees. Initial 150 0:13:40.059 --> 0:13:43.720 word emails had a 1% reply rate, but by cutting 0:13:43.720 --> 0:13:47.110 the text to 37 words, it went to 4%. And 0:13:47.110 --> 0:13:50.660 when they went to 14 words, it went to 8% 0:13:50.660 --> 0:13:54.470 14 words. Last month, Shane Mack offered everyone at his 0:13:54.470 --> 0:13:58.160 company $25,000 to quit and six people took it. Yeah, 0:13:58.160 --> 0:14:01.160 I think this is part of the Alaskan fishing boat 0:14:01.160 --> 0:14:03.740 thing that I wrote a while back. Companies basically want 0:14:03.740 --> 0:14:07.460 fully dedicated murderers is all they want. They want people 0:14:07.460 --> 0:14:11.330 who eat, live, sleep, think and are obsessed with the company. 0:14:11.330 --> 0:14:15.270 That's why they want return to office. That's their way 0:14:15.270 --> 0:14:18.210 of filtering for people who who think of the company 0:14:18.210 --> 0:14:20.700 as a religion. I mean, they can't say that, but 0:14:20.700 --> 0:14:22.229 they can say you have to come to the office 0:14:22.230 --> 0:14:25.500 and that's an automatic filter for it. Right? So this 0:14:25.500 --> 0:14:28.020 is the way that management and managers and the whole 0:14:28.020 --> 0:14:32.490 whole system can basically look for these obsessed people, which 0:14:32.490 --> 0:14:36.420 are likely to be in certain demographics, right? Certain ages, 0:14:36.420 --> 0:14:40.930 you know, certain groups that are awfully likely to look 0:14:40.930 --> 0:14:44.620 kind of similar to each other. Probably young, probably without kids, 0:14:44.620 --> 0:14:48.040 probably male, who are just grind, grind, grind, don't care 0:14:48.040 --> 0:14:51.520 about anything else. Yeah, whatever. Work life balance don't care. 0:14:51.520 --> 0:14:54.040 I just want a code or whatever it is. Right. 0:14:54.040 --> 0:14:57.160 So that's what these companies are looking for more and more. 0:14:57.160 --> 0:14:59.890 And that's why I think and this is just my 0:14:59.890 --> 0:15:03.380 hypothesis here, I don't, you know, we need more data 0:15:03.380 --> 0:15:07.940 for all this. But my pet hypothesis here is that 0:15:07.940 --> 0:15:11.000 this is a factor for all of these layoffs. It's 0:15:11.000 --> 0:15:15.590 like this awakening across all of business that you know 0:15:15.590 --> 0:15:21.170 what I want hardcore crazy people, religious people about this company. 0:15:21.170 --> 0:15:23.660 And I want them to be a-players. And I want 0:15:23.660 --> 0:15:26.720 them to be really good at AI, and they're going 0:15:26.720 --> 0:15:28.950 to help us do even more with AI because they're 0:15:28.950 --> 0:15:31.080 going to bring the AI on and blah, blah, blah. 0:15:31.080 --> 0:15:33.330 So it's like, I'm going to hire a bunch of 0:15:33.330 --> 0:15:36.570 these crazy people, and a team of ten of them 0:15:36.570 --> 0:15:39.030 is going to be like having a team of 1000 0:15:39.030 --> 0:15:43.080 or 2000 people sometime in the future, in the near future. 0:15:43.080 --> 0:15:44.880 Whereas if you get a bunch of people who are 0:15:44.880 --> 0:15:48.270 just straight out of college, they're entitled. They think they 0:15:48.270 --> 0:15:51.210 are owed something even worse. They think that they're about 0:15:51.210 --> 0:15:54.550 to receive training on the job because they don't know 0:15:54.550 --> 0:15:57.010 how to do the job. And it's like, okay, well, 0:15:57.010 --> 0:15:59.470 now train me. Now, teach me how to do this job. 0:15:59.470 --> 0:16:02.500 And all these leaders at these companies are like, I 0:16:02.500 --> 0:16:05.350 do not want you. I don't care what degrees you have. 0:16:05.350 --> 0:16:07.810 If you can't do the job on day one, or 0:16:07.810 --> 0:16:11.440 you can't learn instantly, like just by seeing it once, 0:16:11.440 --> 0:16:13.150 and if you're not obsessed about it and want to 0:16:13.150 --> 0:16:16.150 sleep under your desk, we have no use for you. 0:16:16.160 --> 0:16:20.510 And unfortunately, that's like 80% of the workforce, I'm guessing, right? 0:16:20.510 --> 0:16:24.410 It's like 52, 90% of the workforce, let's call it that. 0:16:24.410 --> 0:16:27.920 And what that means is they are looking for that 10%. 0:16:27.920 --> 0:16:31.400 They're looking for that 5%. They're looking for the A 0:16:31.400 --> 0:16:36.320 players who are dedicated like religious people. And I believe 0:16:36.320 --> 0:16:39.620 this is what we're seeing more than anything now. You 0:16:39.620 --> 0:16:41.910 add AI on top of that. Now you see why 0:16:41.910 --> 0:16:43.920 there's so many layoffs. Now you see why there's so 0:16:43.920 --> 0:16:47.160 many open positions. But nobody's hiring for them because they're 0:16:47.160 --> 0:16:51.060 kind of like fake positions. And this is multiple hypotheses 0:16:51.060 --> 0:16:54.210 all rolled into one. But you get the vibe. This 0:16:54.210 --> 0:16:56.880 is the basic vibe of what I think is happening. 0:16:56.880 --> 0:17:01.980 Journalist Evan Gershkovich was among a group of Americans and 0:17:01.980 --> 0:17:06.550 Russian dissidents released from Russia in a seven nation prisoner swap, 0:17:06.550 --> 0:17:10.389 largest ever since the Cold War. Researchers at the University 0:17:10.390 --> 0:17:13.750 of California, Santa Barbara have developed an AI model called 0:17:13.750 --> 0:17:17.080 shark AI to help prevent shark attacks. The model uses 0:17:17.080 --> 0:17:21.369 drones to detect sharks with greater accuracy than humans. I 0:17:21.369 --> 0:17:24.550 love this, I love this every time I go to Maui, 0:17:24.550 --> 0:17:27.310 I'm stupid and I read the stats about like, shark 0:17:27.310 --> 0:17:29.780 attacks and they're like, oh, actually, right next to you 0:17:29.780 --> 0:17:32.149 is the most dangerous place. And I'm like, cool. I 0:17:32.150 --> 0:17:34.130 didn't want to go in the water anyway. Why did 0:17:34.130 --> 0:17:37.040 I read that right before I went on vacation, where 0:17:37.040 --> 0:17:39.740 I'm supposed to swim in the water? But anyway, if 0:17:39.740 --> 0:17:42.170 I were able to look up and maybe they're so 0:17:42.170 --> 0:17:45.050 high up you can't hear them, maybe it's not super annoying. 0:17:45.050 --> 0:17:47.750 But anyway, if I know that there's ten of these 0:17:47.750 --> 0:17:53.000 drones sweeping back and forth and you know they're being recharged, 0:17:53.000 --> 0:17:55.530 they go back on rotation and they look down. They 0:17:55.530 --> 0:17:58.470 could see very clearly if there's a shark in the water, 0:17:58.470 --> 0:18:01.050 I assume that it wouldn't work. Maybe if the water 0:18:01.050 --> 0:18:04.380 was muddy, but maybe you wouldn't be swimming anyway because 0:18:04.380 --> 0:18:07.380 it would be dangerous water anyway. Usually in a lot 0:18:07.380 --> 0:18:09.659 of places you could see right through the water. It's 0:18:09.660 --> 0:18:13.560 very easy to see a shark from above, and they 0:18:13.560 --> 0:18:18.520 just call the lifeguard station and trigger an alert, and 0:18:18.520 --> 0:18:20.980 they blow the whistle and everyone gets out of the water, like, 0:18:20.980 --> 0:18:24.130 that's going to be amazing. Love it. Treating failing eyesight 0:18:24.130 --> 0:18:26.410 and high cholesterol are two new ways to lower the 0:18:26.410 --> 0:18:30.130 risk of developing dementia, according to a major report. The 0:18:30.130 --> 0:18:35.500 Lancet Commission's latest findings suggest that addressing 14 health issues 0:18:35.500 --> 0:18:39.820 could theoretically prevent nearly half of all dementia cases worldwide. 0:18:39.820 --> 0:18:43.040 And I believe from reading this that essentially they're talking 0:18:43.040 --> 0:18:46.550 about things that just exacerbate it and make it worse. So, 0:18:46.550 --> 0:18:49.640 for example, if you can't really see things, you generally 0:18:49.640 --> 0:18:52.040 maybe you don't go out a lot. If you can't 0:18:52.040 --> 0:18:55.010 hear conversations, you're not involved in conversations. So I think 0:18:55.010 --> 0:18:58.130 a lot of this might be related to social interaction, 0:18:58.130 --> 0:19:01.250 which once you start to get isolated and you're not 0:19:01.250 --> 0:19:04.400 consuming media, you're not reading, you're not like, there's no 0:19:04.400 --> 0:19:07.830 new inputs. Um, again, this is my hypothesis. I believe 0:19:07.830 --> 0:19:10.109 this is based on some solid science I've already read, 0:19:10.109 --> 0:19:13.770 though is basically, once you get isolated in that way, 0:19:13.770 --> 0:19:17.340 your brain starts like shutting down, and it really accelerates 0:19:17.340 --> 0:19:21.180 the dementia. So that that would make sense if, uh, 0:19:21.180 --> 0:19:23.669 that's what they were saying in this paper. Self control 0:19:23.670 --> 0:19:27.990 is about 60% heritable, meaning genes explain roughly 60% of 0:19:27.990 --> 0:19:32.170 the differences in self control among individuals. So I think 0:19:32.170 --> 0:19:34.930 this could be devastating if it's supported in further studies. 0:19:34.930 --> 0:19:39.159 I worry about the narrative that both IQ and self-discipline 0:19:39.160 --> 0:19:43.000 are mostly genetic, thus giving people an easy ramp to 0:19:43.000 --> 0:19:46.629 write off individuals or even groups if they have lower 0:19:46.630 --> 0:19:50.950 averages of these things. And I think even if it 0:19:50.950 --> 0:19:55.400 were true, the groups don't define the individuals and the 0:19:55.400 --> 0:19:58.550 study mentioned individuals here. It's not talking about groups, but 0:19:58.550 --> 0:20:01.159 you know, people are going to people. Right? So the 0:20:01.160 --> 0:20:05.210 the other thing is there's likely a lot of slack in, say, 0:20:05.210 --> 0:20:09.260 the 40%, which is environmental, assuming those numbers are correct, 0:20:09.260 --> 0:20:13.490 like we're probably getting whatever, 10 or 20% of the 40% 0:20:13.490 --> 0:20:17.240 we're supposed to be doing. So if we were to increase, 0:20:17.240 --> 0:20:20.520 you know, the efforts of, you know, training and culture 0:20:20.520 --> 0:20:23.639 and all the environmental things we can control, I think 0:20:23.640 --> 0:20:27.570 that would raise, you know, the bar for, well, everyone, 0:20:27.570 --> 0:20:30.600 but especially the bottom quite a bit. So I'm not 0:20:30.600 --> 0:20:33.570 sure this is really anything too much to despair about 0:20:33.570 --> 0:20:37.980 other than making it easier for certain negative narratives. A 0:20:37.980 --> 0:20:40.620 new study reveals that people tend to alter their appearance 0:20:40.619 --> 0:20:44.899 to match their names. Researchers found that adults faces often 0:20:44.900 --> 0:20:49.400 align with a social stereotype associated with their name, while 0:20:49.400 --> 0:20:52.790 children's faces do not show this pattern. A key protein 0:20:52.790 --> 0:20:56.720 called reelin may help stave off Alzheimer's disease. A number 0:20:56.720 --> 0:20:59.720 of new studies suggest that reelin helps maintain thinking and 0:20:59.720 --> 0:21:04.639 memory in aging brains, and when its levels fall off, 0:21:04.640 --> 0:21:08.689 neurons become more vulnerable and people are starting to obviously 0:21:08.690 --> 0:21:11.780 work on drugs for this. Wizards of the coast will 0:21:11.780 --> 0:21:16.910 release the 2024 Dungeons and Dragons rulebooks under a Creative 0:21:16.910 --> 0:21:21.200 Commons license, which is fulfilling a promise they made after 0:21:21.200 --> 0:21:24.440 the backlash over attempts to change the Open Gaming License. 0:21:24.440 --> 0:21:27.950 If novelists wrote Your bug Reports imagines how famous authors 0:21:27.950 --> 0:21:32.100 would describe software bugs in their unique styles, Ernest Klein 0:21:32.100 --> 0:21:35.100 likens a screen flicker to scenes from back to the 0:21:35.100 --> 0:21:40.590 Future and Ghostbusters, while Ursula K Le Guin philosophizes about 0:21:40.590 --> 0:21:45.540 the existential pain of coding errors. Ideas. More analysis on 0:21:45.540 --> 0:21:48.870 how bad the results were of the recent UBI study 0:21:48.869 --> 0:21:52.950 done by Sam Altman. It looks pretty bad, just like 0:21:52.950 --> 0:21:55.859 we talked about last week, and got a link here 0:21:55.869 --> 0:21:58.629 to go into that in depth. And really cool idea 0:21:58.630 --> 0:22:02.950 from Jonathan Hite about free range kids. And a cool 0:22:02.950 --> 0:22:06.400 idea for giving them freedom is to create a play 0:22:06.400 --> 0:22:08.919 street once a month where you close off a street 0:22:08.920 --> 0:22:12.550 for two hours, give time for kids to play in 0:22:12.550 --> 0:22:15.699 the street safely, and then the whole time that the 0:22:15.700 --> 0:22:20.170 parents are there watching, like around the edges and, you know, whatever. 0:22:20.180 --> 0:22:23.060 But the neighbors are also meeting and talking, and he's 0:22:23.060 --> 0:22:27.320 saying it has transformative effects on the neighborhood and just 0:22:27.320 --> 0:22:30.470 good times all around. I really love ideas like this. 0:22:30.470 --> 0:22:34.100 Discovery Farmbot is an open source farming machine for growing 0:22:34.130 --> 0:22:37.250 food in your own backyard. Super memory an AI powered 0:22:37.250 --> 0:22:41.600 platform to organize, search and utilize saved information acting as 0:22:41.600 --> 0:22:47.190 a digital second brain. Friend is Avi Schiffman's new AI pendant, 0:22:47.190 --> 0:22:51.030 and it's designed to combat loneliness by sending you reassuring 0:22:51.030 --> 0:22:56.310 or playful text based on what it overhears so it 0:22:56.310 --> 0:22:59.730 doesn't have a speaker. It actually sends you notifications. Kind 0:22:59.730 --> 0:23:02.760 of interesting way to do that. Daniel Kosman walks you 0:23:02.760 --> 0:23:07.830 through installing fabric and open source AI framework by Daniel Missler. 0:23:07.859 --> 0:23:10.540 That's weird. I wonder if I wrote that because I 0:23:10.540 --> 0:23:13.060 don't talk about myself in the third person. Fleet is 0:23:13.060 --> 0:23:15.970 an open source version of fleet DMs tool built on 0:23:15.970 --> 0:23:22.270 OS query for vulnerability monitoring, MDM detection engineering and more. 0:23:22.270 --> 0:23:27.460 Soc2 policy Templates collection of templates for Soc2 policies and procedures. 0:23:27.460 --> 0:23:32.020 Clutch security is a platform providing visibility into all non-human 0:23:32.020 --> 0:23:38.610 identities within an organization, helping them identify associated risks and 0:23:38.609 --> 0:23:41.250 the recommendation of the week. If you're at Blackhat this week, 0:23:41.250 --> 0:23:44.160 remember that ten and 20 years from now, you will 0:23:44.160 --> 0:23:47.070 not remember the talks that you saw this year, but 0:23:47.070 --> 0:23:50.970 you will remember spending that time with your friends. So 0:23:50.970 --> 0:23:55.590 prioritize friend time over presentation time. Not only is the 0:23:55.590 --> 0:23:58.620 friend time more precious and valuable, but you can get 0:23:58.619 --> 0:24:00.959 the talks later if you really want to. And the 0:24:00.960 --> 0:24:03.570 aphorism of the week friends show their love in times 0:24:03.570 --> 0:24:07.320 of trouble, not in happiness. Friends show their love in 0:24:07.320 --> 0:24:14.760 times of trouble, not in happiness. Euripides. Unsupervised learning is 0:24:14.760 --> 0:24:17.880 produced and edited by Daniel Miller on a Neumann U87 0:24:17.880 --> 0:24:21.970 AI microphone using Hindenburg. Intro and outro music is by 0:24:21.970 --> 0:24:25.179 Zomby with the Y, and to get the text and 0:24:25.180 --> 0:24:27.550 links from this episode, sign up for the newsletter version 0:24:27.550 --> 0:24:33.190 of the show at Daniel miessler.com/newsletter. We'll see you next time.