WEBVTT - UL NO. 424: Raising Security's Floor 0:00:01.230 --> 0:00:04.470 Welcome to Unsupervised Learning, a security, AI and meaning focused 0:00:04.470 --> 0:00:07.380 podcast that looks at how best to thrive as humans 0:00:07.380 --> 0:00:11.580 in a post AI world. It combines original ideas, analysis, 0:00:11.580 --> 0:00:14.790 and mental models to bring not just the news, but 0:00:14.790 --> 0:00:22.569 why it matters and how to respond. All right. Welcome 0:00:22.570 --> 0:00:25.030 to unsupervised learning. This is Daniel Missler going to do 0:00:25.030 --> 0:00:28.570 an Andrew Huberman Lex Fridman type ad read here. And 0:00:28.570 --> 0:00:31.420 you can of course just jump forward. But like they 0:00:31.420 --> 0:00:34.750 say it would be appreciated if you listen and take 0:00:34.750 --> 0:00:37.059 a look at the sponsor. Definitely helps the show. When 0:00:37.060 --> 0:00:39.430 you go through airport security, there's one line where the 0:00:39.430 --> 0:00:42.250 TSA agent checks your ID and another line where a 0:00:42.250 --> 0:00:45.640 machine scans your bag. Same thing happens in enterprise security, 0:00:45.640 --> 0:00:48.730 but instead of passengers and luggage, it's end users and 0:00:48.729 --> 0:00:51.400 their devices. These days, most companies are pretty good at 0:00:51.400 --> 0:00:54.550 the first part of the equation where they check user identity, 0:00:54.550 --> 0:00:57.850 but user devices can roll right through authentication without getting 0:00:57.850 --> 0:01:02.110 inspected at all. And in fact, 47% of companies allow unmanaged, 0:01:02.110 --> 0:01:05.230 untrusted devices to access their data. That means an employee 0:01:05.230 --> 0:01:07.179 can log in from a laptop that has its own 0:01:07.180 --> 0:01:09.880 firewall turned off, and hasn't been updated in like six 0:01:09.880 --> 0:01:13.180 months or worse. That laptop might actually belong to a 0:01:13.180 --> 0:01:17.440 bad actor. Using employee credentials, collide finally solves the device 0:01:17.440 --> 0:01:20.440 trust problem. Collide ensures that no device can log on 0:01:20.440 --> 0:01:24.460 to your Okta protected apps unless it passes your security checks. Plus, 0:01:24.459 --> 0:01:27.760 you can use Collider and devices without MDM like your 0:01:27.760 --> 0:01:31.720 Linux fleet contractor devices in every BYoD phone and laptop 0:01:31.720 --> 0:01:40.839 in your company, visit collider.com/unsupervised learning that's collide.com/unsupervised learning to 0:01:40.840 --> 0:01:42.850 watch a demo and see how it works. All right 0:01:42.850 --> 0:01:46.810 let's get into it. So some really cool patterns for 0:01:46.810 --> 0:01:50.050 fabric this week. Got one called Create Better frame. So 0:01:50.050 --> 0:01:53.320 basically any place where somebody is putting forth an opinion 0:01:53.320 --> 0:01:56.080 about anything and it basically makes it so that when 0:01:56.080 --> 0:01:59.020 you send that thing in, it looks at the negative 0:01:59.020 --> 0:02:01.240 way to look at that content, and it looks at 0:02:01.240 --> 0:02:03.820 positive ways to look at that content, and that's the 0:02:03.820 --> 0:02:06.310 output that it produces and the negative ways to look 0:02:06.310 --> 0:02:09.940 at the content. It's actually like in escalating levels. So 0:02:09.940 --> 0:02:13.480 if it's about dating for example, it'll be like, oh, 0:02:13.480 --> 0:02:15.910 the dating scene is really bad. And then it kind 0:02:15.910 --> 0:02:19.480 of like catastrophizing that. And the next one is like, 0:02:19.480 --> 0:02:22.120 I'll never find anyone in the last one is like, 0:02:22.120 --> 0:02:25.480 why go on living if I'm always going to be alone? 0:02:25.480 --> 0:02:28.630 So it's like negative frames of thinking that could be 0:02:28.630 --> 0:02:31.960 generated from a given article or something. And what this 0:02:31.960 --> 0:02:35.230 does is it also adds the positive ones at the 0:02:35.230 --> 0:02:39.160 end and says, no, it's more like, okay, dating apps 0:02:39.160 --> 0:02:42.669 aren't very good, but the final one there is, look, 0:02:42.669 --> 0:02:49.540 I already wanted to prioritize in-person meet ups and relationships anyway, 0:02:49.540 --> 0:02:53.680 so this is just support. This article about bad dating 0:02:53.680 --> 0:02:56.169 apps is just support for the fact that I wanted 0:02:56.169 --> 0:02:59.620 more in-person connection anyway, so it's a positive frame for 0:02:59.620 --> 0:03:01.690 a thing. So the idea here is that this can 0:03:01.690 --> 0:03:05.860 function essentially like a positivity filter, which eventually will be 0:03:05.860 --> 0:03:08.320 like a function of days. So that's why I'm super 0:03:08.320 --> 0:03:11.350 excited about this one. Next one is create Academic paper. 0:03:11.350 --> 0:03:14.080 So this one takes any bullet points article, essay or 0:03:14.080 --> 0:03:16.840 anything you've written and turns it into like a LaTeX 0:03:16.840 --> 0:03:20.530 formatted academic paper. And it's pretty cool. It works great. 0:03:20.530 --> 0:03:23.320 And you just paste it into like an online latex 0:03:23.320 --> 0:03:26.890 renderer and you get the results. Okay. Next one here 0:03:26.889 --> 0:03:29.440 is summarize git changes, which is a great way to 0:03:29.440 --> 0:03:33.160 see and share updates on recent progress for a git project. Yeah, 0:03:33.160 --> 0:03:35.530 so really cool. So you basically just paste in the 0:03:35.530 --> 0:03:38.560 front page of a GitHub project and it outputs like 0:03:38.560 --> 0:03:42.070 exactly how to install it or whatever. Oh actually that's 0:03:42.070 --> 0:03:44.740 a different one. That's a different one. That is explain 0:03:44.740 --> 0:03:48.100 project is actually the one I just described. This one 0:03:48.100 --> 0:03:51.640 is different. This one is called summarize get changes and 0:03:51.640 --> 0:03:53.590 this is the output from it. So you basically do 0:03:53.590 --> 0:03:57.700 this git log command. And that pulls out like 500 changes. 0:03:57.700 --> 0:04:01.540 And then it produces this type of a summary of 0:04:01.540 --> 0:04:03.580 the actual project. So it's like here are all the 0:04:03.580 --> 0:04:06.190 features we added like all that kind of stuff. Also 0:04:06.190 --> 0:04:11.050 thresholds first commercial product or Ul's first commercial product called 0:04:11.050 --> 0:04:15.010 threshold is imminent. I'm already using it and making final 0:04:15.010 --> 0:04:17.770 tweaks now, and it's going to launch in preview, which 0:04:17.770 --> 0:04:20.260 means they're going to be like tons of updates in 0:04:20.260 --> 0:04:23.770 like the first couple of weeks and months, but can't 0:04:23.770 --> 0:04:26.260 wait to share this thing. Wrote a piece called Personal 0:04:26.260 --> 0:04:29.710 Eyes will mediate everything, and it's just talking about, you know, 0:04:29.710 --> 0:04:34.180 what happens when Daz, your digital assistants are actually doing 0:04:34.180 --> 0:04:36.580 all the interaction with the products around you. And this 0:04:36.580 --> 0:04:38.560 is kind of like a picture of that. In fact, 0:04:38.560 --> 0:04:40.480 we'll just go and look at that real quick. So 0:04:40.480 --> 0:04:42.970 it's like if you look at this picture, you go 0:04:42.970 --> 0:04:46.300 into a city street or whatever, you're surrounded by thousands 0:04:46.300 --> 0:04:50.620 upon thousands of APIs, demons from the things all around you, 0:04:50.620 --> 0:04:53.049 all the people, all the different buildings, all the all 0:04:53.050 --> 0:04:55.870 the streets and the cars and everything. And, you know, 0:04:55.870 --> 0:04:58.990 like the the driver service or Uber or whatever. And 0:04:58.990 --> 0:05:02.350 all of these are actually like products, and you might 0:05:02.350 --> 0:05:04.570 want to order some food or something. And it has 0:05:04.570 --> 0:05:07.420 a menu and that's a product. And normally you would 0:05:07.420 --> 0:05:09.849 be going there and looking at it and flipping through. 0:05:09.850 --> 0:05:13.810 And as a result, people build applications in a certain way. Right. 0:05:13.810 --> 0:05:17.710 And with AI happening, an API is happening and DHA 0:05:17.710 --> 0:05:19.870 is happening. Humans aren't going to be the ones going 0:05:19.900 --> 0:05:21.930 to the webpage. They're not going to be. Ones looking 0:05:21.930 --> 0:05:24.270 at the catalog, they're going to know our preferences. They're 0:05:24.270 --> 0:05:26.160 going to be doing all the looking. And then what 0:05:26.160 --> 0:05:28.050 they're going to do is they're going to pull all 0:05:28.050 --> 0:05:31.229 this content back via the API, and then they're going 0:05:31.230 --> 0:05:34.740 to use your own preferred way of looking at things 0:05:34.740 --> 0:05:37.320 inside of AR, and that's going to be shown inside 0:05:37.320 --> 0:05:41.370 your glasses or your lenses or whatever the the modality is. 0:05:41.370 --> 0:05:45.300 So you're going to have your own preferred UI interfaces 0:05:45.300 --> 0:05:48.060 that you just like the best. Like it might look cyberpunk, 0:05:48.060 --> 0:05:51.690 it might look solar punk, it might look minimalist, or 0:05:51.690 --> 0:05:55.560 it might look like super pretty and gaudy and extravagant 0:05:55.560 --> 0:05:58.080 or whatever. And that's going to be how you see things. 0:05:58.080 --> 0:06:01.890 But again, like all these are company APIs, right? And 0:06:01.890 --> 0:06:04.979 her Da is now looking at these things and parsing 0:06:04.980 --> 0:06:07.740 all that information for her. So the real question is, 0:06:07.740 --> 0:06:11.219 as a builder of a product like Google or like 0:06:11.220 --> 0:06:13.710 a product catalog or like Amazon, what does it mean 0:06:13.710 --> 0:06:17.159 to your company if nobody's coming to the website because 0:06:17.160 --> 0:06:19.289 the Das are coming to the website and there's going 0:06:19.290 --> 0:06:21.780 to be separate third parties that do the display, and 0:06:21.779 --> 0:06:25.529 that's going to change a whole lot about product development. 0:06:25.529 --> 0:06:29.040 And I don't know, really the economy in general. All right. 0:06:29.040 --> 0:06:32.340 Had a really cool sponsored conversation with Jason Miller of 0:06:32.339 --> 0:06:36.510 One Password. And, uh, he was the CEO at collide 0:06:36.510 --> 0:06:39.360 and is still over that product inside of one password. 0:06:39.360 --> 0:06:42.360 And that is the sponsor slot that we read earlier 0:06:42.360 --> 0:06:46.440 as well. So thanks to Callide for sponsoring and, uh, security. 0:06:46.440 --> 0:06:49.049 So got a set of deepfakes here. Actually I'm going 0:06:49.050 --> 0:06:51.150 to click on these, right. I mean, this is the 0:06:51.150 --> 0:06:53.940 advantage of doing video, right. So I'm going to show 0:06:53.940 --> 0:06:56.640 a couple of these. Um, this one is the most 0:06:56.640 --> 0:06:57.630 scary one to me. 0:06:57.660 --> 0:07:00.599 World from art to music. Today we're diving into how 0:07:00.630 --> 0:07:03.510 AI is transforming the creative world from art to music. 0:07:03.510 --> 0:07:06.510 I'm the deepfake version of Marc Andreessen, made by Argo, 0:07:06.510 --> 0:07:08.940 and I think we need more and more realistic deepfake. 0:07:08.940 --> 0:07:11.730 It will empower new form of content creation. You should 0:07:11.730 --> 0:07:12.510 check it out. 0:07:12.510 --> 0:07:16.470 I mean, that is insane. That is completely insane. So 0:07:16.470 --> 0:07:19.650 what I'm basically saying is we need a Snopes. Okay, 0:07:19.860 --> 0:07:24.090 we're going into 2024 election in the US. Video deepfakes 0:07:24.090 --> 0:07:27.000 that are that good, that could be Biden saying something 0:07:27.000 --> 0:07:32.520 or Trump saying something. Those situations are so bad that 0:07:32.520 --> 0:07:34.170 we need to be able to account for them. So 0:07:34.170 --> 0:07:36.210 this is what I wrote. We we need like a 0:07:36.210 --> 0:07:38.730 global Snopes platform. So you get a bunch of people 0:07:38.730 --> 0:07:42.300 left people, center people, right people, and you build a 0:07:42.300 --> 0:07:45.660 platform that does like Snopes, which is like it tests 0:07:45.660 --> 0:07:48.330 internet claims and basically says, is it real or is 0:07:48.330 --> 0:07:51.180 it crap? And it could be like a collection point 0:07:51.180 --> 0:07:54.120 for arguments, right? So it's something like, okay, there's a 0:07:54.120 --> 0:07:58.470 video of Obama saying we're going to attack Mylanta. The 0:07:58.470 --> 0:08:01.140 video is currently being analyzed, and then you have like 0:08:01.140 --> 0:08:05.250 streaming in. You have oh, it's created by oh, looks 0:08:05.250 --> 0:08:08.100 to be fake. So Fox News says, oh it's fake. Oh. 0:08:08.100 --> 0:08:11.040 And then says there's no evidence that it's fake. Also 0:08:11.040 --> 0:08:13.200 known as they believe it's real or they want you 0:08:13.200 --> 0:08:16.410 to believe it's real. See some analysis. They're like, oh, 0:08:16.410 --> 0:08:18.780 this is definitely a deep fake. And all these are 0:08:18.780 --> 0:08:22.020 links to go actually look at their analysis. Right. And 0:08:22.020 --> 0:08:24.750 Breitbart is like, no, he probably said it or whatever. 0:08:24.750 --> 0:08:28.020 So current conclusion is, given all the evidence, we are 0:08:28.020 --> 0:08:31.500 almost certain and that this is a thing roughly similar 0:08:31.500 --> 0:08:34.590 to what the CIA has used in the past to 0:08:34.590 --> 0:08:38.280 assess likelihood. So it's a good sort of scale to 0:08:38.280 --> 0:08:42.000 use for how certain you are about a certain thing. So, um, 0:08:42.000 --> 0:08:44.520 almost certain that this is a deepfake. So this is 0:08:44.520 --> 0:08:47.460 the conclusion of the site, and you've got all the 0:08:47.460 --> 0:08:50.130 evidence up here of like all the different analyses, they've 0:08:50.130 --> 0:08:53.310 all been collected in one place. So some people are 0:08:53.309 --> 0:08:54.929 just not going to believe it. I mean, they're going 0:08:54.929 --> 0:08:57.390 to believe whatever they want to believe about the video 0:08:57.390 --> 0:09:00.510 because of predetermined bias. But we need to have a 0:09:00.510 --> 0:09:05.939 service in which somebody can see somewhat objective analysis, somewhat 0:09:05.940 --> 0:09:12.870 multi subjective analysis. And then finally a verdict is this 0:09:12.870 --> 0:09:15.450 thing real or not. And it could be like chances 0:09:15.450 --> 0:09:18.990 about even we don't know impossible. We also need to 0:09:18.990 --> 0:09:21.330 put like we just don't know. There's got to be 0:09:21.330 --> 0:09:24.060 one of the options. But we absolutely need the service. 0:09:24.059 --> 0:09:26.309 And Dan Kaminsky used to be fond of saying we 0:09:26.309 --> 0:09:29.100 have the technology right. We can actually build this if 0:09:29.100 --> 0:09:32.219 we want to. And basically the deepfakes are too good 0:09:32.220 --> 0:09:35.640 to not have this. In the election season 2024, they 0:09:35.640 --> 0:09:38.939 were supposedly a data leak of 71 million AT&T customers 0:09:38.940 --> 0:09:41.970 are saying it wasn't them. Someone build a active in 0:09:41.970 --> 0:09:46.229 the middle attack using some Cloudflare worker stuff. Really cool 0:09:46.230 --> 0:09:51.179 bypass of MFA against Microsoft accounts. Leaked documents about a 0:09:51.179 --> 0:09:54.750 hacking group that was going after like tons of foreign governments. 0:09:54.750 --> 0:09:58.260 Just really nasty thing. SpaceX is contracted to build a 0:09:58.260 --> 0:10:02.070 spy satellite network for US intelligence agency. So Elon is 0:10:02.070 --> 0:10:05.250 building spy satellites for the government. That's interesting. I mean, 0:10:05.250 --> 0:10:07.860 who better to actually get satellites in space? I don't 0:10:07.860 --> 0:10:13.020 know of anyone. Rohan Pandit modified lama to to unredacted 0:10:13.020 --> 0:10:18.839 an email from Elon to Ilya. So, uh, yeah. Unredacted content. Interesting. 0:10:18.840 --> 0:10:22.150 Burglars are starting to use Wi-Fi jammers. Oh, actually, this. 0:10:22.150 --> 0:10:26.020 This is our second sponsor. Hardly strictly security. I'm actually 0:10:26.020 --> 0:10:28.420 doing a talk here, so shout out to them in 0:10:28.420 --> 0:10:31.569 that conference. Yeah. Burglars are now using Wi-Fi jammers to 0:10:31.570 --> 0:10:35.770 knock out security cameras, and evidently it's starting to happen more. 0:10:35.770 --> 0:10:38.530 I've got a few cases. I think this was in Chicago. 0:10:38.530 --> 0:10:43.179 Fortinet has disclosed a critical SQL injection flaw. Fortinet has 0:10:43.330 --> 0:10:47.380 had some issues. All right. Technology. Stephen Howe gave Devin 0:10:47.380 --> 0:10:51.100 access to his work stuff, which I'm like, how do 0:10:51.100 --> 0:10:53.920 you do that? That's, uh, that's an agent with full 0:10:53.920 --> 0:10:57.040 access to your account. This is like I've been talking about. 0:10:57.040 --> 0:10:59.530 This is the number one threat from AI, in my opinion, 0:10:59.530 --> 0:11:02.380 is people giving too much power to agents that have 0:11:02.380 --> 0:11:06.550 access to really strong APIs, such as posting on slack 0:11:06.550 --> 0:11:10.000 on behalf of the human right. So, Devin, is this 0:11:10.000 --> 0:11:13.209 really controversial thing. It's basically an agent that writes code. 0:11:13.210 --> 0:11:15.459 There's a million of them out right now. This one 0:11:15.460 --> 0:11:18.640 is particularly well packaged. They had a good launch interview, 0:11:18.640 --> 0:11:21.730 and I think that's why they're getting so much pushback. 0:11:21.730 --> 0:11:23.920 It's like it just launched. Well it got a lot 0:11:23.920 --> 0:11:26.470 of PR. So that's why people are more mad at 0:11:26.470 --> 0:11:28.870 it than other things that they don't know about. That's 0:11:28.870 --> 0:11:33.100 my theory anyway. Um, it can basically do tasks. It 0:11:33.100 --> 0:11:35.770 sets out a thing of tasks it needs to do 0:11:35.770 --> 0:11:39.189 to like write an application or something. Right? Well, if 0:11:39.190 --> 0:11:42.700 it gets stuck, it can go and search many different 0:11:42.700 --> 0:11:45.520 avenues to try to get unstuck. One of the avenues 0:11:45.520 --> 0:11:49.270 that Steven Howe gave it was to go on slack. 0:11:49.270 --> 0:11:51.580 So it actually goes on slack. I'm going to click this, 0:11:52.090 --> 0:11:56.500 look at this, this thing. It goes on slack. And actually, um, 0:11:56.500 --> 0:12:00.130 posts in slack to ask a question that would solve 0:12:00.130 --> 0:12:02.770 its problem. And it got back a response, and it 0:12:02.770 --> 0:12:06.310 used that response to go and continue writing code, got 0:12:06.309 --> 0:12:09.040 unstuck and actually finished the project. And yeah, I was 0:12:09.040 --> 0:12:11.620 talking about this like so much hate towards Steven. I'm 0:12:11.620 --> 0:12:14.950 not sure why there's a million different platforms doing something similar. 0:12:14.950 --> 0:12:19.570 And uh, Midjourney new character reference feature finally makes things 0:12:19.570 --> 0:12:21.880 look the same. So check this out. This is all 0:12:21.880 --> 0:12:24.370 from one prompt. So basically you can have the same 0:12:24.370 --> 0:12:28.090 person in different settings. And uh, that was not possible before. 0:12:28.090 --> 0:12:31.929 So that's really really cool feature now in Midjourney uh 0:12:32.230 --> 0:12:36.699 open source grok. Not really. Um, in my opinion, the weights, 0:12:36.700 --> 0:12:40.059 the data and the full training methodology are used for 0:12:40.059 --> 0:12:44.080 saying that a model is fully open source and grok 0:12:44.080 --> 0:12:46.750 did not get there. They did not release the methodology 0:12:46.750 --> 0:12:50.890 or all of it. At least Co-variance is launching RFM one, 0:12:50.890 --> 0:12:55.630 which is bringing ChatGPT like capabilities to robots. AI is big, 0:12:55.630 --> 0:12:59.650 robots are big, but the biggest is AI in robots. 0:12:59.650 --> 0:13:03.100 This I think might move faster than we think everyone 0:13:03.100 --> 0:13:05.320 is thinking. You know you can't crawl under a house, 0:13:05.320 --> 0:13:07.900 you can't look at the piping from an old house 0:13:07.900 --> 0:13:10.210 and actually figure out how to fix this, like a 0:13:10.210 --> 0:13:13.870 an electrician or a plumber. And if you've seen what 0:13:13.870 --> 0:13:16.210 I can do, you've seen what some of these robot 0:13:16.210 --> 0:13:19.540 demos can do. And then you see how nimble these 0:13:19.540 --> 0:13:22.719 little things can be. And like the dexterity of their 0:13:22.720 --> 0:13:27.160 arms and their hands. Combine that with local models and or, 0:13:27.160 --> 0:13:30.880 you know, pinnacle models that are available via the cloud. 0:13:30.880 --> 0:13:34.390 I'm not sure that this thing can't take pictures or 0:13:34.390 --> 0:13:40.390 stream video of pipe fittings and washers and, you know, 0:13:40.840 --> 0:13:43.300 be able to use wrenches and all this different stuff. 0:13:43.300 --> 0:13:46.000 I'm not saying it's going to be easy. I'm saying 0:13:46.000 --> 0:13:49.780 we should not be so arrogant as to think it's invulnerable, 0:13:49.780 --> 0:13:53.800 like we currently seem to be thinking it is. All right. Um, 0:13:53.800 --> 0:13:57.670 Finland has a giant sand battery to store heat in winter. 0:13:57.880 --> 0:14:01.090 I didn't know this was possible. That one really excites me. 0:14:01.090 --> 0:14:02.710 And the other one that excites me is I'm trying 0:14:02.710 --> 0:14:07.660 to figure out is storing energy in water upstream. I 0:14:07.660 --> 0:14:10.030 can't remember how this works. I think they store the 0:14:10.030 --> 0:14:13.390 water and then they release the water through like a 0:14:13.390 --> 0:14:16.270 turbine that the water moves through as it goes downhill. 0:14:16.270 --> 0:14:20.710 Something about the gravity and potential energy of all that weight, 0:14:20.710 --> 0:14:26.500 I think. Anyway, another way of like generating and storing energy. Yeah. 0:14:26.500 --> 0:14:31.030 Nvidia getting into human robots. Yeah. This thing, it honestly 0:14:31.030 --> 0:14:33.220 reminded me of Black Mirror. I saw a video of 0:14:33.220 --> 0:14:36.430 this thing jumping up like 3 or 4ft high and 0:14:36.430 --> 0:14:39.880 just like instantly just going over this giant obstacle at 0:14:39.880 --> 0:14:43.330 the Nvidia event. And it was terrifying. Hong Kong is 0:14:43.330 --> 0:14:46.630 implementing a new Beijing driven, stringent security law that goes 0:14:46.630 --> 0:14:49.780 after treason and other types. And it's got like life 0:14:49.780 --> 0:14:53.020 in prison, I think maybe even worse, maybe even death. 0:14:53.020 --> 0:14:56.080 I mostly saw life in prison, though. Midjourney is blocking 0:14:56.080 --> 0:14:59.260 AI images of Trump and Biden. US added more jobs 0:14:59.260 --> 0:15:02.770 than usual, but unemployment went slightly up to 3.9. That 0:15:02.770 --> 0:15:04.990 is such a low number, but it trips me out 0:15:04.990 --> 0:15:07.239 that that number does not include people who don't want 0:15:07.240 --> 0:15:10.330 to work. So I mean, if you counted that, I mean, 0:15:10.330 --> 0:15:12.370 it would be a massive number. A really good thread 0:15:12.370 --> 0:15:15.130 here on Hacker News about experienced programmers not being able 0:15:15.130 --> 0:15:19.300 to find jobs, super strict policies at private schools in England, 0:15:19.300 --> 0:15:23.710 rigid routines and discipline. And this is for disadvantaged students 0:15:23.710 --> 0:15:26.080 to help them succeed. And they've had really good results. 0:15:26.110 --> 0:15:29.020 I don't know, I feel like this is probably going 0:15:29.020 --> 0:15:31.960 to be a trend, not just for disadvantaged students, which 0:15:31.960 --> 0:15:34.660 I think it's great for, but for everyone really reminds 0:15:34.660 --> 0:15:36.729 me of, like those man camps that are happening right 0:15:36.730 --> 0:15:40.000 now in like, uh, Appalachia mountains where you go off 0:15:40.000 --> 0:15:42.850 and you do camping and hunting like all these other 0:15:42.850 --> 0:15:46.150 manly things. And it's because people want to build character, right? 0:15:46.150 --> 0:15:49.180 And you can't build character when everything is easy. So 0:15:49.180 --> 0:15:52.239 these schools or these man camps or whatever, it's a 0:15:52.240 --> 0:15:56.020 way of like getting to like this resistance training, this 0:15:56.020 --> 0:16:00.100 calculated suffering that builds character. Uh, women are drifting more 0:16:00.100 --> 0:16:02.770 left and men are not. So there's like a bigger 0:16:02.770 --> 0:16:06.370 gap than ever. Former Boeing whistleblower, the I think I 0:16:06.370 --> 0:16:09.160 don't know if he was the 1 or 1 of them, 0:16:09.160 --> 0:16:13.030 but found dead amid a lawsuit and, uh, some drama 0:16:13.030 --> 0:16:15.820 around that. And Toronto police are now saying, leave your 0:16:15.820 --> 0:16:18.640 car keys at the front door so that when people 0:16:18.640 --> 0:16:21.070 break in, because we're probably not going to be able 0:16:21.070 --> 0:16:23.500 to help you. Basically, when people break in, you want 0:16:23.500 --> 0:16:25.540 to make it easy for them to take your car 0:16:25.540 --> 0:16:27.400 and so they won't come in and hurt you. And 0:16:27.400 --> 0:16:29.920 I think this is like how you get Republicans elected 0:16:29.920 --> 0:16:32.320 because Republicans are going to be like, I got an idea. 0:16:32.320 --> 0:16:35.500 How about we shoot the the attacker? How about we 0:16:35.500 --> 0:16:37.510 don't allow this to happen? How about we have more 0:16:37.510 --> 0:16:41.080 police so we don't just accept break ins. And this 0:16:41.080 --> 0:16:44.680 is how you get like more right leaning people eventually 0:16:44.680 --> 0:16:48.250 getting to to like far right people like we're seeing 0:16:48.250 --> 0:16:53.290 all over like Sweden and, um, you know, Scandinavian countries 0:16:53.290 --> 0:16:56.470 where you have like huge like migrant crime problems. And 0:16:56.470 --> 0:16:59.260 the left is like, no, immigration is fine, which of 0:16:59.260 --> 0:17:01.960 course it is. But it's like when you're not addressing 0:17:01.960 --> 0:17:05.290 the issues, you give the power of the truth to 0:17:05.290 --> 0:17:08.350 the far right. And like liberals cannot do this. And 0:17:08.350 --> 0:17:11.200 if they do, they're going to be consequences. So I 0:17:11.200 --> 0:17:13.389 think we need to, you know, do some sort of 0:17:13.390 --> 0:17:16.840 hybrid of like, you've got to enforce laws strictly and 0:17:16.840 --> 0:17:18.879 then you have to be very liberal in trying to 0:17:18.880 --> 0:17:20.889 make it. So that doesn't happen in the first place. 0:17:20.890 --> 0:17:24.280 So conservative in some ways, liberal in other ways. Uh, 0:17:24.280 --> 0:17:27.790 recent Boing incidents have got a far right conspiracy going 0:17:27.790 --> 0:17:32.439 that diversity is causing intentional failures, and it's part of 0:17:32.440 --> 0:17:36.520 a plot to undermine Western civilization and promote communism. I'm like, 0:17:36.670 --> 0:17:40.180 somebody explain this to me. That seems fringe. Using tap 0:17:40.180 --> 0:17:42.280 water in the neti pot is bad. So this is 0:17:42.280 --> 0:17:44.470 the thing that cleans out your sinuses, which I need 0:17:44.470 --> 0:17:47.650 to do again right now. I did yesterday, but you 0:17:47.650 --> 0:17:51.280 should use sterilized water for this. I use my reverse 0:17:51.280 --> 0:17:55.540 osmosis water from the filter. You do not want nasty stuff. 0:17:55.570 --> 0:17:58.120 You're shooting it up like very close to your brain. 0:17:58.119 --> 0:18:01.450 Like your sinus cavity goes way up inside your your head. 0:18:01.450 --> 0:18:04.570 So you do not want stuff up there that's like 0:18:04.570 --> 0:18:07.929 alive or poisonous or whatever. And if you look at 0:18:07.930 --> 0:18:13.030 Huberman recent episode on water, he says tap water is nasty. 0:18:13.030 --> 0:18:15.130 He's he's like, I don't want to alarm people, but 0:18:15.130 --> 0:18:18.850 it is alarming. It is nasty what I found in 0:18:18.850 --> 0:18:22.240 these tap waters. So don't shoot that up into your brain. 0:18:22.240 --> 0:18:24.730 How about that? This analysis is saying people used to 0:18:24.730 --> 0:18:29.290 consume more calories without gaining as much weight. Skeptical, but 0:18:29.290 --> 0:18:33.490 thought it was worth sharing. Uh, fentanyl poisoning the leading 0:18:33.490 --> 0:18:37.630 cause of death for Americans 18 to 45 fentanyl. What 0:18:37.630 --> 0:18:42.010 is going on? 2000 newspapers have closed since 2004, and 0:18:42.010 --> 0:18:44.649 car washes are popping up because, uh, they make a 0:18:44.650 --> 0:18:52.900 lot of money. All right. Ideas and analysis. Uh. Epiphany. Epiphany. 0:18:52.930 --> 0:18:56.409 What is that? Uh, that, um, that shows you I 0:18:56.410 --> 0:18:59.470 don't use AI to to write the newsletter. I wouldn't 0:18:59.470 --> 0:19:02.350 have made that mistake. Yeah. So last couple of months, 0:19:02.350 --> 0:19:04.119 I'm not going to go into this one. This this 0:19:04.119 --> 0:19:07.750 is a big topic by itself. Basically, I'm learning a 0:19:07.750 --> 0:19:10.689 lot about the power of framing. And it's kind of 0:19:10.690 --> 0:19:13.929 turning into like a, you know, universal like unified theory 0:19:13.930 --> 0:19:15.939 for me for a lot of things. But worth a 0:19:15.940 --> 0:19:19.270 separate discussion. All right. Really interesting back and forth with 0:19:19.270 --> 0:19:23.410 Dino de survie about the cybersecurity flaw and sealing. This 0:19:23.410 --> 0:19:25.060 one is a deep one. I'm not going to go 0:19:25.060 --> 0:19:27.670 into it. You should definitely go check it out. If 0:19:27.670 --> 0:19:32.619 these types of deep philosophical security things are your jam. Basically, 0:19:32.619 --> 0:19:35.379 I think security is so important to innovation in daily 0:19:35.380 --> 0:19:38.530 life in most situations, and that it falls to a 0:19:38.530 --> 0:19:41.830 minimum as a result. And basically we should guard our 0:19:41.830 --> 0:19:45.970 mental health against thinking people are steering us wrong and 0:19:45.970 --> 0:19:49.810 that's why we have low security. But in fact, where 0:19:49.810 --> 0:19:52.900 a lot of security people think it's a horrible situation, 0:19:52.900 --> 0:19:56.320 it's actually not that bad because if it were that bad, 0:19:56.320 --> 0:19:59.350 we would have fixed it. If it were urgent, we 0:19:59.350 --> 0:20:01.540 would know because it would get fixed immediately. And if 0:20:01.540 --> 0:20:06.250 it's not fixed immediately, then it's not urgent. And there's 0:20:06.250 --> 0:20:09.310 just like, I don't know, it's it's my view of 0:20:09.310 --> 0:20:12.800 security at this point. Um. And doesn't mean you can't 0:20:12.800 --> 0:20:15.770 make progress. You absolutely can make progress. It's just that 0:20:15.770 --> 0:20:20.570 progress tends to happen slowly and reactively. And a lot 0:20:20.570 --> 0:20:23.630 of security people are really upset about. They don't realize that. 0:20:23.630 --> 0:20:26.720 So they're very upset because they think something can be 0:20:26.720 --> 0:20:30.199 done when they don't realize that they're hitting the ceiling 0:20:30.200 --> 0:20:34.040 of the difference between what can be done and what 0:20:34.040 --> 0:20:38.240 needs to be done. And security by default is a 0:20:38.240 --> 0:20:41.900 very efficient thing. So I should probably write a piece 0:20:41.900 --> 0:20:43.790 and like name it something like that. I think I 0:20:43.790 --> 0:20:48.980 saw a comment like security efficiency theory set or something. 0:20:48.980 --> 0:20:53.149 So it's like security wants to be as efficient as possible, 0:20:53.150 --> 0:20:56.600 which is why it's easy to pick locks all over 0:20:56.600 --> 0:20:59.840 the country and probably all over the world. It's because 0:20:59.840 --> 0:21:03.620 that is the exact point of security that you need 0:21:03.619 --> 0:21:06.500 to have for it to be somewhat useful, but not 0:21:06.500 --> 0:21:10.159 to be an overspend. And that magical point I'm arguing, 0:21:10.160 --> 0:21:14.270 is usually way lower than what security professionals think it 0:21:14.270 --> 0:21:16.880 should be. They think the bar should be way up here, 0:21:16.880 --> 0:21:20.540 and I think that's what Dino was talking about. Let's 0:21:20.540 --> 0:21:24.110 raise the minimum bar. And my whole point I said 0:21:24.109 --> 0:21:25.910 I wasn't going to go into this, but here we are. 0:21:25.910 --> 0:21:29.389 My whole point was you can't move that bar. That 0:21:29.390 --> 0:21:34.520 is security's efficiency baseline. And it's determined by reality. It's 0:21:34.520 --> 0:21:38.780 not determined by what should be or what security people 0:21:38.780 --> 0:21:41.899 want it to be. And it's related to this framing thing. 0:21:41.900 --> 0:21:45.290 So we'll leave it there. All right. Really cool. You 0:21:45.290 --> 0:21:48.410 all meet up this week. Had a member named John 0:21:48.410 --> 0:21:52.010 just blow our minds about custom keyboards. And I now 0:21:52.010 --> 0:21:54.410 have one on the way. So I'm going to be 0:21:54.410 --> 0:21:59.750 having my tinted, like side keyboard programmable. Really cool. And 0:21:59.750 --> 0:22:03.800 I ordered a kit from Jose and Martinez, so that 0:22:03.800 --> 0:22:07.400 should be fun. And Voyager one is one light day 0:22:07.400 --> 0:22:10.370 away from us and it still it keeps breaking. We're like, oh, 0:22:10.369 --> 0:22:14.690 it's finally dead, you know, salutes, hats off to you. 0:22:14.690 --> 0:22:17.720 And then it'll be like, I woke up. I'm okay, 0:22:17.720 --> 0:22:20.869 I'm okay, I'm okay. And we're like, how are you? Okay. 0:22:20.869 --> 0:22:23.840 It's like, what? I figured it out. We're okay. And 0:22:23.840 --> 0:22:27.530 they're talking about having to troubleshoot this thing from 24 0:22:27.530 --> 0:22:30.980 light hours away. It roughly, it might be closer or 0:22:30.980 --> 0:22:34.669 further at this point, but roughly a day. How far 0:22:34.670 --> 0:22:37.070 away does the thing have to be for the speed 0:22:37.070 --> 0:22:40.070 of light, which is that fast? Eight minutes, by the way, 0:22:40.070 --> 0:22:43.520 to the sun, 45 minutes for Jupiter, and this thing 0:22:43.520 --> 0:22:47.570 is 24 hours away and we have to troubleshoot it. 0:22:47.570 --> 0:22:50.630 So imagine like a vim session and you're typing keys. 0:22:50.630 --> 0:22:52.310 And by the way, that's why a lot of the 0:22:52.310 --> 0:22:54.740 commands are the way they are. I think Bill Joy 0:22:54.740 --> 0:22:58.189 created VI and it was for interacting with very distant 0:22:58.190 --> 0:23:01.129 things I think satellites. So you'd have this slight delay 0:23:01.130 --> 0:23:05.360 and low resources on each computer. Well, imagine when a keystroke. 0:23:05.359 --> 0:23:08.780 I imagine they batch things, but imagine a keystroke taking 0:23:08.780 --> 0:23:11.869 a day just to get there and a day to 0:23:11.869 --> 0:23:15.230 get back. And final thought on on Voyager one. What 0:23:15.230 --> 0:23:18.530 a hero. All right, discovery hack trails Golang client allows 0:23:18.530 --> 0:23:22.160 you to query security trails, API and do super useful 0:23:22.160 --> 0:23:25.939 stuff for bounty hunters. And it's made by my buddy Luke. Uh, 0:23:25.940 --> 0:23:29.840 super cool guy. Open API Tui lets you interact with 0:23:29.840 --> 0:23:34.280 APIs with a terminal user interface, which I love. I 0:23:34.280 --> 0:23:37.489 stopped loving Captain Kirk. Really good piece. Solar punk is 0:23:37.490 --> 0:23:40.609 a new cyberpunk. Do It Now by Steve Pavlovna, one 0:23:40.609 --> 0:23:43.879 of the first pieces that I read on productivity. Uh, 0:23:43.880 --> 0:23:48.830 2005 really takes me back that one minimal viable system. 0:23:48.830 --> 0:23:51.859 Ben Koon shares why and how to blog which skills 0:23:51.859 --> 0:23:54.920 are at least likely to be replaced by AI. Amanda 0:23:54.920 --> 0:23:58.340 Haskell talks about why cloud three is so good. By 0:23:58.340 --> 0:24:02.240 the way, cloud three opus I'm only using opus. It 0:24:02.240 --> 0:24:06.320 is crushing GPT four in the most advanced tasks that 0:24:06.320 --> 0:24:09.050 I have, which is like deep analysis of like the 0:24:09.050 --> 0:24:12.440 most interesting and surprising ideas. So it's like the pinnacle. 0:24:12.440 --> 0:24:15.139 It's like the top end of the most difficult things 0:24:15.140 --> 0:24:18.619 for humans and AI to do, to pull out complex 0:24:18.619 --> 0:24:24.139 ideas and distill them from a corpus, an entire book 0:24:24.140 --> 0:24:26.570 or an entire video or whatever. To be able to 0:24:26.570 --> 0:24:29.389 do that is the hardest thing that I've seen I 0:24:29.390 --> 0:24:34.010 do anywhere. And, um, at that type of skill, opus 0:24:34.040 --> 0:24:37.220 is crushing GPT four. I mean, it's like two or 0:24:37.220 --> 0:24:40.520 3 or 4 times as good in. It's hard to 0:24:40.520 --> 0:24:44.210 measure that. So I'm making up those numbers. But intuitively, 0:24:44.210 --> 0:24:47.030 I feel like the quality of GPT four is somewhere, 0:24:47.030 --> 0:24:49.909 and this thing is like so much better now in 0:24:49.910 --> 0:24:54.110 most other tasks. Uh, GPT four is as good. Um, 0:24:54.109 --> 0:24:57.980 in some places it's actually still better than opus, but 0:24:57.980 --> 0:25:00.410 and it's way faster. It's probably like three times faster, 0:25:00.410 --> 0:25:02.780 I would guess. Yeah. Really interesting to see opus doing 0:25:02.780 --> 0:25:06.170 so well. Spreadsheets as simulation tools. And the Getty has 0:25:06.170 --> 0:25:10.490 released 88,000 art images for anyone to use. Recommendation of 0:25:10.490 --> 0:25:14.720 the week. Let grow this thing. Oh this project. I 0:25:14.720 --> 0:25:21.870 love this project. Can I click this? It's about resilience. 0:25:21.900 --> 0:25:27.900 Independence is the key to developing happy, well-adjusted children. 0:25:28.920 --> 0:25:31.710 Absolutely love this thing. It is a four minute video. 0:25:31.740 --> 0:25:34.050 You've got to watch this thing. And I heard about 0:25:34.050 --> 0:25:38.550 it from Jonathan Haidt. And reduce anxiety in elementary and 0:25:38.550 --> 0:25:42.900 middle school is the let grow experience. It's all about teaching, 0:25:42.900 --> 0:25:48.840 you know, self not so much discipline, independence, independence and resilience. 0:25:48.840 --> 0:25:53.790 Trying falling down, getting up that sort of vibe. Absolutely. 0:25:53.790 --> 0:25:57.300 Love this project. Okay. And the aphorism for the week 0:25:57.600 --> 0:26:00.000 I am not what happened to me, I am what 0:26:00.000 --> 0:26:02.699 I choose to become, I am not what happened to me, 0:26:02.700 --> 0:26:07.680 I am what I choose to become. Carl Jung Unsupervised 0:26:07.680 --> 0:26:10.199 Learning is produced and edited by Daniel Missler on a 0:26:10.200 --> 0:26:15.060 Norman 87 I microphone using Hindenburg intro and outro. Music 0:26:15.060 --> 0:26:18.149 is by zombie with the Y and to get the 0:26:18.150 --> 0:26:20.220 text and links from this episode, sign up for the 0:26:20.220 --> 0:26:25.869 newsletter version of the show at Daniel missler.com/newsletter. We'll see 0:26:25.869 --> 0:26:26.500 you next time.