WEBVTT - UL NO. 425: The Efficient Security Principle 0:00:01.260 --> 0:00:04.530 Welcome to Unsupervised Learning, a security, AI and meaning focused 0:00:04.530 --> 0:00:07.410 podcast that looks at how best to thrive as humans 0:00:07.410 --> 0:00:11.610 in a post AI world. It combines original ideas, analysis, 0:00:11.610 --> 0:00:14.850 and mental models to bring not just the news, but 0:00:14.850 --> 0:00:21.870 why it matters and how to respond. All right. Welcome 0:00:21.870 --> 0:00:25.919 to unsupervised learning. This is Daniel Missler. All right. Number 425. 0:00:25.920 --> 0:00:28.620 So you all know the deal here. Tons going on 0:00:28.620 --> 0:00:31.500 this week prepping for some paid talks. Going to be 0:00:31.530 --> 0:00:34.500 traveling to a few different states. Honestly really excited to 0:00:34.500 --> 0:00:37.290 be alive right now at this moment in human history, 0:00:37.290 --> 0:00:40.200 like all the different opportunities there are that are just 0:00:40.200 --> 0:00:43.200 popping off and I feel like there's always a conference, 0:00:43.200 --> 0:00:45.659 there's always people. There's so many people building right now, 0:00:45.659 --> 0:00:48.210 especially here in the Bay area. And I'm just excited 0:00:48.210 --> 0:00:51.330 about the optimism. Like the optimism has its own gravity, 0:00:51.330 --> 0:00:54.510 if that makes sense. And yeah hack build grind appreciate. 0:00:54.510 --> 0:00:57.360 That's pretty much the current vibe. So I wrote this 0:00:57.360 --> 0:01:00.240 piece called Efficient Security Principle. And I'm going to open 0:01:00.240 --> 0:01:03.600 this up in a new tab. And it's essentially explaining 0:01:03.600 --> 0:01:06.990 why a lot of security people are so upset, I think, 0:01:06.990 --> 0:01:09.030 over the course of their career. So one of the 0:01:09.030 --> 0:01:11.580 hardest things about being in security is just the frustration. 0:01:11.580 --> 0:01:15.930 And this came from the conversation with Deno, where he 0:01:15.930 --> 0:01:20.010 basically posted about raising the security minimum. And I was like, well, 0:01:20.010 --> 0:01:22.200 you know, the minimum is there for a reason. And 0:01:22.200 --> 0:01:25.290 I kind of wanted to capture that into a principle. 0:01:25.290 --> 0:01:28.110 And ideally I would have this be one item, but 0:01:28.110 --> 0:01:29.820 it was getting kind of long. So I broke it 0:01:29.819 --> 0:01:34.350 into these statements. Security baseline of an offering continuous downward 0:01:34.350 --> 0:01:39.209 pressure from customer excitement or reliance on a given thing. 0:01:39.209 --> 0:01:42.330 And then the baseline will be set at the point 0:01:42.330 --> 0:01:45.690 at which people will stop using it if it's any worse. 0:01:45.690 --> 0:01:47.490 But it can't go higher than that. And then the 0:01:47.490 --> 0:01:49.230 third piece of it, which these are all kind of 0:01:49.230 --> 0:01:51.780 saying the same thing, is that the better the offering is, 0:01:51.780 --> 0:01:54.900 or the more important it is, the lower the security 0:01:54.900 --> 0:01:59.340 baseline can go without losing people. And that's really, really important. 0:01:59.340 --> 0:02:02.460 So all this together basically means that the more essential 0:02:02.460 --> 0:02:04.590 a thing is, and the more awesome it is, the 0:02:04.590 --> 0:02:08.070 more people will put up with. And as we'll see 0:02:08.070 --> 0:02:10.410 when we move through this, it's the reason you shouldn't 0:02:10.410 --> 0:02:13.290 get too upset when the baseline is low. So the 0:02:13.290 --> 0:02:15.570 way we know something has the right amount of security 0:02:15.570 --> 0:02:19.500 is people just keep using it despite how insecure it is. Right? 0:02:19.500 --> 0:02:23.429 So online companies constantly getting hacked. My friend Sasha Geller 0:02:23.430 --> 0:02:26.130 basically talked about this one where he said that basically 0:02:26.130 --> 0:02:28.800 the biggest example of this is actually email, where it's 0:02:28.800 --> 0:02:32.100 like email gets everyone hacked. It's like the number one 0:02:32.100 --> 0:02:34.170 way to get hacked is to have email. But guess what? 0:02:34.169 --> 0:02:36.450 You need email. There's a reliance on it. It's like 0:02:36.450 --> 0:02:39.510 our national you know, our global infrastructure of it is 0:02:39.510 --> 0:02:43.380 basically emails to online banking. Everyone's getting constant fraud all 0:02:43.380 --> 0:02:46.230 the time. Door locks and the internet in general is 0:02:46.230 --> 0:02:49.050 pretty nasty. It's a cesspool, right? We use these things 0:02:49.050 --> 0:02:53.160 anyway because the value they provide massively outweighs the security risks, 0:02:53.160 --> 0:02:55.769 at least according to us in our own minds. And 0:02:55.770 --> 0:02:58.620 the moment people stop using something due to security being 0:02:58.620 --> 0:03:02.040 too bad, the baseline goes up. And Sasha had another 0:03:02.040 --> 0:03:04.680 good point in that. SMS is actually an example of 0:03:04.680 --> 0:03:07.770 this where it was deemed too nasty. And then we 0:03:07.770 --> 0:03:10.680 started moving to like app based tufa. So the way 0:03:10.680 --> 0:03:14.100 to actually use this is if you're a security tech professional, 0:03:14.100 --> 0:03:17.430 like down in the weeds, realize that it's not that 0:03:17.430 --> 0:03:20.250 level that matters at a bigger scale. So you want 0:03:20.250 --> 0:03:23.910 to like uplevel your thinking and realize it's not about us. 0:03:23.910 --> 0:03:26.760 It's not about the technical part of it. It's about 0:03:26.760 --> 0:03:29.910 how important is the offering. And that's what actually sets 0:03:29.910 --> 0:03:33.600 the baseline. And most importantly, do not take it personally. 0:03:33.600 --> 0:03:36.510 Don't take it personally that the baseline is low and 0:03:36.510 --> 0:03:38.730 you're showing them how to raise the baseline, but they're 0:03:38.730 --> 0:03:41.280 not doing anything about it, right. That's really key. And 0:03:41.280 --> 0:03:43.680 if you're a security leader, you know, you can have 0:03:43.680 --> 0:03:46.680 the same sort of problem and frustration of a technical person. 0:03:46.680 --> 0:03:49.470 And maybe you're also a technical leader. It doesn't really matter. 0:03:49.470 --> 0:03:51.780 But the point is like you can get frustrated as well. 0:03:51.780 --> 0:03:54.990 And the recommendation is first, make sure the security baseline 0:03:54.990 --> 0:03:57.960 is actually where the customer thinks it is. Maybe they're 0:03:57.960 --> 0:04:00.030 not complaining because they don't really know how bad it is. 0:04:00.030 --> 0:04:02.970 So once those are in alignment now, now you have 0:04:02.970 --> 0:04:05.970 sort of the same situation as the technical side where 0:04:05.970 --> 0:04:10.230 it's like look fine, innovative ways to raise the baseline 0:04:10.230 --> 0:04:14.280 without making it obvious or cause friction because people might 0:04:14.280 --> 0:04:16.350 want to higher baseline, but only if it doesn't cost 0:04:16.350 --> 0:04:18.810 them anything because they're only going to go as high 0:04:18.810 --> 0:04:22.950 as it can go without sacrificing that experience of this 0:04:22.950 --> 0:04:25.530 thing that they love. And I just think it's really powerful. 0:04:25.529 --> 0:04:28.050 So yeah, only as good as it needs to be 0:04:28.050 --> 0:04:30.720 to keep people from abandoning the service. And the more 0:04:30.720 --> 0:04:34.020 popular or essential the offering, the lower the security can be. 0:04:34.020 --> 0:04:37.050 This is in the summary here. And progress is still possible, 0:04:37.050 --> 0:04:40.440 especially with policy change and regulation, because that kind of 0:04:40.440 --> 0:04:44.580 like forces a baseline raise. But security experts loudly calling 0:04:44.580 --> 0:04:48.000 out how low the baseline is and gesturing wildly towards 0:04:48.000 --> 0:04:52.200 the solution seldom results in change. And most importantly, this 0:04:52.200 --> 0:04:55.410 is kind of the big takeaway is experts struggling with 0:04:55.410 --> 0:04:59.910 low security baselines should absorb the truth above and not 0:04:59.910 --> 0:05:03.540 let their job satisfaction and their mental health suffer. And 0:05:03.540 --> 0:05:05.550 got a bunch of notes here. But that's that's basically 0:05:05.550 --> 0:05:07.589 that piece. All right. A couple of really cool patterns 0:05:07.589 --> 0:05:10.140 here put into fabric this week. So the first one 0:05:10.140 --> 0:05:12.990 is like it'll analyze your prose according to my favorite 0:05:12.990 --> 0:05:15.840 writing book, which is The Sense of Style by Steven Pinker. 0:05:15.839 --> 0:05:18.600 And here's what it actually looks like after a run. 0:05:18.600 --> 0:05:22.529 This is. The analysis of me actually doing that, that essay. 0:05:22.529 --> 0:05:24.480 So I put that essay through here and this is 0:05:24.480 --> 0:05:28.080 the result I came back. So style analysis it's multiple 0:05:28.080 --> 0:05:32.520 different styles put together positive assessment. Critical assessment I talk 0:05:32.520 --> 0:05:35.160 a little bit too much about the writing itself I 0:05:35.160 --> 0:05:38.279 guess is what this is saying. It gives me examples 0:05:38.279 --> 0:05:40.859 of good writing from it. Things I can improve. No 0:05:40.860 --> 0:05:42.900 spelling mistakes, which is cool. I'm going to be using 0:05:42.900 --> 0:05:45.270 this to catch spelling from now on, and you start 0:05:45.270 --> 0:05:48.180 off with a 100 score. Oh, and it gives recommendations 0:05:48.180 --> 0:05:50.160 as well. But it starts off with 100 score and 0:05:50.160 --> 0:05:52.710 then deducts. So I got an 85. I got an 0:05:52.710 --> 0:05:56.070 85 here because it took 15 points off for these things. 0:05:56.070 --> 0:05:59.190 So that's what that pattern does. And then this one 0:05:59.190 --> 0:06:03.390 is extract book recommendations. So this is insane. This is 0:06:03.390 --> 0:06:06.960 essentially any book you could think of that was probably 0:06:06.960 --> 0:06:10.290 brought in the training of a given model. And it's 0:06:10.290 --> 0:06:11.910 most that you could think of like a New York 0:06:11.910 --> 0:06:14.520 Times bestseller. Any big name book that you can think 0:06:14.520 --> 0:06:16.529 of is probably going to be in here. So open 0:06:16.529 --> 0:06:18.960 this one up and look at it. So this is 0:06:18.960 --> 0:06:22.800 a man search for Meaning by Viktor Viktor Frankl. And 0:06:22.800 --> 0:06:25.919 these are the recommendations that it gave me for what 0:06:25.920 --> 0:06:29.310 to do right to like improve your life. I mean 0:06:29.310 --> 0:06:33.029 look at these, these these are absolutely insane. Find meaning 0:06:33.029 --> 0:06:36.360 in life by dedicating yourself to a cause greater than yourself. 0:06:36.360 --> 0:06:40.380 Like even the prioritization here is really, really good. Take 0:06:40.380 --> 0:06:44.040 responsibility for your own life and choices. I mean, this 0:06:44.040 --> 0:06:48.600 is gorgeous and this all just came from running. I 0:06:48.600 --> 0:06:52.740 literally said Echo Man's search for meaning pipe into extract 0:06:52.740 --> 0:06:56.159 book recommendation and this is what it produced, right? This 0:06:56.160 --> 0:06:59.099 is so powerful. I'm just blown away every time we 0:06:59.100 --> 0:07:01.170 make one of these things. So those are two of 0:07:01.170 --> 0:07:03.000 the biggest ones that I put in. I actually put 0:07:03.000 --> 0:07:05.039 in more, but those are two of the big ones. 0:07:05.040 --> 0:07:07.860 So security stuff. Yeah. It looks like you can extract 0:07:07.860 --> 0:07:10.860 keys from M-series chips. I'm sure they'll fix that in 0:07:11.340 --> 0:07:13.590 the next iteration. After that, a lot of people are 0:07:13.590 --> 0:07:16.470 saying it's not really packable and that it's kind of 0:07:16.470 --> 0:07:19.320 a big problem. So and that if you do do 0:07:19.320 --> 0:07:22.860 a workaround, it might affect performance. But we'll see. And 0:07:22.860 --> 0:07:24.570 here's the other piece. We'll have to wait and see 0:07:24.570 --> 0:07:27.179 how real world these attacks are to know if it's 0:07:27.180 --> 0:07:30.870 like actually bad and the integration of drones with digitized 0:07:30.870 --> 0:07:34.290 command and control systems. Yeah, this is a US war 0:07:34.290 --> 0:07:39.720 fighting effort essentially, and they're calling it transformative trinity for autonomous, 0:07:39.720 --> 0:07:43.920 like very dangerous drones, which saved dystopian movie writers a 0:07:43.920 --> 0:07:46.650 lot of time. So that's that's good news. And yeah, 0:07:46.680 --> 0:07:48.690 along the lines of this, there's never been a better 0:07:48.690 --> 0:07:51.720 time to read Daniel Suarez's kill decision, which is all 0:07:51.720 --> 0:07:55.110 about autonomous drones. Actually, you should check out that book, 0:07:55.110 --> 0:07:58.200 a really, really good book. And Jess has a new 0:07:58.200 --> 0:08:02.100 strategy for tackling AI risks. Got the PDF here. Someone 0:08:02.100 --> 0:08:04.170 in China got arrested in New York for trying to 0:08:04.170 --> 0:08:08.520 sell Tesla's secret battery tech to undercover agents. I'm not 0:08:08.520 --> 0:08:10.680 sure if he thought he was selling it to China. 0:08:10.680 --> 0:08:13.170 I couldn't quite get that from the story. Canada is 0:08:13.170 --> 0:08:16.380 rethinking its ban of flipper zero, trying to go after 0:08:16.380 --> 0:08:19.470 misuse instead of just banning it outright, which I'm glad 0:08:19.470 --> 0:08:23.310 they listened to the feedback there. Avonte critical bug. This 0:08:23.310 --> 0:08:27.330 just continues to happen. Earth crushing Kraken. Yeah, I'm not 0:08:27.330 --> 0:08:30.150 sure how to pronounce that one, but basically a Abt 0:08:30.150 --> 0:08:35.010 group backed by the Chinese government hit over 70 organizations, 0:08:35.010 --> 0:08:40.140 mostly focused on governments, and Atlassian patched a critical SQL vulnerability. 0:08:40.140 --> 0:08:42.990 So myself and a bunch of other AI builders have 0:08:42.990 --> 0:08:48.990 found something really weird where anthropic smallest model haiku has 0:08:48.990 --> 0:08:53.130 been performing extremely well. And this one I did. What 0:08:53.130 --> 0:08:55.770 is this? So this is extract ideas. So I did 0:08:55.770 --> 0:09:00.600 extract ideas on a recent YouTube video of Jonathan Haidt 0:09:00.600 --> 0:09:03.270 coming on Rogan again. And if you look at these 0:09:03.270 --> 0:09:06.660 results here, they are remarkably similar, similar to each other. 0:09:06.660 --> 0:09:09.480 So like GPT for the first one is opus, which 0:09:09.480 --> 0:09:12.060 is like the best anthropic model. The second one is 0:09:12.059 --> 0:09:15.000 GPT four, which is the best OpenAI model. But this 0:09:15.000 --> 0:09:19.050 is the worst. It's the third level down of anthropic models, 0:09:19.050 --> 0:09:23.730 and these top level like quotes here really good. Not 0:09:23.730 --> 0:09:27.570 not quotes, but findings, ideas extracted. I mean, they're not 0:09:27.570 --> 0:09:30.030 obviously worse than any of the other two. And this 0:09:30.030 --> 0:09:32.040 has happened to me multiple times. So I'm really impressed 0:09:32.040 --> 0:09:34.350 with that. And a lot of people are reporting GPT 0:09:34.380 --> 0:09:36.990 five will be out this summer. And they're pointing to 0:09:36.990 --> 0:09:40.530 this interview that Sam Altman did with Lex. And but 0:09:40.530 --> 0:09:44.459 what he actually said was something good would come out soon, 0:09:44.460 --> 0:09:46.530 maybe like this summer. I think he might have said, 0:09:46.530 --> 0:09:48.870 like this summer. And some people are saying, well, maybe 0:09:48.870 --> 0:09:51.720 it'll be like a 4.5 release, but I don't think 0:09:51.720 --> 0:09:54.210 it's going to be five before the end of the year. 0:09:54.210 --> 0:09:59.969 That's my guess. Nvidia is patterning. Patterning. See, now you know, 0:09:59.970 --> 0:10:02.220 I don't use AI to write my stuff. I wrote 0:10:02.220 --> 0:10:06.720 this and that is misspelled. That's unfortunate. Yeah. Partnering with 0:10:06.720 --> 0:10:11.429 Hippocratic AI to introduce AI nurses for virtual patient care. 0:10:11.429 --> 0:10:13.709 So you basically looking at an iPad and it's like 0:10:13.710 --> 0:10:16.800 a face telling you what to do $9 an hour. 0:10:16.800 --> 0:10:20.140 It's actually an AI. Yeah. So there's a little point 0:10:20.140 --> 0:10:22.030 here most of the benefit will get from AI in 0:10:22.030 --> 0:10:25.719 the first few years. Won't be from doing work better 0:10:25.720 --> 0:10:29.229 than good humans, right? It's going to be doing work 0:10:29.230 --> 0:10:31.809 that otherwise would not have been done at all. So 0:10:31.809 --> 0:10:34.570 the great examples of this are therapists. We need way 0:10:34.570 --> 0:10:39.040 more therapists. We need billions of more therapists, like hundreds 0:10:39.040 --> 0:10:42.160 of millions or billions of therapists, potentially. We don't want 0:10:42.160 --> 0:10:45.760 to overindex on therapy, but we need so many therapists. 0:10:45.760 --> 0:10:48.939 We need billions of tutors, continuous tutors, that you can 0:10:48.940 --> 0:10:52.780 ask anything. Why? Well, because of this. Why? Because of this? Why? 0:10:52.780 --> 0:10:54.790 How many parents get tired of that question? How about 0:10:54.790 --> 0:10:57.250 somebody who doesn't get tired of that question? Right? And 0:10:57.250 --> 0:10:59.740 I'm not saying parents shouldn't be there to answer those questions. 0:10:59.740 --> 0:11:03.130 Of course they should. And so should teachers. But it's 0:11:03.130 --> 0:11:06.850 nice when somebody is shy or introverted or they're just 0:11:06.850 --> 0:11:11.469 alone and they have questions, you know, therapists, tutors and yeah, 0:11:11.470 --> 0:11:14.350 we don't have enough things watching for asteroids in the sky. 0:11:14.350 --> 0:11:18.939 It's extra eyes. Think of trillions of extra eyes and 0:11:18.940 --> 0:11:21.730 a little bit of brainpower tied to that skin cancer screening. 0:11:21.730 --> 0:11:24.340 How many people in the world have access to a 0:11:24.340 --> 0:11:27.370 doctor to say, hey, look at this mole? Very few, 0:11:27.370 --> 0:11:30.670 very few percentage of the entire planet have that ability. 0:11:30.670 --> 0:11:33.250 But now you can with AI, right? So that's what 0:11:33.250 --> 0:11:37.480 I'm excited about. US Department of Justice 16 state and 0:11:37.480 --> 0:11:42.520 district attorneys, district attorney generals. Actually, I think that's supposed 0:11:42.520 --> 0:11:47.559 to be pluralized attorneys general. Yeah. Lawsuit against Apple. Bottom line. 0:11:47.559 --> 0:11:52.449 And I've got a pro Apple biased, very fanboy leaning. Biased. 0:11:52.750 --> 0:11:55.929 I'm all about not avoiding bias, but actually just calling 0:11:55.929 --> 0:11:57.910 it out and naming it, which I think I saw 0:11:57.910 --> 0:12:01.360 Destiny talk about, which I thought was pretty smart. And yeah, 0:12:01.390 --> 0:12:04.750 Apple's now making iPhones in Brazil. And Tim Cook also 0:12:04.750 --> 0:12:07.150 just went to China to tell him everything's okay. And 0:12:07.150 --> 0:12:09.370 don't worry. Don't worry about the fact that we're trying 0:12:09.370 --> 0:12:11.920 to make everything not in China because you guys are crazy. 0:12:11.920 --> 0:12:14.650 He probably didn't say that, but I did. Apple and 0:12:14.650 --> 0:12:18.250 Tesla are losing market share in China. Meanwhile, as national 0:12:18.250 --> 0:12:21.939 loyalty rises, and my sort of spicy take here is 0:12:22.059 --> 0:12:26.320 one of China's greatest strengths is actually its nationalism. Right. 0:12:26.320 --> 0:12:30.160 They have too much I would say. Um, and I 0:12:30.160 --> 0:12:32.710 would say that the left in the US has too little. 0:12:32.710 --> 0:12:34.900 I would say like the far right in the US 0:12:34.900 --> 0:12:37.929 has too much. Can we can we have some medium, 0:12:37.929 --> 0:12:41.320 can we have some, uh, some moderation, please? That would 0:12:41.320 --> 0:12:44.830 be nice. Israel's government is reportedly running covert ops at 0:12:45.190 --> 0:12:48.790 US universities. This turns out to be a little bit 0:12:48.790 --> 0:12:51.550 kind of like a biased story. But that's why I 0:12:51.550 --> 0:12:53.950 put this in here. It's like, what's the difference between 0:12:53.950 --> 0:12:59.470 marketing and counter propaganda information operations? The difference is your perspective. 0:12:59.470 --> 0:13:02.620 That's that's the difference. Like, we should consider ourselves always 0:13:02.620 --> 0:13:06.160 being hit with propaganda and marketing or whatever. And like 0:13:06.160 --> 0:13:08.830 you use the different term based on what you think 0:13:08.830 --> 0:13:11.890 their goals are and how morally aligned you think their 0:13:11.890 --> 0:13:15.370 goals are with, like good or bad. Measles was eliminated 0:13:15.370 --> 0:13:18.850 in the US in 2000, and now it's back because 0:13:18.850 --> 0:13:21.640 people don't want to take vaccines. As long Covid brain 0:13:21.640 --> 0:13:26.320 fog could be from damaged blood vessels. That's a new study. People. 0:13:26.320 --> 0:13:28.600 Young people are a lot less happy. And we actually 0:13:28.600 --> 0:13:30.610 because of this, we fell out of the top 20 0:13:30.640 --> 0:13:34.030 happiest nations in the United States. Uh, Alzheimer's might stem 0:13:34.030 --> 0:13:37.570 from a fat buildup in brain cells. UC Berkeley professor 0:13:37.570 --> 0:13:40.180 getting attacked for telling a student to get out of 0:13:40.179 --> 0:13:43.870 artillery range of San Francisco and San Jose if he 0:13:43.870 --> 0:13:46.210 wants to find a girlfriend, and he's just getting blasted 0:13:46.210 --> 0:13:49.270 for this weather forecast. Massively improved. This is a pretty 0:13:49.270 --> 0:13:53.380 cool seven day forecasts are now actually pretty accurate, and 0:13:53.380 --> 0:13:56.829 four day predictions now are as accurate as one day 0:13:56.830 --> 0:14:00.790 predictions from 30 years ago. Interesting. Yeah, four days versus 0:14:00.790 --> 0:14:03.850 one day. You imagine the level of complexity there. Germany 0:14:03.850 --> 0:14:08.980 legalized recreational marijuana. Blu ray is coming back because you 0:14:08.980 --> 0:14:10.720 always have access to it. You don't have to worry 0:14:10.720 --> 0:14:13.059 about it being pulled out of streaming. And it's still 0:14:13.059 --> 0:14:17.560 the highest quality. Cancer rates under 50 have surged by 80%. 0:14:17.559 --> 0:14:19.600 I think this might be in the US, but from 0:14:19.600 --> 0:14:23.620 1990 to 2019. So that's what 30 years, basically 31 0:14:23.620 --> 0:14:27.370 years married people are thriving way more than unmarried. According 0:14:27.400 --> 0:14:31.330 to a ten year long Gallup study. Bidets getting way 0:14:31.330 --> 0:14:34.780 more popular. Got some recommendations here for you. And, uh, 0:14:34.780 --> 0:14:36.910 a new way of thinking about the economy. I don't 0:14:36.910 --> 0:14:39.790 want to read this whole thing. Uh, just. I think 0:14:39.790 --> 0:14:42.520 everything is starting to come down to framing for me. 0:14:42.520 --> 0:14:46.120 It's like my universal theory. So, uh, the economy might 0:14:46.120 --> 0:14:49.630 be a thing where if you believe in optimism, if 0:14:49.630 --> 0:14:52.960 you become optimistic about the economy that has, like its 0:14:52.960 --> 0:14:55.870 own gravity, it has its own weight, its own power, 0:14:55.870 --> 0:14:58.360 because it turns into action, and then the action turns 0:14:58.360 --> 0:15:00.850 into reality. So it's like, once again, you have this 0:15:00.850 --> 0:15:05.830 link between understanding or framing or viewpoints that translates directly 0:15:05.830 --> 0:15:08.530 into what actually happen, which I just find fascinating. I 0:15:08.530 --> 0:15:11.619 think we should rethink how we use the term, uh, 0:15:11.620 --> 0:15:15.130 the word technical, which I laid out in a tweet thread. 0:15:15.130 --> 0:15:19.590 And this mostly affects women who. Often are the only 0:15:19.590 --> 0:15:22.560 people at a company who could do multiple things right. 0:15:22.590 --> 0:15:26.070 We're talking about being able to, like, navigate a business process, 0:15:26.070 --> 0:15:30.270 get people from one place to another, actually do deeply 0:15:30.270 --> 0:15:35.310 technical technology related work, but also deeply technical, like law 0:15:35.310 --> 0:15:39.180 related work or business related work, other things that are 0:15:39.180 --> 0:15:43.740 like super high end, very, very difficult stuff. And the 0:15:43.740 --> 0:15:46.230 the system would not be able to function. Nobody could 0:15:46.230 --> 0:15:49.680 explain things. Nobody could actually get the work done if 0:15:49.680 --> 0:15:52.560 it wasn't for these people who are there helping. And 0:15:52.560 --> 0:15:56.790 because it's not necessarily oriented around a piece of technology, 0:15:57.240 --> 0:16:01.800 they get really diminished. So I think we should redefine that. Yeah, 0:16:01.800 --> 0:16:04.590 I'm getting all the different AI devices coming in. I 0:16:04.590 --> 0:16:08.040 really just want something that record conversations and transcribe them 0:16:08.040 --> 0:16:11.040 and get them into my AI workflows. Going back to 0:16:11.040 --> 0:16:13.710 original style of the new summary, which is like a 0:16:13.710 --> 0:16:16.230 single sentence which which I did a lot up here, 0:16:16.230 --> 0:16:18.420 rereading the Sense of style. That's why I did that 0:16:18.420 --> 0:16:21.030 pattern really helps my writing. Every time I read it. 0:16:21.030 --> 0:16:22.890 I try to do it every like 2 or 3 years. 0:16:22.890 --> 0:16:26.010 Resubscribe to the Twitter API so I could do things 0:16:26.010 --> 0:16:29.100 like this. Boom, go down here, Twitter boom. So these 0:16:29.100 --> 0:16:32.700 are actually Twitter feeds that are in Feedly because I 0:16:32.700 --> 0:16:35.400 have the connection there and the API key put in. 0:16:35.400 --> 0:16:37.980 So that's pretty cool to be able to consolidate, bring 0:16:37.980 --> 0:16:40.560 everything in. You could also do it with Twitter lists 0:16:40.560 --> 0:16:43.560 which I use quite a bit. Yeah. Analyze pros talked 0:16:43.560 --> 0:16:46.860 about that one and getting into the three body problem 0:16:46.860 --> 0:16:50.550 on Netflix. Really enjoying this. The books were fantastic. Found 0:16:50.550 --> 0:16:54.780 this service called Opus Clip. Um, Jason pointed me to it. 0:16:54.780 --> 0:16:57.660 It's basically you take a video like this video, it'll 0:16:57.660 --> 0:17:00.300 be output and it cuts it up into pieces. And 0:17:00.300 --> 0:17:01.710 you're going to see some of these. You're going to 0:17:01.710 --> 0:17:03.989 see clips from like this will go out in my 0:17:03.990 --> 0:17:06.900 feeds and it'll be like, you know, a 32nd or 0:17:06.900 --> 0:17:10.470 92nd or whatever little clip of a cool little section, 0:17:10.470 --> 0:17:12.570 but it does it automatically. Normally you would send this 0:17:12.570 --> 0:17:14.850 to like an editor, and it's really difficult because they 0:17:14.850 --> 0:17:17.160 have to know the subject matter to know what are 0:17:17.160 --> 0:17:19.380 the coolest parts of what you said. Well, this thing, 0:17:19.380 --> 0:17:22.680 you don't have to do that because it's actually automatic. 0:17:22.680 --> 0:17:26.080 You just upload it extracts like whatever, three or 5 0:17:26.080 --> 0:17:29.010 or 10 different clips out of it. Rag Tune Tool 0:17:29.010 --> 0:17:34.140 for tuning and Optimizing rag Pipelines by Misbah, said collects, 0:17:34.140 --> 0:17:36.750 takes a web page, returns all the URLs, and switch 0:17:36.750 --> 0:17:39.750 back to this view. Open dev an open source project 0:17:39.750 --> 0:17:43.470 to clone and improve Devon Unreal Engine I mean, look 0:17:43.470 --> 0:17:49.619 at this. This is. Come on. This is content from 0:17:49.619 --> 0:17:50.430 Unreal Engine. 0:17:50.940 --> 0:17:52.230 They're combing the streets. 0:17:52.230 --> 0:17:55.230 It looks like sort of video I mean searching helps to. 0:17:55.230 --> 0:17:55.710 How real. 0:17:57.930 --> 0:18:00.390 If they arrest you too, they will take you to 0:18:00.390 --> 0:18:03.149 their headquarters and you will not return. 0:18:05.920 --> 0:18:08.109 I'm more concerned. It looks like a six foot cat. 0:18:08.650 --> 0:18:13.330 It's got claws that can cut through. Vibranium alloy. Slightly tongue. But. 0:18:14.280 --> 0:18:18.389 Let's just go forward. Okay. So let's boom down and 0:18:18.390 --> 0:18:21.720 take a closer look at this environment. Yeah, sure. Look 0:18:21.720 --> 0:18:25.679 at this. So we're talking about nanites. New adaptive tessellation feature. 0:18:25.710 --> 0:18:28.260 Look at this. So while I let you or something. 0:18:28.260 --> 0:18:28.650 Yeah. 0:18:29.160 --> 0:18:31.320 There you go. For artists to create a series of 0:18:31.320 --> 0:18:34.109 layers in the environment and then build up. So one 0:18:34.109 --> 0:18:34.560 angle that. 0:18:34.560 --> 0:18:36.869 We're approaching this from. This is another. Speaking of. 0:18:36.869 --> 0:18:38.910 Set dressing. Let's go check out that. 0:18:39.570 --> 0:18:43.500 In terms of like creating content that's completely controllable in 0:18:43.500 --> 0:18:47.580 creative but realistic look as a skill. The secret to 0:18:47.580 --> 0:18:51.240 a meaningful life I love this so much. A long term, 0:18:51.240 --> 0:18:55.620 ambitious vision that pushes you to grow smarter, wealthier, and 0:18:55.619 --> 0:19:00.629 mentally stronger. That is very concise. A guide to espouse 0:19:00.630 --> 0:19:04.470 and use. Prompt injection and jailbreaking are not the same thing. 0:19:04.500 --> 0:19:08.430 By Simon Wilson and Recommendation of the week. Think about 0:19:08.430 --> 0:19:12.750 your hedonic baseline. So this is like well I'll just 0:19:12.750 --> 0:19:16.230 keep reading. So it's thinking about my life just walking 0:19:16.230 --> 0:19:19.260 upstairs or walking to the car. It's a stoic technique 0:19:19.260 --> 0:19:23.460 of imagining that you're not able to do these simple things. 0:19:23.460 --> 0:19:26.730 And as you start removing them from your life and 0:19:26.730 --> 0:19:30.840 imagining yourself without them, you can imagine how unhappy you 0:19:30.840 --> 0:19:33.750 would be now, like a Jedi or a monk or 0:19:33.750 --> 0:19:36.449 something would say, well, you shouldn't care about that. You 0:19:36.450 --> 0:19:38.730 shouldn't care about losing anything. But I like to appreciate 0:19:38.730 --> 0:19:40.920 the things that I do have, including the ability to 0:19:40.920 --> 0:19:44.550 walk outside in the sun and not not be assaulted 0:19:44.550 --> 0:19:47.970 or something to breathe fresh air like all these things. 0:19:47.970 --> 0:19:52.169 I'm trying to micro appreciate a thousand different things in 0:19:52.170 --> 0:19:55.020 my life, which is a very stoic thing. You imagine 0:19:55.020 --> 0:19:57.689 the things that you have, you imagine them gone, and 0:19:57.690 --> 0:20:01.350 then you get appreciation when you realize that you still 0:20:01.350 --> 0:20:05.220 have them. Right. So relationships, mobility, the ability to think, 0:20:05.220 --> 0:20:07.950 being in the Bay area at this crazy moment with 0:20:07.950 --> 0:20:13.200 AI and stuff, and it's like actively cultivate appreciation by 0:20:13.200 --> 0:20:16.710 lowering my baseline, what is the minimum thing that I 0:20:16.710 --> 0:20:19.980 need to be happy? And ideally that would be nothing. 0:20:19.980 --> 0:20:22.650 And it's just something I'm thinking about. So think about 0:20:22.650 --> 0:20:26.250 what you have and think about those things by subtraction 0:20:26.250 --> 0:20:30.000 and the aphorism of the week. Contentment is natural wealth. 0:20:30.000 --> 0:20:35.040 Luxury is artificial poverty, contentment is natural wealth and luxury 0:20:35.040 --> 0:20:40.050 is artificial poverty. Socrates Unsupervised Learning is produced and edited 0:20:40.050 --> 0:20:43.740 by Daniel Missler on a Newman 87 AI microphone using 0:20:43.740 --> 0:20:47.730 Hindenburg intro and outro. Music is by zombie with a Y, 0:20:48.570 --> 0:20:50.760 and to get the text and links from this episode, 0:20:50.760 --> 0:20:53.010 sign up for the newsletter version of the show at 0:20:53.010 --> 0:20:57.649 Daniel missler.com/newsletter. We'll see you next time.