Welcome to Unsupervised Learning, a security, AI and meaning focused podcast that looks at how best to thrive as humans in a post AI world. It combines original ideas, analysis, and mental models to bring not just the news, but why it matters and how to respond. All right, welcome to unsupervised Learning. This is Daniel Miessler, and I'm building AI to upgrade humans. This is episode 458 of the podcast Unsupervised Learning. And let's jump into it. So just created my first fabric stitch, which is a combination of fabric things all piped into each other to produce a result. In this case, the name of the stitch is called rate AI result, and it's actually running a pattern called rate AI result as well. But what it is doing, I will just click over there real quick. What it is doing is actually see here it is actually using a smart AI to rate a less smart I. So this is the infrastructure that it's using. I'm actually going to do a full write up on this. Well this is the full write up. But I'm going to do a standalone video on this. So you don't have to necessarily worry about not getting everything here. So basically you have your smartest AI on top. It's judging the judged AI and the input and the prompt that it's given and it's producing a result. Then you take the input, the prompt and the results, and you send all of that to the smarter AI, which is the judge. And then what it does is it gives you an output like this. It tells you is this at the level of an uneducated human, a secondary, educated human, high school level, bachelor's, master's, PhD, world class, human or superhuman. And it's giving you that rating for whatever that task is. What's cool is that this is universal. So it's not a just a summarization one or just a threat modeling one. It looks at the task that you've given it, and it looks at the input, and it looks at the output, and it judges universally for any task that you've given it for any particular AI. It's basically saying that somebody at this level would have been able to do it about that. Good. So that was pretty exciting. I'm happy with the results so far. Look forward to anyone being able to improve it. Okay. Um, let's see here. Going back to the podcast. Yeah, I'm going to be sending out the newsletter from newsletter at Unsupervised learning.com instead of Daniel at Daniel. Com so just go ahead and add newsletter at Unsupervised learning.com to your contact list just so it won't get forwarded to a bad folder. And I'm entering the fiber world. This is pretty exciting. Upgraded to five gigabit fiber to the house. And so I'm doing a whole bunch of stuff with, uh, ten Gigabit Ethernet. So you got to do a bunch of stuff on your switching, right? Because if you just have it coming into the house, it doesn't mean it's going to make it to your computer. So you have your switches and your firewalls and your wall wiring, and all the way to your device has to be able to support that faster speed. So ten gigabit is pretty much what most consumer hardware can go up to now. The Mac studio that I use actually already has ten gigabit in it. And for something else you can get, like if it has USB and it's fairly modern, you can get like a USB connector, like USB-C connector that is actually a ten gigabit, uh, translator. So to be able to handle that, ten gigabit, even if your Nic card natively built into the system cannot. But anyway, here's what I'm getting down and up now, basically 4.6 gigabit. So it's supposed to be five. I'm going to try to get it up to five. But uh, 4600 is not bad at all. And I'm going to keep upgrading stuff like in the walls and different switches and stuff like that and see how fast we can go. And had a great sponsor conversation with my best bud, Jason Haddox with flare. We talked about a whole bunch of stuff, but especially why he likes flare and why he thinks it's the best out there for threat intelligence. So definitely go check that video out and let's jump into security. So six critical flaws have been found in Olama AI framework, potentially allowing denial of service or model theft or poisoning attacks. And remember, don't put your olama online. A lot of people know that you can olama natively. It publishes APIs so you can call the different models that you create. So if you go get a model from hugging face and you put it in a llama and you spin it up, well, now there's an API that can listen and answer requests, and you could do some cool stuff there, but do not make that public unless you really, really know what you're doing and you have auth and all that kind of stuff because it could be bad as we see here, we actually have vulnerabilities here. FBI is warning about a rise in hacked police emails being used to send fake subpoenas and emergency data requests. That's the trick one, right? Emergency data requests. So it looks like it's coming from a reputable place that has authority, like a police department. So of course you're going to turn over the data, right? I think this is a pretty nasty general attack here, which is you find a low security organization that's like understaffed in cyber, like a police department. Then you send emails from the police department to a place that you want to get data from, and then they turn it over thinking that they have to. So I think this is a general thing, and I think police departments, also law firms, might be likely to respond to this, although maybe they'll be smarter about asking more questions anyway. I think it's a general interesting type of attack. Google's AI security assessment tool, Bigsleep, found a zero day vulnerability in SQLite, and it's the first time we've really seen. I find something that in production software that hasn't been found via another method. Right. Because they've done fuzzing on this thing or fuzzing has been done on this thing, I'm sure, many times. And this is a good example of a AI actually finding something. So I mean, I always thought that this was inevitable. I still think it's inevitable that this is going to become like the main way of finding things. But yeah, Google found it in the real world. FBI is asking for public help in identifying Chinese hacker, uh, hackers in groups like apt 31 and APT. 41. CrowdStrike has launched a new AI Red team service to identify vulnerabilities in AI systems. So they've spun up an AI security consultancy. Kind of strange inside of CrowdStrike itself, which does endpoint protection, Synology is telling its users to patch a critical zero click RC bug affecting millions of disk, station and Photos NAS devices. Why are the NAS devices online? Friends don't friend. Friends don't let friends put NAS devices online. It is really nasty. It's like the worst thing you could put online. Not because they're the most vulnerable, which they are quite vulnerable as we've seen. But that's not the real reason. The real reason is it's a combination of how vulnerable is it versus how sensitive is it a NAS network attached, attached storage? It tends to have lots of really sensitive stuff in it. So you need to be really careful with what you do with that thing. And putting it on online is not really careful. All right. Nokia is investigating potential breach after a hacker Intel broker claimed to have stolen their source code from a third party vendor. Canada has ordered TikTok Technology Canada to shut down for national security reasons, but doesn't actually block Canadians from using TikTok, which is what I think most people think when they see this story. It's actually just shutting down the Canadian business operations. Researchers from GMU, George Mason University, have introduced Mantis framework that uses prompt injections to hack back against prompt injection. I'm not usually a fan of hack back stuff, but I do like the exploration here. US is tightening rules on foreign real estate deals near military bases, adding 60 more installations to the list. And this is after basically the military found China buying a whole bunch of land around military bases to do crypto mining. And they're like, uh, I don't think we like this. So there's now laws in place to stop them from doing that. In some cases, AI and tech robotic dogs are now patrolling Mar-A-Lago to help protect president elect Trump. These high tech hounds are part of the Astro program equipped with surveillance tech and sensors. So if we open that one up. Yeah. Look at these things. This is just like the Black Mirror episode. Look at that. Scary little things. Now, if it's just one and you're like, oh, there's probably dumb. It's only 2024. What happens when it's 2026 and it's got guns on it? And what happens when it's not one, but it's like 700 and you could just like point a laser pointer on something and be like, get em. And they just swarm and start firing. Sounds like sci fi. It's not really going to be sci fi. And China is crushing in the drone market. They're building these things so fast. They're really good at them. Look at DJI. Its number one drone company in the world, at least on the consumer side. So this is something we have to watch. Yeah. What I said here is 2025 and 2026 are going to become some serious utopia slash dystopia years. Lots of sci fi happening except for in actual reality, Nvidia surpassed Apple to become the world's largest company by market cap 3.43 trillion. And that might have gone up or down. Since that was written, OpenAI has introduced new feature called predicted outputs lets you send expected content to speed up API responses. I like this stuff similar to the caching stuff. I like the little tweaks that are helping you day to day, especially save some money. Waymo has launched its robotaxi service across 80 square mile area in and around Los Angeles. What is going on? Why can't we have this for the whole Bay area? They did San Francisco and they just left. Come on, we need it for the whole Bay area. Apple's adding a new feature to find my in iOS 8.2. So basically if you lose your bag or your bag, your luggage doesn't get to your destination in some foreign country, you can actually send an AirTag for your bag to the airline so the airline can actually find it inside the airport. I think that's super cool. Apple Vision Pro. Vision Pro's Vision OS 2.2. Yeah. Look at this thing.

Welcome to Unsupervised Learning, a security, AI and meaning focused podcast that looks at how best to thrive as humans in a post AI world. It combines original ideas, analysis, and mental models to bring not just the news, but why it matters and how to respond. All right, welcome to unsupervised Learning. This is Daniel Miessler, and I'm building AI to upgrade humans. This is episode 458 of the podcast Unsupervised Learning. And let's jump into it. So just created my first fabric stitch, which is a combination of fabric things all piped into each other to produce a result. In this case, the name of the stitch is called rate AI result, and it's actually running a pattern called rate AI result as well. But what it is doing, I will just click over there real quick. What it is doing is actually see here it is actually using a smart AI to rate a less smart I. So this is the infrastructure that it's using. I'm actually going to do a full write up on this. Well this is the full write up. But I'm going to do a standalone video on this. So you don't have to necessarily worry about not getting everything here. So basically you have your smartest AI on top. It's judging the judged AI and the input and the prompt that it's given and it's producing a result. Then you take the input, the prompt and the results, and you send all of that to the smarter AI, which is the judge. And then what it does is it gives you an output like this. It tells you is this at the level of an uneducated human, a secondary, educated human, high school level, bachelor's, master's, PhD, world class, human or superhuman. And it's giving you that rating for whatever that task is. What's cool is that this is universal. So it's not a just a summarization one or just a threat modeling one. It looks at the task that you've given it, and it looks at the input, and it looks at the output, and it judges universally for any task that you've given it for any particular AI. It's basically saying that somebody at this level would have been able to do it about that. Good. So that was pretty exciting. I'm happy with the results so far. Look forward to anyone being able to improve it. Okay. Um, let's see here. Going back to the podcast. Yeah, I'm going to be sending out the newsletter from newsletter at Unsupervised learning.com instead of Daniel at Daniel. Com so just go ahead and add newsletter at Unsupervised learning.com to your contact list just so it won't get forwarded to a bad folder. And I'm entering the fiber world. This is pretty exciting. Upgraded to five gigabit fiber to the house. And so I'm doing a whole bunch of stuff with, uh, ten Gigabit Ethernet. So you got to do a bunch of stuff on your switching, right? Because if you just have it coming into the house, it doesn't mean it's going to make it to your computer. So you have your switches and your firewalls and your wall wiring, and all the way to your device has to be able to support that faster speed. So ten gigabit is pretty much what most consumer hardware can go up to now. The Mac studio that I use actually already has ten gigabit in it. And for something else you can get, like if it has USB and it's fairly modern, you can get like a USB connector, like USB-C connector that is actually a ten gigabit, uh, translator. So to be able to handle that, ten gigabit, even if your Nic card natively built into the system cannot. But anyway, here's what I'm getting down and up now, basically 4.6 gigabit. So it's supposed to be five. I'm going to try to get it up to five. But uh, 4600 is not bad at all. And I'm going to keep upgrading stuff like in the walls and different switches and stuff like that and see how fast we can go. And had a great sponsor conversation with my best bud, Jason Haddox with flare. We talked about a whole bunch of stuff, but especially why he likes flare and why he thinks it's the best out there for threat intelligence. So definitely go check that video out and let's jump into security. So six critical flaws have been found in Olama AI framework, potentially allowing denial of service or model theft or poisoning attacks. And remember, don't put your olama online. A lot of people know that you can olama natively. It publishes APIs so you can call the different models that you create. So if you go get a model from hugging face and you put it in a llama and you spin it up, well, now there's an API that can listen and answer requests, and you could do some cool stuff there, but do not make that public unless you really, really know what you're doing and you have auth and all that kind of stuff because it could be bad as we see here, we actually have vulnerabilities here. FBI is warning about a rise in hacked police emails being used to send fake subpoenas and emergency data requests. That's the trick one, right? Emergency data requests. So it looks like it's coming from a reputable place that has authority, like a police department. So of course you're going to turn over the data, right? I think this is a pretty nasty general attack here, which is you find a low security organization that's like understaffed in cyber, like a police department. Then you send emails from the police department to a place that you want to get data from, and then they turn it over thinking that they have to. So I think this is a general thing, and I think police departments, also law firms, might be likely to respond to this, although maybe they'll be smarter about asking more questions anyway. I think it's a general interesting type of attack. Google's AI security assessment tool, Bigsleep, found a zero day vulnerability in SQLite, and it's the first time we've really seen. I find something that in production software that hasn't been found via another method. Right. Because they've done fuzzing on this thing or fuzzing has been done on this thing, I'm sure, many times. And this is a good example of a AI actually finding something. So I mean, I always thought that this was inevitable. I still think it's inevitable that this is going to become like the main way of finding things. But yeah, Google found it in the real world. FBI is asking for public help in identifying Chinese hacker, uh, hackers in groups like apt 31 and APT. 41. CrowdStrike has launched a new AI Red team service to identify vulnerabilities in AI systems. So they've spun up an AI security consultancy. Kind of strange inside of CrowdStrike itself, which does endpoint protection, Synology is telling its users to patch a critical zero click RC bug affecting millions of disk, station and Photos NAS devices. Why are the NAS devices online? Friends don't friend. Friends don't let friends put NAS devices online. It is really nasty. It's like the worst thing you could put online. Not because they're the most vulnerable, which they are quite vulnerable as we've seen. But that's not the real reason. The real reason is it's a combination of how vulnerable is it versus how sensitive is it a NAS network attached, attached storage? It tends to have lots of really sensitive stuff in it. So you need to be really careful with what you do with that thing. And putting it on online is not really careful. All right. Nokia is investigating potential breach after a hacker Intel broker claimed to have stolen their source code from a third party vendor. Canada has ordered TikTok Technology Canada to shut down for national security reasons, but doesn't actually block Canadians from using TikTok, which is what I think most people think when they see this story. It's actually just shutting down the Canadian business operations. Researchers from GMU, George Mason University, have introduced Mantis framework that uses prompt injections to hack back against prompt injection. I'm not usually a fan of hack back stuff, but I do like the exploration here. US is tightening rules on foreign real estate deals near military bases, adding 60 more installations to the list. And this is after basically the military found China buying a whole bunch of land around military bases to do crypto mining. And they're like, uh, I don't think we like this. So there's now laws in place to stop them from doing that. In some cases, AI and tech robotic dogs are now patrolling Mar-A-Lago to help protect president elect Trump. These high tech hounds are part of the Astro program equipped with surveillance tech and sensors. So if we open that one up. Yeah. Look at these things. This is just like the Black Mirror episode. Look at that. Scary little things. Now, if it's just one and you're like, oh, there's probably dumb. It's only 2024. What happens when it's 2026 and it's got guns on it? And what happens when it's not one, but it's like 700 and you could just like point a laser pointer on something and be like, get em. And they just swarm and start firing. Sounds like sci fi. It's not really going to be sci fi. And China is crushing in the drone market. They're building these things so fast. They're really good at them. Look at DJI. Its number one drone company in the world, at least on the consumer side. So this is something we have to watch. Yeah. What I said here is 2025 and 2026 are going to become some serious utopia slash dystopia years. Lots of sci fi happening except for in actual reality, Nvidia surpassed Apple to become the world's largest company by market cap 3.43 trillion. And that might have gone up or down. Since that was written, OpenAI has introduced new feature called predicted outputs lets you send expected content to speed up API responses. I like this stuff similar to the caching stuff. I like the little tweaks that are helping you day to day, especially save some money. Waymo has launched its robotaxi service across 80 square mile area in and around Los Angeles. What is going on? Why can't we have this for the whole Bay area? They did San Francisco and they just left. Come on, we need it for the whole Bay area. Apple's adding a new feature to find my in iOS 8.2. So basically if you lose your bag or your bag, your luggage doesn't get to your destination in some foreign country, you can actually send an AirTag for your bag to the airline so the airline can actually find it inside the airport. I think that's super cool. Apple Vision Pro. Vision Pro's Vision OS 2.2. Yeah. Look at this thing.

The full size scopes with everything else. It's like having a two monitor set up. Look at this. Better create a lot more space for those in my I've used this here. Almost have a full size video and a full size scopes with everything else. It's like having a two monitor set up, but better. So let's take this to the logical extreme and go ultra wide. Look at that.

The full size scopes with everything else. It's like having a two monitor set up. Look at this. Better create a lot more space for those in my I've used this here. Almost have a full size video and a full size scopes with everything else. It's like having a two monitor set up, but better. So let's take this to the logical extreme and go ultra wide. Look at that.

Wow, it's bent around you.

Wow, it's bent around you.

This is kind of amazing. Actually. I don't know that using completely crazy.

This is kind of amazing. Actually. I don't know that using completely crazy.

Completely crazy. And actually the ultra one is too large. Excuse me? You can't actually, you can't actually look at this thing. You have to turn your head all the way around it. So just the the bigger one, like the middle one is actually the one that I use. And they've made it so that it's not always in focus wherever you, wherever for the whole screen, because that would be too difficult. Right. So what happens is when wherever you move your eyes, it comes into crystal clear focus. And I'm looking at a really high quality monitor right now. Actually it's the Apple monitor. It looks just as good as that. It really does. It looks just as good as like the six K Apple monitor once you focus on it. Because I mean it's it's sending light right into your eye. Right. So it's pretty good at it. And it's crazy how fast it is too. Because when you're moving your eyes around very quickly, first of all, it knows exactly where you're looking. And it just takes that spot and just crystallizes it. Really, really cool. It's like it's like the most practical thing that I've done actually with the Vision Pro, because mostly it's entertainment and like playing with different apps. But this is the killer app I've been waiting for. Is productivity an actual super wide screen? TSMC is opening up its fab in Arizona in December, So that's going to be cool for Onshoring in the US. TSMC is halting the supply of advanced AI processors to Chinese clients starting in November. Well, already happened a couple of days ago following an investigation showing chips were ending up in Huawei devices. Humans dollars at its highest point in two years. Stock market going crazy. Uh, yeah. Bitcoin over 90,000. Yeah, I did predict that Trump was going to win and that investors were going to go crazy, but I didn't think it was going to be this crazy and this soon. Andreessen Horowitz is backing AI powered parenting tools, with Justine Moore highlighting she's a partner on this thing. New wave of parenting co-pilots I like this, I like this, so my kid is doing this. What should I do? And what I like about that is like, okay, what would a helicopter parent do? What is like an eastern way of doing this? What's a European way of doing this. What is a modern US West Coast way of doing this? So it's like you're hearing a bunch of experts because nobody really knows how to raise kids, right? It's very, very customized. Very there's not a perfect one way to do it. Right. So I don't have kids, but if I did, I would want to know, what do the people who tend to raise the best kids, what is their long term cultural way of handling this particular problem? And AI is amazing for that. So that seems pretty, uh, pretty cool. My buddies participating in a real life bug bounty. Actually, it's actually a treasure hunt, which is very similar to a bug bounty. So my buddy, he's been going out to an island, and he's meeting his other top bug bounty friend out there, and they're actually looking for treasure. The treasure is worth half $1 million. And this book that's coming out, I think it's already out, actually, um, I'm going to click on this book. Yeah. It's called There's Treasure Inside. Okay. So it's all about this person who's launching these treasure hunts with tons, like millions of dollars of prizes. And they're really hard to find. And you have to travel there and you have to follow clues. It's like puzzles. And I was talking to my buddy about this, and it's like it's very similar to Bug Bounty, where it's like, there's this joy of, you're the first one to find this thing, and you get paid money as a result of finding it. It's like very it like, revives your childhood to be chasing something in this way. So I think it's really cool. And he agreed with that assessment. Genetic discrimination becoming a real thing as we knew it would. Insurers use DNA data to deny coverage or to raise prices. So this guy, Bill, who was a healthy 60 year old, was denied long term care insurance after he submitted his genetic stuff and it found a genetic mutation linked to ALS. So he didn't have the disease, but they were like, yeah, you can't get this coverage because your genetics say you might get it. Companies are already moving production out of China as Trump plans massive tariffs. So Steve Madden cutting China made products by 40 to 45%, shifting to Vietnam and Cambodia. Stanley Black and Decker reworking its supply chain and H&M manufacturing. And crews are eyeing increased U.S. production. Seems like the tariffs could work in multiple ways. It definitely has a deterrent. The way I'm excited about it is people wanted to move to us onshoring, or they wanted to move out of China or whatever, which everyone in the West kind of wants to, but they're like, it's just too we can't do it yet. We can't do it yet. I want to do it, but I can't do it yet. And this just pushes people. It pushes people to be like, okay, well, now I might as well because I'm going to I might get charged all this money to, to have to use China. So now it might actually be worth the money to switch. So hopefully the new administration doesn't go crazy with this because it actually will cause inflation because it'll just be more expensive for everyone to buy things. And Elon talked about this recently and he's like, you can't do too many of these and you can't do them too quickly or you will break things. So you have to be careful with them. But just the idea of being out there, I think is going to have the effect that they actually want. NASA's Juno spacecraft just completed 66th flyby of Jupiter, sending back stunning images. Look at this stuff. This is crazy. Look at these. That's ridiculous. Isn't that gorgeous? Is that the spot or is that just a different spot? I thought the spot was red. Well, these are different colors anyway. But look at that. Isn't that amazing? Look at this. These are all storms. Yeah. Just gorgeous. Man, I love astronomy. Yeah, that's. These are all spots. But I don't think. I don't think maybe these are the big red spot. In this case, blue spot. But that's definitely not it. That's a different one. Yeah, really cool DNA Dikmen's Leaving and Waving is a brilliant and touching photo series capturing her parents waving goodbye over the years. Look at this thing. Look at this. Isn't this cool? All these captures of them waving goodbye. And presumably this is like, while she's growing up. Oh, look at that. Switches from black and white to color. Yeah, really, really cool. I love this kind of stuff. New study from Ben-Gurion University shows that controlling blood sugar can slow brain aging. Yeah, sugar, inflammation, brain aging. These are getting more and more linked. And yeah, the science here is is getting pretty solid. Astrobiologist Sara Imari Walker explores the complex questions of what life truly is in her book, Life as No One Knows It physics of Life's emergence. This is on my list might be a candidate for the UL Book Club as well. She says that modern science is yet to develop a theory that fully integrates life into the universe's description. Yeah, I love this life emergent stuff. A mom in Georgia was jailed after her 11 year old son walked alone to town, despite the fact that she was doing this on purpose because she believes in free range upbringing. I'd love for this libertarian mindset of like, yeah, just leave me alone to manage whatever to actually come to these places, right? Because here they are in Georgia, which should be libertarian, right? And no, she went to jail. The mom went to jail. I mean, I grew up in the 80s. I used to walk around all over the place. I could ride my bike anywhere. I just had to be back before the streetlights came on. Or you get your ass whipped. I mean, that's that's that's the way it's supposed to work. Not really. You would get in trouble. Our average age of US home buyers jumped to 56. 56 years old for an average home buyer. It was 49. This needs to be, like 29. What is going on? 56 that is way too old. Oliver Sacks explores the meaning of life through love and despair in his letters, emphasizing that meaning is something we create, not find. Marginalien. This is Maria Popova project. Really, really cool and this is a great write up here. Excuse me. Ideas. Crypto is back, but it's mostly as gambling and money stores, and not so much on what it was promising with like Ethereum. Although maybe Solana might be still like an Ethereum type vibe. It's basically like fast Ethereum. And I got a prediction here about this whole election thing, which I'm purposely not talking too much about right now. I actually think a lot of good can come from the whole Elon Andreessen, Thiel, Trump, uh, Vivek Ramaswamy, like all this stuff. I think all this stuff could go horribly bad. I voted for Harris, by the way, because of election denial essentially is the main reason um, wasn't happy with Harris or the campaign or anything like that, but I yeah, anyway, I'm not going to go into that too much. The point is, now that this has happened, I'm trying to be optimistic. I see no alternative other than being optimistic. And I believe that a lot of these people Thiel, Elon Andreessen, um, Rogan these are Bernie loving people. These are populists. These are long time liberals. Trump is a long time liberal. Trump is a previous Democrat, been a Democrat his whole life. And he switched over. In my opinion, a lot of these people switched over and kind of like got triggered and went crazy because of the crazy left. So my hope is and this is wishful thinking, I understand that. My hope is we're going to see a lot of good come out of these folks. Hopefully. Okay. So watch this reduction of government, this whole dodge thing. It's kind of a gimmick, but I think they might actually do some cool stuff with it. Like, what if a whole lot more of our money that we were spending in taxes? What if that became 50% more effective or 25% more effective, or 75% more effective? I think we could have maybe clean streets, maybe nice roads, fewer potholes, actual infrastructure get built. My hope and my belief, optimistic belief is that these people, including possibly even Trump, are actually trying to raise the bottom. They are actually actually trying to raise everyone, not just their friends. There's no question that a lot of their decisions are very selfish and very focused on their business and very opportunistic. No question. Everyone knows that. I mean, that's most people do that. The question is the balance between these things. So my thing is, if they're sitting on a razor's edge somewhere in the middle or the middle, right? And they could become crazy, right? People, or they could fall back towards the center and actually be positive and try to lift everyone. And a lot of these AI people, they're trying to create free abundance. Okay. Free shelter, free food, not free, but very, very cheap. Okay. Easy to make very, very cheap things. Why? Because of AGI? Because of ASI. Because a lot of these things have the ability to remove how expensive it is for the whole world to have what the best people in the world have. Right? Um, so, so basically you have like the best health care, right? The best, um, the best education. So you have Harvard education, you have the best health care policy you could possibly have, like Kaiser or whatever you have, um, PPO or something like that. The best of everything. Why can't everyone have that? Why is only why do only a few people go to Stanford or Harvard? Right? So the promise of this AI stuff is to be able to give that education to everyone via like an AI agent, something that's always with you. Abundant energy. Right. Not having these wars over oil because they're solar farms everywhere. And we already have enough energy. It's done. Solved problem. I helped us with that. AGI helped us. Maybe ASI helped us. Doesn't matter. The point is, you never have to need something. Okay, now, now we have a meaning crisis. We're going to have to figure out what we're actually going to do with ourselves after that. But the point is, I think a lot of these people are actually going for that better future. A lot of these people, I think, are actually going for the Star Trek The Next Generation model, which is what I'm going after. And again, maybe it's about to get really bad because it's going to go full authoritarian. It's going to go full fascist. Maybe it will, and I'm very worried that it will. Okay. It's very possible that they got triggered and they're never coming back. Okay. I actually talked to Sam Harris about this. Um, and I'm not going to say anything that he said that he didn't say in his podcast, but I responded to him after he did that reckoning podcast. And, um, he basically said, I think they're gone in the podcast. I think they're gone and it's going to get really bad. And so I challenged him on that with basically this argument, and he thinks I'm wrong. He is my favorite thinker. He is the smartest person I believe out there right now in terms of like his ability to see things and articulate them. He's just like God tier to me, um, always has been. I feel like I've caught up in a lot of ways, but he still sees things extremely clearly, and the fact that he doesn't believe me is troubling to me. I actually can't find anyone who believes me about this particular thing. They're like, oh, so are you going right with them? I'm like, no, I'm staying where I'm staying. I am a Star Trek Next Generation liberal. This is where I'm going to be. But I do believe in conservative methods of achieving liberal goals. Okay. So I'm trying to find optimism here. And here's what I'm offering to you. Here's what I'm asking of you. If you are like me and you were in this liberal center, so you've moved away from the left because they're crazy, you're not going to the right because they're crazy. You're here in this, this first principles. Think through the problem situation in the center somewhere. I'm asking you to consider that these people might actually be trying to be doing good things. And I think this comes down to what I think a big problem in our politics is right now is assuming the absolute worst, assuming that they are going, they're just sitting somewhere right now doing Mr. Burns like, conniving, like, how can we ruin the entire bottom 90%? We're just going to make this amazing world for the top 10% of all our friends. And it's just going to be rich. People rule everything, and we're going to do, you know, eugenics. And it's just I don't think they're trying to do that, I really don't. I think they're capable of doing that. Especially Trump is capable of doing that if pushed in the wrong way. If ego gets involved, like we've already seen him do some horrific stuff and very stupid stuff. Bottom line, I'm trying to be optimistic, and I'm asking you to consider trying to be so as well. It's it discovery. Security is a useless controls problem. This is a must read. You got to read this thing even if you don't. You don't agree. Chain forge, open source visual programming tool. This is like Yahoo pipes for running evaluations against prompts. Have it messed with this fully. In fact, I'm going to open another tab. But from what I looked at before, uh, looks pretty cool. And I love the Yahoo pipes like design mechanism. How do you run away from an army of these? Yeah. You don't watch this thing. Look at this thing. Dir. Dir. Dir. This is exactly like the Black Mirror episode. Except for this thing is faster and scarier. Okay, it could do backflips. Not scary at all. Yep. So just imagine you're a soldier. You went through basic training. Some eight. You did some. What's it called? You go out and you see, like over the mountain. You just see thousands of these things pour over, except for they're bigger with bigger muscles, bigger battery packs. Oh, and they got guns. Cool, cool. It's gonna be a problem, I'm telling you. I just put it on. Uh, x the other day. We really need to be thinking about Anti-drone stuff. And there's a company called Anduril. Andrill not sure how to pronounce it, but they are definitely thinking a lot of drone stuff. Um, some people sent me some links saying the Navy's thinking about this, because here's the problem. F-35s, carrier fleets, aircraft carriers themselves, they're very vulnerable to being swarmed by small drones, especially autonomous drones. That can't be, like, jammed. A very serious problem to worry about. And by the way, I keep recommending this book. It's called Kill Decision by Daniel Suarez. Definitely want to check that out. Um, he basically talks about a lot of the stuff in sort of a fiction sense, but it gives you a vibe for why it's pretty scary. Toolkit smart name set of scripts that allow new sub commands to get. I cluster using Mac Minis and XO labs. This is probably what I'm going to build my next AI rig on. Maybe. Maybe it depends how the whole game is. But uh, yeah, you could do like Thunderbolt five and connect these things together. Then you use this project called XO labs and it actually links them all together. How I ship products at Big Tech companies. Diagrams tool for creating diagrams as code. Everything I've learned so far about running local LMS, Patrick or no, I keep wanting to say the author, but Patrick McCormick encourages readers to spend less time doomscrolling and more time reading books, drawing audio. New musical sketchpad using web audio API. Create music directly in your browser. Recommendation of the week. The CEO of anthropic thinks AGI is coming within a couple of years. Sam Altman thinks it's 25 or 26. I've already said it was 25 to 26. I said that in 2003. So these numbers are probably getting real. Of course, it does depend on your definition, but my recommendation to you, given that AGI is coming, start getting ready. Know your life mission. Know your goals. Fill in in practice your most important sentence. Which is this right here. I believe one of the biggest problems in the world is X, which is why I'm building, creating or doing Y. So fill in that sentence for yourself. Start building your telos file. Get really good with your AI tools like fabric ChatGPT whatever you're using. Get your website up. Commit to reading 50 books in 2025 and hit me up if you need book recommendations. And most importantly, start writing. Even if you think you don't have anything to say. Once you start reading this many books, you're going to have something to say, especially if you have your Telos file up. You know your goals. You know your life mission. It's going to be like a particle accelerator. Things are going to smash into your brain. Ideas are just going to start flying off, and then you just start writing online and you're like, hey, wouldn't it be cool if this hey, what about this? Someone's like, hey, that's stupid. This is why it's stupid. You're like, oh, that's true. Yeah. Good point. Pretty soon you're learning publicly, live in real time and you can just build a lot of new intelligence based on that. And the aphorism of the week, if you were offered $1 million not to wake up tomorrow, you would not take it. Which means waking up tomorrow is worth more than $1 million. So treat it that way. If you were offered $1 million not to wake up tomorrow, you would not take it. Which means waking up tomorrow is worth more than $1 million. So treat it that way. See you next time. Unsupervised learning is produced and edited by Daniel Miessler on a Neumann U87 AI microphone using Hindenburg. Intro and outro music is by zombie with a Y. And to get the text and links from this episode, sign up for the newsletter version of the show at Daniel miessler.com/newsletter. We'll see you next time.

Completely crazy. And actually the ultra one is too large. Excuse me? You can't actually, you can't actually look at this thing. You have to turn your head all the way around it. So just the the bigger one, like the middle one is actually the one that I use. And they've made it so that it's not always in focus wherever you, wherever for the whole screen, because that would be too difficult. Right. So what happens is when wherever you move your eyes, it comes into crystal clear focus. And I'm looking at a really high quality monitor right now. Actually it's the Apple monitor. It looks just as good as that. It really does. It looks just as good as like the six K Apple monitor once you focus on it. Because I mean it's it's sending light right into your eye. Right. So it's pretty good at it. And it's crazy how fast it is too. Because when you're moving your eyes around very quickly, first of all, it knows exactly where you're looking. And it just takes that spot and just crystallizes it. Really, really cool. It's like it's like the most practical thing that I've done actually with the Vision Pro, because mostly it's entertainment and like playing with different apps. But this is the killer app I've been waiting for. Is productivity an actual super wide screen? TSMC is opening up its fab in Arizona in December, So that's going to be cool for Onshoring in the US. TSMC is halting the supply of advanced AI processors to Chinese clients starting in November. Well, already happened a couple of days ago following an investigation showing chips were ending up in Huawei devices. Humans dollars at its highest point in two years. Stock market going crazy. Uh, yeah. Bitcoin over 90,000. Yeah, I did predict that Trump was going to win and that investors were going to go crazy, but I didn't think it was going to be this crazy and this soon. Andreessen Horowitz is backing AI powered parenting tools, with Justine Moore highlighting she's a partner on this thing. New wave of parenting co-pilots I like this, I like this, so my kid is doing this. What should I do? And what I like about that is like, okay, what would a helicopter parent do? What is like an eastern way of doing this? What's a European way of doing this. What is a modern US West Coast way of doing this? So it's like you're hearing a bunch of experts because nobody really knows how to raise kids, right? It's very, very customized. Very there's not a perfect one way to do it. Right. So I don't have kids, but if I did, I would want to know, what do the people who tend to raise the best kids, what is their long term cultural way of handling this particular problem? And AI is amazing for that. So that seems pretty, uh, pretty cool. My buddies participating in a real life bug bounty. Actually, it's actually a treasure hunt, which is very similar to a bug bounty. So my buddy, he's been going out to an island, and he's meeting his other top bug bounty friend out there, and they're actually looking for treasure. The treasure is worth half $1 million. And this book that's coming out, I think it's already out, actually, um, I'm going to click on this book. Yeah. It's called There's Treasure Inside. Okay. So it's all about this person who's launching these treasure hunts with tons, like millions of dollars of prizes. And they're really hard to find. And you have to travel there and you have to follow clues. It's like puzzles. And I was talking to my buddy about this, and it's like it's very similar to Bug Bounty, where it's like, there's this joy of, you're the first one to find this thing, and you get paid money as a result of finding it. It's like very it like, revives your childhood to be chasing something in this way. So I think it's really cool. And he agreed with that assessment. Genetic discrimination becoming a real thing as we knew it would. Insurers use DNA data to deny coverage or to raise prices. So this guy, Bill, who was a healthy 60 year old, was denied long term care insurance after he submitted his genetic stuff and it found a genetic mutation linked to ALS. So he didn't have the disease, but they were like, yeah, you can't get this coverage because your genetics say you might get it. Companies are already moving production out of China as Trump plans massive tariffs. So Steve Madden cutting China made products by 40 to 45%, shifting to Vietnam and Cambodia. Stanley Black and Decker reworking its supply chain and H&M manufacturing. And crews are eyeing increased U.S. production. Seems like the tariffs could work in multiple ways. It definitely has a deterrent. The way I'm excited about it is people wanted to move to us onshoring, or they wanted to move out of China or whatever, which everyone in the West kind of wants to, but they're like, it's just too we can't do it yet. We can't do it yet. I want to do it, but I can't do it yet. And this just pushes people. It pushes people to be like, okay, well, now I might as well because I'm going to I might get charged all this money to, to have to use China. So now it might actually be worth the money to switch. So hopefully the new administration doesn't go crazy with this because it actually will cause inflation because it'll just be more expensive for everyone to buy things. And Elon talked about this recently and he's like, you can't do too many of these and you can't do them too quickly or you will break things. So you have to be careful with them. But just the idea of being out there, I think is going to have the effect that they actually want. NASA's Juno spacecraft just completed 66th flyby of Jupiter, sending back stunning images. Look at this stuff. This is crazy. Look at these. That's ridiculous. Isn't that gorgeous? Is that the spot or is that just a different spot? I thought the spot was red. Well, these are different colors anyway. But look at that. Isn't that amazing? Look at this. These are all storms. Yeah. Just gorgeous. Man, I love astronomy. Yeah, that's. These are all spots. But I don't think. I don't think maybe these are the big red spot. In this case, blue spot. But that's definitely not it. That's a different one. Yeah, really cool DNA Dikmen's Leaving and Waving is a brilliant and touching photo series capturing her parents waving goodbye over the years. Look at this thing. Look at this. Isn't this cool? All these captures of them waving goodbye. And presumably this is like, while she's growing up. Oh, look at that. Switches from black and white to color. Yeah, really, really cool. I love this kind of stuff. New study from Ben-Gurion University shows that controlling blood sugar can slow brain aging. Yeah, sugar, inflammation, brain aging. These are getting more and more linked. And yeah, the science here is is getting pretty solid. Astrobiologist Sara Imari Walker explores the complex questions of what life truly is in her book, Life as No One Knows It physics of Life's emergence. This is on my list might be a candidate for the UL Book Club as well. She says that modern science is yet to develop a theory that fully integrates life into the universe's description. Yeah, I love this life emergent stuff. A mom in Georgia was jailed after her 11 year old son walked alone to town, despite the fact that she was doing this on purpose because she believes in free range upbringing. I'd love for this libertarian mindset of like, yeah, just leave me alone to manage whatever to actually come to these places, right? Because here they are in Georgia, which should be libertarian, right? And no, she went to jail. The mom went to jail. I mean, I grew up in the 80s. I used to walk around all over the place. I could ride my bike anywhere. I just had to be back before the streetlights came on. Or you get your ass whipped. I mean, that's that's that's the way it's supposed to work. Not really. You would get in trouble. Our average age of US home buyers jumped to 56. 56 years old for an average home buyer. It was 49. This needs to be, like 29. What is going on? 56 that is way too old. Oliver Sacks explores the meaning of life through love and despair in his letters, emphasizing that meaning is something we create, not find. Marginalien. This is Maria Popova project. Really, really cool and this is a great write up here. Excuse me. Ideas. Crypto is back, but it's mostly as gambling and money stores, and not so much on what it was promising with like Ethereum. Although maybe Solana might be still like an Ethereum type vibe. It's basically like fast Ethereum. And I got a prediction here about this whole election thing, which I'm purposely not talking too much about right now. I actually think a lot of good can come from the whole Elon Andreessen, Thiel, Trump, uh, Vivek Ramaswamy, like all this stuff. I think all this stuff could go horribly bad. I voted for Harris, by the way, because of election denial essentially is the main reason um, wasn't happy with Harris or the campaign or anything like that, but I yeah, anyway, I'm not going to go into that too much. The point is, now that this has happened, I'm trying to be optimistic. I see no alternative other than being optimistic. And I believe that a lot of these people Thiel, Elon Andreessen, um, Rogan these are Bernie loving people. These are populists. These are long time liberals. Trump is a long time liberal. Trump is a previous Democrat, been a Democrat his whole life. And he switched over. In my opinion, a lot of these people switched over and kind of like got triggered and went crazy because of the crazy left. So my hope is and this is wishful thinking, I understand that. My hope is we're going to see a lot of good come out of these folks. Hopefully. Okay. So watch this reduction of government, this whole dodge thing. It's kind of a gimmick, but I think they might actually do some cool stuff with it. Like, what if a whole lot more of our money that we were spending in taxes? What if that became 50% more effective or 25% more effective, or 75% more effective? I think we could have maybe clean streets, maybe nice roads, fewer potholes, actual infrastructure get built. My hope and my belief, optimistic belief is that these people, including possibly even Trump, are actually trying to raise the bottom. They are actually actually trying to raise everyone, not just their friends. There's no question that a lot of their decisions are very selfish and very focused on their business and very opportunistic. No question. Everyone knows that. I mean, that's most people do that. The question is the balance between these things. So my thing is, if they're sitting on a razor's edge somewhere in the middle or the middle, right? And they could become crazy, right? People, or they could fall back towards the center and actually be positive and try to lift everyone. And a lot of these AI people, they're trying to create free abundance. Okay. Free shelter, free food, not free, but very, very cheap. Okay. Easy to make very, very cheap things. Why? Because of AGI? Because of ASI. Because a lot of these things have the ability to remove how expensive it is for the whole world to have what the best people in the world have. Right? Um, so, so basically you have like the best health care, right? The best, um, the best education. So you have Harvard education, you have the best health care policy you could possibly have, like Kaiser or whatever you have, um, PPO or something like that. The best of everything. Why can't everyone have that? Why is only why do only a few people go to Stanford or Harvard? Right? So the promise of this AI stuff is to be able to give that education to everyone via like an AI agent, something that's always with you. Abundant energy. Right. Not having these wars over oil because they're solar farms everywhere. And we already have enough energy. It's done. Solved problem. I helped us with that. AGI helped us. Maybe ASI helped us. Doesn't matter. The point is, you never have to need something. Okay, now, now we have a meaning crisis. We're going to have to figure out what we're actually going to do with ourselves after that. But the point is, I think a lot of these people are actually going for that better future. A lot of these people, I think, are actually going for the Star Trek The Next Generation model, which is what I'm going after. And again, maybe it's about to get really bad because it's going to go full authoritarian. It's going to go full fascist. Maybe it will, and I'm very worried that it will. Okay. It's very possible that they got triggered and they're never coming back. Okay. I actually talked to Sam Harris about this. Um, and I'm not going to say anything that he said that he didn't say in his podcast, but I responded to him after he did that reckoning podcast. And, um, he basically said, I think they're gone in the podcast. I think they're gone and it's going to get really bad. And so I challenged him on that with basically this argument, and he thinks I'm wrong. He is my favorite thinker. He is the smartest person I believe out there right now in terms of like his ability to see things and articulate them. He's just like God tier to me, um, always has been. I feel like I've caught up in a lot of ways, but he still sees things extremely clearly, and the fact that he doesn't believe me is troubling to me. I actually can't find anyone who believes me about this particular thing. They're like, oh, so are you going right with them? I'm like, no, I'm staying where I'm staying. I am a Star Trek Next Generation liberal. This is where I'm going to be. But I do believe in conservative methods of achieving liberal goals. Okay. So I'm trying to find optimism here. And here's what I'm offering to you. Here's what I'm asking of you. If you are like me and you were in this liberal center, so you've moved away from the left because they're crazy, you're not going to the right because they're crazy. You're here in this, this first principles. Think through the problem situation in the center somewhere. I'm asking you to consider that these people might actually be trying to be doing good things. And I think this comes down to what I think a big problem in our politics is right now is assuming the absolute worst, assuming that they are going, they're just sitting somewhere right now doing Mr. Burns like, conniving, like, how can we ruin the entire bottom 90%? We're just going to make this amazing world for the top 10% of all our friends. And it's just going to be rich. People rule everything, and we're going to do, you know, eugenics. And it's just I don't think they're trying to do that, I really don't. I think they're capable of doing that. Especially Trump is capable of doing that if pushed in the wrong way. If ego gets involved, like we've already seen him do some horrific stuff and very stupid stuff. Bottom line, I'm trying to be optimistic, and I'm asking you to consider trying to be so as well. It's it discovery. Security is a useless controls problem. This is a must read. You got to read this thing even if you don't. You don't agree. Chain forge, open source visual programming tool. This is like Yahoo pipes for running evaluations against prompts. Have it messed with this fully. In fact, I'm going to open another tab. But from what I looked at before, uh, looks pretty cool. And I love the Yahoo pipes like design mechanism. How do you run away from an army of these? Yeah. You don't watch this thing. Look at this thing. Dir. Dir. Dir. This is exactly like the Black Mirror episode. Except for this thing is faster and scarier. Okay, it could do backflips. Not scary at all. Yep. So just imagine you're a soldier. You went through basic training. Some eight. You did some. What's it called? You go out and you see, like over the mountain. You just see thousands of these things pour over, except for they're bigger with bigger muscles, bigger battery packs. Oh, and they got guns. Cool, cool. It's gonna be a problem, I'm telling you. I just put it on. Uh, x the other day. We really need to be thinking about Anti-drone stuff. And there's a company called Anduril. Andrill not sure how to pronounce it, but they are definitely thinking a lot of drone stuff. Um, some people sent me some links saying the Navy's thinking about this, because here's the problem. F-35s, carrier fleets, aircraft carriers themselves, they're very vulnerable to being swarmed by small drones, especially autonomous drones. That can't be, like, jammed. A very serious problem to worry about. And by the way, I keep recommending this book. It's called Kill Decision by Daniel Suarez. Definitely want to check that out. Um, he basically talks about a lot of the stuff in sort of a fiction sense, but it gives you a vibe for why it's pretty scary. Toolkit smart name set of scripts that allow new sub commands to get. I cluster using Mac Minis and XO labs. This is probably what I'm going to build my next AI rig on. Maybe. Maybe it depends how the whole game is. But uh, yeah, you could do like Thunderbolt five and connect these things together. Then you use this project called XO labs and it actually links them all together. How I ship products at Big Tech companies. Diagrams tool for creating diagrams as code. Everything I've learned so far about running local LMS, Patrick or no, I keep wanting to say the author, but Patrick McCormick encourages readers to spend less time doomscrolling and more time reading books, drawing audio. New musical sketchpad using web audio API. Create music directly in your browser. Recommendation of the week. The CEO of anthropic thinks AGI is coming within a couple of years. Sam Altman thinks it's 25 or 26. I've already said it was 25 to 26. I said that in 2003. So these numbers are probably getting real. Of course, it does depend on your definition, but my recommendation to you, given that AGI is coming, start getting ready. Know your life mission. Know your goals. Fill in in practice your most important sentence. Which is this right here. I believe one of the biggest problems in the world is X, which is why I'm building, creating or doing Y. So fill in that sentence for yourself. Start building your telos file. Get really good with your AI tools like fabric ChatGPT whatever you're using. Get your website up. Commit to reading 50 books in 2025 and hit me up if you need book recommendations. And most importantly, start writing. Even if you think you don't have anything to say. Once you start reading this many books, you're going to have something to say, especially if you have your Telos file up. You know your goals. You know your life mission. It's going to be like a particle accelerator. Things are going to smash into your brain. Ideas are just going to start flying off, and then you just start writing online and you're like, hey, wouldn't it be cool if this hey, what about this? Someone's like, hey, that's stupid. This is why it's stupid. You're like, oh, that's true. Yeah. Good point. Pretty soon you're learning publicly, live in real time and you can just build a lot of new intelligence based on that. And the aphorism of the week, if you were offered $1 million not to wake up tomorrow, you would not take it. Which means waking up tomorrow is worth more than $1 million. So treat it that way. If you were offered $1 million not to wake up tomorrow, you would not take it. Which means waking up tomorrow is worth more than $1 million. So treat it that way. See you next time. Unsupervised learning is produced and edited by Daniel Miessler on a Neumann U87 AI microphone using Hindenburg. Intro and outro music is by zombie with a Y. And to get the text and links from this episode, sign up for the newsletter version of the show at Daniel miessler.com/newsletter. We'll see you next time.