All right. Welcome to unsupervised learning. This is Daniel. All right. So I found the best coffee recipe basically in existence. And a whole bunch of baristas are super freaking out about this. So excited. And this is the guy I think he's invented, like, a whole bunch of really popular recipes. And I think he's kind of famous for the Hario switch as well. But anyway, he's got a new recipe
that is just unbelievable. And I have tried it three times so far, and it's worked perfectly all three times, and I've not even executed it properly, except for maybe once. And that time was even better. But I was having problems with like inconsistent, like coffee quality, which I think is a in terms of the cup, not the coffee going in. He basically says that this can make any being taste amazing and I honestly believe him. It is really, really that good. So I highly recommend you check this
thing out. I'll tell you really quickly what it is. It's basically a combination of pour over and, uh, immersion. Okay. So basically what you do is you do like a 32nd steep with the hario switch closed. Then you do a pour over where it actually goes through for a certain amount of time. And he has the recipe, of course, um, goes through a for a certain amount of time, then you close it again. So he's made multiple iterations of
this thing to make it even better and better. So you close it again, then you pour water in and this time you pour colder water, you actually add some cold to it. So you start off at like 90 degrees. This is Celsius, by the way. And then I forget what you go down to, but I just added a little bit of colder water and then you do your second immersion steep for X amount of time, which is in the recipe with that colder temperature to avoid extracting
too much harshness. And then, uh, yeah, now you've got essentially three different phases, and then you mix it all up, you let it sit for a little bit, and it is unbelievably good. So it's essentially very round, very fruity, lots of body. It's just it's the best coffee I've been making. Um, now he has mentioned that he's really good with the Aeropress and I just got the metal Aeropress. The new one. Uh, so it's metal and glass. I
haven't tried that. That used to be my favorite. So. But I've learned a lot about coffee since then, so we'll see how this compares to that. I do know one thing. The Aeropress takes longer, so I'm more likely to use this hario switch version anyway. But if you are into coffee, you must check this out. All right. This guy is super funny. I'm not going to click on there. There's too much content. This guy is just absolutely hilarious. He does these skits making fun of computer people. Um,
obviously he is one, so no harm there. Been grinding on my predictions page. This is what it looks like now. And this is essentially where I'm putting like all my different predictions about different things. And I've got a long way to go on this thing, but I've got quite a few added And, uh, I've got a table here. So the table is going to maintain like, uh, the different. Just a collection of them, right. And this is going to actually be fairly long, probably, I don't know, 10
or 20. I'm probably going to keep up here at a given time. Who knows, maybe even more. And then down below I have more detail about the particular one. And this is all sort of in the vein of this book, Super Forecasting, which is just a brilliant way to think about like acting within a world that's uncertain. So really excited about working on that project. Got a fabric call out. Really appreciate. All right. Security. New study
shows AI powered spearfishing is extremely effective. Finding AI tools achieves 54% click through rate versus 12% for regular phishing emails, and matched human expert performance at 1/30 of the cost 1/30. And I was like, well, you don't know about the methodology. You don't know how good it was. So first I ran some fabric patterns against the study itself, and it came back and said it was a good methodology. Then I found out it was basically Bruce Schneier was part
of this and it's out of Harvard. So I think it's a pretty good finding. US launched a new cyber security label program called US Cyber Trustmark to help people know which smart home devices are secure. There's a series of books called Project Russia Reveals. Edit there's a series of books called Project Russia that essentially outlines what Putin is trying to do. So taking down Western democracies, spiritual warfare, Economic collapse. And this came out in 2005, 2010. So
maybe it's been updated since then. Maybe he's different, maybe he's better, maybe he's worse. But, uh, I wanted to get my hands on all of these. And what I really want to do is write some fabric patterns that basically. Okay, if you want to know what like the FSB is trying to do, just, okay, you know, their campaigns are going to orient the world if they're effective towards the direction they want to go, right? So watch this, find out the direction that they want to go, which turns
out is written down in these books. Then if you want to know if someone is being influenced by them, there's not a guarantee, by the way. It's not a guarantee. There could be overlap that's sort of accidental or circumstantial, but in general, if they're doing a good job, then all you have to do is look for spokespeople who are talking the exact book of what they're what they're trying to do, which is like spiritual warfare, economic collapse.
Destroying Western democracies. And I'm like, yeah, find people who are speaking that exact book. Like, for example, Tucker Carlson Taiwan says Chinese ship might have deliberately cut one of their undersea internet cables. I wonder if this could be like the start of a fragmented internet. So, like the Axis and allies have their own internet and they kind of nobody can rely on the undersea cables anymore. I've always wondered about undersea cables. It's it's like been a
thing for me. I have this thing where I wonder why people don't go to overpasses and throw giant rocks off of them into cars that are moving 80mph. And I'm like, there are thousands of these things all over the place. If somebody wanted to do that, they could. Um, and there's just so many opportunities for like, why aren't these things happening? And the answer is that there's some sort of barrier that stops people from doing bad things. There is a very strong barrier. Otherwise it would be
mass pandemonium all over the place. But another one of those examples, similar to the, uh, the overpass thing, is. Okay, we have these giant undersea cables, and the the ocean is essentially outer space. Like, there's just you can't possibly patrol all of it. So why wouldn't somebody who wanted to disrupt billions of dollars of Western commerce? Why wouldn't
they just go out and cut these cables? And I was thinking about this years and years ago, and I was like, there's no way these cables are going to be safe. Like, they're just there. Like, it's. They're easy to find. Maybe not easy to get to. And, like, if it's deep enough. Maybe not easy to get to. But there are probably places where it is at it. There are probably places where it is easy enough to get to. Right. And sure enough, we're seeing this happen
all over the place. We're seeing Finland's complaining about it. We've got Estonia complaining about it. We've got Taiwan complaining about it. So yeah. Anyway, it could be that we are more connected via satellite, which is a tiny, tiny fraction of the bandwidth that would be over an undersea cable. But I don't know, maybe we have a more isolated internet in the future. CISOs released data on how well like 7800 critical infrastructure orgs were adopting their security goals.
It looks like healthcare care and water systems and communication and government facilities are doing the best. Got a Chinese drone company, DJI has released an update. They released an update that removes its geo fencing so it's able to fly next to restricted areas like the white House or airline. Uh, airlines? Airports. Like, how could they possibly. I don't get this. How could they possibly do this and not immediately face, like, threats of being banned? Are they just not worried about it?
Is the story wrong? Like something doesn't seem right here. Arctic Wolf found a likely zero day being used to hack Fortinet firewalls. New zero day and Avantis VPN products. I don't know who's running Avanti VPN, but you should probably stop. Sonicwall is rushing to get customers to patch a serious authentication bypass bug in their SSL VPN. Greenland is becoming increasingly critical for US military strategy in the Arctic,
because Russia already has 50 Arctic bases there. And if you talk to anyone in Finland, that is a new hotspot up there because Russia's up there as well. White House and some intelligence agencies are now saying there's a 5050 chance that a foreign actor did target U.S. personnel. It's actually reminds me of Covid a lot, where it's like common sense says, yeah, it was an energy weapon. Common sense says, yeah, it came out of the Wuhan lab. And then the the experts come out who I think
want to be careful and basically politically correct. And they're like, yeah, no connection whatsoever to the Wuhan Covid lab. And no, Russians are definitely not shooting energy beams at our, you know, diplomats. Definitely not. And here we are years later, after the fad of ignoring obvious things has sort of passed. Now people are now saying obvious things. They're like, hey, maybe it was actually Russians shooting energy weapons. Hey, maybe it
was actually the Wuhan lab. I in tech, Apple's actually building the AI that we want. This is this is crazy. This is going to go on the predictions page, by the way. I actually need to write that down. I'm going to forget it. But. What have we been wanting from Siri all this time? For the last ten years we've been wanting Siri to like, just understand things, be able to do things for us. And it hasn't gotten there.
It's actually behind most of most of the other eyes, in fact, so much so that a lot of brands have solidified on Google and Alexa and not on Siri, because it just wasn't that good. And it kind of still isn't. But I'm telling you, I am staying on all the betas. Um, I've always installed the betas, like the day they get released. But look at this update
for beta for 18.4. Look at number two. Like a real life assistant, you'll be able to reference texts you received, past calendar events, and more personal data to get truly intelligent assistance from Siri. So everyone is saying, okay, where is Apple? Apple is being is being left behind by all this AI stuff. You know they're not competing. They're going too slow or whatever. And I've been saying the whole time, and I did work there for a few years by the way. Um, but this is not coming
from that. This is coming from the fact that I've been a religious fanboy since 2007, and I kind of have been watching how they work since like 2010. So I study them very closely. And then, of course, I also work there, but I'm not giving any insider information whatsoever. But the way they work, okay. And this is not insider, this is just my analysis. And a lot of people already agree with this. The way they work is they
think about the distant future. Okay. I've been saying for a long time they are building life os they have been for a long time education, health, you know, fitness, um, you know, work life, personal art, creativity. They're they have been doing slow pieces, iterative pieces in all these different areas, and now they're pulling them together and they're unifying them. Then they build their cloud. First of all, they have Secure Enclave. So they have the secure local system of
any anyone out there okay. Their hardware is just the best. Then they just built that secure cloud infrastructure, which is essentially the same as a secure enclave, but it's in the cloud. So things that have to go to the cloud for AI operations that they can't do locally, they can they could do that in the cloud now. Okay. So they build all this infrastructure. That was a year and a half ago. And now they're finally starting to
pull it together. New app actions. Apple says Siri will be able to perform hundreds of new actions in Apple apps, without the need to even open those apps. App intents will bring the same capability to supported third party apps as well. And what's this? Siri will know what's on your display, so you can easily ask it to take action on whatever you're looking at. I'm telling you right now, this is what Apple often does. Now, they didn't do it with Siri until now, and I would have bet
against that and I would have been wrong. In fact, I need to put that on my predictions. Wrong about Siri. Sorry for taking note. I just don't want to miss it. That was one. If you had told me ten years ago, would we be at a place ten years from now where Siri has been behind the market for five years? I would have bet good money against that and I would have been wrong. I think this is what undoes it. I'm not often wrong about Apple stuff, by the way.
I'm like Point 9%, right? But I would have been wrong about that one. And I think this is what finally undoes it now. It's going to be hard. It's going to be difficult. There's going to be problems. But I feel like they're finally, finally getting it. I hope they are all right. Open A's zero one keeps randomly switching languages while solving problems and nobody knows why. So it sometimes switches to Chinese, Persian, and other languages right
in the middle of reasoning. Even when the question was asked in English. You know what I hope this is? I don't know what this is. I haven't looked into it yet. But you know what I hope it is? I hope certain types of problems lend themselves better to certain languages. Right? We know that certain languages are like analytical, like German or something. Then you have like, I don't know, is it romantic or love based or, I don't know, maybe there are certain types of analysis. Maybe math. Maybe
math is better in Persian. I don't know, maybe math is better in Chinese. I have no idea, but that would be a cool reason if that turned out to be the reason. I don't know that it is. Zuckerberg says meta will start using AI to replace mid-level software engineers in 2025. He says they'll have AI systems that can effectively work as company engineers writing code, eventually planning
to automate all app development at meta. Yes, this reminds me of what Benioff was talking about for Salesforce and Agent Force. He's also selling models. They're open source models, but ultimately Zuckerberg is also selling AI. Um, I have to think about this a this a little more, but. And it's definitely less the case than Benioff, who is really selling trying to sell Agent Force. But I partially
believe Zuckerberg here. But I also think he's also just trying to hype up the company and get people to buy more stock. So hard to know how much of a balance there is, but I think both are probably true. In video released a blueprint for creating digital twins of robot fleets in factories and warehouses. Virtual environments where companies can test and optimize their robots. Robots are going crazy. I think I have a comment down here, but robots
are going crazy right now. OpenAI released their new economic blueprint that outlines how the US can win the AI race against China. OpenAI is starting to build its own robotics team, hiring for hiring first for its hardware roles, and Dell is changing their laptop names just to Dell, Dell Pro and Del Pro Max, which is copying Apple on multiple different things. So first it was the form. Uh, the form factor of phones. Everybody copies. Everyone's phone looks
like an iPhone now. All all packaging now looks like Apple packaging. And now everything's a pro and a Pro Max. Like, buy anything on Amazon. And like, half of it has those, uh, product lines now. And Apple might launch a new app called invites. Hopefully better than like, things like Calendly, although I think Calendly is pretty good, but it just seems like there should be a more professional way to handle that.
I don't know why nobody has yet. Stack overflow is seeing a massive decline in new questions, dropping 77% Since 2022. Anything crazy happened in 2022? Not sure. Not sure what that could have been. Amazon is winding down some of its Dei programs and integrating others into existing processes. Meta killed off most of their programs and their entire Dei department. My analysis here is that the corporate world continues to
move towards the Alaskan fishing boat model. There is no die on an Alaskan fishing boat, and if robots could do the work, there wouldn't be any cruise. There wouldn't be any people on Alaskan fishing boats either. Expect this to accelerate in the next four years, because essentially the whole concept of Dei. Um. Yeah, the whole concept of Dei kind of lost big time on November 5th. Humans might have just made huge changes to their content policies,
including allowing hate speech against many different groups. I think my opinion on this might be different. A lot of people around this area. I actually want platforms to display full reality, including, uh, really nasty speech and stuff that I do not want to see. I simply want the ability to filter it out. I want to be able to take that stuff and block it and say, I don't want hostility, I don't want sexism, I don't want racism. I don't want conversations about this. I want to block, uh,
politicians with, you know, that rhyme with frump? Like, I want to have that control, but I don't want the Form to make those decisions for me. I want to see reality. If I want to. And this is especially the case if I want to do like pure research on like how often how much is racism or sexism going up or down in certain communities? Well, you can't
do that if the platform itself is blocking it. Um, so this is all part of a discussion, and I got a piece here about this, but, um, it's all about a discussion of like, okay, what actually is a platform, right? Is it a public thing or is it a private thing? And that's really what we have to ask ourselves. Economy added 256,000 jobs in 2024. Unemployment. Unemployment rate at it. The unemployment rate dropped to 4.1%. What if they counted the people who are now looking for jobs, though? it'd
be like 41%. I'm exaggerating. Slightly rare seven planet alignment is coming to Earth's night skies in February 2025. I tell you what, these planets are already out. I think I mentioned this last week. You've got to go see these things. Like right now, like Mars is like straight above me. And Jupiter is moving off the axis towards the west a little bit, which makes sense. Oh, and also, you know what's really cool? Do you know why the planets are always kind of like right in the middle?
And they move left to right through the sky. It's because that is the axis of the plane of all the planets. And we rotate in line with that axis, that or that plane that all the planets are rotating on. So the flatness of the plane of all the different planets is the same axis that we rotate on. We're at a slight angle, but that's why as we're turning, this would be more the case. Um, at the equator. But. We're oh, and they're going towards the west, which means
we're turning counterclockwise from the north. I was going to ask I about that a second ago. In fact, I'm going to do it right now. Are we moving? Counter clockwise. This is probably a stupid question. I haven't spatially figured this out. I don't know why I'm doing this live. Are we moving counterclockwise relative to the. It depends on if you're top or bottom, right. Like Australia versus Canada. Are we moving counterclockwise relative to the. Oh, no. What
I want to know is, are we moving counterclockwise? Are we moving opposite? Are we rotating? Opposite. To the. Direction we are. There is no actual opposite. Okay, so if we're going this. If all the planets are going. Top and bottom. This is so stupid. If all the planets are going clockwise, the question is, is the planet? Is Earth turning with that direction or against that direction, and I guess the inside would be against the outside would
be with. Are we rotating opposite to the direction that we are orbiting sun or. In line with it. And watch this tab. Enter or no. Alt. Enter. Counterclockwise on its axis when viewed from above the North pole. And orbits the sun in the same counterclockwise direction. Okay, so that's. That's how it's thinking. Top down from the North Pole. I should have thought of that. Top down from the North Pole. Counterclockwise on its axis, and it's going around
the sun in the same counterclockwise. Okay. Interesting. All right. Sorry. Sorry to bother you guys. All right. We'll cut that out. We'll use I. Whatever. All right. Germans are using do it yourself balcony solar panels to cut their electricity bills. Young people are increasingly avoiding relationships, and it's happening all over the world. Japan's already low marriage rate dropped another 12%
since 2019. A third of 18 to 34 year olds globally say they're just not interested in dating or relationships. I don't know what kind of problem that's going to be, but it's going to be a problem. In 1965, the US government tried replacing Mexican farm workers with the American high school, with American high school athletes. And this is this is back when like, Americans were super hard working. Right? And it failed spectacularly. 18,000 teenagers, only 3300 worked in
the fields, but they quit within weeks. My dad actually talked about this. He talked about he went out for, like, one day. And this would have been, uh, this would have been in the 60s. Yeah, it would have been in the 60s or the 70s. Yeah, 60s or 70s. So right about this time, he went out and, uh, worked in the field for like, one day, and, or maybe it was like a week and just, like, quit too hard. Yeah. Fafo f around and find out. That's
what we're going to do. If if we if we, uh, kick out, like, all these workers, like the country will just stop and they'll be like, nah, just kidding, you can come back. My thoughts on Zuckerberg and Musk showing us raw internet. Basically, I want that. But with the ability to tune out what I don't want to see. This is what we just talked about. All right. Discovery. This little snippet. I do this all the time GP
and then text. I don't even have to use quotes, just GP and then text to, um, get commit with that text. And yeah, this is a cool little call to action on tweet. New Nvidia free courses for AI. A researcher scanned the entire IPv4 internet on port 80 using mass scan and then analyze the results with nmap. Yeah, this one was cool. See if I can pull it up. Look at that. 42 million IP addresses on port 80 discovered buckets. Digital ocean spaces. Like lots of different data here.
Pretty cool project. I can't optimize your way to being a good person. Your obsession with moral optimization and quantifying ethical behavior might be making us worse people. Yeah. Julia Evans explains what's needed for a modern terminal experience. This is a really good write up by Julia. You got to check this out also. I'm digging the typography. Mad respect for the website. Uh, using fish terminal emulator neovim. There you go. There you go. Neovim base 16 framework. Hmm.
Not familiar. Some out of the box options. Anyway, she's got a full, in-depth thing. Quite good, quite good. Philips Hue users. Uh, I am one of them. We'll be able to create custom scenes using voice or text commands like give me a scene for a garden party, and it'll either use the garden party scene or it'll just go make one that it thinks is appropriate, which I think is super cool. Recommendation of the week. Don't clutter your mind with anything before you start work. Treat your
morning clarity as a precious resource. You want the energy to go into difficult work, not distraction fragmentation. So the way I think about this is like you start off with like 100 energy points. When you wake up in the morning, let's assume you had a perfect night of sleep and you had exercise before, and you're just like, you feel really good. So you wake up, you're at 100. Or worst case scenario, you didn't sleep perfectly. You're out of 90. Let's say you're top shape for whatever it is.
What I've noticed, and this is um, Karpathy was talking about this as well. And actually, I think I originally heard this maybe a couple years back from Alex Hormozi. Um, and it really rung true for me, and I've been thinking about it since, um, basically, if you get into email or especially if you get into social media, those things will consume your energy and it's really, really bad. The worst possible thing that I can do is I
don't get up and get out of bed. I wake up, I reach for my phone, I look at my phone and I start clicking through emails, and I start clicking through social media. And then I check the news. So what happens is my brain starts going crazy with news, crazy with social media, crazy with emails. And then it starts essentially doing work. My brain starts, it spins up a bunch of like avatars, and it just starts going and working and working and working. I still haven't got
up yet. My brain is getting tired. And this is like the worst possible thing you could do because you are extremely creative with that 100 creativity and energy points. Extremely creative. You just cleaned your brain. You just cleansed your whole brain for the night by sleeping. That's what sleep does. You wake up refreshed. Your brain is like, let's do something cool. Let's do something cool. And instead of doing something cool, instead of getting into really hard, difficult,
creative work, you waste it. You wasted on news or social media or something like that. So don't do that. Don't waste those points on trash. What does Karpathy say here? Uh, well, what he said earlier is he wake up and go directly to work. Do not check anything. No messages, no email, no news, no nothing. There's something destructive, distracting about checking the outside world that I don't fully understand. I think I just explained it. I think what I just described
is the explanation for it, roughly. And the aphorism of the week. I can bear any pain as long as it.
Is.
At it. And the aphorism of the week I can bear any pain as long as it has meaning. I can bear any pain as long as it has meaning. Haruki Murakami.