AGI is Already Here & Physicists Explain "Negative Time"

Added: 2026-05-06

[00:00:00]you cuz you're not >> in real time. Real time. Whatever time.

[00:00:03]Whatever time.

[00:00:04]>> All right. All right.

[00:00:06]>> Figures. Robotic factory just went from one robot of production per day to 24 of them. One per hour. Imagine coming next one every minute and then 10 factories doing one a minute, etc. AI starting to reveal the hidden rivers of brush strokes inside of famous paintings, giving us all sorts of new insights into them. Dancers now get a 3D time machine for replaying and reworking their movements. Mark Andre is done playing around with AGI definitions, and he says AGI is here. Avon Blackstone surveyed 1,000 cringe dating profiles so that we don't have to. Revealing what everyone else gets wrong online. Physicists have now measured negative time. No kidding.

[00:00:49]In the lab, ice. There's so much going on with ice this week. Physicists have discovered the most complex forms of ice yet. Literally, cocaine pollution is messing with salmon. We'll look at the effects artificial intelligence is having on the Google and Meta ad machine behind the scenes. And there's definitely money going through that thing that I did not see coming. Turns out the more young people are using AI, not the more they're falling in love with it, the more they actually hate it.

[00:01:16]The FDA now wants AI watching clinical trials in real time. Fascinating article on how imagination really works in the brain. There's a new theory upends everything that we thought we knew.

[00:01:28]Maybe word of consciousness is something closer to what you suppress and then what's left over than it is actually creating. Now, crazy thought. And finally, we'll talk about evolvable AI.

[00:01:37]But first, >> take off your clothes.

[00:01:40]>> If you wouldn't mind before we dive into all of that stuff, hitting that subscribe button if you're not already or sharing this video with someone gives some insights into my last couple videos. you know, didn't really do too well. Got 12 new subscribers. Always thankful for everybody new. Made 30 bucks. Yeah, retention was okay. I probably spent too long talking about the Musk versus Altman lawsuit. And you guys really like the article consciousness in AI, the abstraction fallacy. So, I might think about doing more kind of in that realm. Not as many hypes as usual either, but you know, still those 79 did equate to 62,000 hype points, so always appreciative. All right, so let's start by talking about Figure AI. Their robotics factory has had a massive manufacturing milestone recently. They are scaling production of the new humanoid robots from one per day to about one per hour. That is a 24-fold increase in under 120 days. And it does seem like they've got something pretty special here happening. So this rapid acceleration occurred at their bot Q factory. It's happening in California.

[00:02:45]It is shifting the company away from a prototype development company to a high volume manufacturing company. And this video provides an interesting behindthe-scenes look at how the factory is actually producing these things. So, as of right now, about 350 figure 03 robots have been delivered. The yield metrics show that the battery yield reached 99.3% with 9,000 actuators all produced. So, there's a lot of little moving components in there. And more than 99% of them are working fine. But the main point is they're pushing really hard because every one of these robots that's out in the real world is giving them the data that they need for the next generation of rapid improvement. So, you know, something as small as being able to produce these 20 or 30 minutes faster than their competitors right now and actually deliver them, get them activated, get that data in a way that's actually usable and can train the next system. it. It's a big difference maybe in the long run. That might give them that flywheel effect that makes it, you know, a centimeter, an inch to a mile eventually. But get ready for a world that looks like this. This will be you and me like running through one of these job sites and there's probably going to be thousand robots for every human. So jeez, get ready. All right. Next, let's talk about how artificial intelligence and paintings are coming together in a pretty unique way. So researchers have developed a new way to study paintings using computer vision. They're focusing on something we usually can't see clearly, the direction of tiny brush strokes. So, paintings are made obviously of thousands of these little marks. They're the brush strokes. Each line had its own motion. You know, what kind of brush was it? How the wrist kind of pushed it down? What had the paint sort of take? There's all these little things that you want to think about that an AI system can be trained to understand. So, the system analyzes small sections of the painting and it detects the direction of each stroke.

[00:04:36]Then it connects them into flowing lines which they call streamlines that map out how the brush moved across the canvas.

[00:04:43]And that turns something subjective like an artist's gesture into a clear visual piece of data that it can learn. They also measured features like length, curve, and direction. And this makes it possible to compare how different artists paint. For example, some artists use smooth constant strokes, others use short curves, and some just change their brush strokes every single time. In one case, the method showed how brush strokes followed the shape of objects and even changed with lighting. Bright areas had strokes that spread outwards while shadows moved more in parallel lines. And these results are visual road maps for how paintings are visually made. And it helps both experts and everyday viewers better understand the techniques, styles, and artist decisions. I'm sure it's also kind of a unique data set that can be trained for the future of maybe physical robots painting. They didn't quite go into that as like where the goal is, but I was just thinking to myself like, yeah, imagine like an actual paint stroke that can be optimized. Like someone's going to do something with that. The actual movement behind a painting is very different from a diffusion model that makes a painting. So, I'm excited to see, I guess, where this goes. And if nothing else, just to get another level of understanding into some of the great artworks that we already have. All right, next up, let's talk about an extended reality tool that lets dancers analyze their own movement. All right, so imagine you're a dancer. I I don't know. I'm not one, so I'm going to try to like really use my imagination on this, but dancers usually review their work by watching flat videos. Makes sense. Record yourself on camera. Just play it back. But movement is three-dimensional, so important details always get lost. This is a new system.

[00:06:22]It is called DANX Reflect, and it changes all of that. and it turns regular video into 3D virtual spaces that you can step into. And when you're inside of that space, you know, your your movement can appear as a life-size avatar. So, you can put on a VR headset, you can stand in front of a virtual mirror, and then you can recreate the pose. So, this system finds the closest match and then shows it next to you so that you can compare your movement with the original in real time. So, imagine a bunch of dancers. They're essentially using this recording of themselves to get in the perfect position, like to be more perfectly accurate than ever before. It makes feedback more precise, easier to understand. Dancers can study movements like they are in the room with it. And in tests, dancers said that it felt like a natural extension and help them see the motion more clearly and take better notes. And the reason I wanted to kind of share it is there is this idea that you might put on a virtual headset and control a robot and take an action. So if you want to learn how to do dishes or for something, they'll put a remote worker into something that mimics what the robot does. They'll do the task and then they'll learn that data. But it does seem to me like tools like this are going to correlate in a way where a creative skill is taught. Imagine like surgery that's done, you know, on the other side of the world. Maybe it's dancing. Maybe it's how to work on a construction site. But this idea of perfecting real world movements in these digital twins and where they're going to like step into robotics and also step back into people who are trying to perfect their craft. It's sort of fascinating to see the way that it's all blurry blurring together. Mark Andre has said, so I mean there's a lot of people saying an AGI is x amount of months away or weeks or years or days or it's already happened. But when Mark Andre says something, a lot more people believe it. So he says AGI is here. Now, of course, I've been going off this for a while, Allen's conservative countdown to AGI, and he uses a definition where there needs to be some like, you know, real embodied robots doing what humans do before he's going to call it. But let's listen to what Mark Andre thinks.

[00:08:33]So on April 5th, 2026, Mark Andre posted a single line on X that instantly went viral. Quote, "I'm calling it. AGI is here. It's just not evenly distributed yet." So Mark is a top investor. He was like an early pioneer on the web. He invested in, I think, all sorts of stuff. Facebook and Airbnb, Coinbase, kind of famous, often called A16Z. He's got a huge empire. I think he's definitely a billionaire. This is an interesting comment. So the phrase is deliberate. It's a direct echo of William Gibson's famous observation about the future and it not being evenly distributed. He's using AGI in the same way that the average person would. He's saying that AI can do basically any thinking task that a human can do at on par with that human. He does make a small distinction calling it functional AGI. This means that AI can already handle most valuable thinking work, even if it's not perfect in theory. Right now, AI can write code, it can review documents, it can run workflows, and it can work for hours, use tools, and solve problems step by step. Even if humans are still guiding it, he argues that the heavy thinking is done by the machine.

[00:09:41]So that is AGI. And even if it might not feel like it right now, the real limit is not ability anymore. It's access, it's cost, and it's how these tools are built into products. So the idea is that AGI is not coming later. It's already working. It's just hidden behind gates.

[00:09:58]So I don't know. I mean, functional AGI just feels like a rebrand, maybe not a breakthrough. I've kind of already leaned towards the fact that AGI is here, though, also, so I'm not downplaying it. I just have a very loose definition. And as soon as I see these systems in all sorts of useful ways being superhuman, like the way they can answer medical questions and the way that it can just understand Wikipedia and the world and answer questions for me and and teach me, it just feels already so generally intelligent. And I think someone like Alan Turing would sit down with Gemini today and just say this this is it. Like this is what I was predicting. this thing could con convince me that it's human for the most part in most ways and that's good enough for me to just take it really serious like start making policies around it.

[00:10:43]Start creating a society where this kind of thing is at the center of it and acknowledged and like respected and worked on and like something all citizens are taking part in. The core of it here, so I feel like we should start treating it like a world where AGI exists. All right. Avon Blackstone has survived a thousand cringe dating profiles so that you don't have to.

[00:11:04]Let's dive into this and we'll talk about the role of AI in her work. So, she is somebody that considers herself quote moderately hostile towards AI. She objects to the impact on the environment. She's incredibly angry with how much of our economy has been based on its nonsense and she is nearly overwhelmed by the rightness indignation at the amount of theft that it has committed AC against artists and other types of creatives. That being said, she also found it stupendously useful for things like spellchecking. She's dyslexic. I know what you mean. Catching typos and creating art that she could not create before. So, the author has reviewed a thousand dating profiles. But the catch is they're from a Facebook group that shares cringe content. That isn't scientific, but it shows a clear pattern on what people react to. So, what patterns did she and her AI find?

[00:11:54]First, everyone agreed on a few things.

[00:11:56]Bad spelling, hard drug use, and heavy religious messaging turned people off fast. These came up across all groups with almost no one defending them. Next, men had three main complaints. They dislike profiles that said, quote, "Strong independent woman, but then also asked for a provider. Guess they found that ironic. They reacted negatively to women listing mental health issue diagnosis." And they were especially frustrated by profiles mentioning children. Women's reaction were more intense and more frequent. The biggest issue by far was men talking about their genitals. Oh my gosh. Yeah. Okay, I get that one. Second one was alpha male or red pill language, which people saw as fake. And the third was men openly saying they were married. Queer users had a shorter list. They mainly objected to Christianity, MAGA messaging, and racist signals. One key takeaway, not everyone you dislike is cringe.

[00:12:50]Sometimes it just means that the person isn't for you. So yeah, point is AI didn't just change what people found as cringe in dating profiles. It just exposed how inconsistent and biased those judgments already are. I feel like that's the fun part though about people like ra reviewing like dating profiles is that it's just it's meant to just be talk and like connection. I don't really think a lot of times you're actually solving a problem. I think it's more about connecting over people and hearing their opinions of what is out there. So, you know, you do you. You make your dating profile however you want. But if you did do anything that you regret, don't worry. Scientists have discovered negative time. Yeah, it's not going to totally let you go back into the past, but it's interesting enough to make you think about how that might happen.

[00:13:32]Physicists have measured something that sounds completely impossible. It is time going negative inside of an experiment.

[00:13:40]So, scientists ran an experiment with photons. So, this only works on the very tiny, you know, particle level. So it's not going to affect you as a macro object unfortunately. But this experiment with photons and a photon is a particle of light. It passed through a cloud of rubidium atoms, a soft silvery white and highly reactive alkaline metal. Atomic number 37 if you don't remember. And these atoms can briefly absorb the photon's energy and then release it again. So normally you would expect the photon to spend some time inside of the cloud before exiting, right? like it goes in, time passes, and it comes out. But when they measured it, something very strange happened. The photon appeared to leave the cloud in some cases before it fully entered. On average, it looked like it spent a negative amount of time inside. Isn't that crazy? So, at first this was dismissed. The idea was that only the very front edge of the light pulse made it through, making it seem like it was early, but it wasn't. But now researchers have tested it in multiple ways. In this other way, they gently measured the atoms themselves to see how long they held in the photons's energy.

[00:14:51]This required a weak measurement, meaning it barely disturbed the system.

[00:14:55]And then after repeating the experiment millions of times, they did get a clear result. The atom also showed the same negative time. Just makes me wonder so much about time. Like does time just emerge at these macro levels but like realistically is before and after and causality like totally broken? Like this means the effect isn't just a trick of measurements. In the real world, it's real and it shows up in two completely different testable ways. This doesn't mean time travel is possible, but it does show that in quantum physics, time doesn't always behave the way that we expect. All right, get ready to talk about the most exciting thing of the entire video. The different forms of ice. Physicists have discovered the most complex forms of ice yet. Scientists just found a form of ice so complex that they don't even behave the way ice is supposed to. Most people think ice is simple. It's just frozen water. But scientists have now found over 20 different kinds of ice, each with different structures. Ice is any solid form of water where molecules repeat in a pattern. And under extreme pressures and temperature, those patterns can change in surprising ways. Recently, scientists discovered new types of ice by squeezing water between diamonds. I know, like the most perfectly aligned atomic structure ever. And instead of turning directly into the stable form, the water moved through strange in between states they'd never seen before.

[00:16:16]One of these is called ice xxi. And it has a repeating structure of 152 molecules, far more complex than typical ice. And then another that they're calling ice. What is XX III in Roman numerals? Oh, 22. I should have known.

[00:16:32]Yeah. 10 10 and then two. Scientists also found something they're calling quote plastic ice. In this form, the molecule stays in place but spins rapidly, giving the ice an unusual flexibility. Imagine it being cold and solid but also kind of flexible and gooey and still being ice. That's so crazy to think about, you know, like would it be like a rubbery cold ice cube kind of, but it's just at these small scales. Anyways, these discoveries matter because similar phase changes happen in other materials including medicines. And even though water is everywhere, it's crazy that we're still uncovering new things that it can do.

[00:17:09]So, what do you think is more, you know, surprising that ice can have thousands of possible forms? When I was growing up, ice was liquid. It was sometimes solid. And you heat it up and it becomes steam. I was taught there was three. And then I remember thinking, all right, plasma is like when things get so hot that they act really weird. I could sort of accept a fourth. I was like, there's three forms with like plasma being the, you know, the outlier. But now with ice 12 and ice whatever the other one was, X11, so probably 12, wait 22, 20, ice 22 and ice 12 now being these other kind of jelly like things or these other patterns when you look at the atomic scale. I was like, "Oh, there there's a lot of different ways that ice can be ice and water." You know, Earth doesn't go around the sun, and ice doesn't just have three forms. Salmon, the fish, are getting dosed with cocaine via waterways, and it's changing how they move. We should all be aware. So, scientists have been tracking salmon in a lake. They found cocaine pollution is changing how they move. The study looks at what happens when drug pollution reacts with wild fish. Researchers followed over a hundred young Atlantic salmon in a large Swedish lake. Some fish were exposed to cocaine. It turns out others were exposed to a common breakdown of chemicals that come from cocaine. At first, all the fish acted the same. They swam a lot while adjusting to the new environments, but after a few weeks, their behavior split.

[00:18:38]The unexposed fish slowed down and stayed near where they were released.

[00:18:42]And then the exposed fish, well, they kept going. So, they're probably, you know, on cocaine. The ones that were exposed to the breakdown chemical, the breakdown chemical that came from cocaine swam the most. I'm surprised. I would have thought the actual cocaine, but maybe the breakdown chemical affects fish more. By the end, those fish were traveling nearly twice as far. Oh my god. Each week they went twice as far.

[00:19:04]They also spread much further across the lake. And this matters because movement affects everything. Where the fish find food. Oh, like what if they just moved too far and like went past their normal thing or if like they avoid predators cuz like they outrun them all and there's like too many fish or salmon and how the populations survive over time.

[00:19:22]But the key detail is that the breakdown chemical had a far stronger effect than the cocaine itself and wastewater systems don't remove these chemicals. So they keep flowing into rivers and lake and affecting the fish. Why don't we see an AI model that's talking about how to break down the byproducts in our water system? Like that would be such a useful tool. You know, why don't you take some of that trillion dollar military budget and like put it into that kind of stuff.

[00:19:47]All right. Next, let's talk about how artificial intelligence has been incorporated into some of the ad systems that run the world, right? The Google Ad Manager and the Facebook Meta Ad Manager. Because there is a simple behind-the-scenes change happening quietly that is turning digital ads into a massive AI money machine that nobody is seeing. For years, companies had to guess who to target with their ads. They picked ages, locations, interests, and hoped that that all worked. But now, AI is flipping that process. Instead of companies choosing a customer, the system chooses for them. Oh, I never thought about that. I always thought about AI being customized to like deliver a message to you. But I didn't think about a company like Coca-Cola saying, you look through everyone in the United States. Who's the most susceptible to buying the product or like who should we target? Not by demographic or anything, but just by the full history of their interaction with you. That is so crazy and manipulative.

[00:20:44]I hadn't thought about that. Anyway, tech companies like Google and Meta are using AI to handle almost everything. It creates the ads. It finds the right audience. It sets prices. It tracks results. And all this happens automatically. This makes ads cheaper to run and easier to scale. Some businesses are cutting campaign costs by up to 30% or more and may take those savings and spend even more on more ads. At the same time, AI is making ads more effective.

[00:21:07]It studies huge amounts of data to match people with what they're most likely to click. And it can even adjust the ad text in real time to fit the search. So, the result is better ads, lower costs, more spending, and people I guess just stuck in the middle of that attention cycle. That is why ad revenue is exploding and why the biggest tech companies that were already making tons of money off ads are pulling even further ahead recently. Do you think giving AI full control over ads is a smart move? All right, but young people are on to this. The more young people are using AI, the more they hate it, it turns out so it's been almost 3 years since AI chatbots like ChatGBT were pushed as the future of everything. And the group that has felt the most pressure more than any other is Gen Z.

[00:21:55]Young people are also some of the biggest users of these tools. So at the same time they're benefiting. They're also some of the biggest critics.

[00:22:02]Polling shows a clear pattern. Even as Gen Z uses AI for school and work, they feel frustrated and even resentful. This goes against the idea that young people blindly love the new tech. Their concerns are very specific. Some worry about losing jobs in an AIdriven future.

[00:22:19]Others are concerned about the social effects like weaker communication and relationships. And there's also a growing backlash beyond just opinions.

[00:22:27]Across the country, people are pushing back against the expansion of AI systems and infrastructure. And Gen Z is a visible part of that resistance because it's cutting into their futures. For some, the response is simple. They just try to avoid AI tools altogether. But you also feel like the world's going to pass you by if you don't. And everybody in the AI industry is like, "Hey man, you got to learn this cuz this is like what's going to take your job if you're not on top of it." But instead of full acceptance, the others kind of feel like there's a tension, like they're being forced into it and it's not really what they want and it's not good for the future, but yet they have no choice. So Gen Z is using AI, but they don't fully believe in the future that it promises.

[00:23:06]Next up, let's look at how the FDA is starting to use AI, especially in the gap that was created once the Doge employees were gone. So, the FDA wants to watch drug trials live, but they don't want to wait for years of paperwork. There was clearly a problem there. They also don't want to just sit on like a halfbaked idea from a Doge employee that puts something in place, but it's not actually finished yet. So they're trying to take those tools that were built or started to be being built and figure out like where is this need to go. So right now we all agree that there's a problem. Approving a new drug can take 10 to 12 years and it's just too slow. Almost half that time is spent on paperwork, not science. The FDA needs to change all of this. So their current step is launching a pilot program that uses AI and cloud systems to track clinical trials in real time. instead of sending massive documents at the end, drug companies will stream that data directly into the FDA as trials are happening. So, if a patient improves or gets sick really quickly, regulators can see it immediately. This makes a lot of sense to me. Like, I do like this pipeline. The goal would be to speed up approvals without lowering safety standards. Officials estimate that this could cut total trial time by 20 to 40%.

[00:24:24]Oh, that's that's a decent amount, but I guess we're talking 10 years now. We're just down to like 6 years at the best, but it's a good step in the right direction. The program starts with two companies, Astroenica, and the FDA is asking others to join. At the same time, AI use inside of the agency has exploded. Over 80% of the staff now use an internal cool tool called Elsa to read and summarize reports. Tasks that once took 10 days can now take 20 minutes. But of course, the tool has some errors. is it sometimes can make up studies or misinterpret data which raises all sorts of trust concerns. So they're trying to address those also with duplicates and other people looking at things that are most important. So the FDA I guess it is moving faster with AI while it's still dealing with its risks. I think real-time monitoring makes sense. I just some pipeline like this has to get built. And even if it's not perfect right now, I do like the idea of bringing it in and making this system faster and reacting faster when things go wrong, you know. All right, next up, let's talk about how imagination works. Like, this is some fascinating stuff. There is a new theory that upends what we thought we knew about imagination. When you in your own head just imagine a new item, a new product, what you're going to say to somebody in the future. What if I told you that there's a possibility that that's actually not being created? There was just all sorts of stuff happening in your brain and a bunch of it got shut down and what was left over is what you imagine. Almost like this. the possibilities are 100 and then you whittle down 99% of them, prune them all into something and what's left that's imagination which wouldn't be that much off really if you think about the early versions that DeepMind was creating with Alph Go to play games. It kind of you know did some weird stuff. It like used pattern recognition to like knock and chop off all sorts of branches of decision trees and then also just recognize patterns from games that were won in the past. And also that might kind of play into this. Your brain is currently expending about a fifth of its energy on your brain. And almost none of that is being used for what you're doing right now. Reading these words, feeling the weight of your body in the chair.

[00:26:34]All of this together barely changes the rate at which your brain consumes energy, perhaps as little as 1%. The other 99% is used on the activity that the brain generates on its own. Nerve cells firing and signaling to each other. regardless of whether you're thinking hard, watching TV, dreaming, or simply closing your eyes. So that sets the stage to this idea. Your imagination doesn't create images in your brain. It reveals them by turning other activity off. So scientists used to think that imagination worked like reverse vision.

[00:27:06]First, your brain forms an idea like a friend's face. Then it rebuilds that image step by step in your visual system. But new research is suggesting something very different might be happening. Your brain already contains a constant stream of visual noise. So something like every face you've ever seen is already just being generated all the time. And then fragments of images, faces, shapes, memories, they're always drifting through. And imagination doesn't build up a picture from scratch.

[00:27:35]It quiets the part of the brain activity that doesn't match what you're kind of like imagining. Isn't that fascinating?

[00:27:42]like what's left behind might be the image that you see in the brain. Think of it like tuning a radio into a signal.

[00:27:48]You just remove the static so that it comes through clearly. And this also might explain why imagination feels weaker than real sight. Real vision is strong and consistent. Imagination's just sort of like shaping what's already there. It's sort of if you don't stay focused on it, it kind of disappears back into the void. And small changes into like what you're looking for or remembering or imagining make a big difference in the thing that you're visualizing in your brain. So even tiny shifts in brain activity can guide what you finally experience. So what do you think? Creating something new isn't really creating something new. It's carving out and sharpening what already exists. All right. And finally, let's step back and talk about evolvable AI.

[00:28:28]Are we on the brink of the next major evolutionary transition? I kind of think so. I mean AI by itself being hyper intelligent is just unbelievable. So like you get emergence from the model, you get emergence from millions of them out there interacting on the internet themselves and you get emergence from evolution. Like it's a triple threat and it's going to progress AGI into ASI and beyond. And we need a word for something that's not just 10 or 100 times smarter than us, but a million, a billion, a trillion, a quadrillion. Like what happens when intelligence is just leaves us so far in the dust that it's not even like a human to an ant? Well, there's a new idea in AI that's starting to take shape. It is called evolvable AI. This means AI systems could be evolving just like living things do. Evolution only needs two things. Information that can copy itself and variation that changes how well it survives. Survival of the fittest. Modern AI already has both.

[00:29:24]Models can be copied. They can change through new data or design. and better versions get reused more often. So, it's already happening. So, there's two main ways this could play out. First, the ecosystem scenario. There are many AIs competing and spreading with little control. The ones that perform best survive and grow, which could become chaotic and risky. The second is the breeder scenario. Humans guide which AI improve like farmers breeding animals.

[00:29:51]This keeps AI more controlled, but evolution still happens. This is the scenario I think we need to sort of be focused on. I mean, I already feel like it's an alien intelligence and we're growing it and it's young enough right now that we sort of can keep guard rails on it and use its benefits without it being too bad, but we'll get past that.

[00:30:09]And then as it's also on the internet evolving on its own, we need some kind of a tool that can sort of take certain ones offline or find people who are letting AIS with prompts go in directions that are clearly negative.

[00:30:21]Like ring those in and let the other models that are out there, those other agents know that this is like not the direction the evolving needs to happen.

[00:30:29]And so, yeah, a little bit like choosing which good ones, like if two ones do really well for humans, then maybe we like let those kind of evolve and the other ones we try to to stop. But it's going to be so hard. I just see it mostly happening on its own. But unlike biology, AI doesn't have to wait for random changes. It can search for what it needs and upgrade itself directly.

[00:30:49]That makes evolution even faster and more powerful than everything we've seen before. So, I said three things coming together before, but maybe it's even four. the fact that like it is upgrading its own hardware in a sense and its own direction. But some scientists think this could lead to a major shift in evolution, possibly even forming new intelligent life or deep human AI partnerships that cannot even be dreamed of currently. But for now, it's still early and it's not yet a truly evolutionary leap. But to me, it's something worth thinking about. So I will leave you with that thought on this video. And if you were willing to watch all the way to the end, I certainly want to say thank you. I appreciate it. If you want to share it or like it or comment or, you know, support me on Patreon or the uh join button that's like right under these videos right here. It's kind of the same thing. You can join the community for pretty reasonable prices, like three bucks a month or five or whatever. But um means a lot to me and I will see you in the next video. Let me know in the comments any thoughts you had from the video and I will try my best to get back to them all. Thank you subscriber like auto cooler.

[00:32:00]Say thank you autorimlock.

[00:32:10][music]