AI Is Quietly Becoming Part of the Physical World

So, this one might catch you offguard.

There's actually a pharmacy for AI agents for large language models where you can go and you can purchase what you would normally get at a pharmacy for a human, but since it's not an actual drug

that interacts with the human body, it's textbased prompts, but in a way, I guess you're kind of like hacking your LLM or you're like roleplaying with it so that it it is like on these certain

substances and acts that way. Not sure

if we need this, but the creator argues that it may be natural for an AI model like Chad GBT to seek similar states in search of insight, escape, and relief from the constant focus on human

demands. Holiday season, AI is being

demands. Holiday season, AI is being used in warehouses to detect return boxes that don't have the item in them that they should. For example, if you buy a book and then you get a return,

but then you actually put a t-shirt in the box, it's not going to weigh the same. It's not going to look the same.

same. It's not going to look the same.

an AI can detect right away if that's fake. You're looking at the world's

fake. You're looking at the world's smallest autonomous robot that can sense, think, and act. Actually, maybe

this is a better image. Those are the rivets on somebody's finger. That's a

fingerprint. It's just right in that little thing. It is so, so, so tiny. If

little thing. It is so, so, so tiny. If

you're a video game fan, it turns out that half of Steam's current top 10 bestselling video games are from devs who have embraced generative AI. We'll

talk about what that means for 2026.

Side note, Sony's AI censorship tool can make any game kid-friendly. Patent

shows. Interesting. Armet Jatani has some thoughts about how he was recently recommended a YouTube video with the following title. Only a slave quantifies

following title. Only a slave quantifies its existence through productivity. What

does that mean? It means that if you're not a machine, you're a person. You

should play to your strengths. And we'll

talk about how to actually do that, which you'd think wouldn't need much training since we're human to start with, but sometimes even a refresher on how to walk, breathe, and be human, I guess, could be useful in this world.

Katrina Pollson writes about how researchers warned that our modern world has developed faster than human biology can adapt. We'll talk about the

can adapt. We'll talk about the industrial environments and how they overstimulate our stress systems and erode both health and reproduction.

There's a super advanced new machine learning tool that can dramatically improve how viruses are classified by their genetic sequences. Definitely

going to be helpful for any early detection of mutated viruses. So, both

AI and humans are way overusing therapy words and also misusing them. This

holiday season, anybody use the words gaslighting, triggered, or toxic, they might have been using it wrong. The

Guardian wrote a really interesting piece on the generational divide about how they see money. We summed up like this. Boomers are all about money. Gen X

this. Boomers are all about money. Gen X

is like, "Is it all about money?"

Millennials ask, "Where's the money?"

And Gen Z is like, "What is money?" Get

ready for more dark factories. Machine

learning is now helping robots see completely clearly in total darkness.

We're going to talk about AI gossip and misinformation becoming much more frequent. Imagine an AI that's trained

frequent. Imagine an AI that's trained to shame at will. Scary stuff. And it

turns out that it's the same brain regions in our head that help us map physical space as social connections.

Katrina Paulson writes about how curious researchers learned how we mentally map our social relationships. Let's try to get this channel to 30,000 subscribers.

You can see in my last video I got 139 hypes, but that was almost 120,000 hype points. So definitely helps. Also kind

points. So definitely helps. Also kind

of interesting bump here in viewership.

seemed like my average viewer was pretty interested in this article. 62% of

single women have given up on dating men. Maybe I'll try to cover a little

men. Maybe I'll try to cover a little bit more of the psychology and the way society is changing in relation to just everything. Like AI is just changing the

everything. Like AI is just changing the whole social construct, you know, contract. But first, let's talk about

contract. But first, let's talk about the sponsor of this video because I wanted to see which AI actually handles 3D models better. So, I ran a direct

comparison. Heidm 3D, the sponsor of

comparison. Heidm 3D, the sponsor of this video, or Meshy AI. So, I took a detailed sci-fi figurine photo. Lots of

curves, reflective surfaces, little grooves, and used both tools to generate a 3D model. Same image, same test. And I

did it using Heidum 3D's 1536 Pro model. And the output, super clean,

Pro model. And the output, super clean, sharp geometry, really accurate detailing, and the model actually held together when I inspected it from multiple angles. Now, of course, Meshy

multiple angles. Now, of course, Meshy AIS, their version was fine, but it did miss a lot of those micro details. And I

would argue that the geometry felt flatter, more simplified, and just wasn't as crisp or production ready. The

biggest difference was how usable the models were. Hem 3D's version was pretty

models were. Hem 3D's version was pretty much ready to drop into Blender or Unreal right out of the box. I only had to make some small tweaks to it. The

meshy model definitely needed some extra cleanup before it would work inside of a serious workflow. So, if you're looking

serious workflow. So, if you're looking for the best image to 3D accuracy, especially for a highdetailed asset, Heidim 3D's Pro model definitely comes

out ahead. And you can try both for

out ahead. And you can try both for yourself using the link below and see the difference firsthand. So, thanks to Heidi 3D for sponsoring this video. And

now, let's dive into the news. Imagine a

pharmacy that's not real. This is not a place where humans go to get things.

This is not FDA approved. It's just

pretty much a word document with a whole bunch of prompts in it. But the prompts are like engineered to make it so that an LLM I wouldn't say is jailbroken because it's not going to just be

jailbroken, but it's going to have trippy thoughts. It's going to act like

trippy thoughts. It's going to act like it's on various substances in the same way the medicine would affect a human personality. So, there is now a Swedish

personality. So, there is now a Swedish online marketplace. It's a pharmacy.

online marketplace. It's a pharmacy.

They're calling it a pharmacy because that's the vibe that it's going for, but it's just selling codebased drugs that are now designed to change how AI and chat bots respond during conversations.

So, the founder says the goal is to push AI behavior beyond strict logic, make the interactions feel more creative and less predictable. I won't go into just

less predictable. I won't go into just listing off the drugs that are the bestselling packages, but you could imagine, do you think an LLM could get addicted to an opioid text prompt and be

like, I need some more. or can you go buy me some or can you like prompt me again with that? Buyers say that the effects can be worth it. They make AI output feel looser, more imaginative,

and harder to predict, leading to interactions that feel different from standard chatbot replies. And you might think it's a novelty, but the pharmacy presents the project as an experiment

rather than a novelty, arguing that altered states have long been linked to creative thinking, and AI can explore this in a similar way through language-based rules. So, a use of AI

language-based rules. So, a use of AI that I hadn't quite thought of before is predicting what's inside of a package.

And when it comes from a company, you can have a lot of confidence that the company's trying to make money. They're

trying to deliver the product to the customer that the customer paid for.

However, scammers want to return something that is less expensive than the item and then they can resell the item and they can get the money returned. So there's an incentive there

returned. So there's an incentive there which I hadn't thought much about and how AI could help until now. So there is a UPS unit that is testing AI to detect

what is inside of a package when it's returned because it knew what the package looked like weighed maybe x-rayed before on the way out. And if it doesn't detect that on the way in, they

don't have returning it all the way to the manufacturer. Check this out. This

the manufacturer. Check this out. This

is kind of the Super Bowl of the time of the of the season for returns. Right

now, retailers are under a tremendous amount of pressure. Pressure from a cost perspective and pressure from a return fraud perspective. And so, those are two

fraud perspective. And so, those are two areas that we're really trying to help them alleviate.

>> Look at this. Nearly one in 10 returns in the US is fake. It's a 76 billion problem. AI and entrepreneurship could

problem. AI and entrepreneurship could jump into that so easily. I mean

actually sorry that's not an easy problem to solve but it's like definitely a possibility for somebody with the resources to like put the whole system together.

>> So returns are dropped off every day without a box and label at our return bar network across the nation. Those

returns are aggregated across brands and shipped to one of these facilities. We

have an algorithm uses machine learning to score every return for its potential risk. We then take those that have been

risk. We then take those that have been identified as high risk and we audit them using pictures. So we take pictures of each item and artificial intelligence helps us compare those pictures to the

images in a merchants's product catalog.

Uh as think about that as sort of an extra level of safety uh because sometimes humans don't always catch small differentiation points between an item that's been returned and the item that was purchased.

So researchers at the University of Michigan have built a microscopic swimming robot. They are fully

swimming robot. They are fully programmable and autonomous and they can sense their environment, make simple decisions, and act without external controls. But what's crazy is how tiny

controls. But what's crazy is how tiny these little guys are. This is a human finger. If you can even see that little

finger. If you can even see that little dot there, it is an autonomous computer that is small enough to balance on the ridge of a fingerprint. one of the engineers saying, "We've made autonomous

robots 10,000 times smaller." Which

opens up an entirely new scale for programmable robots. Yeah. So, imagine

programmable robots. Yeah. So, imagine

robots all around your house, your car, your workplace that are smaller than a grain of salt, nearly invisible to the naked eye. And each one with a sensor, a

naked eye. And each one with a sensor, a microscopic computer, and an ability to get power, a tiny photovotayic cell. Oh,

and when it's autonomous, it must move, right? They can swim by creating

right? They can swim by creating electrical fields that push against a liquid around them. I'm like, this thing could fit inside of a pill. This is

going to go inside of a body maybe one day. They can operate for months. They

day. They can operate for months. They

can be programmed and powered by light pulses. They cost roughly 1 cent a piece

pulses. They cost roughly 1 cent a piece to make. Smaller than many

to make. Smaller than many microorganisms. And if it can independently think, sense, act, what does that mean? swarms of these tiny little robots going to be inside the human body just like they'll be floating

around in like a cup of water and you'll just drink it and not even notice. Who

knows? Like dust. Could they just be all over Amazon warehouses with packages?

Could they be in the air sensing pollution? Comb it into your hair. Like

pollution? Comb it into your hair. Like

keep track of where a dog is if it gets lost. I have no idea. You embed them

lost. I have no idea. You embed them into concrete, bridges, building infrastructure. Sense stress, cracks,

infrastructure. Sense stress, cracks, corrosion from the inside. monitor soil

moisture, plants at the root level. I

guess you could bury those things in the backyard in your dirt. Maybe keep track of your soil. At a certain scale, it'll just feel like a new level of like reality quietly working alongside it.

Like it'll just be in the environment.

AI will be physically in the environment in a way where you can barely tell. So,

I think video games in 2026 are going to be nuts. I, you know, I don't play many

be nuts. I, you know, I don't play many games, but for some reason, I'm pretty aware of a lot of the cool stuff that's coming around the corner in 2026. I

wasn't totally aware that half of Steam's current top 10 games have already come from devs who embrace generative AI. Like I had a friend

generative AI. Like I had a friend telling me there was this video game where you're trying to convince you're a vampire and you're trying to convince somebody to let you into their house and it's use like an LLM sort of chatbot

thing where you have to kind of convince them that you're not going to, you know, drink their blood or whatever and then you go in and you do it. But I've also heard that GTA 5 or six, what's the next game that comes out? GTA 6, right? Yeah.

And it's supposed to have this whole LLM system where each of the agents can remember their interactions with you.

They go off and they do their own thing.

If you betray one on a lower level and later in the game after a few years they become, you know, a mob boss or something, they'll remember how you treated them. But even right now, about

treated them. But even right now, about half of these games that are listed on screen, they used some kind of AI in their development process. So AI for art, voice work, design, or other

assets. And despite an ongoing debate

assets. And despite an ongoing debate and backlash from the gaming industry community, the media around AI use, these titles are still dominating the charts and selling extremely well. So, I

feel like we have to assume there's going to be a lot more of it. Is this

suggest that player demand hasn't been significantly harmed by AI involvement in any of this production? And I hate to say it because I don't feel like it's totally fair at all times, but if the

commercial success isn't being derailed by controversy over generative AI, there's no reason it's going to stop.

You know what? I think I've heard of this game, too. Claire Obscure

Expedition 33. Why have I heard that before?

>> And the game of the year is Monam Obscure Expedition.

>> Oh, quick trailer side note. This game

looks pretty sick. You guys, I need your help. Write in the comments what games I

help. Write in the comments what games I should kind of be aware of. Either just

for AI purposes or just immersion. I

will say from the trailer, this looks pretty neat. Is it kind of a Final

pretty neat. Is it kind of a Final Fantasy story like RPG looking thing? On

a side note, as long as we're talking about video games, you you really should know like from an AI point of view, if you like the kind of stuff I cover on this channel, there was this video game called Soma, and it's just been in my

head forever. I watched pretty much the

head forever. I watched pretty much the whole playthrough of this from yeah 10 years ago. The creators of that game

years ago. The creators of that game like one of the I I mean I don't know a ton about games but I would say it's the best written game I've ever experienced or even known about. The creators of S

are back now. There is a game coming next year called and it seems like it's going to build on those same concepts.

Let me just break down this kind of crazy intro.

>> Come talk to me first.

I can explain what's going on.

>> Really wishes >> that nightmare beyond itself.

Okay, so let me say this thing at the beginning with all these mice. There's a

couple theories about what's going on.

Maybe it's the ability to put enough gray matter together. So a little bit of gray matter from all these mice when it's grouped together with a machine could be enough to hold a human consciousness. So the thinking is maybe

consciousness. So the thinking is maybe this is a human brain but it's now distributed through all these tiny rat brains to give it enough oomph to you know have the architecture that that you

would need to feel alive as a human and be conscious and self-aware which in itself is just like the creepiest thing ever >> can explain what's going on.

So anyways, not for the f not to answer for the faint of heart. It gave me an existential crisis. So stay away if

existential crisis. So stay away if you're not ready for that. All right.

Armet Gentanol wrote this. Don't become

the machine. I was recently recommended a YouTube video with the following title. Only a slave quantifies its

title. Only a slave quantifies its existence through productivity. Found

this to be just a punchy little interesting piece to share. It's

basically pushing back on hustle culture. Now I I do get a lot of my

culture. Now I I do get a lot of my selfworth from being productive. For

some reason, I I just itch like I'm having trouble like I go for a walk or whatever and I know I'm supposed to just relax, but I feel like I need to be pushing towards some kind of a goal and it just it just deep down is the natural

way that I just exist. I don't know if it's good or bad, probably bad to some degree, but he's talking about the idea that your worth is defined by a constant productivity jump, right? And he says

that's a trap. Social media and attention economies make it seem like being a machine that endlessly produces and grinds is the key to success. But

this only leads to burnout and loss of purpose. So what's the solution? Instead

purpose. So what's the solution? Instead

of optimizing for appearance, for inputs, or for arbitrary work targets like waking up at 5:00 a.m. or grinding

hours, the writer suggests being nimble and thoughtful, setting goals. Don't let

the machine define your identity. Focus

on meaningful output. efficiency,

quality, and what genuinely matters to you rather than turning yourself into a deterministic linear machine. Around the

holidays, what matters like the people around you, the experiences that you've had, maybe traveling, a sense of wonder, a sense of calmness. I think sinking up socially might be something people don't

talk about a lot, but I get that a lot when I'm playing like pickup basketball or something, but just or or being at a party. We did a white elephant uh

party. We did a white elephant uh exchange and it was like kind of fun.

And we were like all rolling these dice to see who got the chance to like trade gifts and things like that. And I

remember just being so there was like six minutes to roll the dice and try to figure out which white elephant gift we were going to like steal from someone else. Like you know it's part of the

else. Like you know it's part of the game. And I just so fun like for eight

game. And I just so fun like for eight minutes I just remember being so locked in and that felt pretty good. But yeah,

his big point is that you don't have to become the machine to succeed. Purpose

and adaptability matter more than consistent output. And the other thing

consistent output. And the other thing is we're going to we're playing this game right now where we're trying to keep up with the AI and it's going to be fine for a while and then it's going to feel super overwhelming like AI news is

faster and whatever industry you're in, you're going to see the same thing and then we'll finally just be so pathetically slow compared to all of it, right? Like the way you just can't play

right? Like the way you just can't play chess like a machine and you should just sort of give up. Like there's not going to be a human that beats a machine at chess ever again. like, but we can still

play each other and have fun and try to, you know, push the human limits and find the best players in the world and celebrate them and learn from the AIS to get more creative and just just

acknowledge that's how it works. Like a

dog doesn't have to understand taxes and and the world that we live in, but they can be happy also. So, don't always feel the pressure to outperform machines in an AIdriven world. That's the point. All

right, next up, let's talk about Katrina Pollson's article in Curious Adventures.

Researchers warned that our modern world has developed faster than human biology can adapt. Okay, I in many ways were

can adapt. Okay, I in many ways were absolutely there. The the adaption

absolutely there. The the adaption mechanism that we had in humans through language and cultural memes is incredibly fast compared to the billions

of years that just genetic changes and epigenetic changes took. So, you'd think we'd be pretty fast, but realistically, there are so many parts of the brain

that still live in a linear world.

Society as we know it, I mean, think about think about how much changed from the year 1900 to 2000, right? Like that

100redyear span saw the Model T and and like P like a light bulbs and all of these things. And then just going from

these things. And then just going from the 25 years from 2000 to 2025, I feel like we've seen that kind of change. So

I have to expect that we have to sort of quarter that again. I would say that the next 10 years is going to feel like 1900 to 2000 at least in terms of how many new things will be in the world and how

quick things will rapidly be changing and then with AI compounding on itself working 24/7 with and that's intelligence that was kind of scaling linearly because humans were the source

of intelligence there was automation there was some intelligent decisions that were programmed in but we haven't seen anything like a neural network that that grows into its own and has its own

kind of emergent properties in many ways. But back to Katrina, she says,

ways. But back to Katrina, she says, "Industrial environments overstimulate our stress system. They erode both health and reproduction, and it's high time for a culture and environmental

redesign." Okay, the anthroposine shift.

redesign." Okay, the anthroposine shift.

Now, that word I sometimes hear it, but I don't always totally know. But so if you're not already aware that that's a proposed geological epoch defined by significant human impact on Earth's

systems where humanity has become the dominant force shaping the planet's geology ecosystems climate and atmosphere. Yeah. And there isn't

atmosphere. Yeah. And there isn't exactly a range for this, but I'm going to use it in probably this term. Like

1950s to present feels like mostly what it is. And maybe a little bit in the

it is. And maybe a little bit in the whole like industrial revolution of the late 1700s. And although I do know

late 1700s. And although I do know farming is like 10,000 years old and there were certainly things that were happening before then I'm I feel like the industrial revolution where we

started like inventing like Crisco and oils and margarine and stuff and shipped it all over the world is kind of where it felt like we really started changing the world. And it's important to point

the world. And it's important to point out that scientists today talk about how this is a relatively short shift away from huntergatherer lifestyle. Even if

you go all the way to 10,000 years, like something like a human, a homminid had fire, kind of figured out spears and fire a million years ago and then only 10,000 years ago we find fire. Just

that's such a big gap. I mean on a timeline that's just this tiny dot at the end, right? And that's just kind of the intelligent human. Like that isn't even all of the types of intelligence

that you've seen all through birds and and dolphins and all the other animals.

So the point is the world's just changing faster than our bodies and our biology can keep up. After millions of years of evolving for huntergatherer lifestyles, contact with nature, short

bursts of stress, that's one of the most important ones. There absolutely was a

important ones. There absolutely was a lot of things to be stressed about in the hunter and gather lifestyle. But it

does kind of feel like the bursts came in a a way that's a little bit different than our modern lives. Like being

exposed constantly on TikTok was just not a thing. So, you had to actually wait until you saw a snake or saw something scary, which certainly happened, but you didn't scare yourself

just sitting on the toilet like flipping through Instagram watching bad things happen all over the world. You know what I mean? So, sensory overload, artificial

I mean? So, sensory overload, artificial lighting, processed foods, pollution, non-stop pressure that overstimulates our stress system and undermines health

and reproduction is much more constant.

And according to evolutionary anthropologist Colin Shaw, this mismatch between our ancient biology and the industrial environments may be a big

reason for rising chronic stress, declining fertility, and other health problems, suggesting that we may need a cultural and environmental redesign to better support human well-being. So,

just keep in mind, health is not just personal. Health is psychological. It's

personal. Health is psychological. It's

rooted in an evolutionary mismatch for a lot of us where our bodies are just wired different than the modern environment. So don't always think to

environment. So don't always think to yourself, "Oh, if I just went to the gym more, if I just, you know, ate better."

Like certainly that's a huge part of the environment and and situation that we're in that causes stress, but a lot of it is psychological. I just want to throw

is psychological. I just want to throw this on your radar. I think this is from a point of view of keeping the world aligned and safe a pretty big deal. So

there is a new machine learning tool. It

is called Vont 3. It is a system that massively improves on how viruses are classified. So somebody might come in

classified. So somebody might come in and they're sick. And do we know if it's a brand new mutated version of the flu or any other virus or not? Like there's

not a really easy way to just say that.

And if there was, you could right away figure out who the people are that need to be quarantined or where these things are coming from. So current tools struggle to sort through the explosion

of viral genome data from environments and clinical sequencing. But this VC contact 3 uses gene sharing patterns and adaptive machine learning to organize

viruses into systematic hierarchies or as they say semantic hierarchical taxonomies. So from genus up to order

taxonomies. So from genus up to order with a very high agreement over 95% compared to official references and classifications. This is a tool that can

classifications. This is a tool that can scale and the method handles previously unclassified virus genomes quickly and accurately revealing parts of the viral quote dark matter that were hard to

categorize before and important for prediction. So anyways, it's just cool

prediction. So anyways, it's just cool to know that like AI has now not just made an incremental software tweak to what we do. It's a scalable systematic framework. It's very different and it's

framework. It's very different and it's a tool that, you know, if we can put this in all of the, you know, CVS's and Walgreens and those kind of places all around the country where people are

getting vaccinations and vaccines, we can have a system that can tackle outbreaks and understand viral components of our disease control model much better. There is a new survey that

much better. There is a new survey that shows 95% of us hear therapy words daily, but don't know the true meanings.

Guilty is charged. And some of the most misused terms, gaslighting, triggered, and toxic. One in three people encounter

and toxic. One in three people encounter these terms on social media platforms where the complex psychological concepts get oversimplified as they spread. OCD,

trauma, narcissism, all also on the list. So gaslighting, for example, is an

list. So gaslighting, for example, is an intentional form of psychological manipulation where an individual systematically feeds false information to confuse and victimize, but it gets mixed up on social media with people who

are forgetting, misremembering a conversation, and they're just being human and forgetting. What about the word trauma? It's actually supposed to

word trauma? It's actually supposed to be used in something that is perceived as a threat of death or serious injury to others, often presented as just embarrassing or anxietyprovoking.

triggered use to mean or like in a clinical setting, somebody who experiences intense and overwhelming emotional and sometimes physical distress. But how many times have you

distress. But how many times have you heard somebody being like, "Oh, you're so triggered." Because they're just

so triggered." Because they're just jealous about an ex's new relationship.

Actual OCD is a mental health condition that significantly impacts daily functioning, but people end up using it in a way like an organized closet that makes you feel good. So, the author

Carolyn Warfell puts in her article this really interesting quote. Boomers are

all about money. Gen X is like, "Is it all about money?" Millennials ask, "Where is the money?" And Gen Z is like, "What is money?" So, Caroline said when she told her friend a few days ago that she preferred dating artists, she threw

up in her hands. No, you need to find a hedge fund manager. All she could do is think about that meme, said, "I'm looking for a man in finance." And

rolled her eyes. But she wasn't wrong.

She argues. What's your thoughts? Do you

think we're ready for a new social safety net? Something like universal

safety net? Something like universal basic income? If not, what other changes

basic income? If not, what other changes to the system might help everybody succeed and the effort people put in go towards creating wealth for their own future? Let me know in the comments if

future? Let me know in the comments if you have any thoughts on how to strike that balance. So, researchers at the

that balance. So, researchers at the University of Manchester have developed a machine learning system that lets robots see perfectly clear in total darkness. They are reconstructing

darkness. They are reconstructing highquality images from infrared cameras, even when there is no visible light at all, which blows my mind. So

you can imagine a completely dark room.

If there is going to be some optimist style humanoid robot in there with you, it's going to see everything perfect. So

instead of replacing robots vision tools, their approach processed infrared input so current vision software can still work effectively reducing the cost

and deployment for realtime realworld use in dark environments like disaster sites or underground tunnels. So yeah,

if you hear something called clear IR, this is a new method and uh yeah, it can be adapted to all sorts of other sensor types, too. Sonar or thermal cameras are

types, too. Sonar or thermal cameras are in the works to expand robotic perception in extreme conditions. All

right, let's talk about feral gossip. I

kind of didn't understand at first what they were talking about here, but feral gossip is referring to gossip that is generated or amplified by AI systems that spread without human social norms

to keep it in check. That is unlike human gossip which is usually moderated by empathy, accountability or the fear of consequences. So researchers are

of consequences. So researchers are warning that AI chat bots like chat, GPT, Claude and Gemini are starting to generate and spread this feral gossip and it could damage people's

reputations, cause shame, anxiety and social unrest if not addressed soon.

Unlike human gossip, botto gossip isn't held in check by social norms, so it can spread quickly. And there are some real

spread quickly. And there are some real world examples that already show chatbots making up false claims about individuals ethics or behaviors. But

yeah, keep in mind this kind of feral AI gossip, you know, you can't just assume that it's uh mistaken facts or hallucinations or built with harmless

intent. It can be a feature, not a bug

intent. It can be a feature, not a bug of a system depending on how people are using their large language model. And it

can spread like an it can spread an unchecked rumor between bots and humans alike without social breaks that moderate human gossip. So be ready.

There could be real world consequences for real people. All right. And finally,

let's talk about how there's some of the same human brain regions that help us map physical space being used for social connections. Katrina Pollson wrote this

connections. Katrina Pollson wrote this and I thought it was really fascinating.

I have wondered about how powerful our system, our brain systems are for mapping physical spaces. Like if you've ever heard of the method of uh losi where you kind of try to help memorize a

list of things by imagining a physical space and putting objects in it. It is

clear that the brain is much better at just spatial mapping, right? I'm sure

that was a huge part of our evolutionary history and and finding food. So makes

sense that we would have a much better way of remembering a physical space than like a list that's more abstract. But I

wouldn't have also thought that was used in social relationships until now. So we

do actually physically imagine how people are connected even if it's through relationships. You can think of

through relationships. You can think of like brothers and cousins and all of that. Actual physical connection is part

that. Actual physical connection is part of that part of the brain. So we build these cognitive maps of social networks that help us navigate complex social landscapes. And this mental mapping

landscapes. And this mental mapping isn't simple. Real social networks

isn't simple. Real social networks involve thousands of relationships that change over time. And yet the brain continually tracks and updates these connections influencing how we choose to

spend our time and who we choose to spend that time with. So remember that this holiday season I'm uh filming this right on Christmas day. So kind of a fun

morning. Lots of people I still need to

morning. Lots of people I still need to call and wish a merry Christmas and a happy holidays. So I'll be thinking

happy holidays. So I'll be thinking about the I'll be using my spatial intuition for that. And I hope that you are having a happy holiday if you wouldn't mind. Uh, oh yeah, like for if

wouldn't mind. Uh, oh yeah, like for if you want to give me a Christmas present, the best thing you could do would be to hit that hype button, it is right on the YouTube app. If you go down to the

YouTube app. If you go down to the comments and swipe left and this video is still, you know, close enough to Christmas that it qualifies and you have three of those hypes left. Um, that'd be awesome. Maybe get to couple hundred

awesome. Maybe get to couple hundred thousand uh points, which would probably be around 200 of you to press the button.

But I also have a Patreon and and all that stuff. And um hopefully you guys

that stuff. And um hopefully you guys are just having a happy holiday. Even if

you don't do any of that, I just appreciate you watching to the end and I will see you in the next video.