L.O.G. 188 Assimilation

I think you're underselling our knowledge on the topic from one standpoint, right? So you and I spent almost two years working on a DARPA (Defense Advanced Research Projects Agency) project for Ursa, and it's supervised autonomy. So our history with, I know you think about that for a minute, but look up the program and you'll see what we had to do. So the three basic machine learning paradigms are supervised, unsupervised, and then reinforcement learning. Right? So why is that important? Because in reinforcement learning, the focus is always on finding a balance between exploration of uncharted territory and then exploration or exploitation, rather, of our current knowledge state, which is hugely important in what we do.

The "thing" versus the "think."

Yes. So your comments yesterday got me excited because the accuracy was being criticized. And remember, in any form of social exploration, the only real way to collect information is to interact with it.

Yeah.

You've been with me long enough, you know, that's exactly how I learn. I go into the environment, I test the environment, right? I'm asking people questions all the time, and then I'm updating my current models constantly. So so what's the difference? So my thing was, when we started talking about that, we were having a, not a dilemma, you and I were talking about what would be, how would we bring this into human behavior pattern recognition analysis? And my thing is that we assimilate parts of culture and environment and sociology and psychology and biology all the time. And so what scientists are trying to do through these literally a limited objective experiment with language. Okay? And known, look, a computer has knowns and unknowns. Yeah. So we use knowns and unknowns, and what do we do? We put them against the baseline, and that gives us the logical start point for an intervention. Quick interaction to learn or understand something that you're unsure about, don't know, doing the head tilt. Little kids encountering for the first time, yeah, encountering for the first couple of times.

Okay, so so when we assimilate, we don't always assimilate correctly, and that's what struck me as important about your comment. So so let me let me lead you, folks, that are listening and watching. So Brian is saying that the people that are bitching about it or complaining about it or making comments on it are saying that, "Hey, there's a lot of inaccuracies." Okay? So you're a little kid walking around in your environment, and your parents are trying to teach you, and they see your dog, and they go, "Dog, doggy, dog." For the rest of that semester that you're learning as the little kid at home, anything that's furry with four legs is a dog. Right? Okay. So now, the little bear cub who comes walking in, or you see on TV, do you get what I'm trying to say? A thing, and that becomes a doggy because it's on four legs and it's close enough where cognitively close enough. So we value cognitive flexibility, but our environment has to nurture us. How does that happen? Because your neighbor goes, "Now you little buckethead, that's not a dog, that's a goat." Do you get what I'm trying to say? And now we learn, "Wait a minute, okay, dogs and goats." So everything is doggy or a little baby goat. Then all of a sudden my mom or dad, you know, the grandparents, come in and go, "No, no, that's that's not a goat, that's a cow." Okay? So what are we doing? We're building, and we're starting to create this file folder. So now we've got smaller, we've got smellier, we've got cute, we've got scary. Do you see what I'm trying to say? So so what they're attempting to do with their Greg bot and Brian bot is exactly what we do with the human brain. We just protracted and over time, and from genetic material passed on through the birth process, what we do then is we create sort of this, we create sort of a profile, but we create a picture using attributes.

Right? So hairy, four legs. Okay, well, hairy and four legs covers the dog and the goat. But those are two completely different things, and many of the things will, but they don't need to be at first. You're right. In order to get nuance, in order to arrive at the most correct answer or the most logical, I have to be able to create enough attributes, but then what my brain likes to do, because it'll create different conceptual frameworks not unlike the AI process to grab from those and I can go, "So some of the attributes which correctly describe a dog also correctly describe a bear, a goat, all kinds of different animals." Yes. And now what accounted for the mistake in the example that you gave about, "It's a fish, it's a mammal," it's... Okay, the idea is, based on the information that I have at this point, the best decision I can make is, "Yes, this is it."

Yeah, but it doesn't fit this criterion. Yeah, but the availability heuristic is closing off. Do you get what I'm trying to say? That I'm going to glom onto it, and it doesn't seem like that's going to be a large enough mistake that it's going to send us spiraling off into the sun. And that's exactly what happens to that kid. The kid now, it's close enough, it's a dog. Okay, so everything like that is a doggy. So should I go up and pet that wolverine while Mom and Dad are taking me out of hiking? Probably not. So now we have to have danger cues that are associated with it. See, so our entire research of the environment—nuanced, novel, scary, all that other stuff—is how our brain, how our God through the Vishnu Allah decided to make us investigate our environment. Yeah. And then assimilate what? Assimilate best practices of any environment.

Brian, you and I talk about this all the time. If you go to Mexico and have a little rolled-up meat and cheese sandwich, yeah, it's called one thing in Mexico City. If we were in Hong Kong or Beijing or Saudi Arabia, it's called a different thing. But it's the same food, it's almost the same way. It meets certain meats in sauces on some type of bread-type product, flour-type product. What about noodles if we follow noodles across the world? Or you know, all of these things that, you know, the nature of candy, for example, what happens is sociologically, look, psychologically, assimilation means a cognitive process of making new information fit with your existing picture of the world. We're really good at that, and that's why DARPA reached out for us. Biology just means, "Hey, I absorb the food product, and the best of it stays with me. I poop out the rest." But sociologically, anthropology says that when I assimilate something, I put things, the best of from a group, together, and I claim it as my own. That's why societies fight for years and years. "Oh, no, no, that's our Polish weather person." Yeah, yeah, yeah.

It's like, even, even, who knows, besides the religious, you know, differences of like the Israelis and Palestinians and all that, but like they go over the fight, you know, the shawarma and who did what and where. Exactly. And it's hysterical because you're like, "Oh my God, you guys are eating the same." "No, no, this is the right." And it's just, that's a perfect example of how we do that with food.

It is. But the issue, or not issue, but kind of part of the reason we're talking about that, is you when you're talking about assimilation, like you just said, it, I have to take this information and fit it into my current interpretation of reality. So so mine, so one, that means it's inherently subjective.

Uh, and two, and biased. Critically biased.

And I'm fitting, yeah, I'm fitting it into what I know rather than, rather than, you know, creating some new schema or changing the way I interpret my environment. I, all I'm doing is saying, "Now that that's that can be good or bad." Right? So so yeah, can we talk about that from the learning, from a learning process, that means it, it's, it's, it's simple, but sometimes simple is good. Like we, we keep it for decision-making. Sometimes as, "Okay, I'm in a new environment, Greg. You just threw me out of a plane over the Central African Republic. Never even been to the continent, let alone that exact place." I have to assimilate quickly or I may die. Actually, the environment may kill me if I don't understand. So so rather than having to have a guide that says, "Well, let me show you every different type of poisonous plant and animal here, Brian," which is unsustainable.

Brian and I understand how much would I have to possibly learn, too?

Exactly, right. And this tribe does things this way, and that one does it over here, and this one's friendly, and this one. Like, I have to be able, I can't take the time to do that, right? So so assimilation helps with me rapidly understanding the context of the situation that I'm in to make decisions. But but it goes back to the to the dog or the goat. Like I I might say it's a dog, but no, it's a goat, but but it may work for me in that situation. Is that is that right?

You're spot on. And two important things that we could draw from what you just said, and many important things, but two primarily. Let's go back to U.S. Navy SEAL and great friend John Privatera. Prep. Okay. Prep walked away, and somebody asked him a spontaneous hallway comment, "What's the best part of this training?" And he says, "We focus on different things. We were trained our entire life to focus on the difference, Greg. In this training, focuses on similarities." Why is that important? Because using your example, we have to use that same thing as carbon-based lifeforms when we go to Mars, when we go to the Moon. Right? We have to learn this novel environment. Yeah. And how do we learn that? By skinning our knees, by investigating, by experimenting with it. But your brain has certain things that it's predisposed to experiment with in its environment. Why? To keep you alive long enough to learn so you can update your file. Right? You get where I'm going there? Yeah.

I can give you a quick example of that with your eyes. Visual field, visual acuity is hugely important. The entire backside of your skull is involved in you creating in pixelation from stimulation of light, motion, and edges. So your two types of receptors that are most important in in your retina are your rods and cones, and they work together to take available light and turn it into an electrical signal that your visual processing center can can read. Now, why are there so many rods and so many cones? Because color cones, the big lettered C, is for daylight, and we need better acuity. Why? Because we have to put a hook on a line to catch a fish. Right? You have to put the perndl in to back out of the driveway. So those type of things in in ambient light situations are nuanced to the point that we have to have our fingers and our hands moving and doing everything else. Nighttime, we weren't designed to go outside. Nighttime was that scary place where we heard the, you know, we had the fire in the cave. We didn't want to go up. So we had just enough information that was coming in from our rods so we could determine that's likely an exit of the cave. This is likely a creature moving slowly through the environment. Right? Why? Because it was less light available, less stimulation to the brain, and that's why when you take a look, rods and cones are unequally distributed in the back of your skull for a very big reason, except where? Except the fovea centralis. Why? They're totally rod-free because you ain't got F-in' time for, Brian. Right? You have to have that high visual acuity right now to throw that spear at that fish and adjust for where the fish actually is. And your brain does that without asking you a question. So that goes right back to you. You use an example of a very good example, like 25 broadcasts ago, about how the batter never sees the ball when it leaves the thing, and you have to anticipate. And it was a great show about that. Those type of things are survival strategies where we assimilate almost all of the information, but some of it is left on the editing room floor. Why? Because your brain doesn't need that right now.

It doesn't need it right now. And right now that that's an important, I think, distinction when you're getting into sort of, um, you know, assimilation versus accommodation and learning is a new process or something. But but, you know, this is also, I'll give the example of why we classify things the the way that we do. Sometimes good, because we get a lot of questions like, "Well, you guys were talking about this attack over here, but then this situation, like how are those even this even the same at all?" And I think for example of of this of what we're talking about now, and you tell me, um, is of how we talk about these different, yeah, we'll talk about these different areas for assimilation would be something like Insider threats. Right? So so it, meaning how do I address an Insider threat? Well, there's different types, but it, but it's all the same. So so The Insider threat literally like, you know, US forces in Afghanistan, their counterpart that they're training shows up to try to kill them one day. Okay, that's an Insider threat. But but a school shooter is an Insider threat. But but so is, uh, suicide is an Insider threat. Yeah, suicide is an Insider threat. A coach on a child at that spot as an Insider threat. So the idea is I don't have to become an expert in pedophilia and how to counter that. I don't have to become an expert in suicidology and understand that. I don't have to become an expert in school. I have to understand an Insider threat and how to counter an Insider threat, and then I can assimilate that information, prototypically match what I'm finding. So so that's all I need to learn how to find, and then I don't, I might not know which one I'm going to find. I might not know whether it's the kid that's going to shoot up the school or the kid that's going to commit suicide. But I'll get it right. Exactly.

That's why we focus on looking at it from that perspective. And that's why we go with the lowest common denominator rather than this voluminous explanation that people aren't going to understand anyway. We can be street the science. So let me give you a street science. Let's talk about grooming a potential victim. Okay? So let's talk about how a chef at a fine restaurant grooms you when you come into the restaurant. Okay? So back in the day, mace was popular with cops. I was one of the first cops to become an early adopter of Oleoresin Capsicum, two million Scoville units. You could wash it off. If you sprayed it in a restaurant, you didn't ruin a Walmart or a shopping center. You didn't ruin all the stuff that was on the counter like you did with mace. Mace had these horrible side effects, and somebody from mace.com now is going to call and complain about that because my ass, you do your own research. So what do we need to know about that? Well, capsaicin, okay, is the spice, let's call it, that triggers the heat receptors in humans. Okay? So two million Scoville units get so hot to your brain that your brain thinks you're on fire when all it really is is agitating a new threat that your body senses, and it's trying to counter that threat. So it counters it so quickly and immediately because it senses this threat, Brian, that what happens is you close your eyes, you start sneezing and coughing, and you do everything to expel it. And while you're doing that, you can't shoot me or ram me or protect me or fight me. Okay?

So capsaicin has these uses, but we also find out that because your body is trying to remove the capsaicin by your eyes tearing (Jaren) and your nose running and your mouth going, "Oh, that's too hot! Where's the milk? Where's the water?" Okay. Taste is your mouth triage. Taste says, "I will be able to function with this new food, and I will be able to take good things from it." Now, because you don't need to ingest or metabolize capsaicin, your body fights it. So what does a chef learn over time, Brian? The chef learns that just the right amount of spices will stimulate you and give you different reactions, and sometimes you'll go, "Wow, that was too much!" But other times you'll go, "Wow, that was just enough, that was very adventurous." And what do we do? We stimulate and stimulate different things because those excite us and those create a chemical dump. So just as your body is preparing to metabolize the capsaicin and goes, "Danger, warning, Will Robinson! This is an Insider threat!" I'm willfully trying to put this on board. A chef does that to mess with you. A chef says, "Hey, listen, I'm going to make chocolate with jalapeños or with ghost peppers." So so what the chef figured out through scientific process is we experiment with our environment, and certain things create sense triggers. Do you get what I'm trying to say? And if you manipulate those, so how is that different from a pedophile? So a pedophile's got a joint in the ashtray. A pedophile has a Playboy Channel or some pornographic images that are around, and they say, "Hey, give me a hug! I got that big hair bear hug. Listen, I'm babysitting you today for a couple hours. Why don't you go, you know, I'm going to go get a beer. You want one?" Those type of things normalize. Do you get what I'm trying to say? And normalize is what? It's another word for assimilate. You get it. I'm assimilating this new culture. And guess what? In a minute, I'm going to have your pants down.

Yeah, which is, this is like, I mean, you're, you're talking about historically cultures and, and, you know, how, how civilizations have clashed whether through war or through something like that. Either way, so I can either fight my way in and take over your country and now you force you to assimilate, or I can do it slow, steady over time, something like you just said. It's either got to be the brutal sexual assault, or I can, or I can slowly gain your trust and get this. Which one, which one is going to give me the best return on my investment?

Right. If I brutally abduct you and rape you, I got a couple of life felonies that I've committed, and people are going to search for me like the citizens searching for Frankenstein, the village. Okay? Whereas if I do the slow, steady, a little bit at a time, maybe I'll get away with it. Yeah. Or maybe, "We'll wait a minute, it's okay for them to come over. You can't prove that."

I'm confused. Exactly. No, and that's exactly the point we're making. And and so so, uh, so this happens when, okay, so the assimilation that we're talking about too, and then to kind of sort of tie it back to the original AI sort of machine learning, you, you know, to give that that sort of street example, like you said, you know, when you go on, um, anytime you're on some site, like you're on Amazon and you buy something, or I'm looking at something, and what it also gives you as you scroll down, you can scroll across the bottom.

People who bought that item also bought this.

Exactly, right. And so they're doing the same thing. They're trying to couple it together. And and what it, the machine is doing exactly what your brain is trying to do. It's it's trying to constantly predict, "Okay, what do you need? What do you want? What is it? What is it? What is it?" Right now, right? Because I got to give you an answer and and give you the answer sort of before you arrive at one because I want to show you that. So in the Amazon case, I'm just trying to sell you stuff. I want you to buy more products. I want to make more money off of you. But the your brain's just doing it sort of for survival purposes, right? Yeah, yeah. That's not something I'd never buy. But because of this, um, this either attribute that this has, right, that that attribute algorithm.

You're exactly right. It it, but it pairs it with, "Okay, what, uh, what other things that other people have bought who bought this item as well have bought? Maybe they're interested in that too." And that's the simple learning process. And it is. But what happens if it's not, if you don't buy that? If they suggest that you don't buy that, they update their algorithm. They say, "Listen, these people bought this, and then I went out and I I looked for new people to buy that based on the earlier algorithm, and nobody did. So now I'm going to drop that and go back and glean another evidence item to compare it with." If that's what the chefs did, and that's what a pedophile does, and that's what an armed robber does. So AI has a limited value set. Now, what do I mean by that? Humans have unlimited ways of putting things together. A computer only has the design.

Yes. Because you can't allow complete free play with a computer, you get what I'm trying to say, because you can't either become self-aware.

Right. You get what I'm joking about, but it doesn't, though. It's actually, it won't. It won't do. It'll be confused. It'll give completely random answers. Right?

Which aren't going to help. Which aren't going to help you get to that ultimate goal because if it doesn't have a schema, it doesn't have the mental framework, right, or a neural framework or whatever you want to call it, a pathway to go. But it does have a pathway to make you excited. The reason we're talking about it today is because the computer understands that if it plays with you, you keep playing with it. Yeah.

Okay. That interaction is absolute key. Just like our interaction as humans with our environment, with other people. You're exactly right. The reason that we suck at being with other people is while you're talking, Brian, I'm thinking of what I'm going to say next. I don't give a damn what you're saying. I'm just pausing long enough to go, "Yeah, yeah." Well, the reason is, what are the two things that human brains have to do in this moment right now? How can I assimilate the information around me to fight or flee or, you know, the other F words, or do all those things? Okay, we got that. What's the second part of it? Prediction. What do I need to predict? Yeah. Am I sleeping tonight? Is the sun going to come up tomorrow? Am I going to be warm enough overnight not to die before the sun? We're constantly doing saying that we're balancing and measuring those things in our life. But your friends and my friends consistently think, "Well, I've got to learn to drive at a high rate of speed. I've got to learn to appendix carry and dump mags and do all this other stuff." Brian, while that's important, your brain will assimilate that knowledge, but it's not critical knowledge. So the brain will always separate that from other things. Now, let's go to biology and to dump it over time if it don't continue.

It has to.

So so why? So let's go back to biology. Certain things that you eat your body doesn't need. It doesn't need a lot of things. Right now, my body doesn't.

Oh, yeah, exactly. Is bourbon a food?

So the idea, though, if you think about that, Brian, that's the perfect model. Look, I used to get into fights all the time with the Marines, the combat hunter. They would go, "Yeah, well, atmospherics, you can't use that word because the atmosphere is like around the Earth." "Hey, kiss my balls!" The idea is what you're not doing is you're not using the scientific definition, the clinical definition of the words I'm using. So for example, assimilation can be biological and be sociological or any psychological. Right? And there's a different definition that goes along with those. Yes. So what we're talking about is similar traits are always clustered by your brain, but your brain looks at them in a very specific way saying, "What do I do?" So biologically, when I ingest something that I can't assimilate, it's going to poop it out. So the longer that I go like, like not to get into Maslow again, but but your good friend the the the Gladwellian thinkers came up with this ten thousand hours. Yeah. Where's that damn? No.

Yeah, but show me that, show me how that works.

First of all, you know, I would agree that the more you do something and the more, yeah, rehearsal it, the better you get. But this goes back to to our other friend on coffee table, yeah, that Drew the conclusion when he saw the Civil War rifles that were overloaded two and three times. He said, "It's because they didn't want to shoot their brothers, so in the moment and so afraid of the situation they were convinced they fired it." Yeah, they reloaded and fired again. Never ever fired accurately. That's foreign. But there's fundamental conclusions were completely I completely disagree with and are wrong. So exactly. So when we get to those points, we have to look back at what we're talking about here. The brain assimilates that information that it believes will have the best chance of surviving long-term and changing one of my file folders in my mental Rolodex. If I am going to a place that's so different, my brain allows more influences and limited objective experiments. Why? Because just like your example of being in Africa, this may cost me my life. So intuitively and iteratively, you can work certain things out. Like if you have a set of keys and you've never been to a car before, but you go up to the car and you hit it and go, "They pooped and the door unlocks." "Okay, well, that's a welcome." Now I investigate further. "I wonder where this key goes." You get what I'm saying? And people are going, "What's the key now?" Because when we rent a car, you don't advocate, you get a little block. But the idea is that sooner or later enough monkeys with typewriters are going to write the Magna Carta. Right? That's the same thing with sooner or later you're going to figure out how a car works and figure out the perndl, and you're going to either drive over your own head or you're going to learn to drive on your own. Your brain has the ability to occur. How many people do you know are self-taught on piano? How many people you know learned Spanish or French because to their exposure and environment where they had to? Even being a waiter at a restaurant, even understanding simple things like, like this goes back, people are right now going, "Bing, bong, bing." "No, it's far from it." Okay. Stoic. Uh, I have a principal argument for stoics because when I talk to the people in Greece, when we were teaching that antibiotic, they pointed to Mount Olympus and they said, "Our gods up there were the ones that taught you this information." I principally disagree with that. So the same people that brought you Mount Olympus and the guides also brought the stoicism, so you must investigate what that means to you now. Sitting in a chair or standing in a line in the environment we're in, you can't take that giant evolutionary step backwards because it's not the same, Brian. And that's what assimilation means. Assimilation means that if I'm assimilating something quickly, then it's probably not going to survive or it's the most important thing to survival. So I'll give you an example of that. When you see a kid on her phone, they're not paying attention, situationally unaware, 101. Okay? So we know that sooner or later there's going to be enough deaths in traffic, people checked out, people walking into things, that that has to change. We don't know how that's going to.

Well, we respond with other, um, technological advancements. Right? So so there's there's they hate this situation. Well, the the safety features in your car, cars are safer today than they've ever been, they have different alarms and beeps. Um, they they are man. Oh, it's hilarious too, what I love when car dealers advertise that the car has a backup camera, but it's federally required that every vehicle has a backup camera. But they use it. Right? But it's the same thing. It's we can't do that stuff anymore. I am unaware. So so I have to figure out a way and build that into the design of the product knowing that so so that and that that's that's the the evolution. Some might call that because it takes away from our survival systems. We don't have to exercise them anymore. But, you know, I, you know, all of these these sort of with the examples and stuff you're talking about, uh, uh, sort of the comparison between the the machine learning part. I if I want to create, I have to create an algorithm for the machine. Yes. So I have to sort of, uh, design the the pathways from which it extracts information from. Well, humans, we do that. Our brain does that same process. The problem is, well, it's not the problem. It's just yours are completely different from mine. So so it's it's highly subjective in nature, and you have to, why do we have tribes?

Well, but but you're solving a paradigm too. We have tribes, why? Because we have to associate with different thinkers. Because and this is associated with different thinkers. This is the point about assimilation too, because I fit everything into my current interpretation of reality. Well, if all I do is constantly assimilate information into what I know and how I know it in my own life, why don't I actually increase sort of my cognitive flexibility?

I I can reinforce my own beliefs. But but the the issue with that is because it's it's highly subjective. Um, it can our experiences are different, our memory recall is different, and also what we feel and think is logical is different and how we connect those patterns. So like you're like you always do, like when you bring something up, um, you don't, you don't typically just give one example, right? You give a minimum of three, sometimes up to 30. Right?

But it's a really reckless meant to go. It just back to my Insider threat example was, "Okay, here's a concept. I'm going to give you an example here, an example here, an example here, an example here. And then I'm also going to use that example for another concept to tie that together." So what you just did.

So the idea is, let's start, let's look at this from almost a schematic sort of way of looking at it is, "Here's The Insider threat." We gave the the pedophile, the Afghan, you know, military guy that shows up and kills the U.S. soldier. We gave the suicide and we gave, um, what was the, uh, example?

Well, the school shooter, of course.

Yeah. So then I can also take that, um, uh, uh, the the pedophile, right, and let's jump that and throw that over in this concept of here about access or something else. And now I've linked those things together. So that's sort of neural mapping in a sense. Yeah. How we give out the information. Now the the what you do run into different issues too of of, you know, me looking up at the sky and going, "Hey, that's a, look at those stars. If you look at that one and that one, that one, that's a bull." It's like, "Well, no, it's not." "Well, yeah, it is. I can see it right there. That's a face on the moon. There's a man on the moon." And it's like, "Well, right, no, there isn't." So I'm starting to do that pattern recognition past the point where it serves its purpose. Now I'm sort of making things up. So do you see what I'm getting at here? Yeah. Exactly.

So many of those attributes or I link too much together, it becomes but your brain won't allow the rational person to do that. The brain will not allow you. Smoke some peyote and you're on the field, now all of a sudden everything becomes, yeah, right, yeah. So I'm I'm skeptical. I believe completely in diversity and inclusion. You can't live without it. Okay. On the HMS Beagle they were talking about it. So anybody coming up with that crap today and talking about it, you missed your mark because it's been around since humans have been around. Okay. And in the animal kingdom, in many instances. Now, equity is one of those things that because of free will doesn't always work so well. You got to train equity into a system. So what you're talking about is building an algorithm is brilliant. And and I would tell you this, I would tell you that in certain environments, certain things like diseases, apex predators, find the sweet spot in your algorithm and dwell in the shadows. Why? Because they have to play the long game too. Like a parasite will eat you, the camelback spider, but keep you alive by making you paralyzed in the area that it's injecting you. Yeah. So so we have to see that in society too, Brian. That's what we do. What you and I are trying to point out to people that look, those dangers lurk in the same wonderful places in your environment. Here's how to see them. We don't tell you the danger. Right. We tell you how to find the danger. Do you see what I'm saying? So so I'll take you back to a danger. Okay. So we got a guy that that flipped his wig back in '93, and the only reason it wasn't the story of the century is because during the same time, OJ was going on. So William Kunstler, famous attorney, was one of the guys that I was following because I used to do the legal updates all the time, and I knew Doug Fieger, the singer from The Knack, whose brother Jeffrey Fieger is a big attorney in in Michigan. So anytime we're talking about human rights, I like to get involved because I'm very intrigued. So in '93, Colin Ferguson gets onto a train in in New York, the Long Island Railroad train, and shoots up a bunch of people. Okay. Killed six, injures 19. All these different things are going. Goes to trial and insists when they give him a public defender that, "Nobody knows how to defend me better than me." Right. Now he's got a deficit because he has zero training in the law. But the thing that Kunstler repeated, and and I heard over and over from different interviews, was watching Ferguson from day one of the trial to the second week, and the third week. First couple of times, he was just being a classic obstructionist. Second couple of days, he was getting up and he was getting the motions down, where to stand. He was repeating. Okay. Then what happened is he started going, "Okay, if I object here, that changes the pace, the tone, the feel, the atmosphere." Now, Brian, he wasn't sure what he was doing, but he was able to mimic and assimilate those parts that seemed to give him the best chance. Now, of course, he was convicted. He's not getting out till 2352. Okay. But watching that as an experiment, uh, uh, we've seen that before. We've seen serial killers that do it with greater aplomb. So you see a a shoplifter that can do the same amount of research as a pedophile, and you want to get away for many years. And pedophiles that don't do the amount of research and and go to jail immediately. You see what I'm trying to say? So each of us is different and assimilates at a different rate based on our view.

The Ferguson one is a is a good, um, example to compare to. I don't I don't remember his name, the guy in Wisconsin who drove through the parade, killed a bunch of people. On the trial episode, they said the same thing, "I'm going to I'm going to represent myself." And was trying to do all those things, but the guy had no, he didn't learn. He didn't understand. He was just, uh, literally, uh, "I'm objecting for this." Wouldn't answer the experiments, you wouldn't even ask questions correctly. But he didn't have the cognitive ability to learn, right? Meaning to get better at it. It didn't, it didn't become an iterative process where he went, "Oh, okay, I screwed it up there. I need to change it." He just kept repeating the same thing. He kept doing it. And like, people were like, "I I can't believe." Like there was attorneys and lawyers being like, "Oh my God, just, you know, please shut your mouth and stop doing this. Like you're ruining your own case." Whereas Ferguson, or compare that to even the the Ted Bundy one where, and I would throw in, I would throw in Aileen Wuornos, my favorite.

Oh, yeah, yeah. Assimilated and learned how to play. So so Bundy did it so well, and and we remark on this all the time that even the jurors and even the judge was like, "Oh, man, boy, if you if you decided not to to cut those heads off, you would have stayed in school a little longer, you'd have been a great judge, a great attorney." Yeah. He was sitting there like talking, but but he, that's that's a higher level of intellect, a higher level of organization, a guy who is wicked smart versus a guy who's just, you know, diversity.

So the diversity in our species to clump together when we need each other to solve problems.

Exactly. I'm I'm with you 100, uh, 100%. And the problem is that you had a boob, a nut, a murderer that could pass off as a competent attorney. If you walked in in this the ninth day of that trial, or you walked in and your only observation was the first day, your perspective would have changed, and how you viewed that situation would have changed. So so it's completely important for us to continue to conduct experimentation, and we have to understand that with a computer, the computer is not always going to be right. Neither is a human, but the computer is going to update its view differently than a human is. A human's always going to take the the perspective of egocentric view. Right? How is this going to impact me and my environment? Where you can program a computer to disregard, for example, emotion. Do you get what I'm trying to say? Which you can't with a human. You know, if you remember, have you ever watched, uh, uh, what was that Mel Brooks movie with the Frankenstein, where, uh, uh, Young Frankenstein, and at the beginning he, uh, gets a guy, the old man off the gurney and he knees him in the balls. Well, then the second time he puts a clamp on the back of his neck and he does everything, and the guy doesn't move because now, you know, he's closed those receptors. You can do that with a computer, you can't do that with the human. Humans are constantly updating their set based on slight's injury. When we have a, uh, injustice collector, Brian, you see somebody, "Well, that doesn't stop the rage that's growing and does something stupid." A computer won't do that.

But a human, no. And well, because the computer, if I input something, the computer is not affected right there at the point, right? Because remember like, you know, it bad information in equals bad information out. Right? So so if I, if if you're entering something into a computer, right, it's always going to take what you enter in the same manner. But but if I walk out this morning and I see something, uh, that's going to be different, uh, walk outside my house and I make some observation about someone, uh, pulling up and getting out of the vehicle, that's going to change whether if if I'm, you know, that scary movie I saw last night and I didn't sleep well, so I'm not not even seeing as well. Now I'm a little bit hazy, then I see the same thing. So I can make the same exact observation and and arrive at two complete different conclusions with just the same data set, the same exact one. Whereas the machine on the other hand, it's it's going to, it's going to take in the same thing and look at it the same. Now its conclusion may be different based on how it was programmed to think and learn. But but mine, even just the information in, uh, the way I assimilate that, the way I categorize it, the buckets I put it into in my brain is going to change simply so subjectively from one minute to the next. And that's the difficult part. That's also the dangerous part because that leads that dangerous part. It's it's the, "Well, I could tell by how their voice was on the 911 call whether or not they did it," and it's a fallacy.

So so let's use me an example because I'm always making mistakes. I want you to think out loud, Brian, on how many pictures of sunsets that I've sent you over the years that we've known each other. Yeah, yeah, exactly. So no matter, was, "Take a look at this, this is beautiful." Now, in my brain, every single one is different, and the colors are different, and it's remarkable, and the time, just a few seconds before or after, and the right camera exposure, and all that other stuff matter because it speaks to me because that sunset is just for me. Now, if we go to machine learning and we're reacting with our environment, we understand that it was the forest fire four months ago, and the ice crystals in the air, and the barometric pressure dropping, all of those things. Now, my brain doesn't need that for me to be inspired and send you that photo. Does that make sense? So so the computer is following a very logical, a very clinical path to get to the desired results based on the input, whereas I'm a human, and I don't I don't have to rely on those constrictions. And that's why every human is different, and that's why a person, a mom, can turn postpartum depression into a homicide. That's why the the kid shoplifting turns around and kills the security guard. That's why those things happen because we're inefficient at processing nuanced information, novel information, or voluminous information. And so can a computer help us through that? Absolutely. But AI isn't the answer. You see what I'm trying to say? Yeah.

You can't you can't do the same amount of computation that a human brain can, and you have to program in exactly what you want that to compute. I mean, that goes back to it's following, you know, the the example of the fish and the mammal. What's the fastest one that the person pointed out? Well, the the computer followed a completely what it was programmed to do, which is supposed to be a logical framework, and it came to the wrong answer every time. And when asked to explain, but it defended it. It's flipping in, and it but it but it explained it correctly. It gave the correct attributes. Yes. And yet consistently arrived at the wrong conclusion. It wasn't a goat, it was a bear cub. Right?

It wasn't the dog, it wasn't the goat.

But but that was the thing is, and it's it was great for showing, um, kind of humans can do that if we're programmed incorrectly or I have a corrupt file folder.

I have such a strong that's why we called it a corrupt file folder, because people would get pissed where I would go, "You have a corruption in your file folder." People would take that from me, and they would relearn the information when we were talking about snipers or bombs. You hear what I'm trying to say? And and how a human is going to think through it because, Brian, when we started the LEO program, the Marine Corps, I was dealing with coppers that had my experience level and then some, and FBI agents and everything else. And they came in with these preconceived notions, and I had to break them down, and I'd say, "Look, you got a file folder because you had all of this ability, and you had all of this equipment. This is a conclusion you came to that Marine's not going to be able to rely on that. So you got to dumb it down." And I don't mean no, I mean, "You got to dumb that science down so I understand that these things." You remember over and over the the the a marine with less information that didn't have the combat hunter training would run unreasonable conclusion about their environment, and that costs time, that costs money, and sometimes that costs lives. And that a lot of that came from an Uber trained, there was a guy, and I'm not going to talk about them other than say that was operating around the same time I was, and what he was doing is going in and finding anomalies and pointing them out. "Yeah, that's great. Uh, uh, you're fishing for me, but I'm never going to learn how to fish, Brian." You get what I'm trying to say? And that's why I was so insistent that our program was then and continues to be different. We will show you how to get to the answer, and it's okay that your answer is wrong, even if your answer is wrong half the time because the times that you're right, you're going to be right enough to save a life. And therefore, we're not being sloppy like cognitively close enough means cognitive flexibility. It doesn't mean taking erroneous information and forcing a round peg into a square hole. And that's the key. The key to our program is that we're going to make you more resilient through cognitive flexibility.

It's the the door handle example I give sometimes, right? Once you learn, you're a little kid and you can reach up and you can turn that knob on the door and you understand how a door opens. Well, sometimes it's a knob, sometimes it's a big ball one, sometimes it's a handle with a button you press with your thumb, sometimes it's a push bar. I don't have to be shown how to open every single one of those. My brain can assimilate, can use a prototypical match. It understands how the mechanism, even just theoretically, basically how it works and what it's supposed to do. I don't suddenly show up at a new place and look at a door I've never seen before and go, "Looks like we ain't getting in there today."

Exactly. I'm going to die on the porch. Right? So so the, but but that's that that's a point of sort of what we're talking about. So I even if I come across something I haven't seen before, I I can, it's cognitively close enough and I'll be right most of the time. Now, maybe I might come up to something that's so complex that I can't crack the code on, but that's going to be so rare, it's got to have a movie about it.

Exactly. That's right. That that becomes the thing that then everyone talks about and that's okay. Those are out there, but they're they're they're statistical anomalies. Right? They don't happen very often. It's not it's not it's something that I have to constantly concern myself with. It's it's learning how to apply this this assimilation. What we're talking about is how do I understand this in in every environment? It's the gas station in the Middle East. "Oh my God, what the hell are these guys doing? They got a bunch of." It's like, "Hey, man, see those gas cans stacked up over there and a car comes in and stops for a minute? We can't see everything that's going on, and there's a guy hanging out there, right? That's a gas station. Is it nice and clean with the big bright signs and the handles and all that?" No. But but but the the behavior associated with the area is exactly what you would expect to see in a gas station. Therefore, like the flexibility, buddy.

Yeah, you have to be cognitively flexible to solve for X in new and novel and nuanced environments. And we say that all the time. Why? Because if your plane crashes in the Andes, the assimilation process is flash to bang is going to be a lot shorter. When you get hungry, you're going to eat the frozen guy next to you. Yep. Okay. And people are going to say, "I would never do that." You can't say that because you're not in that situation, and you will. People are going to say, "My son could never kill and dismember," and it happens all the time. So why is denial built in? Denial is built in so we can operate in teams and groups of humans because I have to be able to sit on that plane in the middle seat and smell your funk and not want to murder you because it's not socially acceptable. But if it was a couple hundred years ago, and it was a cave instead of a means of transportation, things might have been different. We assimilate based on current trend thinking. That's why we've got to stop going back to things that didn't survive. If it didn't survive the purge, Brian, yeah, it's probably not good for us. You get what I mean? Yeah.

And that's a philosophical point, and I had no idea that but that philosophical point is important what you just said is the, "Well, every time he hears them, well, if I was in that situation, I would do that." "No, you wouldn't." Like you don't you don't know. Okay. You can't even start there with those 400 people did when they showed up at that scene. Guess what? That's a statistically significant number to go, "I bet you would have done that too if you were in that 400." Like you you can't sit there and say, "I'm the damn hero. I would."

Those are snipers. And and that's the worst. Like again, folks, snipers, I hold them in the highest esteem, but then a couple of days ago, Brian and I saw a sniper in a crowd. And meaning that they were sniping some subject matter experts that were on a stage. Why? Because they wanted to hide their vulnerability, Brian. They were not as smart as the people in the room, and so what they had to do is they had to resort to vocal violence. Yeah. You get what I'm trying to say? To to to try to upset the applecart. So will those people be around for a long time? Well, you're always going to have oxygen thieves. You're always going to have people that sit in a seat. You're going to have that guy that we heard about that volunteered to stand in line for you. You didn't have to pay much because he loved it. Yeah. You remember that? Okay, for concert tickets and, "Oh, yeah, man, cool. I can, yeah, I'm good." Those are so remarkable that they make movies about them, they write books. So if you stay in the mean, you'll understand exactly what we meant by how AI can mimic human learning, and that's why we call it machine learning. You know, yeah.

So and this goes back to my argument that I've had with the government many many years now. I always talk about the human in the loop, okay, and the machine in the loop, and they always get it wrong because they're always saying, "Okay, well, there's a human in the loop." And and we're saying, "No, it should be the machine in the loop." Like, "Why, though?" Why? Because of the the things that we laid out in the last 50 minutes today. Exactly.

And and I may, you know, maybe if you're listening, have questions or interested or something, let us know because I I think sometimes when we take these approaches when we talk about, you know, AI or machine learning and what happens in errors and along the way, it's kind of easier to pick that apart because we're not talking about you as a human. Right? Even though it's a very similar process. So it's a good metaphor to use to go, "Hey, see this problem here? See this problem?" How many times, yeah, to your exact point, how many times do we have to get on here and say, "Boyd's OODA Loop does not have anything to do with your business, isn't going to make you a better salesman." It's a model. It's a model for John Boyd that was trying to explicate his thought process in a very very close set. You see what I'm saying? So that's your point exactly where you're saying that, "Oh, I I would have done this." "No, you wouldn't." And you wouldn't recreated Boyd. Why? Because John Boyd was a guy that hit on something at one point in his career that was brilliant for fight. You got a swim team, but the swim is individual. You know what I'm saying? Unless it's an IM, that swimmer's the one that's competing, not you. Well, yeah.

Looking at a fighter pilot and go, "Oh, that'll be a great speaker at my next event because that has nothing to do with the damn you're doing." Right. But and and of course he was also, uh, his whole thing was obviously he invented energy maneuverability theory, fundamentally changed airplane designs. Headlights on a car, the reason why it came to where it was was because he was always talking about maneuverability and then the Marine Corps said, "What do you mean?" He said, "Well, even in warfare, generally speaking, it's better to be faster and move this way than it is to be big and strong." And that, "Yes, there's certain situations, but if I can out-think and outmaneuver," and it's all about speed and process. But he was talking about that, like the asymmetry, physics, and evolutionary biology. Like he wasn't talking about shooting on a range with your buddies. Like so so the, but but that that that's another good example of sort of, uh, uh, those and people get mad at us because they think that we're downgrading something that they're doing. Yes. If you were in a boardroom and you're using Boyd's OODA Loop to tell the people how to produce better and sell things, I don't want I don't want what kind of magic you're using.

But they're selling a Harry Potter's wand for Christmas. Yeah, you want to get one of them? Well, it's a little too.

I think that's a little too general for that situation. Obtuse, maybe a little open. I'm sorry.

Now you're cute, but now you're obtuse. So you would be better off teaching people how to do business with Family Guy.

Yeah, we certainly do. We do. We certainly live most of our lives on, "I got everything I talked about today. I got it from a Family Guy episode." So I think so let's name episode MacFarlane.

And folks, today was not a Meg episode. No, no, it's not.

Well, I okay, so we sort of covered a lot. I hope that that that made sense. But where the hell do you recap this? Well, the the the idea of of assimilation is, I I think we made the point on that, and how just like the machine learning, we made a valid point. Yeah. Um, you're not always going to be right, even sometimes when it you follow a logical pathway, because your logical pathway is already inherently subjective. So so that that's that's good and bad. Right? I mean, so let me throw this to a CEO in the audience right now, or a teacher, or a human resource person. You're going to go, "Okay, so so what? So so what here is, even though you have the authority to do it, even though you have the money, experience, and background to do it, is doing it moving the dial in the right direction?" Even if it was what you're trained to do in this situation. Yes. Does it actually apply? Is the cost-benefit analysis going to play? That's a simple answer. Do you get what I'm saying? So yeah, so I stopped you and this is a Michigan example, and found out later that the car was a VWOP (vehicle taken without permission), or a stolen. I also found out later that you had an extensive criminal background. I also found out, well, all of those after-the-fact things didn't come into play when you had a drunk guy that was trying to run and get away and he ended up dead. You get the balance, Brian. So even though you had the legal right, even though you're under color of law, even though you had those. So HR person, same thing. And this is why, uh, uh, the the doctor, uh, uh, Peter Langman, uh, yesterday got got into my 'hooda loop' and and because he was saying, "Hey, listen, look at all these mistakes that were made at Oxford and they still didn't fit it into their their paradigm, Brian, into their their, uh, uh, example, into their." Do you get what I'm saying? Look, that's not what we want. We don't want to fill in, uh, uh, a state-of-the-art, a significant events log in everybody's life and every student. What we want to do is we want to say when threshold behaviors, uh, uh, make themselves known, like whack-a-mole, you got to beat them down, negotiate. If the nail sticks out, pound it down. None of those architects.

Is the juice worth the squeeze for this situation at right now, for the good of man and mankind?

No, no, that's perfect. And then prove it. Is the juice worth the squeeze? Well, prove it. Tell me right now, is this worth making this decision right now or going down this pathway or doing that? Or or do you have a second? Do you have a minute? Taking that person into custody, was it worth a homicide? Was it worth a cop dying? Was it worth a student? Is it worth it out? Yes, yeah. So so if we were going to say what was this episode about, I'd play it backwards. I I would do this one. Yeah. Uh, you know, the last couple of minutes and go back to the beginning and play again because that's what it's all about. And and when a machine can get us to have an hour-long conversation on how we interact with our environment, that's a good thing. So so that alone, that joke with you and Stu alone, yeah, uh, spun out into something where I now understand that if I don't constantly update my my knowledge, I don't update my perspective, and I constantly don't sample my environment, that's true situational awareness right there. Yeah.

No, no, I I I agree. I think that's actually a great great point to kind of bring it in for landing on it and none so that's that's like that's what that's what it took us 59 minutes to get to.

Yes, exactly. I'm not sure we should have started that at the beginning and then just know exactly. Yeah, maybe just maybe just erase everything about that last few minutes, folks. So that's hilarious. So that's the so what. But no, I I, uh, we talked about a lot. If you're listening, please, if you're not sure or have a question or you say, "No, you guys don't know what the hell you're talking about, here's why." Just include the end, "Here's why," and we'll answer it and we'll bring it up. Um, and and and reach out to us and and, you know, The Human Behavior Podcast@gmail.com. I answered everyone that reaches. Give us what you're thinking. What's the specific point that you want to make, not just, "Hey, talk more about that topic." Yeah, that's where you're foggy. And of course, we always, we usually go into different details, talk about more about some of the stuff on the Patreon side that we don't want to play. Yeah, we're going there. It's just it's just a it just takes a few minutes today. Yeah, in a gating mechanism, so not put it out there to everyone. But um, we do appreciate everyone, uh, uh, listening in. Thank you so much, and don't forget that training changes behavior.