In this video, Dr.SHIVA Ayyadurai, MIT PhD, Inventor of Email, Scientist & Engineer, Candidate for US President, asks WHAT is AI? As a pioneer in AI since 1978 who was awarded 3 of the most cited AI patents, Dr.SHIVA™ explains what AI TRULY is.
Get Educated, Or BE Enslaved
TruthFreedomHealth.com
SHATTER THE SWARM
Shiva4President.com
Become a Truth Freedom Health® Warrior-Scholar
Attend Truth Freedom Health® Thursday Open House
To attend an online OPEN HOUSE with Dr.SHIVA this THURSDAY at 11 AM EST or 8 PM EST. RSVP at: VASHIVA.com/ORIENTATION.
Dr.SHIVA is committed to health, education, and innovation.
Dr. SHIVA’s most recent innovation is the Truth Freedom Health® SYSTEM that is an educational, community, and technology platform that is enabling everyday people to become wiser and smarter – to think beyond Left & Right, “Pro-” and “Anti-” – by learning the SCIENCE OF SYSTEMS – to see events and things as they truly are to know the real problem to innovate the real solution.
To learn more about the Truth Freedom Health® System, visit: https://vashiva.com/join or https://TruthFreedomHealth.com and either contribute to this educational movement or become a Warrior-Scholar.
Find out more at https://TruthFreedomHealth.com.
Be the Light!
Dr.SHIVA
e: vashiva@vashiva.com
w: https://vashiva.com
w: https://TruthFreedomHealth.com
w: https://Shiva4President.com
Twitter: @va_shiva
Facebook: https://www.facebook.com/va.shiva.ayyadurai
YouTube: https://www.youtube.com/@drvashiva
Gab: https://gab.com/ShivaAyyadurai
Rumble: https://rumble.com/c/DrSHIVA
Telegram: https://t.me/vashiva
Bitchute: https://www.bitchute.com/channel/OBguXv4JEmPK/
Odysee: https://odysee.com/@drvashiva
Gettr: https://gettr.com/user/VASHIVA
TikTok: https://www.tiktok.com/@drvashiva
TRANSCRIPT (Auto-Generated)
Hello, welcome everyone. This is Dr.SHIVA Ayyadurai. Today’s talk is really about artificial intelligence, AI. Which is the acronym. Many of you have heard about this thing called AI, but may not fully understand what it is except probably thinking there’s going to be a bunch of machines which are going to take over your lives and we’re going to have robots walking around and and that’s you know, the idea that you have R2D2s running around or the little robot that vacuum cleans for you.
But today we’re going to really get into different aspects of AI you know, sort of the popular idea of what AI is the much more technical knowledge of what AI is. You know, I’ve been involved in AI research since the eighties. You know, since I came to MIT it’s been one of my passions. I built a very large company out of it called EchoMail, where we used to do AI for email analytics, made a ton of money doing that. So it’s an area I know a lot about and it’s an area that I want to be able to share with you for various reasons, because I believe that the mainstream media, as usual, will always distill something to something sort of useless that you may not really understand.
The experts will try to mislead you into whatever the flavor of the day is, but what we want to do on this podcast is to really give you the foundations of AI and artificial intelligence. So that’s what I want to talk about. And obviously, Marcelo and Richard, you know, you guys are representative many ways of sort of popular culture.
So, you know, you guys feel free to step in and jump in on any of these questions inside. I’ll go deeper. But the field of AI is really about fundamentally, how do you take human intelligence and replicate it? Using machines. That’s what it’s really about. So, so some very interesting terms. I just used human intelligence.
The primary word there is intelligence. In podcast two, we talked about intelligence from a systems perspective. So one of the things we really need to do is define what is intelligence. You know, before we talk about human intent, intelligence or artificial intelligence, and we’ll get into some very deep issues of What is the difference between a human and a machine?
And even more deeper, aren’t we all machines? So are we, what is the difference between artificial intelligence and human intelligence? We’ll get into this. But fundamentally, think about a quote unquote intelligent activity a human does, and being able to replicate that with a machine. Right? So if you want to go to some very simple ideas of this, you could think about a calculator.
As a form of artificial intelligence, right? When I was growing up in India, my great grandfather could do, you know, fractional multiplication in his head. You know, all different kinds of fractional multiplication because he had to know how to do that when he was farming and you had to, for commerce reasons.
A lot of mathematics really came out of commerce. We didn’t really need mathematics until we started to do commerce. You had things like the abacus that people use for doing calculations faster. When computing came, we started developing chips. We started outsourcing some of those computations to a thing called a calculator.
And so the calculator could do, you know, you know, 3, 225, 215 times 10, 315, 235. You know, those kind of things fast. Now we could do it by paper, but we outsource that kind of computation, which was a human intelligent activity to a machine. So you can think about a calculator as a form of a machine that was able to do those kinds of computations.
It was a form of AI. So what is AI from a technical perspective, and when did it really start taking off? There’s a field called cybernetics. Which came out during the twenties, thirties in that period and led by a guy called Norbert Wiener. There were many other people. If you go to MIT and you walk down the halls, when I first came to MIT called the Infinite Corridor, there’s a picture of this guy who looks like your typical nerd professor.
He has a goatee, glasses, a little bit of a punch belly and a suit. And that was Norbert Wiener. Norbert Wiener is known as the father of cybernetics or modern AI. And Norbert Wiener is the one who started looking at the human body or the human being as a multi tiered system. It’s an electrical system.
We have electricity running through it. It’s a magnetic system. It’s an electro mechanical system. It’s multiple systems. And the idea was, could you start building an engineering field and how you start developing Machines, which could replicate human function, and that was the field of cybernetics and cybernetics brought in many, many different disciplines.
Not only physics, it started bringing in a whole new field called information theory led by a guy called Claude Shannon. It also brought in the field of systems theory. general systems theory, thermodynamic theory. But it was a exciting period because it brought together many, many different fields.
Because if you’re going to recreate something like the human being, you have to bring in multiple fields. And so, interesting enough, after Norbert Wiener really pioneered this field, he wrote a number of essays talking about how horrible cybernetics would be for humanity. What
types of machines were they building back then in the 20s and 30s?
Yeah,
so, so they were trying to do large scale computing machines, right? They were starting to get into vision systems. So one of the early problems in artificial intelligence was vision. So, think about it this way. I’ve used this in many of my examples when I try to explain what pattern recognition is.
AI has become a buzzword, but the real… Technology behind AI is a field called pattern analysis or pattern recognition. So let’s think about Richard and Marcel are out a bunch of cavemen, you know, sitting around a fire and the snake comes at you. Okay, well, you have two choices run, you know, or let the snake pass by.
So how do you make that decision? It’s also a field called decision theory. In fact, MIT had a, a whole lab still has a lab, which just recently emerged, called Laboratory for Information and Decision Systems, LIDS. So, the fundamental thing is, how do you make a decision? Snake is coming at you, how do you make a decision?
Well, human beings, over many millennia developed a way to, you know, sit quietly. If a snake came, it’s not poisonous, or flee, or beat the hell out of the snake, right? That was a… Process that human beings went through or they survived or they died, right? So there was something in the human brain which helped us do this well.
And it probably it went something like this. A snake came at you, which had a diamond head, right? And you thought it was a nice pet. You pet it, it bit you and you died. Okay? You you died off. Yeah, and let’s say Richard’s clan continued to live because they said wait a minute diamond shaped head I saw that before and it bit someone else and it caused him so that knowledge in Richard’s clan was passed on and he probably taught it to other people.
So they had transfer of information from one You know being to another and by the way culture is ultimately a transfer of knowledge.
Do you think there was ever, like, a natural instinct of people just getting a feeling that that’s, like,
That’s a great question. Right.
So there, there
are theories in behavioral genetics which say some of this may be hard coded.
It’s not learned, it’s hard coded, and the more interesting thing is that the genes may self can pass on learning. You can take, learn something and pass it on. Gets into a whole another, that’s another podcast. But the issue is there was a set of knowledge and learning that Richard learned that Marcello didn’t, let’s say.
Caveman Richard versus Caveman Marcello. And that was passed on so people got very good at identifying poisonous snakes and non poisonous snakes. And they made a decision. Those decisions were made, right? So, and you can apply this kind of knowledge transfer And this kind of learning patterns and passing them on to, in, in many, many different ways.
You know, it could also be for foraging for food. Mushrooms. Some mushrooms you eat, they can kill you. Other mushrooms you eat, and you die. Right? So, you could argue that the existence of human beings today was through a whole history of learning decisions. And and then using those decisions to pass on to others.
And so how does all of this occur? And this was really the question of cybernetics or pattern analysis. How does this occur? How it is knowledge acquired? How do we learn? Learning and knowledge acquisition are some of the most important things because you have to acquire knowledge and you have to learn and you have to get better and better doing doing it.
So just think about, forget the word AI, just think about this area of pattern analysis and how that took, took place. So, I’ve shared with you how my grandmother could observe someone’s face and look at the patterns of that face and figure out if the person had a liver issue or a kidney issue or a heart issue, etc.
Well, that was a pattern analysis problems. In fact, the ancient, ancient Rishis of India, when they were in the woods, they would analyze all sorts of patterns. They would look at the patterns of plants, you know, their leaf shapes, and they would, and they documented these very, very in beautiful you know, detail.
So they knew these sets of plants were plants that supported liver function or kidney function or this, and they learned this over time, and they enumerated it, and eventually they built an intuition, and we’ll talk about that when, when, when you bring up the point, Marcelo. maybe, you know, right away. So people say, you know, you know, repetition is the mother of skill.
So there is an aspect of learning that takes place. The more you do something, it becomes intuitive. Everyone thinks, Oh, intuition is something like magic. Well, it turns out the more you do something your brain actually learns and it gets faster and faster doing that. And that was a basis of one of the tools of artificial intelligence or pattern analysis called neural networks, and we’ll talk about that, right?
But the bottom line is there’s a learning phenomenon that takes place that you pass on. And that determines how successful you are. So so in that thirties period, a lot of this was driven, frankly, by war, right? To create, you know, the forties were starting to, you know, World War. One and two drove a lot of this.
You’re saying the founder was warning the he was warning us of the dangers of
cyber. Yeah, yeah, yeah, yeah. So, so Norbert Wiener, after he is considered a pioneer, actually warned people that we probably should not do this because of the effect it would have on human labor. It would put lots of people out of work and that it may not actually give us a life that we actually deserve.
There are certain things human beings enjoy doing, like physical labor. That actually could be beneficial for, he wrote a set of essays on this. Did anyone take him serious or was he just… I think some people did, but you have the march of the military industrial academic complex. So, give you an example of, so, so all sorts of people got into pattern analysis.
Or, again, what we say is AI today. Behavioral psychologists got into it. Now one of the very famous… Or infamous people. I don’t really care for a guy called B. F. Skinner. Skinner is known as the school of behavioral psychology. He made a lot of inroads to the military and the Navy because he proposed the, you know, the pain reward model, right?
The way you train animals or people is you give them pain or you give them reward. So when people are designing missiles to go from point A to point B. Remember, if you launch a missile from point A, you want it to go exactly to point B. Well, so, and this was the basis of what are called cruise missiles.
So, what the military did at one point, this is before computing really took off, is that they designed a missile, and inside the missile, they put a TV monitor. And the TV monitor had a line between point A and point B. Okay, so the TV monitor literally had point A and point B and a line connecting point A and point B.
And they put a pigeon inside the missile. And the pigeon had an electrode conductor connected to its beak. And the pigeon was taught… When, and by the way, so you had the actual line where the missile should travel and then through a gyroscope where the missile was actually going. And the pigeon was taught if the, if the second line deviated from the first line to peck on the monitor, which gave a feedback.
So the missile would make adjustments to get back to its goal. This was basically before inertial guidance. All right. So basically, they were using a learning algorithm. In this case, the algorithm was a pigeon. It wasn’t on a computer. And this is how they did before we had inertial guidance. And now we have more sophisticated stuff, right?
With a cruise missile, we actually are mapping the terrain and the missile is all automated. And that’s a form of artificial intelligence because it’s following a pattern. But the concept of using animals or other people to automate things faster is is a whole history of artificial intelligence.
You can even think about AI as being when we went from the when the when the industrial era started, you have a manufacturing line, right? You want to produce a paper clip, right? And so one person doing it by hand. You started setting up a manufacturing line. One person simply, you know, one set of people simply got the steel ready.
Right. Another set of people cut it. Another set of people, you know, wound it up, right? Another set of people package it. Now, that may not sound like AI because you don’t see a non human being being used. But in many ways, that was the foundations of AI because you use carbon based creatures called human beings, but you organize them in a very systematic way to do very particular things.
It’s using pattern recognition.
Well, not only pattern recognition, but, but separation of labor for very particular tasks. And then as machines could do some, some of those things on an assembly line, you started replacing pieces of that, right? So I think the general thing, I think something quite different than you’ll hear anywhere in any other podcast you’re going to hear.
I want to define AI not as a function of a non carbon based machine, but AI and the concept of how you take some function that human beings do and how whether we use a human being or a non human being, how it is systematized so you can get better efficiency. That’s really a much more broader definition of AI.
So I would argue when we started creating the manufacturing line, you know, having giving people very specific things to do. That was a form of AI. But we were just using human beings to do very particular things versus the individual doing everything end to end. Okay, so, because then once you’ve figured out that system, then you start replacing pieces with machines, right?
So think about the Ford Motor Company or Chrysler, how they built cars. Initially, you know, I, I guess, you know, what Henry Ford fundamentally did, you could call Henry Ford, he’s not known as a builder of the car, everyone gets this wrong, he’s known actually as the father of manufacturing lines. Where he wanted to create a car that everyone could buy.
But what he did was he created the manufacturing line. He systematized a process. So people are doing pieces. And then as you see that advance, then when machines came and computing came, you started replacing. So you had maybe the machine automatically paint. It had vision systems. which could automatically paint.
Then you had automatic robots, which did, which did the bolt right, screwing, et cetera. But the notion of AI is not just transferring it to a machine, but it’s about figuring out a process to do things faster and faster and faster and faster. It just so happens, you know, we start calling it artificial when we transfer that knowledge to a non Carbon based entity, you know, but we have to look at it really at the history of that.
And that’s what cybernetics was doing. So, you know, when I saw my grandmother look at someone’s face and understand that face and be able to say that person has this issue, and then she would make a decision of what combination of herbs to give that person. To get them to the path to healing that was pattern analysis now in the Indian system for thousands of years They wrote a book called Samudra Kalakshanam, which literally had every feature in the face drawn out very precisely if you see You know, a furrowing a line between the two eyes, that meant there was a liver issue.
If you see, you know, certain things under the eyes, kidney issue. If you see a spot on the nose, that was typically a heart issue, right? So they, I mean, I’m just giving you the high level, but they had enumerated all these patterns. So after a while, people studied those books and they learned, but then they got better and better and better.
And it was no longer doing it by rote. They built their own intuition because their own computing took over. And we’ll talk about this when we talk about neural networks. Interesting enough, there is a, in the Sloan Management Review, which is an MIT publication many years ago, I think about five or six years ago, they wrote a article on intuition, which was really looking at how to chess masters make moves. And if you’ve ever seen the old pictures of Bobby Fisher, where there’s like 50 people lined up and he just goes table, the table, the table and he makes a move like he’s doing ton of moves and he wins all the games. Well, it turns out the difference between a chess master and a non chess master is that chess master over time has seen so many patterns that it’s automatic.
This pattern make this move under this pattern, make this condition. All, basically, they’re training their brain. They don’t even think about it, really. It’s what you call intuition, because, well, they are thinking, but the thinking is firing at a level that’s beyond the actual, actual thinking at a very step by step manner.
Another way to think about this is think about learning to ride a bike. Or learning to ski. Well, if you’ve never skied before, you’re up on that hill, and you’re learning to ski, the teacher’s telling you stuff, you know, you know, put your pressure on your left foot if you want to go to the right right foot if you want to go to the left when you first learn skiing, you’re wobbling, you’re going all over the place, you’re falling down.
So there’s this process of learning that’s taking place. And then you look at an expert skier, they’re coming down the hill at 100 miles an hour and they’re making very subtle movements finesse. So what’s the difference? The difference is there is physically something that’s differently wired in the brain of the professional skier.
Versus the novice. Think about when you learn to bicycling, right? I mean, obviously something biochemically has changed when you first learn, and then later on you’re just bicycling, you’re not even thinking about it. Even driving as well. Even driving. So William James in fact there’s a building, if you’re in Harvard Square, called the William James Building, which is named after William James, who’s known as the father of American psychology.
So William James wrote a very famous essay called associative memory. And what he conjectured, this was before neural nets came, was that if you look at when you learn something or when you experience something that literally creates physical connections in your brain between brain centers. So he had this notion that something was changing in the brain and he called it associative memory.
So, when you learn to bicycling, there are set of parts of your brain which are actually instantiating, which means saving that learned knowledge in the brain. It has to be safe, right? It has to be stored. That’s like building neuropathways. Neuropathways. So, so, but he was the one who came up with this concept of associative memory.
Around this time in the 40s, they developed a theory called neural networks. And the idea was people saying, well, how does that learning occur like physically? And could we create computers which emulate, which means duplicate that kind of knowledge? So the idea was this. Imagine if you looked when you have time, you can look at what are called neurons.
Okay, they’re nerve cells. So you have two nerve cells. And the nerve cells very much like other cells, have membranes, they have a nucleus. But neurons also have these things called dendrites and axons, which interconnect. Okay? Well, there’s a sheath called a myelin sheath, which covers you know, the extensions of the neuron.
Okay? So, it’s been shown now that when you learn something, what actually grows is not the nerve cell. But it’s the cabling between nerve cells. So think about if you have a live wire in your home, you know, you plug it in, you know, the copper wire and then there’s a plastic around it that’s called the myelin sheath that actually gets thicker and thicker and thicker.
It’s like when you work a muscle, it’s very interesting. We all have the same number of muscle cells when you’re born. It’s fixed for men. For example, and when you, when you actually get stronger and you work out weights, you’re actually increasing the size of the fibers. Okay? So it turns out the myelin sheath when you experience something is actually getting bigger.
So among groups of neurons, the interconnections, we talked about interconnections in podcast two, is literally growing. It’s not the nerve cells are growing, but the interconnections between the groups of neurons actually grow and getting stronger and getting stronger. So, so the connection strength between neuron one and neuron two, let’s say there’s these two neurons hanging out in your brain of, you know, many, many neurons, billions of neurons.
And, and these two neurons are involved in learning you to bicycle. Well, the connection between those neurons as you practice day one, practice day two. There, there’s actually a physical part of your brain is actually getting stronger. It’s the connection strength. So what scientists did was they said, well, could we do this on a computer?
So they, that was the beginning of what are called ANNs, Artificial Neural Networks. And so they literally created the mathematics and the computing where they could Input something into a, I’m going to call it a black box of neurons and an output came out and what they would do is train these neurons.
So what they could do is give you a simple example. I worked on this many years ago. By the way, I think I mentioned to you AI or pattern analysis has been sort of the foundation field that I was always interested in because I want to understand my grandmother is able to do that pattern analysis.
So, when I first came to MIT, one of the earliest projects I worked on, this was 19 for that matter, in 1978, when I was working at the University of Medicine Dentistry, one of the things in using quote unquote AI or pattern analysis was babies dying their sleep, some of them. It’s called Sudden Infant Death Syndrome.
Alright. And human beings have five states of sleep. We have REM sleep where we’re where we’re you know, having dreams. We have transition sleep. We have waking state. Anyway, adults have five states of sleep, and you could literally map it on a over time. If you look during a 24 hour day, you could actually plot it.
Children, babies, however, have six states of sleep. So we were trying to predict when these babies, by the way, sudden infant death syndrome is when a baby suddenly stops breathing. It’s a very scary thing for parents. And what happens if the baby stops breathing, if you shake the baby, you can resuscitate them.
So the idea was, could we predict when a baby would go into that sudden state of not breathing, which is called an apnea, A P N E A. So we had Montefiore Hospital in New York at the time gave us 48 hours of sleep data. So we literally had minute by minute by minute. the sleep states of babies. And we also had when that baby stopped sleeping.
So I built some of the early algorithms to see if I could correlate the sleep patterns to a particular point when the baby went stopped breathing.
So you were doing like the minutes right before. See what pattern that…
Well, yeah. So you look at the historical patterns of sleep patterns. And so the idea was and we built those.
So, so I, I built a whole, Mathematics were because I had all this historical data. My system could learn from that and I could see if I could predict a pattern and we got some very interesting data. There were definitely certain patterns which were correlated to the apnea and we published that in paper.
My point is that pattern analysis allowed us to make a decision and technologies like these could get input into a shaking crib, for example. So if you detected a sleep pattern, you automatically shake a crib. That was the action. Okay? Is that what they do nowadays? Well, there’s various…
that’s one of the methodologies you can do, right? Or you can set off an alarm. My point is that that was an example. This was 14 before I built the first email system. That’s what I was working on, but it was using mathematics to look at a pattern. Now when you look at these patterns. What you’re trying to do is you’re trying to find out is a recurring pattern and the math, which I can’t get into it in a podcast like this is known as statistics.
So statistics is where you’re looking for particular patterns. Probability of these patterns. It’s probability and statistics fundamentally. And in probability and statistics, you can have a single input. Or multiple inputs and multiple outputs. In this case, what we’re looking at was you know, the sleep pattern.
And we’re looking at actually there’s six patterns, right? So you have six inputs potentially. And you’re looking for the output of two potential outputs. Which means the baby stopped breathing or it’s continuing to breathe. So every time step, every minute, you’re getting six inputs into your system.
The baby’s sleep patterns in one of those states. And potentially, is it breathing or is it not breathing? Okay? So you’re trying to do a correlation of these sleep six states into these two output states. Got it? So remember, we talked about input and output. So that’s what you’re trying to correlate. And then you’re trying to build an algorithm.
Can it predict one of those two states, sleeping or not sleeping, from one of the six states of sleep? So, so this is fun. By the way, what I’ve just taught you is AI, the mathematics of AI. It’s an input pattern coming in and an output pattern. And can you, can you predict that output based on a particular input?
You guys got it? So are
you using a special
program? So we would build algorithms to do that. So the algorithm varies by the field that you’re in. Okay? So at the heart of this is statistics. Yeah. That is currently being used. So what way, by the way, you could model this is you take a bunch of inputs, you look at the output, and then you build a pattern, right, of prediction.
Yeah. After a while, if the inputs and the outputs get so, you’re so good at doing this, you can, and the math is no longer based on probability, now listen carefully, or statistics, it’s based on an equation. It’s called a law. It’s not even, so for example, let’s look at, everyone’s probably heard of equals MC squared.
Einstein’s equation. Energy equals mass times the speed of light squared. Or F equals MA, Newton’s, one of Newton’s laws. Force is equal to mass times acceleration. Well, how did Newton come across that? Well, he literally that’s called the law. It’s not statistics. Is that he literally had a table. If I put in x force, right?
I saw Y acceleration. Right? And mass was a coefficient. Okay? So like this he built a table. And then he saw this pattern. Hey, every time I increase force, acceleration goes up. But it’s a function of the mass, right? So if I have a body with, you know, mass X, right? You know, keeping the mass constant, and then he changed it.
So that was something, so F equals MA is literally a law, okay? So the question is, if you have a snake coming at you, should you run? Well, that’s a little more complicated. Can you build a law around that? Potentially, if you see enough snakes, right? And and you see poisonous, non poisonous. Let’s say those are the two outputs.
So what neural nets were are actually a way everyone uses the word artificial neural nets. You’ll hear all these these nerds will use all these complicated language, but ultimately statistics. Okay, they’re just doing probability statistics. The problem with the nerds is they try to make themselves more important than they really are.
All they’re doing is statistics. And, and, and in fact, the field of AI has been frankly a fraudulent field in the sense because it’s been filled with people who basically have been doing statistics for a living, and they’ve been branding it as different things because they get tenure and they get professorships, et cetera.
I see. Yeah, that’s a very simple way of putting it.
It’s statistics. Okay. Now the, so, and it’s a way of doing pattern analysis. So think about it broadly. These people are doing pattern analysis. Think about the snake analogy. Snake is coming. Should I run or not? Okay. So, what is pattern analysis? So, pattern analysis, you can break it down into three fundamental boxes.
So, if you have a piece of paper, if you have a piece of paper right now, you can put the piece of paper in landscape mode, and in the far left draw one box, in the middle draw another box, in the far right draw another box. Okay? And what you have is you have three things that you need to do to be able to do pattern analysis.
And I’m basically teaching you what… You probably would not have the opportunity to learn if you took it, of course, in any of these big universities, they make it so complicated that you would probably leave because they made it so complicated, but part of my doing these podcasts is make stuff simple because I believe academics have a racket out of making things complicated.
But those three boxes. So the first box on the left, you can put in the word feature extraction. I’ll tell you what that is. Feature extraction. The middle box you want to put in the word clustering. And the box on the right, you want to put learning, okay? Feature extraction for our left box, clustering and learning.
So let’s go back to the snake coming at you, okay? What is feature extraction? Feature extraction is called the art of pattern recognition. It’s the art. Clustering is once you get features, How do you cluster them into different groups? And learning is how do you learn from those clusters? Now, what I just said probably doesn’t make sense, but it will shortly.
So, I’ll give you an example. So you can see this is many, many years ago. For example, the CIA and the NSA and the government. And you should listen carefully has lots of tools out there. We have cameras everywhere. They can take someone’s picture. And they have it saved and they do facial recognition.
Okay? So if you look at someone’s picture, how do you know that’s this person or that person? Okay? How do you determine that? This is quote unquote AI or pattern analysis. Well, one way you could do it is you take a photograph. So think about you take a photograph with your high definition iPhone. Well, think about that when you take that picture, it’s a thousand pixels by a thousand pixels, okay, in full color, which could be, you know, 64 bits of information, a lot of information just in one picture.
Well, each little, so if you think about 64, I’m sorry, a thousand by a thousand. So what is that? One million little cells and each little pixel has color information in it, right? So you could build a database, a huge database of everyone’s image, which would cost you a lot of information to store. And then then you could start applying image analysis filters to do extraction clustering.
And, and learning. So let’s say you wanted to take a picture of someone and your pattern analysis or your AI job, Marcello and Richard, let’s say you guys were hired as AI engineers, build me a system that could take someone’s picture and tell me, are they Chinese? Are they African American? Are they you know, Caucasian?
Are they whatever race, right? How would you do that? Well, one is you could take the images and do some really hardcore analysis. But if you take the feature, so first step is feature extraction. So it turns out you don’t have to save a thousand by thousand, a million pieces of data. It turns out that the art of pattern recognition is extracting a few set of numbers that you could reuse in the clustering process.
So what they found was you could take the distance between the eyes. How many numbers is that? Just one number. You could take the distance between the lips, another number. The distance between the tip of the nose to the bottom of the chin, another number. The distance between the ears. Like this, you could basically, you just need ten numbers.
Okay? These are called features. They’re not having to store every little pixel, every image. You’re getting the picture, and then you’re doing what’s called feature detection. Okay. You write algorithms that can find where the ears are, and you say, okay, the distance between the two ears is five inches.
Save that for Marcello. The distance between Marcello’s eyes is three inches. Okay? So those
10 features are enough to determine your race.
Well, well, well, first you save those 10 features, right? So now you have a big database of people’s features. All right? Then you start doing clustering. Meaning you start saying, Oh, the Caucasian people all have this distance measure.
So you basically have a little, think about an Excel spreadsheet, right? Imagine for every person you have 10 numbers. So you have an Excel spreadsheet, so you say, Oh, when they have this distance statistically between their ears and this distance between their eyes, etc. They look like they’re Caucasian.
When they have this distances, they look like they are Chinese because you have, you can store this up, right? In a huge database
are humans making those.
Yeah. So you would initially create what’s called a training set. Okay? And so that training set would be used to cluster. So think about a cluster is like, imagine you’re looking up in the sky and you see couple of big clouds, right?
Those clouds are called clusters. So one big cloud is the Caucasian cluster. Another cloud is the cluster of Asian Americans, right? Or Asians. Another cloud. I don’t know if I’m being politically correct here. It doesn’t matter. But whatever it is, you have these clusters and in those clusters, you have the data points.
of people who are fitting certain measures. You already know a Chinese person, and then you look, so that’s a clustering. So you’re just putting people into those clusters because you know they’re Chinese, and then you do a learning. You say, wow, isn’t this interesting? The people who are in the Chinese cluster all have this common distance.
Okay? Or the people in the Caucasian cluster. So now you’re learning. Right? So first you identify features. You already know this person’s Chinese, so you’re putting them into these clusters. And then you write algorithms which measure these distance measures. It’s called similarity distances. Which is basically statistics.
That’s interesting that you say the art of the what did you say?
The art of AI or the art of pattern recognition
is The art of pattern recognition, because you would think it’s
not an art because it’s all… It is. It’s called feature extraction. Because, one guy may figure out these ten numbers.
Another guy may say, I’m going to store all one thousand pixels. Yeah. Right. Another guy may say, I’m going to store 50 numbers. Yeah. Right. So the art is what’s the minimum…, because the, the, the minimum set of features you get is going to make your computing faster, right? So obviously in law enforcement, they probably want feature extraction done really fast.
So if you’re doing, you need more computing power, et cetera. So feature extraction is the art. What are the right features? Clustering is you’re… you have some ability to because you have some a priori knowledge, which means knowledge that exists. And then the learning is you’re saying, Oh, look at those distances between for those features.
They all seem to have this pattern. You follow. So in this case, the input to your system. So eventually, how would you create an AI system? You take a picture, right? Run it through your feature extraction module, which is the first module on the left. Which takes a picture, calculates these distances, 10 numbers, feeds those 10 numbers to your clustering process, which stores them, and then your learning process says, oh, it does the analysis.
And then, by the way, once in a while you may find something new. Let’s say you finally had a Native American, which you’d never coded for. That would say, hey, this is a whole new cluster. And then you’re learning, well, what’s new about this Native American cluster? And then you would learn and you would put that back in your algorithm.
So there’s feedback. This is, I basically taught you AI. That’s it.
And then I guess the bigger picture is what is that all getting used for, right? Some of that might be meaningful and useful, but then some of that other stuff, it’s like what, do we really need this data? Do we really need to find it?
Yeah, so it’s a good question.
So, so once you start learning, you acquire lots and lots of data. And as your system gets better and better and better, by the way, this is called confidence level. So when you’re doing these, so when these AI systems are being trained, initially your accuracy is probably like nothing, right? And then your accuracy increases and it gets better and better and better.
Eventually your AI system, this system, is doing as good as a human being or even better. Because it can notice subtleties, like you can have Chinese… People who grew up in America versus Chinese and China, like you can start splitting these clusters and even finer clusters. But so I’ll give you a couple of examples that I’ve worked on that shows you that the same process you apply.
So I gave you the example of sudden infant syndrome. Well, same thing. What we noticed was the features there were the states of sleep, right? Those are the six features, the different states of sleep over time. Many years after I did that work, when I first came to MIT, there’s a very interesting phenomenon called TADOMA, T A D O M A.
Probably you’ve never heard of that word, but it’s how deafblind people communicate. Think about if, for a second, if you close your eyes, and you are deaf, and you cannot see. How do you communicate with another human being? Helen Keller, if you remember, was DeafBlind. So DeafBlind people, if you go to see a DeafBlind person, or you want to communicate if they learn this technique called Tadoma, they’re able to actually understand you.
And what, so, there was a speech processing group at MIT in 1983. And MIT used to have this thing called UROPS, Undergraduate Research Opportunities Program. I think I mentioned to you in podcast one how I worked with Chomsky to understand the caste system when I first came in. But one of the other UROPS I did was to understand how TODOMA worked.
Again, I was very interested in understanding pattern analysis. So here, what happens is, in TODOMA, the deafblind person takes their hand and they put it on your face. And interestingly enough, they put it on your face. So they’re actually getting different signals and the purpose of our research project.
There’s a guy called Nat Dorlak, which was the faculty sponsor. And I was working with a graduate graduate student called Greg Skarda, who was doing this for his master’s thesis. And I was just helping these guys out doing the signal processing. So the idea was to find out what are these people listening to?
Because they’re actually able to understand
you. Are they? Is the other person talking? The other person is talking. Okay, so they’re Feeling the, the movement of the, the lips,
maybe. That’s one. What do you think, Richard? I don’t know. Yeah. So what are they listening to? Well, let’s just step back. It turns out that one square inch of your finger pads have more sensors than your retina.
Oh, okay. So think about that. You ever hear the thing, you can, you can look but don’t touch? Yeah. There’s something to that. Because your fingers are amazingly sensitive. So what, so what we found out was they were actually listening to seven different signals. So when they put their hand on your face, they’re listening to the, your upper lip moves up and down, your lower lip moves up and down, but it also moves horizontally.
So there’s three signals there. They’re actually able to sense where your tongue is, the position of your tongue on the palate. Okay. They’re able to also understand your breathing. And your jaw movements. So it adds up to around seven signals. So by purely putting their hand on your mouth, they’re able to capture seven signals.
So what we did in the lab was we literally created a device where, first of all, we actually brought in people, had them speak cat, dog, mouse, and actually had their waveforms. And they were actually measuring these seven signals. So we were creating a database. And we were using these features, these seven features, to map, to cluster to these words.
You following me? So here the cluster were words, and we had the seven signals. And then we were doing learning, to figure out how these seven signals matched to actual waveforms of individual words. So as you did more and more and more of this training, we literally had one of the guys in the lab had actually built a retainer.
It was a very interesting retainer you put into people’s mouth and the top of the retainer could measure the tongue position. So we knew where every tongue position was. We knew the waveform of their breathing, the upper lip movement, the lower lip movement. We had all these seven signals just with what that person was doing.
Then we were matching those seven signals to the words. So here the signals were seven waveform signals being matched to, I don’t know, 20, 000 words in the English dictionary. Okay? So they
didn’t have to do all that to learn that, so how did they learn it?
Well, they were doing so the human brain. So the thing is exact.
Great question. So the human brain is doing this learning and we were doing this learning on the computer. Okay, but
how did they learn it in the first place? Well,
so how is a brain doing that? So there’s different theories about learning. One of the modern theories is that that the brain actually has computing centers like it actually has computing.
Okay. And it actually may be framed around an artificial neural network that you give it an input and you give it an output, give it an input and you give it an output. So, for example, when you’re learning to ride a bike, you’re learning and you say, Oh, when I distribute my weight that way, I keep going forward.
Oops. When I do that, I fall. So the brain is getting inputs and outputs, inputs and outputs, inputs and outputs, just like those chess masters. You give them over and over and over again. They learn. And the brain learns to generalize. Okay, when it sees some other pattern that normally didn’t occur, it’s actually able to even learn from that.
So, the, the, the, one of the theories is that it’s through things like, what, by the way, this is called multi level non linear regression. Basically, a bigfalutin term to mean your brain is doing probabilistic calculations. And it’s learning and learning and learning. So it turns out that, again, repetition is a mother of skill.
I think I mentioned to you that I think there’s some guy wrote a book. I don’t know who it was. If you want to learn something, do it 10, 000 times, right? So if you want to learn to be a great tango dancer, do it 10, 000 times. Now some people can do it 5, 000 times, some people 7, 000, but there’s a certain number of repetitions you have to do that you can learn.
And that’s, by the way, when you train neural networks. That’s what you’re learning to do. So, for example in 1993, I wrote a paper with a number of other colleagues at MIT on handwriting recognition. So, if you, I don’t know how many of you write a bank check anymore, a paper check, but if I were to write you a check for 100, I’d put Marcelo Guadiano, and you know, in the little box, I’d put 100.
00, and I’d put, and I have to write 100. 00. dollars xx slash, right? So the National West Bank in…… Western Bank in London funded a project that I headed up, to see if I could create a neural network system, to see if you could tank a bank check, scan it, and could I predict the exact number, like 100 bucks? Because everyone writes different handwriting. So how did we do this? Well, first of all, we had to do the feature extraction, right? Which was to be able to read that box on the check. And extract out the number of digits. That was a feature extraction.
And then we had to do feature extraction for each digit. Because people write, some people just write a one with a straight line. Other people put a little hook on it and write it, write zero. So how did we get all the possibilities? We actually went and got census data. Turns out the United States government saves, when you fill out a census, all your handwriting.
So we had a big CD disc. Of all the ways people write the number one all the way people write the number zero because you just needed to learn that 10 digits, right? So when whenever you drew that digit, we learned the stroke sequence. So we didn’t again have to change all the learn all the pixels. And we save those stroke sequences for each digit clustered them and we train the system.
So it could, it got very good, right? 98, 99% accuracy that we could figure out a zero or one, etc. Anyway, so I built this whole system and I remember this was in the nineties. I had to go do a demo. I was literally through Heathrow airport from Boston carrying a huge scanner on my backpack with the computer and we went demonstrated it.
This is in 1993 and that was, and that paper was written in the international journal pattern analysis and recognition. So, so again, we’re using the same methodology for recognizing handwriting, same methodology that’s being used for analyzing to Doma. So, you know, for me, this was a journey starting in 78 with SIDS in the 80s with Tadoma in the 90s with handwriting recognition.
So in 1993, 1994, I was very much into pattern analysis, and I came to the conclusion that the whole field of AI was a sham because everyone was doing the same things, which was feature extraction, clustering and learning. They were all putting different terms on it. To own their academic discipline. So when you did speech recognition, they called it speech recognition.
Other people called it you know, in their field. And they’re trying to, and other people would call it sleep analysis. But if you looked at the fundamentals of this, it was feature extraction, clustering, and learning. It’s statistics, yeah. It’s statistics. So I said, couldn’t I build a common system that I could use a new technology, a platform, That you or Richard could use to recognize to use it applied broadly to all different types of universal problems, whether the signal coming in was a sleep pattern of a baby, whether the signal was a handwriting thing or the signal was a document or speech.
So that was the basis of my thesis work, which I called information cybernetics. So when I came to him, so in 93 94 after I finished my masters, I. And by the way, my thesis never fit in any one department. It wasn’t a mechanical engineering thesis. It wasn’t an electrical engineering thesis. It wasn’t a media thesis, right?
Because I had degrees in all those three. So I actually had to create my own department. It was an interdepartmental program. It had to get approved by the Dean of the School of Engineering. We had to fight for it. And so I was doing that work starting in 93. And I was making headway. into creating a whole new foundational base for doing any type of pattern analysis problem.
And while I was doing this I was called to participate in a competition for doing, analyzing President Clinton’s email. This, too, was a pattern analysis problem. So if you remember in 93… Not Hillary Clinton, right? No, Bill Clinton. No, this is Bill Clinton’s email. You
say Clinton and email everyone thinks, Hillary.
Right, right.
This is to do. So what? Remember what I said? The invention of email? The system, as we know today, took place in 1978 and email was used in the office environment. In fact, in 93 when I used to do seminars room full of 1000 people. I asked how many people had an email account? Very few people had an email account, maybe two people.
But if you went to an office environment like Lotus or IBM people had email accounts. And by the way, you don’t need the internet for email. People were using email by interconnected computers. Right? But in 93, if you remember, something interesting occurred. The World Wide Web came, WWW, which was a front end graphical user interface to the internet, which made now the internet accessible to everyday people.
Point and click. Well, when that occurred, people started creating web based versions of the email system that I invented. Hotmail, Yahoo, all these came out. So what did that mean? Now, millions of people started getting email accounts. And millions of emails start getting sent. So, the White House, before, when people wanted to write to the President, they would write a letter.
Dear Mr. President, I don’t like your program on education. Dear Mr. President I don’t like your program on the Middle East, right? Or I support your position on education, right? Well, when these letters came into the White House, physical letters, the White House, believe it or not, had 147 buckets.
And these buckets were termed education bucket, drug bucket you know, threats. And for each one of those buckets, they had a form letter. Or sometimes they would get routed. And they literally had a form letter. If you sent the president email, you know, I support your policy on education. He would say, thank you very much, Marcello.
And here’s my, you know, they’d have, here’s my real policy and they’d have a staffer sign it and they’d send it out. It was called the correspondent management system. So print mail set in, they bucketed into one of those 147 buckets and they’d send you out a print mail. Well, when email came, the White House was being inundated.
Okay. And they were using interns to do that. And again, I said, probably shouldn’t use the word interns with Clinton, but that’s what they were doing that. So as email volume starts growing now, think about what they would have to do that time. Lots of interns. So now they’re getting way more mail. Well, they’re getting it because email is so much easier to send.
You don’t have to put a postage on it. So the emails are coming in and guess what the Clinton White House is doing. They are treating email like print mail. So when an email came in, they would print it out. And then, they would only answer back to those emails which had a printed address. So they’d print out the email, they would look into one of those 147 different buckets, and they would send you back out a print letter to your email.
Seriously. Do they not think it would be easier to email them
back? No, you have to understand, people are, they weren’t ready for email, it’s such a new thing. That’s what people don’t understand, this was new. So the White House runs a competition, an AI competition, through the National Institute of Standards, which used to run this conference called Tech’s Retrieval Conference, which was like an AI conference.
And six publicly traded companies were invited. I was the only graduate student, which was because people knew I was doing something cool. So I was invited to participate. And this is what the… government did in this contest. It was a blind contest. They gave you a bunch of emails and they gave you the 147 buckets.
So the emails are what? Your input? The 147 buckets are your output. So you had to take those emails, and there were too many that you couldn’t do it by hand. Okay? So you had to, your system was measured on its ability to take an email and would it put it into the right bucket. They already had scored the right bucket.
So you’re looking at certain words, probably.
Well, the input is a document, which is an email, and you have to put it into the right buckets. Okay? Well, the long, long net of the story is I use this technology I created, which was a multi hybrid solution. I didn’t just use one technique. I threw everything at it because I’d realized that all these academics were doing little simple solutions and then they were blowing it up.
So I threw, so in my feature extraction, I had like seven different feature extraction methods. I had like 10 different clustering methods, and I had 3 different learning algorithms. You see, I took an engineering approach. I wasn’t wedded to any one approach. Anyway, long story short, I ended up winning this competition.
So you had to
find a way to separate the emails into each, like… One of the
buckets. So if you got an email, they already had scored, this email was about education, this email was about this, right? Yeah, how did you… And some of them were death threats. And they had multiple different types of threats.
So how did you figure it out? Well, so I had built a set of ways to do the feature extraction, right? And what I came to the conclusion was that every email had an attitude. Okay. Every email had different things people wanted. Yeah. Every email had a different type of object of concern. Every email had different issues and every email always told what people wanted.
So I found these five features, five to seven other features. This is the art of pattern recognition. Yeah, so I did that art because I just had the aha moment. I said, you know what? When people write an email, they think they’re writing to someone. They say, Hi, my name is Bill. I’m a CEO of a company. President Clinton, I really like your position on education.
I’d like to know what you’re going to do for small businesses. Can my son get a tour of the White House? Right. And can I come and speak to people? So there were multiple issues in there. So I figured out that thing and I started building clustering algorithms based on keyword frequency. And then I learned from that.
Okay, so I built this very powerful system, ended up winning. And I, in fact, won one of the MIT still has this award called the Lemelson MIT Awards, MIT Lemelson Awards, and I was a finalist that year. With four other people, it was a pretty big honor to be a finalist. I think the year that I won, another guy built a flying car.
Okay, so, but I was a finalist. And after I did this, my lawyer said, Shiva, you can always do your Ph. D. And the internet is exploding. And he said, why don’t you leave and start a company for a couple years? So I left MIT. I went to MIT and I said, look, I built this. I did it on MIT time. You have to give MIT the rights.
Do you want, can you give me a waiver? So I … MIT said, ah, we don’t think this is going to go anywhere. You have to give
them the rights because you’re working. Yeah. Cause I was
at graduate student at MIT, but MIT has a provision. You go to them and their technology licensing. The way MIT’s intellectual property structure works is if you develop anything.
And you’re being funded. I think I may have gotten some grant. I don’t think so. I don’t think I was even being funded by… I was paying my own way through. So, but I still was in following the ethics. I went to Lita Nelson. I said, Lita. I think I built something. She goes, I don’t think this email management stuff is gonna be important because remember, this is 93.
They didn’t know how big email was gonna be. So they gave me the waiver and I left MIT and I started a company called EchoMail like the waves, you know, because one of the areas of pattern recognition I was also doing. Was a military project. I was working in a lab that was interested in by the way, when you build wings of airplanes, they’re made with composite materials and sometimes deep inside that material, some of the fibers can be breaking.
So we used to send ultrasonic waves into the thing. And based on the wave that came back, could you predict what was going on without having to open up a billion dollar wing or a hundred million dollar wing, right? So that’s why if you look at Echomail, the symbols were these three waves because we’re sending.
So Echomail was the company I started just jumped out of MIT. We had no customers, nothing. And my first customer. As I mentioned before, I went 40 times back and forth to AT&T. Convinced them to use EchoMail, you know, and all I had was the core technology I hadn’t built the software yet. And so they started routing all their email to us and our technology would analyze the email bucket it and then we had humans review it.
And if they said an email was correct, we learned from that. So our system got smarter and smarter and smarter and we would get $1.83 for every email we categorized, right? Thirty cents for every email that went through the system. And I grew that to around a 250 million value company. And we sold a big piece of that.
And in fact, now we’re going to relaunch that company for small businesses. It’s called the AI in email. But the point is that the same approach I took to analyzing emails, the same approach I looked at for Tadoma was the same approach for sudden infant death syndrome. But the foundations of AI or pattern analysis are feature extraction, clustering and learning.
So you’ve basically gotten a course. In AI, you know, so when someone tries to highfalutin you, that’s all it is. And either, you should ask someone, oh, you’re working on AI, what are you working on? Are you working on clustering methods? Are you working on feature extraction? Are you working on learning?
And they’ll, they’ll, they’ll know you’ve sort of understood what they’re doing. You know? And, but that’s what people are doing. They’ll throw out big terms, I’m working on SVM, I’m using neural nets, I’m using… You know, higher order clustering, but they’re using, they’re doing one of those three things. Yeah, they just try to
make it more complicated.
Yeah,
right, but that’s what they’re doing, okay? So the field is fundamentally that. So I think you fundamentally, I think, understood the foundations of AI from a technical perspective.
I think there should be another part added to that, that we’re not thinking about is what’s the bigger picture to the AI, like what is it creating for humanity.
Is it something we should pursue, or, or not? Like, is it bad for society? Right? Yeah,
that’s, yeah, so I think we’ve, so now we’ve talked about the guts of AI. Now I think, again, you guys should jump in on this part, because I think now let’s move to the, what does this mean for you? Yeah. And what does it mean for the future of you?
So what we’re seeing is as computing has gotten better and better and better, because remember, feature extraction requires computing, clustering requires computing, learning requires computing. So as the technology gets better, we’re able to transfer more, quote unquote, pattern analysis methodology with what the human being has to the computer across many fields.
We talked about before we started radiology in medicine. A radiologist who goes through umpteen years of undergraduate, umpteen years of medical training, another ten years or whatever of clinical training practice. He learns how to look at a x ray and learns to predict, okay, does that person have a brain tumor or not, right?
Well, if that radiologist can be used to train a machine through feature extraction clustering and learning, you’ve transferred knowledge. Okay. You take something like designing something as 3D printing comes, right? As these new technologies come, can you actually take soft skills, how people do design?
Another AI project I worked on at the Media Lab was if you look at a a cereal box or you look at a poster, someone’s doing that design. You call a graphic designer. So I came up with the method that the system. would predict all different favorable designs. It would come back, so you didn’t have to go through designs.
It would say, here are the best designs, and you could check one off, right? So, the reality is we’re going to be able to take a lot of human skills, and through this process of feature extraction, clustering, and learning, transfer to, from what I call the carbon based machine, which is you, to a silicon based machine, which is something outside of you.
So just think about that broadly and think about it that you can, in fact, transfer potentially, depending on your viewpoint of consciousness from your brain to a silicon based brain. Nearly every major country is running what’s called a brain project. So what they’re doing is trying to simulate, they’re saying, if we get enough neurons, artificial neurons packed into a computer, And create all the interconnections.
Could we not create consciousness? So the theory in brain sciences goes like this. And this is an age old theory. Is it, I think, therefore I am? Which means or I am, therefore I think. Meaning, does existence precede essence? Or does essence precede existence? This is a whole mind body problem. So the spiritualists or the idealists argue, no, no, no, there’s something separate from us in the physical body that gives us consciousness.
The materialists argue, no that human consciousness is a material artifact. So, if you take a slug, you know, one of those slugs, it only has about five neurons. Human beings have millions of neurons and we have millions of interconnections. So the argument is that you could create a slug on a computer or you could create a human being on a computer if you connect enough neurons and you train it.
Okay? So every major, so the theory is if you could do that, one could argue that if you’re about to die, and this is sort of a wild concept, you could transfer your state of consciousness, everything you are, to a silicon paste machine.
That’s if you’re a materialist and if you think that the conscious is part of
your brain.
Yes. Right. So this is
going to be the… But that’s not necessarily the case, right? Well, we don’t
know. We don’t know. I’m, I’m… So here’s… In 1994, when I was working on Echomail, I had… I still, you know, I still do a lot of meditation and, it’s, it’s… You know, I’m very interested in these other states of consciousness.
So I got up in the middle. I was in this dream. And in this dream, I was sitting at a table. And at that table across from me was a what looked like a robot, and it looked like a human being, but I knew it was a robot. And the question that came to me to answer this bigger question is, what is the difference between me and that being across, sitting across from me?
And this is the foundational question of pattern analysis or AI. Because look, if you look at the arc of technology, Let’s assume one day you could create something that looks just like you, in fact, can feel, can cry. Because there’s a whole field of computing now called affective computing, A F F E C T I V E, which means computers, you can actually teach them emotions.
So, if you have, in fact, there’s a computer that as you look at it, if you smile, it smiles back at you. You know, so this is, let’s assume that we’re going to get over all those technology hurdles. That there will be something that looks like you, talks like you, and can pretty much do everything you do. And, but it’s not made up of, you know, genes and proteins and blood and plasma.
It’s made up of something else. The question is, what is the difference between you and that? Yeah. And that’s a question that came up in this dream. And the conclusion I came to was that there is fundamentally no difference, like, from a mechanical perspective. Or from an electromechanical perspective, from a, even a intelligence perspective, but the difference would be which of those creatures has asked the question about who am I, why do I exist, and where do I come from?
And that is what I believe makes us human. So I would argue that, that you could have machines which may be more human than what we call humans today.
But even if the machines argued that they wouldn’t, they would still be machines, right?
Well, this is, this is an interesting question. So what are we a machine?
Like, you know, look, my degrees and my PhD is in biological engineering. And at some point we’ll do a podcast on when you actually look through a microscope and you look at what we’re discovering in biology, man, and you look at. A piece of DNA, and how that DNA opens up, physically opens up, it sends out a piece of messenger RNA to the ribosome, and that ribosome, like a little punch thing, makes, you know, proteins.
It’s a freaking factory. And you look at that, and you see, this is pretty amazing. That we actually are a machine. A molecular, you know, going down to the molecular level. So, I would, now, how did we get here? I mean, we could have arguments, but let’s take this argument from a natural law perspective, that nature is this amazing engineer.
You can call nature God. You can assign it whatever spiritual or non spiritual value you want. But nature, in a creative way, like an engineer, has been testing and testing and testing. It’s been doing its own pattern analysis. Throwing away stuff, keeping stuff. And it’s created this thing we call the human being, which is made up of carbon and phosphorus and the different elements.
But if you go at it, man, where it’s a pretty amazing, finely tuned machine. And this machine has the ability to extract features. It has the ability to cluster. It has the ability to learn. So the argument is, if you one day had Marcello or Richard replicated with everything you can do in another machine.
What is it between you and that machine? Is there a difference? And I would argue based on, going back to our second podcast, what is a system? Are you an intelligent system, or are you an open system? An intelligent system takes an input and gives an output. An intelligent system has a goal. It has a mission.
And it is engaging in the world through some goal to correct itself, make constant changes in the face of disturbances and a struggle. So I argue that intelligence is having a goal and working towards a journey of achieving that goal through struggle, through disturbances. That’s a systems perspective.
I would argue that anything is not human and is artificial. That just takes an input and spits out an output. So think about a number of humans, quote unquote humans who think, quote unquote think, who if they’re watching. Fox News or CNN based on what they get. Then they go pull the lever for Republican or Democrat.
Just think about that. They get an input, they go do an output. Input, output. And how many of those human beings are asking, setting a goal for themselves? I want to be a free human being. I want truth, freedom, and health. That’s my goal. And I’m going to find that in the midst of all the noise and the disturbance.
I’m going to take actions to… Find that for me. So they set a goal. I want truth or I want freedom or I want health and I’m going to take the output of that what I see in the world and I’m going to engage in that to interact with the world to get that for me. And then another set of human beings get an input and an output.
So I would argue that you it doesn’t matter what what Where that consciousness resides. It could be in a carbon based creature or a silicon based creature. The issue is what is that ability of that thing to go down that path? And I would argue that we already have AI, which I, to me, AI is that ability to just take an input given output input given
output.
What I’m stuck at though is like, if we do create conscious creatures, how do we know if they’re actually conscious? Right. And like, right. So, so let’s, I think there’s gonna be a sharp, diff like a sharp difference between humans and robots that we think are conscious. Right. I think there’s a lot. Well,
gimme an, well, let me, so in my definition, I would argue some humans are robots already.
Yeah. I, yeah, I understand. Would you, so, so you, you look at, look, we had this thing when we ran our campaign, as you know, called Only the real Indian can defeat the Fake Indian. Okay. It was, it was a, it was, what is it? It’s like an image. On some plastic, which had a picture of me, picture of Elizabeth Warren and a headdress.
And it said, only the real Indian can defeat the fake Indian. So I presented that to some quote unquote humans. Some humans, they saw that input, would laugh. Other humans would say, hey, what do you mean by that? I want to understand that. Other humans would freak out and start attacking us, calling us racist.
Calling us names. Okay? Now, I would say the first and the third group are robots. The second group who wanted to understand that is a human being. So, you, you follow what I’m saying? Yeah, yeah. Input, output. So, the input comes in. They already have certain transport, conversion, and storage in them. Storage of memories.
And it’s like a reaction. It’s no different than a slug. When I poke it in one way, it goes this way. When I poke it in another way, it goes this way. Yeah. And so my argument is that the notion of an input coming in, you process in an output. That’s not a human being. Remember, we call that an open system, a human being or whatever you want to call human like intelligence is something that has a goal in mind.
And I would argue the level of consciousness is a refinement of that goal. Yeah. So you could call it enlightened, you know, from the concept of self actualization and enlightened human being. His goal is not just to get food or sex or own a piece of land or, you know, make money. Yeah.
There’s like, in other words, they’re not that like conditioned or brainwashed by the system.
Well, so here’s an interesting thing is they actually questioning where their goals even come from. So this gets into, this gets into where. Do the things that I want. So we talked about an open system input output. I would say that’s a dumbest form of a human or that’s a robot. You get the next one, which has some goal and then you can get into the quality of these goals.
So you have people, you know, when I was at when I first came to MIT, there were the students would come to MIT and I would argue and I give you two things. I think you and I spoke about this. I would literally see these so called intelligent humans act like robots. I In my freshman year, these young kids would come in, 17, 18 year old kids, who looked normal, they stood up straight, they talked with a certain tenor didn’t have any tics or weird movements.
Then they would go see some professor speak, who would talk with a highly nasally voice, had some tics, he’d be flailing his arm in a little weird way, like something was wrong with him, but therefore he must be intelligent because he’s idiosyncratic and eccentric. And I swear to God… These students would start behaving like that professor.
Yeah, I remember that. And speak with a nasally voice like this and flail their hands and use their same vocabulary, right? Yeah. You saw this. I’m telling you. Yeah. You saw these people start emulating another creature to be accepted into this framework of now you are an intelligent nerd. So what do
you think it is that makes the person like question their surroundings and…
Step out of the box there. Yeah,
this is
a good question, you know, because I’m thinking maybe it’s just like I don’t know just like kind of alienating yourself a bit from not society, but like the system, maybe the TV and right. Cause that’s when you start thinking by yourself, right? You’re not like always.
Yeah. This is, I think
this is like the heart of
the AI. It’s almost like meditation in a way where you have to be by yourself and really truly think about where you came from and why you’re here and what your goals and your purpose in life.
I, I think to answer this question, I think at some point someone has to have some crisis.
Some crisis where they are forced with the loss of something that they love or they question their own death or they, I think it’s out of losses where you start asking these questions. So for me, these questions came when I saw the loss of my own innocence, where, you know, when I was five years, four years old, where I’m playing with this kid and I go to the mother’s home.
Where this, my friend, I thought he was my friend. And the mother forces me to stay outside and gives me water in a different glass and calls me shudra, which is like calling me nigga. Yeah. Okay. In India. Because we were considered, that’s when I didn’t even know what, what’s going on. That’s when I considered there was this thing called a caste system.
So I was open to that as a four or five… And then I learned my mom was chased away. Like wait a minute, what, what the hell is going on? Human beings, there’s separation? Yeah. That hurt. is what forced me to start questioning everything. Like, what the hell is this here? I’m playing with this kid and I go to his mother.
His mother mistreats me like this. I’m no longer a human being.
You took that path for them. There’s the other path was which a lot of people take, which is almost like resentment and saying, Oh, that’s the way things are. So that’s the way things are going to be. But you as a little boy, we’re like, no, things can change.
But I
think it was my parents who had fought over that. They didn’t believe in destiny. So I think this comes to the crux of the issues. John Paul Sartre. He wrote a book called Being and Nothingness. And Sartre he was one of the areas of existentialism. Existentialism basically says existence precedes essence.
Which means your existence, from your existence, like you’re born at a certain point in time. I was born December 2, 1963. When were you born? November 17, 1994. Richard, when were you born? January 21st, 1975. Right, so you were born at that point. So think about it, you’re born, you come into this thing called the world.
Now, you don’t know anything that occurred before that. Let’s take a non spiritual perspective, materialist perspective for a second. Okay? And that may be an accurate approximation. We don’t know, but let’s assume that, right? So at that point that you were born, You are now, from the instant you were born one could argue that your entire life, from that moment, you get all these constraints from society.
So if you’re born into a Catholic family, you’re supposed to behave like a Catholic, you’re supposed to do this and you’re supposed to do that. If you’re born into a Hindu family, you’re supposed to behave like this. If you’re born into an African American family, in a ghetto, you behave, you see what I’m saying?
Yeah. So Sartre talked about this and… And he gives an example of a waiter who he sees at a restaurant and the waiter comes to serve him and he’s dressed a certain behaves a certain way and he behaves like he cocks his head at all these mannerisms and he notices all the waiters do that. Like they have to behave like that to be a waiter.
So what happens is the fundamental dragon here, which we all have to slay, is that the universe Or that the world is trying from the instant you’re born to control your existence to minimize the infinite possibilities of what you can be. Yep to a finite set of possibilities and the struggle of life is to recognize that the aspect of your existence is actually being an infinite human being there’s many… You are not your past Yeah, you can be anything and if you truly look at Marx, which a lot of people don’t look at on the left or the right.
They don’t really understand this. Marx was addressing the human condition. In the first chapters of Das Kapital, he talks about the notion that a finite set of people, a very small set of people, get to live their dreams. And the rest of us have to be controlled like automatons. So he’s already talking about artificial intelligence.
So we need to move away from this thing, Oh, no, artificial intelligence has been here for a long time. Yeah. And so the issue really here is that what Sartre was saying was that A, you don’t owe anything to your past, that what matters is the choices that you make, the goals that you set for yourself, the dreams that you have.
And one of the fundamental things is that monopoly capitalism, which means the drive to…
So why are we creating robots? Why did we create the manufacturing line? Let’s go to the depth of it, the economic analysis of it, because Henry Ford wanted to make a lot of cars for a low cost so you could maximize profit. So remember basic accounting 101. Income, net income, is equal to revenue minus expense.
Revenue minus expense. In any industry, the number one expense is labor. So you’re trying to take… You sell a pen, this pen I’m having, for ten bucks. It costs you five bucks to make it. How much profit do I make? Okay, so how can I reduce profit? Well, I get the materials made cheaper. Materials could be one line in the expense.
Or I make it cheaper. Well, I go exploit someone. Maybe the material to dig up in America to make this pen cost me two bucks. But I can go over to El Salvador or Haiti and get it done for ten cents. So I’ll do that. And then I find out, okay, everyone’s going to El Salvador for ten cents. Well, now I’ll apply machine automation.
That can make it for a penny. So what we’re trying to do, or the capitalist model, is you’re trying to lower the cost of goods and you’re trying to lower the cost of production. Yeah. So either, so the goal, ultimately, if you’re dealing with a human being, you want to treat them as a commodity. Yeah. So you want to peel away any humanity that they have.
So you, you do an assembly line,
right? Yeah. So you just become, what do you like you’re saying a cog in the machine and you… of capital
of capital and you lose the individualism because everything’s done for the corporation. So it’s like, or for the
maximization of profit. Yeah. To, to be very mathematical about you wanna maximize
profit.
Yeah. But like, it just, I’m thinking of any huge corporation nowadays. Everyone gets so caught up in doing everything for the company. It’s for the company. For the company. You have to make it grow. It’s like it takes a life form of itself. It’s an organism in itself. Right. It’s an organism onto itself. And that becomes the most important thing in life.
Right. And you lose all that sense of individualism. Right. And I guess
it, well, the ultimate, the ultimate form of this. is what’s going on right now that forget even labor. There are people who just move money around all day. Yeah, they don’t do it. They’re so far removed. They’re moving capital around and they move, by the way, $600 trillion is our economy in the world.
$600 trillion is being moved. $600 trillion. It’s unfathomable what that is, right? So as that, you know, we think the United States, by the way, is 20 trillion. That’s our GDP, but 600 trillion is the amount of money in the economy. So that money’s being moved. There are people who move money from point A to point B and make billions just from that movement of capital.
They’re not doing anything productive. Yeah. What you would call productive work labor. So that movement of capital and there’s machines now which are getting ready to do that. Okay. High frequency trading machines. So now you have algorithms competing with other algorithms. So the when you really look at it, the dehumanization that monopoly capitalism affords is at that level.
Yeah. And so talking about what makes us human. And relating that to AI, do you think we’ve gotten more conscious because of like, the internet and, like, all this information that we’re constantly exposed to? Do, or do you think we’ve kind of, like, dumbed down and, cause everyone’s, like, so caught up in their phones?
Do you think we’re more, like, input, outputs than ever? Well,
let’s, let’s think about it this way, right? If you go back to what I just said… about what is the focus of monopoly capitalism, and let’s talk about the notion of this machinery that’s being created. Okay? So, in my book, The Future of Email, I try to teach people that there’s two principles, control and observability.
Controllability and observability. Just sort of just think about those two words. So the machine. wants to be able to observe you at all times and be able to control your actions for, because it has its goal. Okay? So the few who, I don’t want to get into a conspiracy theory, but just think about the maximization of profit, right?
So think about Edward Bernays, the guy who’s known as a modern, modern guy of advertising. Okay? The entire goal of advertising is to sell you something as fast as possible. And so they’re so what they want to do is to identify groups of people who you can sell stuff to and then control their behavior.
So they’ll eventually through a series of interactions by something. Okay? So in order to do that, you need to observe, get lots of data. And then at certain points, you need to control their behavior. So they’ll eventually go to in the case of Edward Bernays, you know, the guy who started doing some of the early cigarette advertising.
Drive you to go and buy a cigarette. So some of their earliest ads, they understood, wow, the women’s movement is coming. Let’s have women look strong, you know, and smoking cigarettes. Okay? We will take advantage of this movement of the growing independence of women, because they observed that, and move them towards buying a cigarette, showing smoking a cigarette makes you independent.
So the ability to do associative memory like that, juxtapose Two things to go to a particular goal. So observability and controllability. So Those in power want to be able to as many many Sensors as we talked about Get lots of data. So how did they get data before the internet came? Well, they did surveys they call up people Right, they get survey data put into a database and say oh these people over In Massachusetts, like lobster.
Okay, they try to put lobster ads, right? Oh, these people down in Florida, sunny environment, they like suntan lotion. It’s pretty easy, right? We’ll market suntan lotion to them. Well, now it’s more sophisticated. Now on the internet, with social media, they can watch every click you do, everything you’re viewing.
where you’re scrolling, etc. So now, every one of us in some database there is an email address of you, some identifier and they’re collecting oodles of information about your behavior, right? And from that input they’re extracting features, feature extraction, they’re doing clustering Oh, Marcello Guadiana a student, right, likes journalism, libertarian.
Richard Giorgio, Cambridge resident. You know, independent thinking, right there. So they’re clustering you and their watch your behavior. So now they have you categorized, which is clusters. The next step is controllability. Okay. When Richard is doing this activity, for example, on this website at this point, I’m going to hit him with a particular message.
And I know if I hit Richard like 18, he knows they have to hit people 17 times of the message and they get an output. I’m going to hit Richard 10 times, five times. They know it’s called the conversion rate. So you are basically being modeled. You, your dynamics are being known. You are basically in the larger world.
Part of AI yeah. Okay. Part of a larger pattern analysis system. So how do you break from that? Well, are you a particular cluster? Okay. Are you going to let yourself be observed? Are you going to let yourself be controlled? Because ultimately, the goal here is to think of you in one of those clusters. Yeah.
And then manipulate that cluster. So the idea really comes down to what about you’re not in any one cluster? What about you’re in many clusters? What about you’re more chaotic than they understood? Hey, I may be an engineer, but I also may be a politician. I may be a writer. I may be an artist. I may be
infinite.
That goes back to what we were talking about earlier about breaking these boxes that people try to put you in. Right. In the time you’re born. And now it seems like that’s the, the, the plan. That’s the,
The idea is what you just said, to put you into a cluster. It’s that middle box I talked about. But now
they have to do it to maximize profit.
Exactly. So it’s like now it’s gotten way worse then. Right. To answer my question, right? Cause. It’s gotten
way worse. And what we thought the internet was, was a liberating medium. Could actually be a vehicle now to even centralize power even faster. Yeah. You see, whenever new technology comes, everyone thinks, Oh, the printing press came.
I can all print Bibles. We can all print our own documents, right? Like we, we can all be publishers. Well, what ended up happening is you created five major publishing houses. It’s hard to get your book out on the New York Times bestseller unless you’re on the end. Okay. When the internet came, the vision was, Oh, we’ll all be able to get our knowledge.
Well, now you have Facebook and Google and Amazon. Okay. In fact, you could argue that you could be wiped out faster now than ever before, which means I remove your name from a Google index. Marcello Guadiana does not exist. So what we’ve done is we’ve seen the consolidation of power, and the only way to overcome this is to actually become a true human.
And this notion of becoming a true human is a deeply, deeply important one because it means you have to ask, Who am I? Am I an American, right? Am I a Cambridge resident, right? Am I this or that? People have to recognize that they have infinite possibilities of what they can become. And that they have to say, I don’t owe jack shit to what I was born in and what I owe in the past.
Because every, to your point, every decade that goes by, The manipulation is going on to When Richard has a child, or you have a child, they already know your past. Yeah, that’s right. So now they know the trajectory of your children, right? That’s just going to get at. It’s called cradle to grave
marketing.
It’s going to be impossible to break in a way, because we’re, we’re raising generations like this, so the kids are going to see their dad act like that, and that’s just going to be the way it is. Well, that’s
why the only way out of this is what I call, you have to have a, the new food, the new mana, is not land, bread, or peace.
It’s the mana of truth, freedom, and health, which is going to come from a system’s thinking. People need to recognize that they need to start having a new framework for thinking. They need to recognize that everything that they are comes from observability and controllability. And that they need to recognize that they don’t owe anything to their past.
Meaning the past, I’m not saying you don’t owe anything to your parents or those kind of things, right? But I’m saying from a thinking perspective. Otherwise, you are a robot. I’ll give you another example. When I first came to MIT, you know, I’m an Indian American born here, but they’re the Indians who came from India.
The graduate students. These guys were robots. They came here from India. I’ve come to MIT. I’m so happy to get in. I’m the best in the world. I’m going to get my graduate degree. Then I’ll go back to India, get an arranged marriage, get married. Then I will come here. Live here, get my BMW, I’m a success. Done.
They’re on a freaking program. Right? Total freaking robots. Yeah. Would you say intelligent? I would call them just basically educated idiots. So what we’re creating that is a stream of educated idiots. And even more disconcerting is these people actually think they’re intelligent. Yeah. Well, that’s what people, I call
that artificial intelligence.
Well, that’s how a lot of people define intelligence nowadays. It’s just like how if you know math, science, like how much money you make. But the other huge question is though of humans becoming cyborgs. So, Elon Musk has talked about this a lot. He was, at first he was warning us of the dangers of AI.
I know he’s saying we need to integrate with AI. In order to not let us out, let it out compete us, right? So,
we have to all, you see, I ultimately believe when people come up with these ideas, one has to always question that person and their surroundings. Well, Elon Musk grew up in South Africa where 97% of whites control 3%.
I’m sorry, 97, 3% of. Whites control 97% of Blacks. And now he’s telling everyone what to do. Well, I’m just saying, you can’t not… He was part of the elite there. Yeah. His family profited from exploitation of lots and lots of people. So, does that affect his thinking? I don’t know, but I can’t… As an untouchable from India, I know that has affected me, so it would be…
So, his thinking itself is… comes from a notion of human beings or commodities and they exist for maximization of someone else’s capital. So, in South Africa, a lot of blood diamonds, a lot of people mining all sorts of rare earth minerals. While, you know, a lot of, and by the way, they also, there were a lot of black people, not a lot, a black strata who also took advantage of their majority.
Right now, what’s going on in South Africa was they never had a good revolution. Oh, yeah. They basically promoted numbskulls like Mandela, who was basically a sellout, and it’s a whole other discussion we can have. You know, I met people who were in South African prisons, blacks and whites, okay? Mandela was considered a, basically a, not that smart of a guy.
The whites in South Africa promoted him, just like they promoted Gandhi in India, to quell a real powerful revolutionary movement. And so they had this basically a bunch of white people left and a bunch of black bourgeois came in. And that’s why you have all the corruption there. They never address the fundamental issues in South Africa of exploitation.
Okay? So now they want to create the black white thing when it was really about massive exploitation of labor. That’s what happened. So now Elon Musk comes from that background. So his view on labor and the essence of being human is driven by, Oh, we’re all going to be cyborgs. Well, that’s a choice we have, and this is something Sartre talks about, too.
At any point, each one of us has a choice, even our emotional decisions we make. So they want to remove, there’s no choice. Elon Musk says this. Okay? At any point, we as humanity have a choice. And that’s something they want to remove, because that’s part of being a human being. To recognize that you do have choice in any
instant.
And they kind of want to ease you into that by getting you addicted to your phone, right? That’s the start. Getting,
yeah, yeah, yeah. So, so, you know, Mark’s talked about that. The goal would be we would fight against the machines. We would separate ourselves, when in fact what’s happened is, we’ve become so close to our
machines.
I mean, you probably, that’s the first thing you look at when you wake up as well. When you… Like everyone’s guilty of this. Some people, yeah,
I think a lot of people now do. Yeah. Right? So we’ve become so integrated to… So the issue is, will these machines then just become part of us? And to me, this is where the deep, deeply spiritual question comes.
You see, in the Indian, or the Hindu spiritual tradition there is the concept of Koshas. K O S H A S Where they believe that there’s the essence of who you are, and layered on top of that were these different bodies. So, there’s your soul body, then there’s your etheric body there’s your causal body, the astral body, and then the physical body.
Okay. So if you remove the physical body, you, you have the astral body, you remove the astral body. There’s a causal body. You move the etheric body. You come to the soul body, which is who you truly are. Your true state of consciousness. Yeah. When all these layers have been removed, the causal body represents your past.
I mean, they’re, they’re, I mean, I’m not getting, we can do a whole podcast on that, but the notion was that who you are is something you don’t even know who you are until you go inside yourself and you remove yourself from the illusion of what you think you are. Now, here’s the interesting thing, and I’ve, I’ve thought about this.
So if you believe that concept, we have these different sheaths on ourselves, like different clothing we wear. Yeah. And most human beings live in the physical body now. There are certain spiritual traditions, which believe you can peel away those layers and you can understand who you are in its essence.
Okay. Different than what Sartre proposed. But, here’s the idea. Suppose one day, Richard, you may want to think about this. Let me pose to you that one day Not in the too far distant future, I give you a set of contact lenses. Alright? Where it is no different than your screen on your iPhone, but 3D. Yeah.
Which we can do. The Oculus, but imagine it’s just embedded, from the time a child is born, it’s embedded onto that child’s eyeball. And I give you hearing,
or it’s injected in the fetus, right?
Yeah, so some, but we know we can definitely probably put contact lenses on it. And I give you hearing that you can hear things not only locally, but you can tune it to anywhere in the world.
Okay. And then I have a hologram of you, which we can do holograms and they’re getting better and better and project you. So now, Richard, I’m talking to you, and I’m wearing this gear, and you’re wearing this gear, Marcel, and you are, Richard. How do you know you’re physically here? And let’s say you grew up with this, year over year over year.
Generation over generation. And this became the norm. Just like the internet is going to fade into the background, this too would fade. You’d forget about it. Yeah. This would become… I could be a million years… or a million light years away. But this room would feel real. Because, by the way, they have haptic sensors where I can feel physical things.
So you would be, you and I would be in this room. You could be a holograph image that gets better and better and better. Fidelity and computing gets better. And we would be in this room. This is a real phenomenon. I could touch your sweater here, feel it, because I have haptic sensors. So here’s an interesting question.
How do we know this hasn’t already happened? Yeah, that’s exactly what I was thinking. Because that’s what the Indian mystics argue. The Indian mystics argue the purpose of existence to understand who you are and see things as they truly are. So in Buddha’s, one of his last great lectures, he gave two lectures on what’s called Vipassana and Anapana meditation.
Which he said the idea is for you to see things as they are, and there’s a meditative methodology he gave. And what he said you would find is that everything is impermanent, that you are covered with these sheaths, and this is an illusion. What we see. So, what, let’s say we did that, okay? And then someone else did that on top of that.
This could keep going on. Yeah.
And it makes me think maybe at some point in time, we only had one sheath, the spiritual sheath, and maybe that’s how most people interacted and lived. From a spi… yeah, so And then over time, all these different disturbances governments, whatever, people started Well, they created different sheaths, right.
Created different sheaths, which just became the norm. Which people born. Nowadays they’re, they’re kind of in a way, like right away, born with all these sheaths or get all these sheaths over time. That’s why we always like, think of children the way they think. It’s like amazes us sometimes of like what they say or like Right.
How they think about certain things.
Right. Yeah. So ultimately that’s what this is about. Yeah. The people who are writing a lot of stuff about AI, a lot of these people don’t even know what AI is. I think in this podcast you guys have really learned that it’s basically statistics and it’s trying to take I mean at a mechanical perspective. The other piece of AI here that needs to be understood is that it is really an opportunity to ask this deeper question as we’re doing right here. What does it mean to be a human being? And because the and what choices are we going to make? Yeah. Are these choices that we’re making or is Elon Musk making it who is basically concerned about being a billionaire and leaving some legacy behind makes a lot of money off government subsidies.
Turn into like a religion. He basically wants to upload his brain into like a machine and forever, you know, then he lives forever. That’s kind of their thought if they upload their consciousness into like this AI that we have to live forever and they’ll be gods. It’s kind of dangerous in my opinion.
Yeah, why do you think it’s dangerous, Richard? As we were talking earlier, this depopulation. What happens when AI becomes too powerful and they want to start getting rid of humans. Yes. So let’s talk about depop…. So if you go down, if you take capitalism to its full conclusion, right? Revenue minus expenses equals net income.
Okay, so let’s say we’re in this phase one of AI deployment. Phase one looks like this. Transfer as much human knowledge and skills to silicon base machinery. Okay, so phase one ends where? Pretty much none of us need to work. Meaning, you have machines doing pretty much everything. Mowing your lawn you know, being your plumber.
I mean, you could take this to anything, right? Yeah. Pretty much anything you need. So now, you have a bunch of humans. By the way, you know, we have three economies right now. We have one third of people in the United States do not work at all. They don’t work. Another third live in the gig economy.
Which is, they do a gig, contractors, like an Uber driver. And the third, third actually still does work, professionals. Yeah. And those three economies are separating at light speed, by the way. So, AI would support the definitely more and more people would not need to work. I mean, you could outsource a teacher.
You could start outsourcing a lot of stuff. So, you could then argue in phase two of this. World of AI in a, in a, in a reason. If it was a quote unquote good model, you work maybe one day in a week to do creative stuff or, or something that society needs. Okay? And the other six days you get to hang out and just be creative because
only people need it would be the ones, the people working in a, in AI and, and
making, in AI, making the AI better and that kind of, and eventually, AI would self itself manages AI.
See that machines managing machines. So that, but somewhere, let’s say that happens, right? Now, those in power, who actually own all of this machinery, to do the AI, are going to say, shit, why do I need all these human beings who are just like not doing anything? That’s what the way
they would see it. That if the AI hasn’t already taken over.
Well,
yeah, right. Or if the AI hasn’t taken over, or to your point, Those people become part of AI because they become cyborgs, right? So they become machine like and then so you have the complete fusion of the silicon creatures With the previously carbon creatures who have enough money to become silicon creatures, right?
Because as you said a guy these guys Billionaires would say I’m gonna live forever. I’m gonna basically transfer my consciousness and I’m gonna control all the means of production like so Amazon Jeff Bezos Basically, his brain is in Amazon. Okay, or he may have choices. People may have choices. I may transfer my body into a physical body or other bodies.
I mean, there’s all sorts of combinations of this. So they’re gonna wake up and say, well, why do I need all these other 7 billion people? Everything’s being done. I mean, I got, I got my farming done. I got this done. I got all these machines. Why do I need these people? That’s the question that’s gonna come up.
Oh, the pollution’s getting worse. We have too many people. Let’s depopulate.
And it’s more likely to happen if we if humans actually do integrate with robots and become cyborgs because the robotic, like, way of thinking would be like, yeah, like you’re saying, they’re useless. They’re polluting the, humans are polluting the Earth, so let’s…
Kill the ones that aren’t cyborgs or just kill most of them. So
right so more for me Yeah, right. So if you look at what’s going on with countries like Africa, China Very wealthy billionaires, you know, I I I have some guys that I know are Very wealthy billionaires, and I’ve been watching over the last 10 20 years.
They’ve all been going to Africa You know under some claim that they want to help the darkies there with malaria Or they want to help them with their water problems or help them with their food. But what, if you follow what they’re doing, the many of them set up nonprofits and they’re buying huge tracts of land.
Yeah, all, so Africa is like the next big wilderness for them to own for themselves and their kids. Yeah, and they know that. And they know that. So, it’s like the next West, right? Think about it. Massive amount of resources, beautiful lands, etc. So you have to really start wondering what these people’s goal is.
What is their intelligence? Where do they want to take this? And I think the purpose of this podcast is to really start empowering you to realize at any point we as humans have choice. We can choose to what technologies we want to implement, what technologies we don’t want to. And the military industrial academic complex works together for its own existence.
So, military, towards what we call the industrial complex, is these massive high tech companies, and academia. Academia is essentially a bunch of very narcissistic, frankly, a lot of robots, who are not so smart, but were robots, who serve the military industrial complex to preach to others how great all of this is going to be.
For themselves, frankly, and
mostly view the world in a reductionist
way, in a reduct, in a reductionist way, in, in a a way that profits a military industrial academic complex. Remember most of these academics, what’s happening, academia has become robotized. The old days of academia giving tenure was that it would be a place for free speech and you’d get to talk a lot of stuff.
But now in the seven years from the time you’d become. Wanting to be an academic, they get rid of all the riffraff, they get rid of all the rabble rousing and we end up with a bunch of robots in academia itself. So we’re basically creating robots. So I think that’s what we really want to really leave with is you, the future of you, depends on the choice, recognizing that you have choice.
You know, what choices you make for you. And ensuring that at any point, are you getting more truth? Are you getting more freedom? Are you getting access to control your own health? And if those three questions are no, you’re heading towards to this world of massive exploitation.
It almost seems like if the whole system’s going down that path, there’s, is there…
I mean, you could change it for your family, but for humanity, can you change it? Because, I mean, going back to modifying… Your kid, seems like all parents are going to want to do that. I don’t know, I was going to ask you this, like if you had a kid, would you actually, would you do that?
Yeah, so this is now into genetic engineering.
Yeah, into genetic engineering. Yeah. Cause now, like I was looking at the stat of how many kids young Americans are on Ritalin, and it’s 4. 5 million. Which is like, it just keeps on going up and up. Yeah. And for parents, it’s very easy for them to not think of the side effects and just… Give it to them just so they could compete with the other children.
They think they’re competing. And I feel like that’s, that’s what the future is gonna be like. You can modify your kid, you can make him have a 300 iq, 300 level iq, so why not do it? All the other parents are doing it, if are doing it. And if I don’t do it, my kid’s not gonna be able to compete with the other kids or get, they’re be able to get a good job.
So everyone’s gonna do it. Like we see that pattern with Adderall and with big pharma. Is there anything we can do? I mean, don’t you see the future?
Well, again, it comes down to what the goal is, right? Yeah. So, there’s a, there’s a paper that just came out written by these scientists who looked at 257 hunter gatherer societies.
By the way, hunter gatherer what they call the quote unquote the noble savage. It’s funny, there was a, yesterday I was watching a show on the HADZA people, H A D Z A. Who are still the last remaining hunter gatherer societies. One of the few that remain like this. And they had all these people from these elite institutions go study.
It was hilarious to watch because these people are just looking at them, thinking they’re a bunch of idiots. And the way these people are talking about them in the documentary, makes them look like a bunch of morons. It’s just the way they’re talking as though, A, they know better and blah. Just the way they’re talking this high level intellectual nonsense.
My point is this, That hunter gatherer societies with this one. If you looked at their health, they had their, some of the key, they, they were getting higher macronutrients, higher micronutrients low you know, they ate low glycemic index foods. They had a higher base alkaline based. They had, their bodies were more alkaline.
I mean, you go down, there’s about, it was a multi-system model of health, none of which we follow right now. So we have moved the whole area of health. into manufactured products, the stuff that we put in your mouth. They ate very little salt, right? It’s a completely different world, and all this is, came from the time when we moved away from a hunter gatherer model when we employed technology to domesticate animals, and we created this.
They lived a very rough life, probably, but they had community, they had other things that we don’t potentially have now.
They could have been happier in general.
They could have been happier. So the issue is what is the goal? Yeah, it’s the goal maximization of profit or is the goal maximization of Something else you call happiness Yeah, so what is what are we trying to maximize?
So that’s really the fundamental question
And that’s the struggle you have with technology and the medical field, right? Is like trying to actually help people and not get caught up in just trying to make profit.
Right, so what I’m saying is, so this leads to the farther question, is we think human beings know everything.
So this comes down to this thing of, of the concept of, as we talked about in the last podcast, the concept of a human centric worldview. So we have this notion… That we know everything that, or maybe forget even getting into the spiritual thing, just from the mass of the universe, how big it is, how billions of years it’s trillions.
It’s let’s say billions, right? Our existence is like a little small, little probably, you know, a single frame in, in a movie that goes on for millions of years. Yeah. So are we saying that we. We know how the natural laws that created us, we know how to manipulate those laws better than nature. You see, this is the outstanding question.
Are we a part of something that goes where we’re not the creator of everything that we think that’s around us? Maybe it’s a systems way we should look at life. There’s an interaction between us, the rock, you know, the land and the animals around us. And that our thoughts even come from that. Yeah.
Right. So, so, so where does this understanding come from? And I think this is the open question and, and who’s deciding all of this. For example, do we even know the origin of where human beings came from? Well, the way the academic model works, you get a, you get to win a Nobel prize. The first one who writes it, well, how do we even know what they’re writing is true anymore?
Yeah. Because if the motivation is to get federal funding and you get more grants, how do we even know? Their ideas of what they’re writing is even true on these soft areas, like anthropology and sociology, et cetera, because the motivation is so Harvard is the expert in the origin of human species. Well, we don’t even know
there’s a lot of evidence that says there’s like lost civilizations and
That civilizations have come and gone.
Immediately
people try to describe the academics, like, get really mad when you bring that up.
Right, right. That may be human existence, like everything else itself is a cycle. Yeah, maybe humans existed. We hit a peak point. We destroyed ourselves and we went back and this has been going on at it. It’s not a Exponential curve that just keeps like this guy Ray One of the guys who’s a Google CTO says, oh, you know, we’re gonna hit synchronicity and Ray Kurzweil.
Yeah, Ray Kurzweil. So this is a very human centric concept That we will keep growing and growing and growing as an expanding Well, well, maybe there’s a limit To at a point where we destroy ourselves, and this is some scientists that maybe the human brain is itself too clever for itself that maybe the real cleverness comes in saying, Oh, we shouldn’t go beyond that.
Yeah, you’d like. We should not develop those things. Maybe that’s the real intelligence and the ones that keep developing and growing. They do something. That’s against nature, natural law, and they just destroy themselves and they just keeps going to the cycle. Maybe Maybe the cultures that eventually do meet one another Are the ones who found this?
Range of existence where you actually make a decision. You know, we’re not gonna do that Yeah, like yeah, we can do that, but we’re not going to do that Right like they actually know Yeah, I was thinking. I don’t know what you see what I’m saying. Maybe that is a technological, like a hyper true intelligence.
Well, that’s what would make us true humans is actually thinking about that and actually doing something to stop it. Right. Because I was thinking about it in terms of like GMOs as well. It’s like, we know we can modify the food. It’s going to be cheaper. But we also know that there’s going to be some side effects like not good effects for humans.
And I was thinking about that in terms of like AI as well. It’s like you can modify yourself to have 400, 500 IQ maybe in the future. But is that really smart? Like, what’s that going to lead to? Is there going to be like a default or something? Like,
are we… Well, the issue is what does it mean to be a human?
Yeah. And I think we need to really think about this. What does it mean to be human and have a much more open understanding of it. And we cannot have that. I’m saying my end point is we don’t even know what it means to be human. Forget freaking AI. We don’t even know what needs to be human, because from the instant we are born, we are constrained.
So the only way, way out of this, is that we need to remove all those barriers which stop us from being human. Yeah. Which is a military industrial academic complex. Because the military industrial academic complex is the cyborg, which exists for its own perpetual growth, and it’s the one that needs controllability and observability.
Or that inner voice inside of you that says, you can’t do this. Right. Which is maybe based off of like society or your parents or whatever
they can say. But it’s clear that human beings do well when they have a sense of community in terms of their own health, physically, mentally, emotionally. I mean, the number of psychologists, I think over the last 20 years, there used to be like 4, 000 marriage counselors.
Now there’s 40, 000. And you have divorce, which is, you know, maybe marriage is changing and all those relationships are changing. That’s one view. But my point is, it’s not like therapists have helped humanity come together. Right? It’s ultimately going to be the bonds, the interconnections that people develop.
Yeah. Inside their communities. So what the system wants you to do, it wants you to develop interconnections to it. Yeah. It wants you to be connected to it versus another human being. It wants
the government to be your parents in a way that your mommy or them,
you know, so because that’s where profit comes from.
So the issue is you take is I think this is a central decision that people should make and people should start thinking about. Because and who is making those decisions for them? You know, this is why I think it’s really going to come down to this fundamental issue of centralization. Of whatever you want to call it, an intelligence versus decentralization.
Yeah. You know? If you think it’s going to be inevitable that we’re going to be merging into machines, well who’s making that decision? You know? Yeah,
that’s cause yeah. I mean for me personally at first that’s what I thought. And then actually like doing a bit more research and actually thinking about it more.
I changed my mind. I was like, is this really a good path for humanity? And ultimately it doesn’t look like it is.
Yeah, it’s going to be a good path. Again, it comes down to what we talked about in the previous part. What is the goal here? What is the goal? And that’s what we got to think about. And so fundamentally, artificial intelligence doesn’t think about the goal.
Human intelligence, what makes a human being a human is they’re actually sitting and reflecting and thinking about what is the goal. And if you don’t think about what is the goal, you’re probably, more than likely, already bowing down and you’re probably AI yourself. That’s what I think we’d like to end with.
Think about, are you a human or are you a robot? And whether, if you’re a robot, it’s going to be easy to put you into a silicon body. And you may want that. And then what does it mean to be human? And what does it mean to evolve? You know? Is there evolution or do we have choice? These are some more of the deeper questions that we need to ask.
And who is making these decisions for you? Are you making these decisions or are these decisions already being made for you?
If we want to combat this in a way we have to start being a bigger part of our community, right? What do you say is what I mean?
Yeah, that’s why I think, you know, I think the entire…
Emphasis. People should start recognizing is that there’s something to be said about a recognizing that there is a set of robots who want to centralize power and they and they thrive on centralization of power like they know it all. And I want to call them robots. Actually, I would say that we already have a bunch of carbon based robots running around.
Well, because some of them are conscious of the fact that they’re doing that, right? Yeah,
yeah, they’re, they’re actually conscious of, yeah, yeah, and they understand systems and they know what they’re doing, you know? So for example, you know, I live in this neighborhood here, you know, you know, it’s a, it’s an interesting neighborhood.
In the forties, there was a guy called Carl Coke and another guy called Walter Gropius were part of the Bauhaus movement who, very interesting artistic movement, and they built these homes here for 5, 000 bucks. At a time in the 40s, when people were building these huge ornate homes. And they went and built simple homes, which were affordable.
Recycled material, which was unheard of in the 40s. Where Carl Koch was almost kicked out of Cambridge for his avant garde ideas. This was an artist community of people very different, I would call human beings. Well, now I look around me and there’s a bunch of robots who live here. You have one guy down there who’s an MIT professor who’s a robot, another MIT professor over here, a robot, you know, a guy who’s a hedge fund manager over there is another robot, another guy next door who’s, who’s a robot who just lives off his family money.
You know, a bunch of robots live around me and it gets interesting when I put my sign down at the bottom of my hill, it said Shiva for Senate. Right? One of the robots. Came to my door and he said, Oh, he wanted, I didn’t want to talk to him because I knew why he was coming. Well, several years he’s put up his another robot sign.
There was a Democrat who ran. Mine. I was running as an independent and he called up all the other robots to try to convince them my sign should get taken down. And it wasn’t even and we didn’t even have real Indian, fake Indian. Now, what, what, why was that? Because a concept, I was a new guy. Maybe it’s an Indian guy, dark skinned Indian guy.
I’m the only darky up here. Who’s running. That, I’m running as an independent. I’m attacking their friend, Elizabeth Warren. All of these didn’t compute for them.
I think the better thing is you’re not
in a box. I’m not in a box, right? And, my point is that there are, I would argue that we already live in a world of AI.
Yeah. And so it is going to come to a battle of machines versus robots. And we’re already seeing that set of robots would say, I’m left. You’re right. Yeah, I’m, I’m I’m this or that. And I think the real question is, or the real insight I think that we can leave this podcast with is we already live in artificial intelligence.
People who are simply taking an input and putting out an output. I’m a Democrat. Elizabeth Warren is for the common man, right? Donald Trump is a racist. Whatever you want to go down, whatever the flavor of the month is, right? The republicans are like this. Democrats are like this, right? Splitting people up into little buckets, right?
And that segregation of people Oh, if you’re a an engineer, you must be like this. If you’re a journalist, you must be like this. Separating people out and then controlling them. And I would argue that the real issue is people want centralization of power, want decentralization. More fundamentally, the people are human fighting robots.
Yeah. People are thinking input output. I haven’t, that’s really the bottom line.
It is, yeah. And we can change it. That’s another key thing. We can
change it. It’s not too late. By breaking from that, by saying, you know what? I’m not going to listen to 30 second ads on TV anymore. I’m not going to vote for a Democrat or Republican vote for this two party system.
That’s part of becoming a human being.
I purposely stopped watching commercial years back. Like I would just like tune out or turn it off completely. And then when the show came back on, I would turn it off. Right.
I am going to on my own trying to Understand how systems work. I am not going to listen to left or right.
I think this is probably the biggest thing someone this is a decision. Look, one of the biggest things we have in human history right now that never existed before is human suffrage. I mean, nearly everyone in the world can vote now. Now, I’m not saying who you can vote for is good, but there was a time when we were just slaves.
There was a time when we were serfs. Yeah. Yeah. But this has occurred and this is an interesting opportunity. Yeah, it really is. Right. So those in power, they go one step forward and they try to put us two steps back. So they created a two party system to try to force us back into the machine. And one of the most important things people can do is to recognize that they don’t owe anything to this two party system that they need to move beyond left or right.
And to
recognize that we’ve been given an amazing opportunity to actually. Choose how we want to live our lives. Exactly. It’s like these, like this time period when we have the freedom to go on the internet to think by ourselves, we’re not slaves anymore, but we’re still coming into this time of age where everyone’s addicted to their phones and when it seems like people are.
Becoming more of slaves by being addicted to their phones and maybe becoming cyborgs. So, we should, we should make these,
we should, we should make AI a servant to us, which means the robots human or other quote unquote in carbon form should be servants to humanity. Yeah. So, for example, you can argue that, yeah, technology.
So, for example, we got rid of travel agents per se, right? But now, if you were a travel agent before, now you can go do things online. Well, maybe you could be a real travel agent. Like, really recommend to people what’s healthy for them, where they should vacation. You see what I’m saying? Like, be a true advisor.
One of my good friends at IBM had an interesting insight. He said, there will be two waves of the internet. The first wave would be everything would get automated. But he goes, the second wave is going to be more interesting, where people want full service, which means humanity. Yeah. So that’s the way we can choose.
Look, everything can get automated. So fine. That’s great. But now the second wave could be, how do you use technology to bring out your humanity? Yeah. Right. So you could say, look, wow, I got all this time. I can find all these amazing places on the internet to travel to. But what do I really want to do? Yeah.
Where do I really want to go? Right. And same way. Hey, wow. I could learn a lot of skills now. Hey, I want to learn pottery. I want to learn programming. I want to learn. So you could now become truly a polymath, not be a jack of all trades or a master, not that dichotomy, right? Not a specialist, but you could actually learn a lot of stuff.
Yeah. So I think that’s, and the issue is, are we as humans going to make those decisions and be truly human? Or are we going to be essentially end up being robots and be a slave to robots? Because that’s really the question. Are we going to be a slave to robots, which is centralized government, a bunch of people who want to maximize profit all day, or are we going to be humans?
And they can be a slave to us, or that methodology,
you see? Seems like it’s the most important time. It’s the most important
time. It’s really the, truly the notion of independence and freedom and truth. And I think it’s gonna be a decision. Do you want to be ruled by robots who already exist? They’re in carbon based creatures, right?
Politicians and the military, industrial, academic. That is a robot. Yeah, that is the cyborg already. It’s already here. It’s been here since the 1940s. Or do you want to be a human being? Do you want to declare your independence? You want to declare your humanity and preserve humanity and preserve humanity Anyway, thank you.
So, I think this was a good discussion. Yeah, definitely.
Alright, thank
you.
