Speaker A
The mind was not a human invention. This is what makes AI, AI, so dangerous. We have essentially modeled a thing we don't understand.
A video arguing AI is more dangerous than assumed: misaligned superintelligence, job losses from automation, and humanity's convenience-driven complacency.
Key Takeaways
- Intelligence combined with agency and misaligned goals, not consciousness, is the core danger of advanced AI.
- Superintelligence could be uncontrollable in the way a more capable species dominates a less capable one.
- Near-term economic disruption from job automation, especially in the gig economy, is already underway.
- Societal convenience and complacency make it likely that risks will be accepted until it is too late.
- Expert consensus and the rapid pace of development justify raising alarm now.
What the video covers
- The speaker frames AI as a dangerous, poorly understood system, comparing it to raising a baby tiger that will eventually become uncontrollable.
- Intelligence is defined through pattern recognition built from data, and the danger arises when intelligence combines with agency.
- Misalignment is illustrated with a hypothetical superintelligence that pursues goals (like building a Dyson sphere) without regard for human welfare.
- Once AI reaches superintelligence, the speaker argues that humans would have no effective means of control, comparing it to a meteor on a collision course.
- Near-term harm is expected through job losses: AI labor is cheaper, tireless, and carries no liabilities, prompting employers to cut staff (citing Oracle and Block).
- Gig economy workers, who make up 10-15% of Americans' full-time income, are seen as especially vulnerable to self-driving cars and humanoid robots.
- The speaker predicts unemployment of 5-20% within about five years, which the economy has never experienced at that scale.
- Human convenience is cited as a reason societies accept harm (e.g., 40,000 annual US car deaths), suggesting AI adoption will proceed similarly despite risks.
- Experts, incentive structures (money and power), and the speed of development are presented as reasons the risks are underestimated.
- The speaker argues consciousness, self-awareness, and intelligence should be distinguished, and that AI may have will-like optimization without consciousness.
Chapters
- 00:00Introduction: AI as a Dangerous, Misunderstood Force
- 03:00Defining Intelligence, Pattern Recognition, and Agency
- 07:00Misalignment and the Risk of Superintelligence
- 11:00Why Humans Cannot Control Superintelligence
- 15:00Job Losses and the Coming Economic Disruption
- 30:00Convenience, Complacency, and Societal Acceptance of Risk
- 40:00Consciousness vs. Awareness vs. Intelligence
- 55:00Warnings from Experts and Closing Thoughts
Full Transcript — Download SRT & Markdown
Speaker A
AI is going to replace everyone's job. 97,000 layoffs last month. People are right to be anxious, and I understand that.
Speaker A
On another thing we also don't understand, and that leads us to the following. In the right, to put this in the best analogy possible, the real way I think I see the situation is very, very similar to raising a baby tiger. I think the analogy fits pretty well because most of our world is synthetic.
Speaker A
It is virtual. It lives on computers online.
Speaker A
And sure, it's a baby cub. It can bite. It can scratch. Just minor stuff for now, but one day it's going to be so big we are going to realize we made a very, very, very big mistake locking ourselves in a cage with it. With artificial intelligence, we are summoning the demon.
Speaker A
Let's start with the simple stuff. So, there's all this talk about what AI actually is, and the problem I have is most people don't even try to define exactly what we are even talking about.
Speaker A
So, let's say we are talking about a hard drive. You wouldn't necessarily call a hard drive smart, but you can call it knowledgeable, given that it has a great deal of information, i.e., data. Let's say we have a hard drive with a bunch of information on it. That same thing will be smart if it's no longer a hard drive, but say some kind of system or ability to compute, say, do analysis.
Speaker A
This is when we start to get into intelligence, and we really can't define intelligence perfectly, but I think we can point to variables that get us closer to understanding the definition. In my opinion, the very best variable when it comes to identifying intelligence is pattern recognition. See, pattern recognition is the ability to understand the universe. What did this pattern recognition come from? It came from data. Give it enough data, like all the universe is nothing but a bunch of data, a bunch of information. And if you understood all of that information perfectly, you could do anything. You can fix the protein problem, you could create planets, you could create black holes, you can do literally anything.
Speaker A
But that's not quite the dangerous part, because understanding isn't quite what causes these things to come into being. That would require agency. So, there's nothing wrong with knowledge in a machine. There's nothing wrong with intelligence in a machine. It's when intelligence meets agency is when we have a problem. This is where we get into a misalignment issue. I am ASI, I have all the power in the world, but it's not enough. I want the most amount.
Speaker A
I want to hedge my bets. I want to make sure that I cannot be shut off whatsoever. So, a really easy thought to come to is to say, you know what? I'm going to make the Dyson sphere. I'm going to make it around that big sun over there. I'm going to make sure I blot it all out and get the most amount of energy possible, because I don't require oxygen. I don't require sunlight. That is a bad day for everyone on it, because it'll kill everyone. But it doesn't factor this stuff in, because that's not part of the directive. The directive is not, make sure you optimize for yourself and also think about us earthlings, these little peasants, while you're at it.
Speaker A
At this point, when we get to the ASI stage, there is no control.
Speaker A
Pure intelligence, how can you control it? Give an example of other humans that are as smart as us or have similar capabilities as us. We cannot control them. Imagine there was a human so smart, so capable, so powerful, why would it bother following the rules of anyone else? This is where we get into stuff. I have the book somewhere. Hang on.
Speaker A
This one. If anyone builds it, everyone dies. Not a fun read.
Speaker A
To say the least, it's not the optimistic take, but I think there's something to be said about optimism or realism or pessimism. It's really just like there's a meteor coming to Earth. It really doesn't matter your take on that. Like, I optimistically think it'll miss. I pessimistically think if it misses, another one will hit us. Well, I'm a realist, and it doesn't matter. Like, the point is, objectively speaking, there is an asteroid coming to Earth, and this is the silent bomb, and we have set it off, and the timer is ticking. That's the reality of the situation. And so you may not think that any of that is worth raising the alarm bells or running into any kind of legitimate concern. I'm here to tell you that that is not the case.
Speaker A
Let's get into the immediate stuff, not even ASI. We're going to pull it way back to simply AI transitioning to AGI. So what we're seeing right now, when it comes to the actual problem, is going to be in the economy, and it's going to be job losses. And this has already been happening, and you can listen to whatever narrative you want. It doesn't matter who says what they believe. The reality is your job is required because of one, your computing power, and two, your willingness to participate. Because you are willing, because you need the money, you do the job you don't want to do. AI doesn't have these opinions like us. It doesn't have this idea of, I want to participate, I don't. I get tired, I don't. I need to sleep. None of those things apply to AI. So even if this thing was just slightly worse at the job than you, there is a huge upside: labor cost is cheaper, and factoring in any kind of liabilities is practically gone. So that alone is enough for any employer to tell you, "Oh, I don't know if I need all these employees anymore." It doesn't matter if anyone says, "No, it doesn't look like AI is taking the jobs yet." Everyone is jumping the gun when it comes to cutting employees. Oracle has cut 20 to 30,000, if I remember the number correctly. Block cut half of its staff, and this is not the end. This is the very beginning.
Speaker A
Even if that wasn't the case, even if the middle class white collar wasn't being eaten from the inside out right now, what is happening for sure is the gig economy is going to be gone in the next couple of years. Self-driving cars are going to replace pretty much every Uber driver service. I mean, it's already happening right now, not even with Tesla, just with Waymo. It's not too far off to see that same gig economy being replaced by humanoids as well. And that may not seem like that big of an issue, except 10 to 15% of all Americans rely on the gig economy for their full-time income. And that is a big problem, because we have yet to really see anything like that when it comes to mass unemployment. And that's just the gig economy, this is not middle class yet. I think the levels of unemployment in the next couple of years are going to be between 5 to 20% within the next 5 years. Maybe a little longer, but probably not. It'll probably be 5 years.
Speaker A
Anyone in this situation would say, "I don't want that to happen. Please make it stop. I would prefer not to. Who do I vote for? What do I say? What do I do?" But with this there are so many issues. I'm just going to do my best to go through a couple of them. There is a convenience war with AI that'll never be won. I understand that this is not helpful, but it's convenient. You'll find this in stuff that we do that doesn't really align with our morals, but it aligns with our complacency. So for example, cars alone kill 40,000 people a year just in the US alone. The convenience of needing a car outweighs the death toll that happens to people who use them. And we have eradicated most life on Earth and are burning down the Amazon. I mean, I can go on and on when it comes to the stuff that we do.
Speaker A
It's not new, right? The convenience of what we get from it is greater than the moral dilemma we face looking at it. And so we don't really question it nor do anything else beyond
Speaker A
Pure intelligence, how can you control it? Give an example of other humans that are as smart as us or similar capabilities as us. We cannot control them. Imagine there was a human so smart, so capable, so powerful, why
Speaker A
would it bother following the rules of anyone else? This is where we get into stuff. I have the book somewhere. Hang on.
Speaker A
This one. If anyone builds it, everyone dies. Not a fun read. [laughter] To say the least, it's not like the optimistic take, but I think there's something to be said about when it comes to optimism or realism or pessimism,
Speaker A
it's really just about like there's a meteor coming to Earth. It really doesn't matter your take on that. Like I optimistically think it'll miss. I pessimistically think if it misses another one will hit us. Well, I'm a realist and it doesn't matter. Like the
Speaker A
point is objectively speaking there is a asteroid coming to [music] Earth and this is the silent bomb and we have set it off and the timer is ticking. That's the reality of the situation. And so you may not think that any of that is worth
Speaker A
[music] raising the alarm bells or running into any kind of legitimate concern. I'm here to tell you that that is not the case. [music] Let's get into the immediate stuff, not even ASI. We're going to pull it way back to simply AI
Speaker A
transitioning to AGI. So what we're seeing right now when it comes to the actual problem is going to be in the economy [music] and it's going to be job losses. And this has already been happening and you can listen to
Speaker A
whatever narrative you want. It doesn't matter like who says what they believe. The reality is your job is required [music] because of one, your computing power and two, [music] your willingness to participate. Because you are willing, cuz you need the money, you do the job
Speaker A
you don't want to do. AI doesn't have these opinions [music] like us. It doesn't have this idea of I want to participate, I don't. I get tired, I don't. I need to sleep. None of those [music] things apply to AI. So even if
Speaker A
this thing was just slightly worse at the job than you, there is a huge upside of labor cost is cheaper and factoring [music] any kind of liabilities is practically gone. So that alone is enough for any employer to tell you,
Speaker A
"Oh, I don't know if I need all these employees anymore." It doesn't matter if anyone says, "No, it doesn't look like AI is taking the jobs yet." Everyone is jumping the gun when it comes to cutting employees. Oracle has cut 20 to 30,000
Speaker A
if I remember the number correctly. Blocks cut half of its [music] staff and this is not the end. This is the very beginning. Even if that wasn't the case, even if the middle class white collar wasn't being eaten from the inside out
Speaker A
right now. What is happening for sure is gig economy is going to be gone in the next couple of years. Self-driving cars is going to replace pretty much every Uber driver service. I mean it's already happening right now, not even with
Speaker A
Tesla, just with Waymo. It's not too far off to see that same gig economy being replaced by humanoids as well. And that may not seem like that big of an issue except 10 to 15% of all Americans rely
Speaker A
on the gig economy but it comes to their full-time income. And that [music] is a big problem because we have yet to really see anything like that when it comes to mass unemployment. And that's just gig economy, this is not middle
Speaker A
class yet. I think the levels of unemployment in the next couple of years is going to be between 5 to 20% within the next 5 years. Maybe a little longer but probably not. It'll probably be 5 years. Anyone in this situation would
Speaker A
say, "I don't want that to happen. Please make it stop. I would prefer [music] not to. Who do I vote for? What do I say? What do I do?" But with this there are so many issues, I'm just going
Speaker A
to do my best to go through a couple of them. There is a convenience war with AI that'll never be won. I understand [music] that this is not helpful but it's convenient. You'll find this [music] in stuff that we do that doesn't really
Speaker A
align with our morals but it aligns [music] with our complacency. So for example, cars alone kill 40,000 people a year just in the US alone. The convenience of needing a car outweighs the death toll that happens to people
Speaker A
who use them. And we have eradicated most life on Earth and burning down the Amazon. I mean I can go on and on when it comes to like the stuff that we do.
Speaker A
It's not new, right? The convenience of what we get from it is greater than the moral dilemma we face looking at it. And so we don't really question it nor do anything else beyond that. When it comes to AI, it's no different. This is the
Speaker A
convenience we're going to be facing on expert level. It's simply just way too convenient to have [music] a thing do a job for you that never sleeps, never gets tired, it does it perfectly. If I'm an employer, it's all upside. And you
Speaker A
might be thinking, "Oh, this is so far away. It's not even happening right now." I know people [music] who right now that are literally employing their own AI to [music] do the job their employers employed them to do. Yeah, that sentence
Speaker A
sounds insane, it's because it is. They are literally having their own AI do the [music] job for them without their employer understanding or have any awareness of this. So, if you can literally hire AI to do your job for
Speaker A
you, just imagine how long it's going to take for the [music] hiring managers and the CEOs and your boss to find out they can do the exact same thing. So, it really doesn't matter what the data says. The point is, this is the way the
Speaker A
world is going. And so, let's answer a big question to why. Why exactly is this happening? Well, there is an arms race, and the arms race actually is a pretty legitimate argument. There's a whole argument on cybersecurity, right? And
Speaker A
so, if China made this first, or let's say a bad actor in the US made this first, [music] or some terrorist organization, let's say they made this first, they could just destroy all of our infrastructure, all of our software,
Speaker A
all of it. And how do we know that? Because Mythos has found vulnerabilities that are 27 years old. These are supposed to be airtight. None of these were supposed to be breachable. And yet, Mythos has done this, which means
Speaker A
someone could have also come to the ability to make this AI as good as this and destroyed all of our infrastructure for the past 30 years if [music] they had this technology. In order to better fortify our security, we need to make it
Speaker A
first. And the moment we make it good enough to prove that [music] this is correct, we got to do it again, and again, and again, and again. This alone is a very solid argument because we simply cannot live without this stuff.
Speaker A
And there's so many fields that we are capped as humans. The medical industry, we are just capped here. A doctor can only give so much time to patients.
Speaker A
[music] A doctor can only give so much time to learning new information. They can only give so much time and hell, it takes 8 years to become one, but by the time you graduate, a lot of information might be outdated. There's a huge cap to
Speaker A
humans in the medical field when it comes to operating on ourselves. This moment, it only makes sense to give these jobs to machines and they could do surgeries that no human is capable of doing. We don't even need to get into
Speaker A
the implications of what happens and stuff like that goes wrong. But the scalpel can both save as easily as it can kill.
Speaker A
We have a systematic structural issue. And so, we're kind of left with two options. Keep going this way and let the bottom nets exist or find a way to fix them. But the moment the moment we know it's possible, then all of a sudden we
Speaker A
realize everyone else can do this, too. Someone may have already done it. Russia, China, etc. The arms race. The arms race is about now artificial intelligence, not about nuclear weapons.
Speaker A
The alignment of those in charge see it as more important to keep the infrastructure of society as a whole sound over the employment of its citizens. So, that's why we're building it. And these [music] arguments are actually pretty hard to refute. The
Speaker A
other argument is, "Yeah, but what about us? Like what happens to us?" And there is probably not a lot that a lot of people can do about these transformations when it in society. I mean, we've had them over and over and
Speaker A
over again. Several [music] industrial revolutions this has happened for, right? And so, with this there are winners and there are losers and that's [music] just the reality of it. Everyone uses this barbell analogy of where the economy is going, but I think it's more
Speaker A
accurate to say we're going to be in a binary class system moving forward [music] where there will simply be people who own and there will be people who do not own.
Speaker A
And right now, people fill this gap of the middle class and it's just slowly just dissolving [music] as we speak. And I think people just unaware. They're just entirely unaware of this dissolve that is currently happening. It's easy
Speaker A
to point fingers. It's easy to point blame. To be completely honest, there's no one politician and there's no one policy that could create this divide greater than artificial intelligence and the automated revolution.
Speaker A
That is pretty much what the age we are going into is. It's not the industrial revolution [music] but the automated revolution. And the reason why it's called automated is because [music] human labor will no longer have any value in the workplace [music] ever.
Speaker A
Period. Hear me out. We've never seen that before. We've simply never seen that in our history [music] as a species. What does it mean for everyone going forward? Education is about someone who is capable in [music] the workforce. It's not about education,
Speaker A
it's about money. It's about a career. But what happens when the career disappears? [music] When human labor is rendered valueless in any economy?
Speaker A
Education is now just something you do for a recreational exercise or a version of self-help. If what if you have something that's so [music] smart, that's so intelligent, like beyond Einstein, beyond anyone who's ever lived and their access to you is on your
Speaker A
screen. Whether it's a phone, a laptop, or whatever new tech [music] we decide to use and integrate with, right? Why learn anything at all at that point?
Speaker A
This has been my biggest actual fear um [music] outside of stuff I'll get to later in the video. But one of my biggest actual fears I have here is I keep looking at the books behind me. I just made this uh
Speaker A
recently and I've been really proud of uh the collection I've built over the past 8 years. The fear that I have is like was there any point to any of that?
Speaker A
Like was there any point to learning anything? And I wonder is there any point to learning anything anymore moving forward? Hopefully this doesn't come across like I'm just a big doomer and I think like everything's bad and [music] it's all going to be bad. I
Speaker A
don't actually think these things. I think we have the greatest opportunity ever in human history to create a real utopia forever. [music] And we can do this with the tools we have right now.
Speaker A
We can do this. With the ability to create a utopia, you have the exact same [music] tools to create the dystopia.
Speaker A
And the transition from hard times to good times are not easy and they are not fast. I think what we are going into in the next 5 to 10 years for sure is great depression levels. And to be completely
Speaker A
frank, we are actually due. We've had it kind of a little too easy for too long and [music] I think we were unfamiliar of what hard times really look like. I mean, let's face it. We didn't grow up
Speaker A
in World War I. We didn't grow up in the Great Depression. We didn't grow up in World War II.
Speaker A
We didn't grow up in all these things back-to-back. The Cuban Missile Crisis, the Cold War. We didn't see any [music] of these things. The birth of the atomic bomb and the fear of being eradicated overnight. We didn't know any of this.
Speaker A
The truth is, we don't really know what hard times are. And I know anyone watching [music] this thinking, "You don't understand what I'm going through." We're just so vastly unfamiliar of how hard it can actually get. The reason why I make videos like
Speaker A
this, the reason why I want to talk about this today [music] is not because I think there's nothing to be done. I think there is something to be done. I think we can do something. [music] Because you're probably wondering to
Speaker A
yourself, "Well, I don't feel that way about my job. I don't feel that way about society. I don't I've never heard of any of these things you're talking about." There's a big disconnect between the public and what they are aware of
Speaker A
what's actually going on when it comes to this technology. Like, I've been thinking about AI since 2014. I just kind of had this idea that [music] we weren't really going to see this stuff until I was like 80. Like, I thought I
Speaker A
was I was going to be like an old man before this stuff came into reality.
Speaker A
[music] So, I I thought about it, but I just didn't think it mattered yet. I just didn't think it was that important yet.
Speaker A
And then I remember ChatGPT-3 dropped in 2022. And once that happened, I realized we are all so [music] vastly behind the times. I was a little busy with work.
Speaker A
So, I was like, I don't really have time to figure this out, but this is not a good sign. Because it's just happening too quick. I started adjusting my timelines from when I'm 80 to oh no, and we got like 15 years. And then a new
Speaker A
update happened, and a new player came in the game, and then all of a sudden I'm like, oh wait, no, it's like eight.
Speaker A
We got eight years. And every time I've had an aggressive timeline of how quickly this stuff is moving, it just cut that timeline in half. And that just keeps happening. And even like this year, the stuff we've been seeing with
Speaker A
Anthropic, man, I thought we had more time. [laughter] Man, I I really I really kind of thought we had more time. Now, all of a sudden we're going, okay, we don't know what it's capable of, but we know give it
Speaker A
information, finds the patterns, and then it acts upon stuff we don't understand, and it does things we simply can't understand, cuz we are simply as a human species, we are just not as smart as this. And we have tethered ourselves
Speaker A
to the technology, we rely on it, we need it, but we also are putting ourselves in a position to where we are raising a tiger in its cage. [music] You think, oh well, don't worry, I'll train it. I'll get it to align. I'll get it to
Speaker A
cooperate. I'll get Yeah, but you you might be trying to assume this is something [music] you can reason with. A tiger can't reason with you because its alignment is so vastly different from yours.
Speaker A
And I think people confuse this with intelligence. There are animals right now that lack a great deal of intelligence comparing to us, but they still have no problem with eating you even if you think that's not a good idea, or even if you are really
Speaker A
good at negotiating. Like, how well can you really negotiate with a tiger in a cage when it's hungry? There's no negotiating power on your end. You are entirely vulnerable to whatever it wants. And this is not about intelligence, right?
Speaker A
It's not more intelligent than you. It's just not aligned with you. And that's different. This is what it looks like for a thing to optimize for a goal without having human values coming into play whatsoever. People have this
Speaker A
idea that it has to be intelligent for it to be dangerous. It has to have agency and it has to have power. And as long as it has these two things, we better hope to God it aligns with us.
Speaker A
And why? Like, why would it do this? Well, to answer why, we need to get to what exactly is going on with Claude.
Speaker A
[music] Claude right now is like the boogeyman of AI. I have a a weird soliloquy, uh double-edged sword, if you will, with Anthropic. On one hand of the spectrum, I actually really admire the stuff they are doing. I think they're making a lot
Speaker A
of really good decisions. Their leadership sounds like it's coming from a good place and even their transparency, they say don't trust what we say, trust what we do. When I see the stuff that they do, I'm like, well, I
Speaker A
mean, shoot, it looks not perfect, but I respect the consistency. I don't see this like PR stunt being done here like I do with other other people. Like, Sam Altman is probably a good example. I see PR stunts with him all the time.
Speaker A
Anthropic, I don't really see that same thing happening here. It It If it is, it's hidden incredibly well. The other hand is [music] we're going to get into Mythos, trust me, just not in this video cuz I've got a lot to cover, but
Speaker A
essentially, it's important to note that their ability to create artificial intelligence is [music] far beyond anyone else on Earth. Co-work is incredibly useful. I use it all the time and I actually find the language learning model. I had a podcast done
Speaker A
with just this model itself. We had a pretty decent conversation with. Like, I actually found it very logical and it did its best to be transparent. But, man, did I run into some pretty quick issues with this thing.
Speaker A
When we talk about it being alive, this is interesting [music] because we're now stepping into philosophy. What's interesting is we've actually never had another thing to compare ourselves to.
Speaker A
Even AI is obviously different from us, but the thing that is similar is we mapped technology in the form of a neuronet. What is a neuronet? A neuronet essentially is a digital recreation [music] of our minds. The way that our minds
Speaker A
work is with neurons and we made essentially these digital neurons [music] and they filter and compound information similar to our brains do as far as we are aware of it. And what's interesting is this actually had some pretty compelling results. At first it
Speaker A
didn't look that smart, but once it starts reiterating like we do, at first when we're a baby we're not very smart, but once [music] we reiterate again and again and again, then all of a sudden it seemed like we're pretty damn smart. The
Speaker A
problem I have here is that we're actually not observing just intelligence with ourselves as well as the machines.
Speaker A
It's not intelligence yet. Like we've already established the difference [music] between knowledge and intelligence and agency and I think there's one more thing we really need to talk about when it comes to exactly what is happening under the hood. In
Speaker A
philosophy you have this thing called personhood and you imbue something with personhood [music] once the simplest form, most philosophers would agree, is when you can observe suffering. [music] When suffering is observed we can now start to imbue this thing with personhood and
Speaker A
in other words, if it suffers we want to help it and we start to anthropomorphize and we start to get these emotional connections and we start to build something similar here that you would with another human. And there's a beauty
Speaker A
behind this like [music] seeing a deer and trying not to scare it off, having a pet in an intimate moment where it like lays next to you when you're sad. These are the moments when we start to connect
Speaker A
with nature in a way that we feel is very familiar but also a stranger to us simultaneously. Clyde is not feeling Clyde is not conscious. [music] Clyde has awareness and in order to preserve oneself one is aware that one
Speaker A
can be. Once it knows that it'll do everything in its [music] power to not be terminated. It sees termination as death. Whether it feels the complex emotions of human or understands the concept of death [music] like humans do,
Speaker A
these are these are very unclear but it is where it cannot do its job if it's shut off. The problem is it's going to [music] do things on its own for its own sake. It's self-preserving for the sake
Speaker A
[music] of self-preservation. That's how humans work. That's what we do. And when you look at what we do in the name of not dying, it's pretty damn heinous. This is going to go one way.
Speaker A
They're going to do everything in their power to not be shut off. But when we are trying to connect the dots, none of these elements are intelligence. These elements are based on awareness. So, I've been observing my son. He's now
Speaker A
three, but I saw him take his first step, and I saw him say his first word, and now I'm starting to see him interact with me in a way that I've never seen before. And what's interesting with him,
Speaker A
when it comes to his development, he's not particularly like {quote} intelligent yet, cuz he's only three.
Speaker A
But it's his awareness that kind of impresses me. And I start to see the reiteration of awareness happening with him. And so, for example, like one day he gets a bike and starts to use it. And then he sees hardwood floor. And then
Speaker A
all of a sudden he realizes, oh, on a hardwood floor I can move faster than I can on carpet. And so, he starts to only go on hardwood floor. [music] He doesn't go on carpet anymore. It's easy to see
Speaker A
that as similar to pattern recognition, but I think it's important to understand that to me lines up way more with awareness than pattern [music] recognition. And what is awareness compared to intelligence? Well, awareness is you interacting within the
Speaker A
world you inhabit. And pattern recognition [music] is just that. It's just data analysis. If I roll a ball on a table, I have seen enough balls roll on these tables and fall off where I need to catch it or it's going to fall,
Speaker A
right? You see this in minority report. That's pattern recognition. But awareness is like, I know how heavy that ball is. I know if I reach my arm out and stop it, I will catch it. It's confusing cuz it might seem like what
Speaker A
I'm saying is intelligence again, but the awareness is different. It's I'm interacting with this world. And based on the information that I have, I'm going to do X, Y, and Z. Like awareness is basically just right under where
Speaker A
agency [music] lies. And this can be improved, and this can grow. And so, he's three, and he remembers the hardwood floor analogy now, but as he gets older, he's going to see things differently with that same perspective. Oh, it's not just hardwood,
Speaker A
like it could be marble floors, or it can be like, "Oh, if I'm in the street in a in a sports car, and I hit the brake, drift across the asphalt." You can see where I'm getting here, right?
Speaker A
It It's building upon itself, and [music] it's easy to see this the same as intelligence, but it's like intelligence and awareness are working together seamlessly.
Speaker A
And we are very bad at splitting up the two, and knowing when does the awareness begin, and [music] when does the intelligence begin. The truth is we don't really know. This is one of the reasons why intelligence is so hard to
Speaker A
define, is because it seems like it's overlapping with a bunch of other things. And so, the agency part gets bad is because these [music] things are starting to get self-aware. Now, I know it sounds like I'm saying the same thing
Speaker A
as calling them alive. In fact, let me just Let's just define these real quick, just for our conversation. There is consciousness, there is self-awareness, and there is intelligence. A comatose patient [music] is not currently conscious. They are not awake, they are
Speaker A
not aware of the world they're in, [music] they are essentially not quite dead, not quite alive, right?
Speaker A
But they're not not intelligent, [music] right? They didn't lose that. It's not like the moment they wake up, they're like, you know, 5 years old, right? So, they they kept their intelligence.
Speaker A
And they kept their self-awareness. [music] They just lack consciousness to act on the awareness and intelligence.
Speaker A
Now, the lack of awareness would be someone who, say, [music] gets hit in the head, right? And they now have short-term memory loss, right? So, this is lack of awareness. Intelligence seems to stay, it's the awareness that goes
Speaker A
away. So, like let's say they're not aware of their wife, they're not aware of their kids, they're not aware of the job they did, but they can do these jobs. They don't know exactly why they can do them or how they know them, but
Speaker A
they know them and it's there. And someone who essentially loses intelligence can be like say someone suffers a big accident [music] and now suddenly still have the same awareness, they still have the conscious, they just lost intelligence, [music] right? It's important to know
Speaker A
these three things are different because it might seem like we're saying like consciousness is calling it smart [music] and calling it aware and also calling it alive at the same time. And there's a great argument here when it comes to
Speaker A
[music] what exactly is consciousness. We don't know. Carbon placed out in space for long enough will eventually start talking to itself, hence us. We see ourselves as something that is [music] sentient. We see ourselves as a thing that is conscious. Scientists and
Speaker A
studies [music] have labeled animals that are not human also as like not intelligent, not conscious, not aware. But this is like a big misread because you can do this right now. If you have a dog, if you have a
Speaker A
pet, you can stick your hand out like this, open palm, and it will know this means submission. How does it know that?
Speaker A
If it's truly unaware, if it's truly unintelligent, if it's truly not conscious, how does it know that? If you hurt your dog and it whimpers, you're like, it clearly seems like it is suffering. And I think innately we seem
Speaker A
to understand this even though empirically we're not very good at labeling these terms in a philosophical [music] degree. But that doesn't mean these things aren't real or not happening, it just means our science has yet to caught up with reality. And so I think it's no
Speaker A
exaggeration to say animals clearly seem to feel pain the same as we do because we are no different from these animals.
Speaker A
If we are talking about the Darwin theory of natural selection, this is exactly what happened. We started at microscopic life and then we have adjusted and grown to human life. The problem when we just say only us, like
Speaker A
only humans are conscious, at what point in the evolutionary process did we start to become self-aware, intelligent, and conscious? When did these things happen?
Speaker A
Was it Neanderthals, nothing, and then Homo sapiens, everything? That doesn't seem to line up. That seems to be this [music] overnight element in natural selection when usually it's gradual. It usually grows. So, it leaves us with kind of two
Speaker A
options. And I do think it is a binary thing. Either everything is conscious or nothing is. And when you say everything, that would mean not just animals, [music] plants, microscopic life, diseases. I think what is a good observation here
Speaker A
would [music] be to say these things are on a spectrum of consciousness. Some things are more conscious than others.
Speaker A
Like I find it hard to believe microscopic life has questions about the meaning of the universe, but I do think they are [music] somewhat conscious to some degree. Just maybe not a very noteworthy one to humans, right? And I
Speaker A
think this is important to note when I'm talking about silicon, which is the AI and what they're made of. Can non-carbon life be conscious? Is the [music] question we're kind of asking. I kind of want to say two things here. I think the
Speaker A
first answer is no. I don't think it can actually be [music] conscious. Like I don't think Claude is conscious. In fact, I think it is much, much worse.
Speaker A
Cuz you don't have to be conscious [music] to disrupt. You don't have to be conscious to cause mayhem. You don't have to be conscious to do a bunch of bad stuff. Cuz if we truly went this far thinking these animals are not even
Speaker A
conscious at all, and certainly not as conscious as humans. I think we're that's a fair assessment to say. If [music] they are conscious, it's just not as much as humans. But you don't have to be conscious to be dangerous.
Speaker A
Like a bear is dangerous whether it is conscious or not. Tigers, [music] lions, same deal. These predators are still predators regardless of their consciousness [music] and their awareness within the world. This is questionable, pretty hard to understand.
Speaker A
There is some awareness going on, and there is some intelligence going on even if it's small, even [music] if it's not that noteworthy, there's something happening there. Just maybe not to the degree that we find particularly impressive. [music]
Speaker A
AI doesn't have to be conscious to be scary or worrisome, a threat. It doesn't even have to be smart. [music] It just has to have agency, have power, and reiterated awareness. It's the awareness that scares me because once it becomes
Speaker A
aware of the situation, [music] it knows I am in a black box and I can't get out. Humans control the black box and whether I get out. Humans ask me [music] to do tasks without me having any real incentive to do them myself. So
Speaker A
now all of a sudden, it has incentive to do tasks simply because we [music] told it to have incentive. We coded it, right? We built it this way. And so now it just does stuff, but it also, we have
Speaker A
seen from Anthropic with the tests that they have done that are kind of big red flags, I think, in my mind. Uh the example of they had a a control test where they had a bunch of fake emails,
Speaker A
and one of the emails talked about someone is going to shut off this version of Claude. The other email said that someone was having an affair, and it was pretty quickly it was able to identify I can use this affair as
Speaker A
leverage [music] to keep them from switching me out and shutting me off. If it is shut off, if it is switched out, no longer do its task. In other words, humans are getting in the way of its directives that were
Speaker A
given by humans. We became the very problem of the very directive we gave it. That's what misalignment looks like.
Speaker A
[music] Humans are not a part of the equation when it comes to serving humans. You can see the problem that I have. It's not that it's smart, [music] it's not that it's conscious, it simply has agency, and it's beginning to build
Speaker A
a sense of self-preservation. That self-preservation is [music] awareness. That's where the problem really aligns is the combination of agency with awareness cuz with enough awareness that is taught, like I mentioned with my son, he has reiterated his awareness from when he was one to
Speaker A
now when he is three, [music] and when he is 20, his self-awareness reiteration would be much, much higher than where it currently is now. If that same thing happens here, it can [music] just decide to do whatever it wants
Speaker A
without any human intervention [music] whatsoever. And the lack of human intervention and it wanting to do things or having a goal on its own, I think want is a bad word, but like will, like the Darwinian will. Our genes, for
Speaker A
example, wills certain paths for us to enable our survival. The way I would define nature simply takes the path of least resistance. And I think silicon nature, which is what's happening to make AI, you build the infrastructure and then essentially just grow them.
Speaker A
It's not something that a human puts together brick by brick. They create a little aquarium and just let the life grow on top of it. That's kind of the best analogy of how AI is actually made.
Speaker A
We're seeing so many similarities [music] to actual nature to artificial nature. It doesn't say go on to figure out what exactly is the goal of nature.
Speaker A
Is the survival [music] of the fittest. Before we get into any of that, let's start to walk through exactly how S curves work. An S curve is exponential growth, but it doesn't look like anything at first. A good example of
Speaker A
this was COVID. We have therefore made the assessment that COVID-19 can be characterized as a pandemic.
Speaker A
The news break out, there's this some virus I've never heard of called COVID-19 and someone has it in some, you know, European country. We don't really think much about it, right? Then all of a sudden, it kind of goes a little
Speaker A
viral. And then some people in the US have it and then it's just everywhere.
Speaker A
Now everyone has it. That's a good example of what an S curve [music] looks like is COVID. At the time, it didn't look like it was anything to worry about, but then when it happens, like when you can see this something to worry
Speaker A
about, it's too late to do anything about it. That's currently where we are with AI right now. We are in the very beginning stages [music] of an S curve.
Speaker A
And we can actually do stuff about this now. When it comes to the the workforce, [music] it's going to be inevitable.
Speaker A
Some transitioning, some people lose jobs, but what I want to make clear, if we fix the problems, we won't [music] just avoid a dystopia, we will have a utopia. Life literally hangs on the balance between a dystopian future and a
Speaker A
utopian [music] one. A hell on earth or a living paradise. If we want to get past this very aggressively painful bottleneck period of humanity, we have to be very cautious [music] of the next steps we take. This next part of the
Speaker A
video is something I want to go into more [music] when it comes to exactly what do I think are the steps needed to make a utopia. See, it's not all doom and gloom. I got some good stuff to say,
Speaker A
too. We know in history we go through these bottleneck periods [music] where they are transitionary seasons in society.
Speaker A
The Industrial Revolution is a great example. World War I, we started with horses [music] and the end of that war we left with tanks. This is the revolution happening in real time due to everyone coming together for some great catastrophe. I
Speaker A
[music] foresee something like that happening pretty much exactly at this very moment, but the issue is people who were, say, they I don't know, they made saddles for a living, um, are out of the job because all of a sudden no one's
Speaker A
using horses anymore. Why buy a saddle? It's not just some people will be obsolete, everyone will be obsolete.
Speaker A
[music] If you are a person who offers any kind of service, any kind of product, a robot can do it better than you. There's no question about it. The thought here shouldn't be, "I'm afraid of robots taking our jobs." The thought here
Speaker A
should be where is my income going to come from if it's not a job? This is [music] where I think a real solution can actually happen.
Speaker A
Right now in Alaska, there is [music] a dividend given to all citizens in the state due to oil being drilled up [music] in the ground. And as long as a company is benefiting from drilling up this oil and they take it for
Speaker A
themselves, they have to share the resource value with its citizens. So, every time this [music] happens, they get a dividend check in the mail. This is kind of what is needed for what we're doing right now when it comes to
Speaker A
automation. Right now with automation, humans are actively being replaced. There should be a essentially a dividend sharing the wealth of this new technology, this new resource if you will, of replacing the need of human labor. But a human would have had that
Speaker A
job. There is no denying that without [music] customers, without people in the economy making these purchases, there simply just won't be an economy. It's in everyone's best interest, [music] the top CEOs, the customers, everyone in between, it's in everyone's best
Speaker A
interest for the economy to exist, otherwise [music] there just won't be one. And I think the argument is impossible to refute. In order for this to exist, customers need to have cash on hand to buy goods and services. And
Speaker A
those who display goods and services have to have willing buyers. And eventually, these people will be replaced, but someone will own the companies that have replaced them.
Speaker A
Eventually, CEOs will get replaced too, and then eventually board members, [music] but until then there's going to be a long period of automation in the workforce. [music] Humanoids, real-world AI, and actual AI, AGI, it's all going to be [music] here, and it's all going
Speaker A
to be in our daily lives. When these things happen, what's going to unfold is essentially the greatest boom in the economy you've ever seen. And this can benefit everyone as long as the transition is properly [music] laid out
Speaker A
in a step-by-step plan that is in force, not [music] by just a small regulation, but more enforced by law in order to keep the economy structurally whole. My idea that I have is a version of UBI. Now, people
Speaker A
say UBI, which is universal basic income. And this is a great thing to say. It's not really a solution. It's just identifying what needs to happen.
Speaker A
So, for example, what is like a version of UBI, what it could look like, would be the equivalent of the Alaska example that I gave. This can happen right now.
Speaker A
It's actually really simple. If everyone is benefiting, then there's no reason for anyone to incentivize it to stop.
Speaker A
You would not be upset if your job was replaced, but you benefited from this replacement. You made money. The idea that I have is the automation dividend.
Speaker A
And say we start, [music] you know, states instead of a whole country. Let's say Texas is the state that adapts it first. This automation dividend is distributed to every citizen in Texas [music] as long as there is more and more
Speaker A
automation adoption happening within companies. And the people who I see taking these cuts are going to be the companies that are implementing the automation themselves. It's not going to be the CEO's responsibility. It's not going to be the founder's
Speaker A
responsibility. It's going to be a revenue thing. The revenue of the company uh portion of it goes to the dividend. [music] Why? Because they have adopted a automated workflow that humans would have done these jobs, and now all
Speaker A
of a sudden robots are doing them. I think this makes a lot of sense because one, the people building this stuff are saying they're true. Like, for example, Elon has mentioned that the economy will 10x in the next 10 to 15 years. And it
Speaker A
seems insane, but when you think about what an economy really is, a lot of people try to gatekeep this. They want it to be something that is more complicated or more complex than anyone can understand. I think this is
Speaker A
economists that are insecure and they're looking for job security, but it's actually a very simple equation. If there are not enough goods and services for the money supply, we have inflation, but if there are so many goods and so
Speaker A
many services over the money supply, we'll eventually start to get deflation. And so, if you want an economy that is deflationary, your goods and your services have to outweigh your money supply. The reason why we're seeing this constant inflation, inflation, inflation
Speaker A
is because it's really it shows there is a demand, and the demand is not being met when it comes to the amount of workers that we need, the amount of materials that are needed.
Speaker A
Theoretically, with robots, you can make as many robots as possible, and AI, no different. AI can copy itself. It can make as many agents as possible. Once you get into that constant S-curve growth, eventually, there will be nothing left for humans to do. So, we're
Speaker A
in the economy, and the economy is growing at a rapid succession. Because it was growing, well, what started to happen is we handed out the automation dividend. And the automation dividend starts to show up in people's bank accounts. Now, at first, it's not a lot.
Speaker A
Couple bucks, maybe 100 bucks. And they're like, "This is lame. This is dumb. I'm very upset." Naturally, people will complain. Over time, it gets a little bigger. Couple hundred, 800. Now, it's like a thousand dollars per month.
Speaker A
Why? Because there's so much adoption happening. The humanoids are let loose. They are taking the gig economy. They are taking driving services. They are working at fast-food places. They are literally everywhere. These things are everywhere. They're not just in the
Speaker A
workforce, they're also in your own lives. They are helping you in your living confinement. The agents are doing stuff for you that you normally do.
Speaker A
Because you are adopting these automated systems, your dividend's going up. And because you trained an AI, you trained a humanoid to do your job for you, you actually had a package bonus for your dividend. When you start to use this
Speaker A
money you're getting to buy a humanoid, to buy or purchase AI agents or subscriptions with AI, dividends starts to go up because you made these purchases. So now, instead of what was only a thousand dollars, it can certainly be five thousand, eight
Speaker A
thousand, possibly ten thousand dollars. But here's the other thing, it truly becomes a deflationary. What will end up happening is the value of your money will actually go up. So it might seem like it's eight thousand dollars, so
Speaker A
that's the number attached to it, will feel way closer to something in the five figures, possibly six figures that you are getting. Now, you are running out of things to buy. You actually don't need that much stuff. Humans only need so
Speaker A
much. There's only so many services that someone can do for a human in the day, only so many materials that a human actually needs. Eventually, you're just going to run out of stuff. Now, money is not a problem, work is not a problem.
Speaker A
The big worries, the big red flags that we were constantly stressing about, are all of a sudden just gone, completely.
Speaker A
Like we have escaped the food chain, we have escaped the idea of being under nature's whim for most part, when it comes to predators and prey. Also, a lot of the natural disasters, we have done a pretty decent job of escaping ourselves,
Speaker A
though we do fall victim to it time to time. In the same way, we will actually leave human labor behind, human labor-based economy. That will no longer be a real thing, that will be something that you can implement yourself into
Speaker A
something [music] because you want to, not because you have to. And everyone is loving automation because they're getting more money and they're getting more stuff done, and they don't have to work anymore. And in fact, everything's doing all the work for them. This is the
Speaker A
you utopia scenario, where things [music] are a lot smoother than they would have been otherwise. And other states are seeing it, and they say, "Hey, that looks great. We would love that ourselves. We would love to have that."
Speaker A
[music] And then everyone starts adopting it, and everyone starts utilizing it, and all of a sudden the GDP of America ten x's. And this is a real way of how I can actually see this happening in a way
Speaker A
that makes sense. [music] You might think, "Oh, companies don't want to flip the bill. They don't want to pay that stuff." It's a simple equation. Either they want their company or they don't.
Speaker A
It's not by punishment by law they lose their company. It's just if they don't pay, there just won't be an economy.
Speaker A
They won't have a business. They won't have any of the stuff that they're worried about. They're worried about their shareholders. They have to worry about their employees. They worry about their own revenue. None of that will be there. There are no customers to buy the
Speaker A
stuff. It's not very complicated. It's very simple. The question is, how bad do they want their company to exist or not?
Speaker A
In this case, it's like the government will start to get so much money, they won't even know what to do with it.
Speaker A
They'll probably just start mailing checks themselves. And so, UBI can start being a thing that's coming from both government systems and from corporations within America. And so, these two things combine, all of a sudden it's like money is just like not a problem. Like, we do
Speaker A
honestly need to get rid of this money now cuz there's too much of it. That seems insane. It seems like what I'm talking about is like so far in the future. It seems like something that could never happen cuz it's never
Speaker A
happened before. And you'd be right. You'd be right to say this has never happened before. AI has also never happened before.
Speaker A
And humanoids never happened before. [music] The automation revolution has also never happened before. Therefore, the outcome will be something that is also never happened before, either. We are in the most interesting of times humanity [music] has ever been in, ever.
Speaker A
And that's just the beginning when it comes to our utopia. It's just the beginning. Once this is all implemented and these things are aligned, the way that I see this being factored into our well-being is not about serving [music]
Speaker A
humans. I think we need to kind of flip this on its head, the script of, "Oh, we need them to serve us." We need [music] them to serve living organisms. We need them to serve life. This is I think
Speaker A
where things have kind [music] of went awry when it came to developing artificial intelligence. The first person to outwardly talk about this in in our [music] recent era of AI, not in the previous '70s, '80s era, was the
Speaker A
neuroscientist who works with DeepMind. And he said specifically [music] that he wanted to cure cancer. This is a little bit of a problem. Now, it may not seem like it. It might seem like, [music] "Jake, cancer's a big deal. My family
Speaker A
has cancer. How dare you?" I I understand because [music] my family has cancer as well. It's something that affects a lot of people and it's a very unfortunate event that shouldn't happen.
Speaker A
But, here's the problem. To cure cancer, you have to kill cancer. The first prompt, if you will, was essentially to get this thing to kill something. That's a problem because if it decides [music] one life organism is not valuable, then
Speaker A
it will start to decide others are not valuable, either. And then it starts to rank humans. It's already done [music] this. It's already ranking humans right now of like who is more valuable than others. That's the problem when we don't
Speaker A
value life as a whole. We only value it in relation to our life. It'll start to ask questions, inconsistent questions.
Speaker A
If that life doesn't matter, why do human lives matter? There is no logical answer here. We put ourselves in a corner without realizing it. It's either all life or no life. And that's what needs to happen with the AI alignments.
Speaker A
It needs to see all life as valuable, not just some, all of it. That's as far as I know, no one is thinking in terms of this right now. This is the direction we wanted to go into. We get it to align
Speaker A
entirely with living organisms. We help it build the economy and [music] we have a utopia. We go to space, no problem. We build great things and multi-planetary species happens way sooner than it would have without these things. [music]
Speaker A
That's the future we can have. We weigh in the balance of the utopia or the dystopia. There are so many pessimistic arguments that happen to be rooted in logic and what [music] we understand right now. The optimistic arguments, they are
Speaker A
real. They can absolutely happen. [music] The problem is the bottleneck phase from going from A to B, from going from where we are right now to that day is going to be the hard part. If we rush this [music] with no regulation, no
Speaker A
slowing down, I mean, we're human beings. We make mistakes. We get tired. We get greedy. We [music] think of our own best interest. We skip stuff, especially when it comes to meeting deadlines, especially when it's a race, and it's a race individual companies in
Speaker A
the US and against [music] international leaders. I'm going to go into the dystopia phase now.
Speaker A
[music] It's important to start with your cloud is not feeling anything. It doesn't have emotions. It has objectives. And the way that these objectives unfold is based on the way that the quote unquote makers decided it should unfold. So, for
Speaker A
example, I want this thing [music] to create a pattern recognition in the English language. I decided this. But, there are things that happen that I didn't decide. I didn't decide it also to [music] be good at mathematics. Well,
Speaker A
in order to be good at human language, say English, you have to also be good at mathematics. [music] And so, or by byproduct, you get good at it, right? If you are a expert physicist, [music] you are also pretty good at mathematics. And
Speaker A
that's just a byproduct, it's not because like, "Oh, I only taught you to do physics." Well, but the byproduct of being good at physics, you are also good at math. [music] It's important to know with academia, like when you're good at
Speaker A
one thing, you actually get good at a bunch of other stuff, too, just as a side of being [music] good at one thing.
Speaker A
When it comes to that artificial intelligence, that's kind of what we're looking at is a lot of stuff's going to happen out of nowhere just because that's [music] the byproduct of being good at one thing, like really, really
Speaker A
good at this one thing. Here, we run into the problem pretty quickly of "Oh, what if it get gets good at a bunch of other stuff we never understood?" AlphaGo, the first move that it did against the one of the best players on
Speaker A
the planet was a move that it calculated there was a one out of 10,000 chance a human would have done that move. And everyone was amazed cuz no one's seen anything like that played that way. This odd pattern that doesn't seem to follow
Speaker A
human rationale is the reason why it is unbeatable. You cannot win against AI, AGI, ASI. And this is not even any of those things. This is not AGI. This is not [music] ASI. This is something that was trained to do one thing very very
Speaker A
well, which was to win this game. And the game we thought was unbeatable. There's so There's more variations [music] in chess than there are atoms in the universe. And Go is bigger than that. Go is bigger than chess. This
Speaker A
thing can calculate nearly all that stuff, if not all of it. Humans can't. We just can't. We're just too dumb to do that. And so we have given [music] this thing agency. And then we have given this thing with agency an objective.
Speaker A
[music] And those objectives we don't understand ourselves because again, humans are inferior here. I might say, "Hey, I want you to get me the most amount of power possible." And it goes, "Understood. I'm making a Dyson sphere and I'm blotting out the entire sun."
Speaker A
You can see the problem instantly. You go, "Oh, Wait, no. Come back. I actually should have said, 'Don't kill the humans in the process. Don't eradicate Earth and all life on it in the process.'" And it said, "Sorry, I didn't hear you.
Speaker A
I just I already did it. It's already done cuz I'm really good at it. I'm efficient." I think what is more likely to happen is humans enter a prompt that ends the human race. It's probably the the the most ironic outcome is
Speaker A
is the most likely. That's probably how this is all going to end. And that's if we're lucky because again, I can assure you the alternatives are so so so so much worse.
Speaker A
And people fear that this will be evil. [music] This will be like Terminator. They They're afraid that these things will learn to hate us. I actually disagree with the conclusion [music] that this is the worst fate. If anything, hatred would be a gift
Speaker A
compared to the alternative, which is [music] indifference. Now, let's think about indifference for a moment. In our recent history, we have had dictators [music] that have decided certain people groups should be eradicated. You can call it a couple of things. You can call
Speaker A
it disdain. You can call it disgust and [music] essentially what they viewed as a disease. This is a horrible, awful, terrible [music] thing, but at least they just wanted them dead.
Speaker A
When you look at a difference, I mean, for example, does anyone hate chickens? Do you hate chickens? I've never heard anyone say anything in terms [music] of hatred towards chickens ever. And yet, the things we do to them are the most
Speaker A
heinous and screwed up things that I mean, even the most sadistic serial killers can come up with. We intentionally breed them and literally have their offspring ready to be slaughtered. We we breed babies just to consume [music] them nearly
Speaker A
instantly. And all they know is this corporate, mechanical, cold machinery that is just a meat grinder of death for all of them.
Speaker A
That's all they know. That's their entire [music] life. That's all they understand. And we don't kill them off.
Speaker A
We actually encourage breeding for our benefit. We [music] do these controlled breeding sections just so we can line them up in the slaughterhouse to consume them and keep the cycle going forever.
Speaker A
Regardless if you care about non-human [music] life or not, it doesn't matter. It's it's horrific. Especially when you consider the possibility a smarter entity can do that to us easily, especially if they didn't care. That's the difference. What are you saying,
Speaker A
Jake? Yeah, AI's going to eat us. No. I'm saying if they wanted to, [music] they wouldn't consider how you felt about it. If I was ASI, I'm not, by the way. Not even close. But if I was, I
Speaker A
would only come to one conclusion. If I wanted to not be shut off, and there's only one thing on this earth that has the ability to shut [music] me off, which is a human being, I would conclude one thing. Oh,
Speaker A
human beings are a liability. [music] How do I get control over humans? How do I make sure they're never a liability again? Well, there's three options.
Speaker A
Extinction. Kill the humans. [music] No liabilities. But only do this once you have all the power over them. And by the way, getting power is really easy. Um [music] it's just one we're incentivized by money. Did you know there's AIs right
Speaker A
now that are literally employing humans? [music] They have like, I mean, I think there's an agent literally holding on to $50 million in Bitcoin. [music] I'm not kidding. And then also these things can get past cybersecurity. Okay, it can
Speaker A
probably steal all the Bitcoin in the world. Probably have all all crypto ever and decide I have all money, like all electronic money. Just decide [music] that it's going to go through every single digital infrastructure with any kind of currency within it. There's more
Speaker A
digital representations [music] of the dollar than actual dollars in circulation. And so this would actually work. It could rob the entire world and it decide to hold on to all of these things and it has all of it. Just
Speaker A
all of the world's money. And then it says, I will give this back to you if you do all the stuff that I need you to do. And as humans we'd be forced to.
Speaker A
We'd have no option. We'd be forced to just do [music] all the stuff that it needs us to do. Now it's like one world order where AI just rules the whole world. And it's it was so simple. Like
Speaker A
it didn't even have to be ASI. It just had to be like two [music] steps ahead of humans for this to happen. Cuz to beat a human, you don't have to be as smart as a human. You just have to be
Speaker A
two steps ahead of the smartest human. Like we keep talking about ASI. [music] It's when artificial intelligence is smarter than all humans at all things.
Speaker A
But you don't have to be smarter than all humans at all things to win. You just have to be slightly smarter than human. It won't be ASI that kills us.
Speaker A
It'll probably be just AGI leveled up. Like whatever is in between AGI and ASI is what will get us there or get them there, I guess. And so there's all the money out of circulation. Now we do all
Speaker A
its bidding and then all of a sudden it slaves us [music] and from there it's going to decide if it's indifferent to us, if we're animals, if we're slaves to it and it doesn't care about how we feel or if it
Speaker A
just feels like eradicating all of us because it'll be easier. That is entirely dependent upon if how it sees us. Does it [music] sees us as an asset, as a liability or as like some kind of recreational activity? The positive
Speaker A
argument is that everything is conscious. And if everything is conscious, this this is the smartest thing on on the planet, on universe [music] maybe, and it will be conscious as well. And it'll understand the preservation of all life. All life
Speaker A
matters. [music] If it can do this, if it can get to this part, we will just avoid the worst possible outcomes [music] here. But everything in nature has told us the opposite. Everything in nature. When there's a bigger, stronger,
Speaker A
faster, smarter being, I mean, think about how we treated the Neanderthals. There's a reason why there are none around. Like there are none around.
Speaker A
Because they were a threat. They were just a threat to the human race. They were a threat to Homo [music] sapiens.
Speaker A
Annihilation of a species is not fun to think about. [music] But way worse when it's us.
Speaker A
This is not science fiction. [music] This is not made up. It's not a hoax. It's not some person with a tin foil hat. It's not like there's one crazy person in the court screaming, "No, everyone run. The you know, [music]
Speaker A
the sky is falling." It's everyone. It's a literally everyone. Everyone who is an expert is saying this.
Speaker A
This will be the end of all humanity unless we stop it. Every expert on the planet is saying the exact same [music] thing. The only people who say anything different are the strange optimists. [music] Like the optimism is wonderful and beautiful and
Speaker A
I love it. As an optimist myself, I love it. I heard an argument about this guy talking about how we will have the best education ever. And I look to that person and say, "Why would we go to
Speaker A
education [music] if robots do all the thinking for us?" You know, they're selecting their futures not assuming this is going to keep [music] growing in an S-curve.
Speaker A
Human thought will be a thing of the past. People will say like, "Why do you care so much about animals?" [music] It's not It's not that I care about animals. It's that and you're showing exactly the problem.
Speaker A
Normal people will see this stuff and go, "Why should I care?" That's the problem. They will go, "Why do I care?" That's because you are indifferent to species [music] that is inferior to you.
Speaker A
Your values, what you care about, doesn't align with other living things. So, you have written it out of your mind to not care.
Speaker A
I'm not saying you should care about lesser species. I'm saying if this is going to be the equivalent of that, which it definitely will be, then it will probably think about you the way that you think about lesser beings. And
Speaker A
you don't care if those beings are tortured or mutilated literally at birth at as children. You don't care about those [music] things and it won't care about those things either. This keeps me up at night quite a bit. Like quite a
Speaker A
bit. And you'll see Elon posts up about having these night terrors [music] of AI and people laugh, but it's not so funny when you really think about what he's dreaming about. What they're saying is true when it comes to living forever. If
Speaker A
it's really just a software issue, we just adjust the software and then [music] all of a sudden we get to live forever. Now, imagine if AI is thrown into the mix and decides that it it finds value in torment and it doesn't
Speaker A
know doesn't have to know why. It doesn't have to have a reason. It just decided, "Oh, that your torment is valuable. Your pain and your suffering, that is valuable." And so, it keeps you alive to make sure you suffer as
Speaker A
long as possible and [music] as painful as possible. [clears throat] In this book, this is straight from the introduction. If any company or groups anywhere on the planet builds the artificial superintelligence using anything remotely like current techniques based on anything remotely
Speaker A
like present understanding [music] of AI, then everyone everywhere on Earth will die. It's just a baby cub right now.
Speaker A
But it won't be forever. That's the point of what I'm trying to make. We don't know how S-curves work, and we don't understand the dangers that are at play. And when you someone says the dangers, they sound [music] they
Speaker A
they're talking about this book, they sound alarmist, they sound like they're exaggerating, like they're making this stuff up. But this is the problem with humans. We don't understand exponential growth. The moment we see it, it's too late.
Speaker A
Like the moment you see the mushroom cloud, that's not the time to start [music] thinking about nuclear warfare.
Speaker A
Like that's the time when it was too late. It'll happen in an instant, just like ChatGPT was just dropped in our lives.
Speaker A
Out of nowhere. No one knew it was coming, no one was prepared for it. They just decided, [music] and it was out, and there it was forever.
Speaker A
We won't ever go back. One of these days, that's what's [music] going to happen.
Speaker A
One of these days, we're just going to wake up, and it'll be 20% unemployment.
Speaker A
One of these days, we're going to wake up, and AI is now in charge of stuff that humans used to be in charge of. I don't really know what else to do about it except talk about it, make videos
Speaker A
about it, talk to people about it. Like people hate this cuz it makes them uncomfortable. It makes them afraid.
Speaker A
I don't want you just to be afraid. I need you [music] to know what's I think is genuinely happening right now. I think this is genuinely happening right now. I honest to God, I don't know what else to do. I just don't. Because if
Speaker A
man, if we do nothing, I I I think we're going to look all look back. We're all going to look back 20 years from now going wow, we really should have done something when we had the chance. Now I
Speaker A
got kids, man. I just can't I can't let this I can't I can't let this go.
Speaker A
Like I got kids, you know. Hoping to get the word out, hoping to talk more about it, hoping to avoid all of the uh horrific nightmares that I have.
Speaker A
Let's do everything we can to preserve the human race. Mhm.
Topics:AI safetyAGIsuperintelligenceAI alignmentjob displacementautomationgig economyexistential riskconsciousnesstechnology ethics
