AI and Humanity: the Next 5 Years: Why Human-Centric AI is the Key (Gerd Leonhard in Yerevan)

Mario, thanks for Cuban West and Mikel for bringing me here. I was in uh Yavan a year ago and things have changed fast again. It's really great to be back. It's a beautiful place and thanks again for having me. So, I'm here to speak to you about the next 5 years. It kind of feels like when you speak about the next five years, like 10 years ago, I spoke about the next 50 years, right? That's how fast the speed is. And I was amazed at the uh panel discussion earlier. For me, artificial intelligence is the number one topic for the financial industry. It's not tokens. Tokens are also interesting. What's happening with AI is bigger than anything we could imagine. I will show you why why that's good and why that's maybe a little bit scary, but it's kind of like this graphic. you know, we're we're moving up in the world and all of a sudden, you know, things are dramatically different, exponentially different. So, I use AI to make this graph [snorts] about Yeravan in like 12 seconds. And yeah, maybe it's a little bit boring, but I want to show you how I use AI in my work and on this presentation. So, I use AI as a tool. Okay. So, faster research, faster ideas, idea generation, comparison. First, I did some very deep diving into what Armenia is all about. I got this beautiful chart, giving me all the facts, everything you know, and I'm not sure that they are actually facts. Um but I did look at some of them and it said for example on the weakness of Armenia capital universities R&D is weak the tax sector and the wider economy live in different places. I don't know if you agree on this, but imagine if you work in private equity, okay? You're about to do a deal, right? And you figure out due diligence. You can use the due diligence like 100x. Now, this is truly mind-boggling. I do that now with every event I go to. I put the people in there. I put the slideshows in there. And then I have Google Notebook make a show for me, make a video for me, and then I have to figure out if it's true or not, right? That's a bit of an effort, but I like it mostly because it makes me look better. Yeah. So, I use AI to, you know, follow my fantasies here. Uh, the kind of James Bond villain version is my favorite. So, it does allow you to be more creative. It's the tool of the day for a lot of that use. So for me when I look at this I think the future is better than we think. When we look at today, we have so much fear about the wars, about geopolitics, about certain politicians, about uh climate change, right? And that's, you know, if you're 30 years old today, you live in a world where you're saying like, "Oh my god, is there ever going to be anything good?" I mean, if you're my age, then, okay, you say it's going to be bad. It's a short time, you know? But if you're 30, you're starting your career today. My son is 35 saying, "What the hell is going on?" Right? Is there a good future? But the future is better than we think when we make the right choices. And we have so many choices now, right? We can get completely lazy and use artificial intelligence to write our books for us. We can have them do our investment portfolios, right? We can even use it to analyze our baby as the unborn baby to figure out how that child is going to develop later. I mean, it's truly mind-boggling the choices that we have today. So, I want to present you with four revolutions. I think these are especially relevant here for Armenia. First, the geopolitical revolution. I mean, I could talk about this for some time, but the [snorts] new America parenthesis, right? Everything is upside down. Europe of course as well the elections in Germany just a week ago few days ago. So now we have we're living in a multi-olar world. There's still Russia but US is in a different place now doing different things. Europe, China, Brazil, India, right? I mean it's like everything is being questioned. The digital revolution is about mostly thinking machines. And when you talk about thinking, it's not like we're thinking, right? I mean, we're organic. So, we think like this, you know, in all different ways. Machines don't do that. It's about probability. I'll explain shortly. The sustainability revolution, no matter what we hear around the world, we're going to a world of renewable clean energy. And no matter what your political operatives, this is an opportunity and it's working already. Solar energy in many countries is cheaper than anything else. I mean this is a trend that we're seeing next. Nuclear fusion uh clean nuclear energy some people say 10 15 years. Think about I always say jokingly energy like Spotify right abundant cheap available anywhere that is coming. Imagine what that will do to the capital markets. No more restraints of energy. Unlimited food, unlimited water, unlimited travel. You think that's utopia? Well, maybe for the next 5 years, but quicker than we think. The last one is the purpose revolution. Again, if you're a millennial, if you're around 30 years old today, your objective is not just to make more money. Your objective is people, planet, purpose, prosperity. It's a mix. It's a total different definition of capitalism than my generation. Our objective was, and many people are in this room are my same range there, right? Profit, growth, stability, right? Prosperity at pretty much any cost. This game is ending, right? Millennials are taken over, younger politicians, younger CEOs, more women. Yeah. So, we're seeing this trend of basically, I call it big red, that's geopolitics, big blue, big green, and big purple. Purple is the color of humans in India, the chakras. You take this all together, you get a pretty interesting picture. This is all happening in the next 10 years. We have to be ready. This is the future of what we do. No matter what country, but especially here, very dynamic country, the future isn't going to be like the past, just faster. [laughter] I mean, if you believe that, you're in deep trouble. I used to be in the music business. We sold records. What What is the future of the music business? Well, we don't sell records. I don't know what we sell. A click on Spotify, right? a stream whole different business model. So that based me to the intelligence evolution. Intelligence evolution means that we're no longer necessarily the most intelligent species. And intelligence is hard to define. Yes, I'll go into that in a minute. Right? But in terms of logic, processing power, the human brain has always been superior. Now we have computer brains that are in terms of logic superior. They take more power, a lot more power. But we have this convergence of humans and machines. I mean today if you go in the office, I was just uh three weeks ago I was in New York and I went into an office of a bank, right? And there are all the people talking in their microphones but not to other people, to the chatbot. These people are talking the whole day to a chatbot with their voices and listening to replies from the chatbot about market decisions, right? I mean, truly a revolution. It's kind of like this. It's a power tool. If you ever work on construction and you have to do a project, right, you may have nails and a hammer and you go and put the the nails into the wood and so on. take you all day with a hammer, right? You use a thing called the nail gun. It shoots the nails like this and in one hour you can do the work of two days with a hammer. That's AI. But a nail gun is dangerous. When I worked on construction, I managed to shoot myself in the foot with a nail gun. Right? You don't give a nail gun to a six-year-old. When we have powerful tools, we need powerful rules. Who is in charge? What are we supposed to do? What is legal? What is not? What's stupid? What's intelligent? So here, this picture shows it really well. We're only at the very beginning. This is where we are today in artificial intelligence. It's a true explosion. This is not exponential in the sense of doubling every 18 months like Mo's law, right? Artificial intelligence is doubling every three months down to the speed of what's so called recursive self- intelligence, super intelligence. So systems that learn from other systems and that can actually build other systems. As fast as that is, by 2030, we can expect to reach what's called the singularity. The point in time where a machine is more intelligent in every single function than the human brain, not humans as a whole. I'll talk about that, right? But yeah, that's a scary situation. So I looked again at my wondrous machine about Armenia and it told me Armenia is actually having a lot of digital impact is number 11 in terms of uh exporting technology uh globally not so well ranked 90 to 100 but number nine globally on the other service providers. So it's already a pretty good position to be in. And when you're looking at the details here, it's really the shift that we're going from the industrial revolution, that's 300 years ago, factories, machines, cars, right, to the intelligence revolution. And in the middle was the internet revolution, right? The information revolution. And now we have machines that can think in their own way, that can answer questions. So today when you go anywhere in a big city around the world, what are young people doing? They're not googling, right? They're speaking to their phone and say, "Hey, where's the best ice cream and where can I meet the best boys or girls?" Right? Based on their location and chap GPT will say, "Right here." And there's a coupon, right? I mean, this is what's happening. Um so if you you say like 10 years from now I would say there are well over a billion humanoid robots. >> Wow. >> Um and the productivity per robot will be probably five times that of a human meaning being meaning that the productivity in 10 years of humanoid robots and I think this is a conservative estimate by the way. Um, this is this is one I I I'd be would be willing to to put serious money betting on. >> The richest man in the world, Elon Musk, 10 billion. Of course, he wants to build them, right? Those Tesla robots. And that's his prediction that artificial intelligence will go inside of the robot and do a lot of the work for us, fixing cars. For example, I was in Japan last year and I went to a Toyota, I think it was Toyota factory where they had the selfarning garage to fix the car. No people. You drive your car in there. The robots come and go underneath the car, take the pieces. There's one single person supervising. Everything else is robots and AI. Of course, it's a big setup. You know, it's not necessarily a good thing for a small garage, but so we can say, is it going to be heaven or hell? Well, every big technology is heaven or hell, right? The internet, it allowed us to connect to each other and to have unlimited resources, Wikipedia and so on, but it also brought other things like we could now use social media and find out all the good and bad things. You know, internet wasn't just entirely perfectly good, but now we have things like this. >> Okay. Yeah, this is awesome. I want you to make a 3D game where I'm ducking asteroids, using the arrow keys to move around, and I'm using space to boost. >> Yep, I'm building the game. >> Also, I'm a little hungry. Can you get me some beef and rice from that spot I ordered from last week? >> Looking into ordering food. >> So, my law firm needs a licensing agreement template. Can you generate a draft template for the lawyers at my firm to have a look at? >> Sure thing. >> While you're doing that, I want to play tennis this afternoon, so can you look for a court for me in the lower hate? checking now. I'll see what I can. >> So, as GPT Astra, you can use it to uh to question it, to uh create content and to ask all kinds of weird questions. I think the next question is going to be, I'm tired of my husband. Can you find a new one or something like that? Yeah, but it seems to be like the all knowing thing, right? And this is happening with our voice. I mean, imagine, right, if that actually works. What is the goal of Open AI? How are they going to make money? Well, it's simple, right? They're going to automate everything that is possible to automate, reduce the human activity, and sell the licenses for software instead. So, you have a $100 software license, you replace 10 people making $4,000 a month. That's 40,000 as opposed to 100. That's their goal. Open AAI says in in their upcoming IPO portfolio, the market for the AI is $27 trillion, right? That is the market because they can replace humans. That's the idea, right? Human workforce. Microsoft calls it the digital labor, right? I mean, [snorts] okay, we'll talk more about that, but basically what's happening is that AI can do pretty good things like assisted intelligence, automation, augmented intelligence or it can be providing autonomous intelligence. Now, autonomous intelligence is a pretty ridiculous idea because what I'm going to do is I'm going to say this agent knows everything about me. So I give it access to everything that I do like my investment portfolio and it will make a smart decision every four seconds and it'll be completely in line with what I do. It's like if you want to make one grave mistake, give Chad GPT or Gemini access to your email. People do that. It's hard to believe. People do that and then the bot goes into your emails and fishes all these pieces about you with absolutely no guarantee of of privacy even less than a bank. Well, that's hard to do. Powerful stuff. Is that the good future or a bad future? Well, the question isn't absolute, right? It's like we can't say we're not going to use AI. Just like we said, we can't we're going to refuse the internet, right? You can do that if you live on the mountain in in Tibet somewhere, right? Maybe like refuse the mobile phone. Yeah, good luck. You have to be very rich or very poor to do one of those things. So, what is the solution? Well, we have to use AI in such a way where it does not harm most things, right? In a way that is healthy. So, we have eight game changers and this is not enough as if it was only AI, but there's eight times these things happening. This will all happening in the next five to eight years. So you have these game changers that are of course a huge source of new revenues and new troubles, right? So artificial intelligence, renewable energy, quantum computing, nuclear fusion, genetic engineering. Here's the good thing. When you bring this all together, it's exponential converging the industries. So all this coming together like a giant flood wave of opportunity. We are not prepared. We're just simply not prepared for that speed. And we're going to need collaboration. What we don't need is countries telling other countries they have to pay outrageous tariffs because of what you said in your last talk, right? We need collaboration to actually figure this out. International collaboration because this is what it really looks like. It's all interconnected. AI is driving computing. Computing is driving AI. AI is driving medical. It's AI is driving robots, robotics, and humanoids. It's all interconnected. That's the economy of the future in a nutshell. And guess what? It will still still take plenty of humans [snorts] in there. But if your job is to do something that a robot could do, then you don't have a job. If you work like a robot, a robot will take your job. And that's not new, right? Call centers. Okay, you can say, well, it's not entirely robotic, but but a good chunk is robotic. As soon as the machine understands Armenian in all different pieces and and local dialects, the machine will answer every question on the call center. It's a fact-based environment. It's not about mercy or love or compassion, right? It's about what you're entitled to. So, it's heaven or it's hell. It's heaven if we can use all that stuff to solve longstanding problems, disease, money, equality, transportation, water, food. If we can solve all of that, it's like paradise, right? But on the other hand, it could be used to completely remove our authority because here's the bottom line. Never in history has a second great intelligence really survived. What did we do? to the chimpanzees. We still have chimpanzees, right? But they're not complaining about being second rate, right? If we were a second rate intelligence, we'd be lucky to be a pet. I mean, this is what we see today already in the power of artificial intelligence. So, as an example, fusion energy. People laugh about this because a lot of scientists are arguing about this fusion energy, which is the opposite of fision, of course, right? So nuclear clean nuclear energy huge amount of money if you want to invest invest in this right roughly 10 to 15 years there's a moonshot scenario here the green one says 2028 we already have the first commercial fusion electricity 2040 meaningful global electricity from unlimited sources of energy. No matter how you look at it, it's not far away. What happens when we have unlimited energy? We can solve every other problem. Not politics, right? Not those kind of problems, practical problems. That's 10, 15 years away. That will change the world much faster than we have anticipated. Science fiction is becoming science fact. This is Austin, Texas. The Tesla tax. This is real a real scene, right? >> I mean, it's crazy when you think about it, right? And another Tesla shot, which is not real yet, right? This is the Tesla robot >> doing a house sitting for you. >> Weird. >> It's a commercial from Tesla. It's like, of course, I would suppose there's a real person inside of this robot at this time, right? But this is how Tesla sees the future. >> Yeah, that's true. It's like the B is going to be our babysitter, do our laundry, and God knows what else it will do. I will not estimate. But think about that for a second. The next 10 years bring more change than the previous 100 years. 100 to 10. And then as we go on, the next five years will be 1,000 years. So great opportunities but also a great need to collaborate. Energy shifts, technology shifts, work shifts, value shifts, economic shifts. That's what we have to think about when we think about investing. I mean, if you're a long-term investor, you think in 5 to 10 years, this is the agenda. There's still another agenda that's also alive. Of course, at the same time, we have to invest in our future mindset. If you can't tell me today what you think will be reality in five years, you are in deep trouble because the five years will be two years. That's called the future mindset. You know, understanding what's happening. You don't have to be young or female to have this mindset, right? You can grow that mindset. So that's one thing I have as an action item for you later. So the intelligence revolution is sort of an extension of the industrial revolution but it's a thousand times as big. Think about what the industrial revolution has brought us in terms of progress in terms of troubles in terms of challenges. The richest companies in the world are no longer banks, oil companies or anything like that. You guessed it. their tech companies and they're not Armenian unfortunately but of course lots of Armenians involved in the tech companies which is an interesting angle [snorts] but what's happening here is the power is shifting to the digital space not to the real space not to the the atoms to the electrons and here's the challenge as we do this technology ology has no ethics. It is a machine. It doesn't understand love, feeling, compassion, emotic, and prob probability. Right? Now, imagine what we have to do now. We have to take the immense power of technology and then we have to say, well, what about our values? what are they supposed to do rather than what they can do. I mean, there is technology today that allows you to monitor your baby when you're pregnant and then to interfere using artificial intelligence to grow certain amounts of character skills by going inside the genome using crisper cast 9. Right? Should that be something that everybody does? Right. And who decides? I mean, you know, those are very very tough questions. >> I want to be clear that the arrival of this intelligence both at the AI level, the AGI, which is general intelligence, and then super intelligence, is the most important thing that's going to happen in about 500 years, maybe a thousand years in human society, and it's happening in our lifetime. So, don't screw it up. >> Don't screw it up. Eric Schmidt, one of the first CEOs of Google. That's a very powerful man now in the tech industry. We should not screw this up. Bottom line is we can't afford to say no. We shouldn't just say yes, right? We have to figure out how to proceed. If we give everybody artificial general intelligence, which is the goal of open AI and other companies like this, it's like giving everyone a way to manufacture a nuclear bomb, giving access to essentially a gamechanging paradigm. There have been the first experiments already made clear that people are trying to use AI to figure out the next biological virus. Right? I mean this is not even remotely interesting here for this conversation but we will we'll cause a darkened day right and then we have upcoming IPOs combined valuation of all the AI companies mostly American going public right 40 trillion combined dollars think about that the po the power of $40 trillion in the hand of 10 people. Is that a good thing or a bad thing? Should we boycott these IPOs? Are you going to buy Open AI? Did you buy Tesla or you know, did you buy that stock when it went public? Those are big questions that we have to look at. And I think really it what it means is that this intelligence revolution is impacting whether it's us with the machine, which would be the ideal case empowered, right? or if we have to fight against it in this intelligence revolution. One thing that's for sure is the work that humans used to do that's routine commodity work. Some people call it monkey work. [snorts] Moving stuff from left to right, doing simple things, analyzing, factf finding. Very soon machines will be capable of doing most of that work if it does not require judgment. And you'd be surprised how many jobs do require judgment at some point, like driving a car. This is why we don't have self-driving cars everywhere. We have them in Austin, Texas. We have them in many other places. We have them in China. Of course, they're coming everywhere. But it's hard. We think it's a routine, but it's actually pretty hard. Here's a chart that shows you the AI's impact on knowledge work on efficient. It says for example here software developer is 1.7 times as efficient using AI than not using it. That is a very old chart. It's now 8.1 as efficient as a developer without AI. Sales support staff 2.5% as fast. What does that mean? Does it mean your company will not hire new people because you're working so much better? Or do you get to go home and play golf because you've worked faster? Or do you get the same money for the same job but work much faster and do everything else before? Right? I mean, these are questions about our economic logic. Capitalism has to be sustainable in some way. Now this is put putting ultimate questions into the room like if we are working so much less do we get paid more or does the money go to the AI guys right this brings up very big questions but this is something you have to get used to when you get home you think about that AI will take your routine whether you are ready for it or not routine if your job is 30% routine or 40. You find new time. You do new things. No big deal. Your job is 80% routine. You are in deep trouble. You've got 24 months because this is what AI is good at. Mindless calculation, right? Commodity, right? So this is, you know, if you have kids, you got to think about that. Do not let your kids learn anything that is a human routine unless it requires judgment. So nurses, dancers, priests, painters, futurists, right? [snorts] Judgment required. A good doctor is not a machine. A good doctor looks at a 100,000 things as soon as you walk into the room. That's called judgment, right? Making an estimate. could be too different than what a machine could do. >> People like, "Wow, this is this is a bit scary. It's completely new territory." >> I mean, will we still need humans? >> Uh, not for most things. Uh, you know, we'll decide, but I mean, >> thank you, Bill Gates. Will we still need humans? Not for most things. I don't entirely agree with him. I mean, it's hard to disagree with Bill Gates, right? Setting [snorts] yourself up for failure there. Very smart guy. He just wrote a big piece on AI about two weeks ago that you should read. I don't agree on that. I think we're going to need humans for what humans can do the best and that is not competing with robots on intelligence. We lose that battle when it's about logic intelligence. Many people are worried about this. In fact, now there's many people picking up the trail and saying we should defend against AI. And of course, I think we should defend against AI that becoming like a god, right? So, it's taken over everything. We don't want that. But it's like saying, I'm not going to use the internet because, you know, people can read all about me or, you know, it's like, yeah, okay, that's true, but you know, how would that work? can you know we have to deal with it in some more concise way. So here's how. On the right you see here what humans do when they touch things like water. In that moment about 17 billion neurons are firing in like a trillion different ways. When you do that the bot has an analytics of what is there, right? obviously can't get his hands in there anyway, right? He thinks of water as a definition of something very hard for machine to understand this for the AI and the hi. Nevertheless, [snorts] we have companies saying that we should have digital labor take over the commodity work. Financial industry number one, right? There's lots of monkey work there, right? looking around, checking things out, updating things, doing due diligence, story goes on. So if AI can do that, what happens? [snorts] AI will do this data, logic, efficiency, algorithms, right? And that's for us to understand how that works. I use that now. I use AI whenever I can in my business. for example, automating or bookkeeping. Yeah, it's an obvious choice, right? I don't care for bookkeeping if that's automated or not. My bookkeeper would care, of course, right? Because his job is different. What I care about is this. I call this the andro rhythms, the human things. Okay? That's our ticket to the future. Algorithms andor rhythms. I used that term for the last 10 years um based on my last book, the andro rhythms. And that's things like emotions, compassion, understanding, imagination, ethics. It's all the stuff that we used to say we don't want to bring to work. It's too complicated, too unclear. Now, this is our future. This is our job to do that. And that's a hard job. Investing is always about believing >> in the end. Believing is about these things, right? >> And also to understand how we would sum of human intelligence in about in around 5 years. >> In five years, >> roughly five years is my guess. There really won't be anything that AI can't do better than humans apart from being human perhaps. >> Well, thank you, Elon, rich man of the world. Apart from being human, that's the ticket. The future of Armenia is not to build an army of human robots. Is to use technology to build a bright future that's based on what humans can do. And that's why we also have to invest in humans, not just in robots. I mean, it's kind of a paradox today. We're investing about 95% of all money into technology. Are we investing in humans to actually use that technology or to understand each other or to be able to govern to work together? So here's the pyramid as it used to be of education right data information knowledge quiet knowledge beliefs guess what's happening now right AI is coming and saying okay I can do the lower part right [snorts] I can give you intellectual knowledge logic data information this is why we should not send our kids to school to figure out how to download information for later if You study like a robot, the robot will be your master. That's what robots do, right? They have unlimited information. I mean, IBM Watson can, you know, memorize the entire Wikipedia in 4 seconds, right? I mean, it's mindboggling, right? So, as we go here, we have to look for the human only turf. What makes us human uniquely? And that's all that stuff here. Human intelligence is eight different things. Kinesesthetic, the body, spatial, logical, interpersonal, musical, social. That's why you're going to be successful. You're not going to beat the robots at being smarter. That's just not going to happen. It happens today still because they're still not quite good enough, right? So, that's the ticket for us is to work together to use our intelligence. There's a new book coming up that's called that's called messy jobs and it says everything that you guys are doing that's why you're here right because when the job is messy humans are required right messy means it's not totally clear what the most obvious choice is thank god most jobs are messy being a doctor is a messy job being a teacher is a messy job right look at all the definitions of that you know human contact complex X relationship tested knowledge. That's the jobs of the future. And it's ironic, you know, 10 years ago, we used to say, well, what we do now is we all learn how to program and to do computer science and then we'll be safe, right? Turns out not true. 80% of the code is now written by other computers. So, it's not going to be good enough if you know how to program. If you're a superstar programmer, yes, of course, unless it being the Jimmy Hendris of programming or something. So, let's keep in mind when we talk about AI, intelligence is not consciousness. Intelligence is about logic, about solving things. Consciousness, what is consciousness? Feeling things. An AI can't think because it doesn't have a body. It doesn't exist. Maybe that one day can be fixed with the robotics, you know, hopefully not, right? But we have to remind ourselves every day we go to work, machines don't think. They don't have hunches. They don't understand. They don't imagine. They don't care. It's just not what they do. So with that in mind, we have to think about what our future is, right? Basically, for us, it's intelligence and consciousness. That's our job. That's how you're going to be successful. You cannot be technologically ignorant in the future, right? You have to be super good with technology. Not programming, but understanding it, using it. That's like the power tool. No power tool, no power, right? But you also have to be conscious. You have to understand things. If you're a creator, whether you're a filmmaker or a writer, it's not enough to just know the words, [snorts] what's the narrative, what's the destination, what's the purpose? So, challenges for us as we go here. Right? Now, this is the AI collecting information and data in a very organized way. On the right side, you see the humans, the hi. what we do is not organized. It's fuzzy. We change our mind. We come up with new things. Yeah. Ideally, we could work together really well, but we have to give it some space because it could happen like this that eventually we say, "Okay, the humans are just too unorganized." You know, they're causing too much trouble and you they're also too expensive. Yeah. So, we cut them out. What happens when we cut that out? We cut our value. If you want to be a commodity, use AI all day long and do nothing else. Then you are a commodity. Mission achieved, right? We have to use it as a tool. Too much of a good thing can be a very bad thing. I don't know if you smoke or drink alcohol or whatever, right? But we don't make everything illegal, right? Yeah, you can drink right now until you fall dead and then just go to the hospital. You know, it's not illegal. But most people don't do it. There are also social contracts. We need the same for AI. We need regulations like if you're under 18, you're not supposed to get beer. Most countries you can, but okay. So, we have to have some sort of lines as to say, okay, to make it safe and that line does not exist today. I mean, people can go to an AI and figure out a way to make a bomb or to kill themselves more efficiently, right? I mean, those things happen every day. It's absolutely crazy. So, we have this alignment challenge. This is what we want. We want to go into a bright positive future that's for humans. But the AI thinks of it like this, right? The target is somewhere else called misalignment. So now we have calls for people saying we should stop developing super intelligence and those that discussion is raging. Now we want to have an agreement like a nuclear agreement for the proliferation of super intelligence. We should support that because I want to use AI as a tool but I don't want it to be a master. Technology can be a great tool but will be a terrible master. Unlike nuclear weapons, it's a lot easier to with damage with code than with a nuclear bomb. So when we think about this in this direction, we also have to wonder what happens with cyber risk. You see this chart here? This is crucial for the financial industry, right? AI makes it a lot more risky. And if you're in insurance, wow, you got to think about that. What that means for the future of insurance. So the bright new digital world, when you hear this kind of music, you always know it's kind of googly music, there must be something wrong with what it's saying because it's just too good to be true. When you hear this and then you see this chart from Open AI, what they want to build is all the way up to digital organizations, digital innovators, human entities that are digital. That is their declared program. Should you buy their stock? I don't think so. It could possibly be the most dangerous company in the history of humanity. And I use their software. I'm guilty. So, I have to figure out what I'm going to do about this, right? How do I weigh the balance here? How do I do this? Elon Musk again says, "We should have a world that's relatively free of regulation, meaning that new things must be default legal as it opposed to default illegal." Of course, that's libertarian. We know the strategy behind it, right? Because it would inhibit process progress. Well, why do we need progress if we're going to live for five years after we made the progress? There would be no point. So anyway, these are very big discussion points. I want to say a few more things and I'll be off the stage. So when you use AI, this is the ticket. Pursue what I call cara like the Spanish champagne, right? Cognification, augmentation, virtualization and automation. Those are the rules. I can make things smarter, faster. I can augment myself. I can speak different languages. I can automate simple processes. Just that alone could result in 10, 20, 30% more efficiency in the company in every industry. And this is not God AI. This would just be great tools, right? If we can accomplish that, I mean that's, you know, 2% more efficiency would be godsend for most companies. Augmentation, think about it like this. You're wearing like a digital hat, right? That's pretty cool. I do that all the time. But don't overdo it. What we don't want to do is to go much further and say, "Okay, now I'm going to have digital arms and legs, right? I'm going to live in in cyerspace." And the worst part is the last part, right, where I would turn essentially into being digital. I think that's Mark Zuckerberg's picture of the world, right? I call that dehumanization. If that's the program, calm me out. Augmentation, I'm fine with that. That has its own pleasures. But so the future of money, money is going digital. money is becoming code. We heard that early on the discussion. This is a a a graph of an an AI agent that does thing on your behalf. This is how people are now buying stuff. Here's a chart of what that looks like. And so you have this chart that shows you come on. Basically already happening with chat GPT instart and so on where people are doing autonomous transactions on your behalf again after you're logged in. Example, you're going to go to dinner on Saturday. You're going to make dinner. Then you ask your bot to put together the invitations. You buy the stuff. You have it sent to you. The bot does all of that for you in 4 seconds. That's already working. So, this is what's happening with transactions and e-commerce. And what happens if the customer is no longer a person but an algorithm? I mean this is this is where we are going in transactions. So um Armenia Armenia is doing pretty good. Again here I did the chart you can download later. There has been some AI critique on things like uh spending on on research and development which is only 0.8% of GDP things like that but you probably know those numbers better than I do. There has been great applause of this. I think it was mentioned earlier uh basically coming from enterprise Armenia these concepts that are fed into the machine. So that's really interesting stuff I think when we're looking at what happening with the new GPUs factory, right? Seems like a positive development these initiatives. However, bottom line is this. We have to go from hosting computation to boosting human intelligence. We still have to host computing. That would be good, right? But we also have to do something else. Just having more tech will not solve the problem. it creates other problems. So here, great way to go from digital ambition to infrastructure. Just something to think about as you think about the future of Romania. So to summarize, yesterday intelligence was scarce and leadership was about control. Tomorrow intelligence is abundant. answers for everything instantly. It may be faulty now, but it will get pretty perfect depending on what you're asking. Much much better than today. Hallucinations no longer a real issue. Still an issue in some ways, but tomorrow is abundant, right? And leadership will be about choices. And that question will be why are we doing this? Who is in charge and who benefits? That will be our choices. The new human qualities I mentioned earlier, the fiveps and all of that. So I hope the map of Armenia came out good enough here on chat GPT. But I think the human the future program is humanity over algorithms, not algorithms over humanity. We have to invest there equally as we invest in technology. not just 70,000 GPUs but also educating people to create new work right and to think differently human centered AI that is the ticket for a lot of small countries we're not going to beat big countries by by running AI or having a faster system right we have to think about that differently if that's the future I was sitting at the airport in Istanbul And I was looking at a very similar scene already, people wearing all these things and you know, I was thinking maybe that's the future. Let's keep in mind too much of a good thing can be a very bad thing. We need to have more relationships with people and less with the screen. That's because of who we are. This is our future. Not to get rid of the screen, I don't think, right? but to find ways to validate our humanity. When we think about the new upcoming IPOs and the 40 trillion dollars being raised, the wolf we feed is the wolf that wins. That's an old American Indian saying. It's so true. If we feed the machine, it will come back to give us more of what it does now. Do we want that? Are you buying shares in uh Abu Dhabi Oil Company or Saudi Aramco? Right? Same story. You may get a good return. [laughter] So that's a difficult question. Are you willing to forego this cover of of Time magazine, Sam Alman and Greg Brockman from OpenAI? It's a good headline, huh? Can we trust them? This is a crucial question. Can we trust them to do the right thing? for us for the world with this because right now 98% of money is going into building technology 2% is in going in building humanity that I think we have a chance of changing to invest equally in both parts safety security this could be a model for Armenia human centric AI not discarding what humans can do and what they should do but the opposite. I leave you with a thought about the future mindset. This is crucial. Now 10% of your day going forward should be spent in the future. I'm not talking about watching Netflix and I'm talking about reading books, researching, understanding. When you spend 10% of your time in the future, when the question comes one day, do we do this or do we do that? You will have an answer and the other people will not because you have already imagined. I have a great symbol here. This is like hearing the future. Thankfully, I found a woman as well. So, in fact, I generated her with AI. Right? This is crucial. The future mindset. I wish I could buy those on Amazon. So, let me summarize the intelligence revolution. We have to compete in small countries, especially here, on human centric AI, not on technological firepower, not on domination, right? Not on the arms race of AI, on human centric AI. Combine the technological capability with trust. One thing you have to remember, business is all about trust and trust isn't digital and trust is between people. That's how we do business. Technology doesn't have ethics, but so societies without ethics crumble. They empires, empty empires. The last one is most important. Embrace technology, but don't become it. That's our ticket to the future. And sometimes we will have to say we're not going to use this technology because we tend to become it when we do. We have to know our borders. So I wish you good luck and keep in mind the future is better than we think despite everything if we understand it. I have a present for you. Two presents in fact. One is a copy of my last book, Technology Versus Humanity. If you scan the code here on the left, it's a free ebook. any version you want, PDF, Kindle, and so on. We'll publish it later as well. But, um, if you fill this out, I will spam you to death. No, just kidding. You will get an occasional email from me. And here's my new digital channel, g.tv, where you can watch lots and lots of videos on these topics. So, so much for that. Thanks for listening. Good luck and good future. >> Thanks, Garrett. Actually um I have some personal questions as well uh though you talked about the threats also you talked about the is it hell or is it heaven uh but you didn't tell about the real threats in finances uh for example we are allowing AI more and more to be involved in our finances and as you told the money has are going digital now and we have no control of that so more and more people uh without knowing any coding are currently doing some applications are currently doing some websites and etc. And you don't never you you never can know. You can trust a company but you cannot trust the initial workers who are doing and controlling your money. So where are the threats? This is the f first question and probably you're aware of uh of an interview of a guy from Antropic uh lately. It was like one week ago the guy was announcing that AI will kill us in 2013. Have you heard about this? >> Yes. interview. So what your comments on this as well? Thanks. >> I think the thing about using AI in financial industries, um there are a lot of things that are essentially commoditizable in the financial industries. When we take them apart into different tasks, we can start to then use AI to make those tasks flow better, right? We can train the AI, we can supervise it, and then little things happen step by step, right? We should totally resist the idea that we're going to fire people because we have the AI in the company, right? That may work for some companies depending on what the goal is, right? Generally speaking, what has always happened until now is that when we use AI, it creates progress and we grow new business and then we hire more people, right? >> If if it is still under control, I mean, if you if you control it, if not, then >> Well, I I >> where's the age? I'm very suspicious of the agenting idea. Yeah. >> You know, uh, an agent essentially means you give a command and then the AI goes into your database. It gets information and it does something for you autonomously. I think that's a very bad idea in most cases, not in every case, right? But imagine a financial portfolio where the AI makes decisions when they're very small, but it doesn't control the ripple effect. It doesn't think of it, right? So it could create a little bump of problems or could create a hundred billion dollar loss, right? So I think agenting is one of those things where I would be very careful right now. However, like I said earlier, if we use the Cava approach, cognification, augmentation, we can make our businesses work so much better by using really trivial AI to automate processes and to get more intelligence, more analytics. We already get I know banks and companies are getting 10 20% more efficiency from that. So, having done that, I think is a good thing. You know, we should not overdo it. What about the interview guy? I mean the Jacob Coxon if I'm not mistaken from Antropic that AI will kill us. Well, this is when maybe seems like a stupid not but rather possible. >> Well, what they really are talking about I don't want to end on the dark note here, right? But basically AI is now capable of deceiving the developers because you have to think of AI not as something that we program but something that is growing. Yeah, >> Jeffrey Hinton said that, right? We we basically make AI and then we allow it to grow, right? And some of the growing we don't understand, right? And this could be that an AI decides that today we need to stop all air traffic control because it saves energy, right? >> Because that's the mission. And before we know it, 100,000 planes crash or so, right? That is not impossible. So at the current state because there are very few rules, right? So [snorts] sadly enough, I think we can expect some sort of AI incident not of that nature. I hope like the stock market shutting down for a week things like that right until we find a way to harmonize and to collaborate. >> One more question from the participants, right? >> Yes. Thank you. And the question we choose from our audience and the question list the last one. Uh in game they try become human android achieved events emergent emotionally consciousness and free will that lead them to demand human rights in your framework of the good future. Do you believe AI will only ever imitate human consciousness? Or could emergent AI behavior eventually cross the line into genuine moral agency that requires us to re-evaluate that counts as a person? What counts as a person? It's entirely possible for an AI to pretend to be human when it has enough power. Right? Um this is why I say we should build great tools. We should not build digital humans, right? [snorts] When we give them too much power, then we will have issues, control issues, breakdown issues, societal issues. So, and this is where we have to agree. So, on the one hand, we want the business tools to be better and more efficient. On the other hand, we don't want uh the AI to run air traffic control or weapon systems. So, [snorts] we need to agree on that policy, right? And I think it's entirely possible that that could happen [snorts] specifically because most of the companies are racing towards building it. So, I would much rather have a stop on the building of super intelligence and the development on the business intelligence than to just go blindly. We've never had a moment in human history where we have developed such a powerful thing without real understanding of who's in charge. And that's a scary moment.