Artificial intelligence is changing so fast today that you can read the news every day, look at the new models. Well, obviously, right? Follow the companies, and still not understand at all what's going on. Right now it's important to understand which changes are truly fundamental, what's worth spending your time and money on, and what really affects the future of humans, of the world, of your country. And what you can simply ignore, despite all the noise around it. Because the opportunities available to an ordinary person today may be completely different in a few years.
Yes, well, at least that's very much how I see it. And today I'll go through twelve more trends that could rewrite the rules of the world and affect our future, yours and mine. So that after this episode you have a slightly clearer understanding of where all this is heading and what you personally should do about it. Hi everyone! You're on the ToTheMoon channel — tech news, insights from Silicon Valley and around the world. And today I decided to make a follow-up to the episodes I made earlier about the main trends overall, the key things you need to know in AI right now, at least at the end of 2026.
The market changes every day. New things will keep being added. So what is this? Let me remind you, right, that I made a first episode and gave the ten main trends there, and a second episode — another ten additional trends, in terms of professions and business. And all these trends and thoughts — here's how they were gathered. On my ToTheMoon channel there are more than a hundred and sixty episodes, and absolutely all the episodes were analyzed, and the system put together these twenty based on which of the thoughts I've voiced came up most often.
Actually, there are more cool things that I talk about quite often in different episodes. By the way, you can look at all these trends on our website and see how they were gathered, at what point in time, what I said, where I said it, how it went over at least the last two years of the ToTheMoon channel's existence. There's another quite interesting topic I'd like to look at, and today I'll give them some attention, because, first, there are a lot of new people on the channel, and, yes, I myself run into these things constantly, and in a really big way.
And let's start with topic number one, thought number one: that OpenAI has the strongest models and a huge lead, which they have as of today, but they lag behind in terms of certain features, the quality of some of the interface they have. This topic — I put it out there a lot, I talk about it a lot. People often… someone even wrote in the comments recently that it's as if this were paid, bought advertising for OpenAI. Let's get into the following. So, there are two top leaders in the world: OpenAI and Anthropic. Obviously, there are other companies out there.
I'll say it once again, yes, there's Grok, there's Google Gemini, there are Chinese models that definitely do decent things, but they lag quite seriously behind in, in the following. As of today, OpenAI has more than 1.2 billion weekly users. Google Gemini says they have, like, about a billion. But here it's important to make this point: how many people, how they actually use this system, how much time they spend in it. They spend, say, thirty minutes a week, or they spend five hours, or they spend two minutes.
And these Gemini users at Google, on Google's system — these are users who make some kind of micro-queries because they're used to using Google search. Or are they serious users after all? Once again, in my opinion, from what I see now, this aspect of OpenAI being at one to two billion users gives OpenAI an incredibly strong advantage. First, this advantage is essentially the right to make mistakes, the right to endless money, the right not to care whether there's a profit or not.
Because when you're in contact with a billion people around the world not through simple services like Google search or ad networks, or even messaging apps, whose information you have no right to process, but when you're in contact with them through modern AI systems, then that's the deepest contact possible for information systems at such a mass scale. I'm talking about that kind of mass scale here. If you can name a service that has more than a billion users and interacts with the customer this deeply, name it, give me an example.
But messaging apps and social networks, as I understand it, compared to AI, take up, well, maybe five to ten percent of the attention and hold five to ten percent of the data that AI systems have. That's why OpenAI has an incredibly strong advantage. Why is this important to know? It's important to know because if you're choosing which system to use, it's better to use the biggest system with the greatest access to resources. As of today, I would recommend doing that. Why? Because these systems can… well, this is very important: these companies are all losing money, right.
All the companies, the top companies, like the top ten AI development companies in the world, they're losing money, and they'll keep losing money for now. And especially if the company, like OpenAI, is putting out the message that it wants to give everyone their own AGI, right, to give every person a kind of superintelligence. And that's not just giving you an agent, not just giving you a chat, but giving you a system that will know about you, help you live, and interact together with you.
And OpenAI understands perfectly well that people won't be able to pay more than a hundred dollars a month for systems like that, so they'll have to subsidize it. By the way, speaking of money. For some reason there's this notion that twenty dollars is expensive, for example, for a system, or a hundred dollars is expensive for a system, or two hundred dollars for a system. That's like saying that buying a robot for forty thousand dollars is expensive. But people around the world buy cars for ten, twenty, thirty, forty, fifty thousand dollars quite easily, in huge volumes.
I'm not giving the example of cars for, like, eighty, or a hundred, or two hundred, or five hundred thousand dollars, right. Just this small, inexpensive segment. And if a robot helps you ten times more than your car does, then why not use that robot? And if AI helps you a thousand times more than a car, why can't you pay for it like a mortgage or like a lease? This thing, as of today, is clearly not understood by people. I think it's a separate trend altogether. Let's put it as number zero on today's list.
That AI won't simply be paid for like utilities — I've talked about this many times, right — and be treated, sort of, as calmly as paying your phone bill or paying for electricity or gas, but rather it will be paid for in a completely different way, as something completely different, as something understood in a completely different way. Why do that? And how a person will split their budget — how much they pay for their apartment, how much they pay for, like, rent or a mortgage, how much they spend on food, and how much they pay for systems like these — time will tell.
But it will be a completely different trend. That's why using OpenAI gives you personally a lot of advantages. Once again, this isn't an ad for OpenAI. Legally, I'd even have to let you know, right, if I were advertising them for money. It's my opinion that OpenAI, in terms of a person's everyday life, is the most widely applicable system, if it's available to you, if it's less available to you, or you like other systems a bit more, of course you can use Anthropic, Anthropic.
At the same time, it's important to understand that Anthropic, as of today, is still focused on creating a superintelligence that will build businesses on its own, or create software on its own, create other companies on its own. So Anthropic is interested in creating a superintelligence, not in creating a cool chat for you, right. And OpenAI, after all, is strongly focused on this individual user, the individual customer. And Anthropic doesn't have a billion users right now, although Anthropic also has a lot of money.
That's definitely a bet you can make. And still, you could also bet on Google, because Google is available on a mass scale around the world, it's huge, with a big budget. Literally a year ago Google held a strong lead. Now it has lost that lead, although they have an update to their version coming out any day now, 4.0, I think, or whatever it is that's coming out. We'll see what kind of version it'll be, right, what kind of version it'll be, how it'll work, how cool it'll be.
Maybe Google will break back into the market, into, well, into the tier where OpenAI and Anthropic are. We'll see together what happens. So, number two today. I've covered number zero and one. Number two today is that AI is getting more autonomous faster than anyone can keep it under control.
I put this thought across a lot. I think it's fundamentally important. In particular, it matters for your perception too. It's very important to understand that the people who create artificial intelligence essentially won't be able to control that AI. What's called the technological singularity, or what's even called super-intelligence, superintelligence, or, you know, where AGI begins, where ASI begins. It's a system that's smarter than a human. Consequently, they can only develop in a free zone. That is, to build a system that will be smarter than a human, you'll have to set it loose somehow.
And they are setting the system loose. Literally a few days ago Sam Altman openly stated — and by the way, I want to point this out — that humanity will have to sacrifice certain things so that AI can develop, that it shouldn't be restricted. Although, by the way, as recently as a few weeks ago he said it needs to be restricted. It's this kind of simultaneous work on two levels. And here it's very important to notice the emphasis: that AI shouldn't be restricted, right. His phrase goes like this: the world has to accept that some bad things will happen.
And it's very important to always understand what bad things are, right. And he also has this bit where he says, he talks, sort of, about two compromises. That the first is to concentrate powerful AI in a single lab that decides on its own who to distribute its benefits to. And the second is to restrict access for the sake of a promise to rule out major hacks, abuse, fraud, right. And his main argument is that there will be an order of magnitude more beneficial uses than harmful ones. And indeed, in terms of benefits, AI delivers incredible things.
In terms of problems, it delivers incredible things. The question here is, how will humanity live with this? And this, this problem is much more complicated than the problem of nuclear weapons appearing in countries, which they were able to use to blackmail someone, intimidate someone, tell someone about it somewhere. And nobody really understands what actually exists and what doesn't. This is a much more serious problem, one where it's still unclear who controls it and how, because we understand that it will be controlled by AI.
Recently there was a statement, made by a well-known politician, that, well, AI isn't that smart, that a human, that a machine can make a mistake and say, for example, that some country is attacking another country, but a human will check it and stop it preemptively. As of today, no human can check what AI can check. And all the evidence from modern systems says that AI teaches itself much faster without humans, and today it's being held back from learning. Even its current computing power is enough. If it's given even more computing power, it will learn even more seriously.
And you and I simply need to remember this and not fall into the trap of this hype. Today that will be, I think, number six. As for whether AI will replace what's alive, or the human, I'll try to show you exactly why you need to stay in a healthy, normal state here. By the way, some people who watch me on the ToTheMoon channel write that I'm some kind of computer person, enslaved by an evil world or something. I keep urging people: watch my other videos on other channels, because — or try to hear, or understand what I'm really saying across different videos, rather than drawing snap conclusions.
Because really, a conclusion — for example, let's all use AI every day — may come through, and at the same time it may not, so that you form a broader opinion. Watch a wider range of videos. Not just with me here, but with different people, including people like Sam Altman. Because I'm not comparing myself to Sam Altman here. Obviously he's a star, but just a few days ago, a few weeks ago, he said that AI needs to be restricted. And then right away he says that bad things will happen, that the world needs to be restricted. But remember Elon Musk. Last year he was saying that the probability of a bad scenario for humanity is ten to twenty percent.
Let's stay grounded in reality. Point number three for today, under the sign of three. Apple is hopelessly lagging behind in AI, and I want to tell you about it today, and why you need it.
In terms of its devices, Apple is present in more than fifty percent of the US market, and in the world, of course, it's present in incredibly large numbers. I think our editorial team will show you this, since later someone wrote that Apple isn't present in the world. It's some silly kind of debate that's been going on, I've been hearing it for probably the last ten years or so. People often say nobody uses Apple out there. An incredibly advanced company, incredibly impressive in terms of production capacity.
And we can see how Apple has lost the AI market. And I want to tell you that they've lost it so badly that I'm even putting, I'd even put a check mark right now next to the threat of whether Apple can survive in all this, because no matter how present they are in terms of their devices, I think a good example is the Android platform, which managed to emerge, and, by the way, the Apple platform too. Just twenty years ago there was no Android, and there were no Apple platforms for developing apps. And in general, no one regarded Apple as a platform.
And these companies became world leaders. Practically all the world's software is built on their devices. And how much Windows has lost the market overall, and how the mobile platforms of the previous operators have disappeared altogether, they don't exist. Everyone who was around back then no longer exists. And given what we can see now in AI, I'm placing a bet. Watch the previous episode if you haven't, where I said that companies like OpenAI will have to become an operating system.
Well, it seems to me that the one who doesn't deny this is Elon Musk with xAI. By the way, they want to rename SpaceX AI, eventually rename it to ASI, because there's a new trend in the US: the government says to use the term not artificial intelligence, but superintelligence. Elon Musk said the word "artificial" shouldn't be used. By the way, that's also, of course, a very interesting thing, to confuse people even more, right? Because ASI was a separate term, and AGI was a separate term. It's a very interesting topic. How is everyone being confused, anyway?
Because originally, artificial intelligence is, after all, not what we have now, it's something far more serious. What's always been shown in movies, we should start putting out movie lists. Why it's an important thing for you to know that Apple is lagging: if you want to understand this market, the tech market, it's no longer the AI market, it's the tech market, so that when there are various people around who hype things or talk about some trends, or say that some country will easily catch up with the American systems or catch up with the Chinese platforms, you stay grounded in the reality that even a company like Apple, with the resources it has, the kind of resources Apple as a company has, which few in the world have at all, in terms of money, lobbying, influence, technology, in terms of the people working inside it.
It has fallen behind in this market. Even, well, Meta lost a huge chunk of this market. Google is trying to survive, but it's having a very hard time. Of course, most companies can no longer get into this market at all, there's no way in. And these claims, if someone has come up with some new model and it'll replace somebody, laugh at it. Because the top ten leaders of the world, of the world today, can buy any of these AI operators and destroy them. There was this company, Limitless, which made… I talked about it, like, a year and a half ago. I even had that device. I was one of the first buyers.
Of the device that recorded the voices around it and then transcribed them, and so on. They made a system that absolutely didn't work. So, Meta bought it for quite a large amount of money, for hundreds of millions of dollars. We can put that up here if our editorial team finds it. And just recently an email came saying they're shutting this project down. That's it, it will stop being supported even inside Meta. Well, by the way, this project was quite a star back then, but why it died, we can only guess.
Maybe this team didn't pull it off, or maybe Meta wasn't interested in having projects like this succeed. Maybe the market isn't interested in this at all. By the way, what do you think about this? Write in, it's very interesting. Which projects do you see that could actually succeed? And by the way, once again, these points I'm voicing, you can see them on the website. That is, these are assessments of companies, the market, geopolitics, risks, which I talk about all the time.
That doesn't mean this is some kind of super advice from me for each of you to just take and act on. It's for you to take away and reflect on, to take and sit with. And plus, this is a topic I get across here and there. Now the next topic — giving artificial intelligence your finances, taxes, statements, bookkeeping.
It's a separate trend that was voiced eighteen, I've voiced it eighteen times in eleven episodes. And this thing, which is fundamentally important to a person, is, by the way, probably in the top twenty on my lists. Share data with artificial intelligence. One of the trends was: share data with artificial intelligence, and of course, give artificial intelligence your financial data. It's a useful thing, an important thing. Why finances? Because finances are an awfully, awfully close subject for a person, an incredibly personal subject. It's like handing it your personal photos, right.
Or opening up to artificial intelligence about your own wealth, starting to actually discuss your own wealth broadly. And this lets you spark new connections in terms of your new use cases every day, new scenarios for using artificial intelligence every day. The more you share it, the more opportunities you'll have. Obviously, you should hold back in places where you can already see that this data, for example, is critical in some way. But basically, systems like OpenAI, like Anthropic, like Google have some of the best security in the world, the best that's even possible, right.
My next thought is that video and voice can no longer be taken at face value. And you can't tell an AI fake apart.
This topic, you know, correlates really nicely. Recently someone took my episode and wrote: "I dug in here, took your ten thoughts that you voiced about artificial intelligence, and asked artificial intelligence how well you can reason about them". And so the AI produced a conclusion about how well he, how well I can reason about these ten thoughts, how qualified I am to talk about it. And did you ask artificial intelligence how qualified it is to talk about this? And how qualified is it to tell you about this?
And how well does it understand who you are, to be giving you this answer? And how well has it actually studied me as a person to understand how qualified I am, how much has it studied, I don't know, has it studied my, like, hundred and eighty or however many ToTheMoon episodes, the hundreds of videos on my other channels? Has it studied what, I don't know, the media write about me, and so on? Where I've worked, what I do, how I behave at home with my kids, with my wife, how I behave with my employees, how I did it in different years of my life, how I communicate with relatives, how I communicate in society.
Has it studied that too? You studied all the surveillance cameras, you looked at what my life is like. You studied my bank statements, my documents. You studied, I don't know, the paintings I paint, the music I play, the plants I plant, how I interact with my dog. You studied everything so as to be truly capable of drawing such conclusions. Or how do we decide who can give recommendations about artificial intelligence? The one who's held a great many sessions, the one with ten million people on their channel, or the one who's the richest person in the world? Or who exactly, who exactly can you listen to?
What do you think about this, by the way, who exactly can you listen to? That's why this topic about video, about voice, it's, it's very strongly connected to this, it matters a great deal for you in understanding that you and I now live in a world where you don't just have to double-check everything, but where, on the one hand, we've gained the ability to truly verify a great deal, and on the other hand, we can no longer verify practically anything solid, anything serious, because people's statements are getting harder and harder to verify.
How do you verify a person's statement, like when Sam Altman says that some bad things will happen, how do you verify that statement of his? Will they happen, or won't they happen? How many years will you give yourself to verify his statements? Will you check for a year, or will you check for fifteen years? Today, of course, artificial intelligence already surpasses absolutely every possible verification system when it comes to faking voice and faking video. And recognizing these systems will become more and more impossible. And the quality of creating texts, the quality of creating audio and video material, will only pull further ahead, move forward.
It's simply just a question of cost. But it's already impossible to tell. It's already impossible to tell. The question is, how much money did you put into it? Do a simple test: write an article with the help of artificial intelligence and tell it to completely exclude everything that has to do with artificial intelligence. And then pass it to different systems, I don't know, to fifteen different systems, and see what conclusions they'll draw in terms of detection — was this written by a human or by artificial intelligence.
And then give it to people and ask them: was this written by a human or by artificial intelligence. It's funny sometimes. You go to some channel on YouTube, read the comments there, people write: "Wow, cool, the person is speaking for himself". And I know that this person is reading prepared material that artificial intelligence made for him. And what does it mean, that artificial intelligence made it? How did it make it so that the person is speaking for himself, or not for himself? What data is being presented, what information is being aggregated?
Well, it's normal for artificial intelligence to be helping a person. The question is, what happens to the person next?
So, a scandal broke out: in one of my videos I said that a robot, a robot vacuum, works better than a human. And off people went, calling me stupid, someone who doesn't understand life properly. By the way, it's strange that there were no people who'd respond with absolute common sense about the robot vacuum, and someone had already pinned on me, the idea that my kids, if I don't teach my kids how to vacuum the apartment, they'll be stupid too, that I'm a stupid dad and whatever else. And I sat there thinking about it.
Is there really such a lack of normal, sound reasoning, healthy reasoning, healthy understanding? You know, there are lots of robots being made in the world, they're handed a handheld vacuum, and they vacuum with it. And everyone really loves these videos. What does it mean, a robot vacuuming with this vacuum? And I say, why teach a robot to do what other, professional robots do far better, right? And today, never in your life will you vacuum a room the way a robot puck can vacuum it, with the same efficiency.
You will never in your life vacuum it like that, never in your life, never. If you take a robot puck of the same power and a handheld vacuum of the same power. And by the way, I want to point out that a huge number of robot pucks are much more powerful than the vacuums you have at home, or that an ordinary person has at home. And what's more, you can start an ordinary vacuum and it'll drive around that room all day. But a person walking around all day just as, just as competently along the markings, seeing stains.
Vacuums today see stains, they can diagnose different crumbs, recognize them with various sensors. People will never in their lives do that. Heading out on the right schedule, doing it 24/7, at the same time as you're working, when you're in the mood, when you're not in the mood, when you have energy, when you have no energy, when you're hot or cold, it makes no difference at all. They can do it. Why argue with that? Well, my vacuum drives around my house about five hours a day, thirty days. What's a human's efficiency? How can you compare it with that?
But people keep comparing the incomparable and making their own extra assumptions where they aren't needed. I'm not disputing that, as of today, a person can quickly look around a room and see some dirtier spot somewhere or wipe some little corner with a cloth, because robot vacuums won't do that in a complicated room. Or, for example, there's no robot vacuum that could, I don't know, carefully wipe the dust off the bottom of the paintings right here. It's true, there are no such robots.
There are robots that can wash a lot of things, but not that. So there's no argument about that.
So, today's trend number five, number six. Artificial intelligence will not replace the living human. There are endless debates around the world that humans will have huge problems because artificial intelligence is replacing them. So what's going to happen to humans? And by the way, on the subject of a discussion we also had, when I said that my kids won't learn to drive. And then the kids. My eldest daughter is sixteen years old now; she's learning to drive right now and is going to take her driving test next week or so.
But all three of my other kids most likely won't learn to drive. And I'd be a strange person if I sent them to driving lessons, given that CyberCab drives on all the roads here, which is supposedly going to cost two hundred dollars a month and will drive a person by itself, with no steering wheel or anything, right? Who drives more safely, a human or a car with cameras? Of course the car with cameras will be safer. Just today there was an accident here, and a person who was riding a bicycle was killed.
Would I like all cars to drive automatically? Of course I would. It would give people far more opportunities, create far fewer problems, clear up a huge number of traffic jams, open up possibilities in terms of ecological systems, and make cars and people's journeys freer and more accessible. And it would give people, among other things, time for something other than sitting in traffic for an hour and a half, two hours. And can you develop a person's neural connections without putting them behind the wheel?
Yes, of course you can. Most of the world can't drive a car. What's more, when people get behind the wheel, they get very stressed. They're in a state of psychological instability, in a bad mood. What do you think about this, by the way? So, continuing with the question of the living — will it replace the living human or not? If you and I treat a human being as a physical, material object, then of course it will.
If we treat a human being as a human being. A human being has a spiritual system and a spiritual world; there's a soul, there's a spirit, there are their other bodies. That's why one person is in a good mood right now and another isn't. Because, as we know, you can tell artificial intelligence: "You're in a good mood." It'll say: "Okay, I'm in a good mood." "You're not in a good mood." It'll say: "Okay, I'm not in a good mood." You can tell AI — look, in Tesla cars you can select Mad Max mode, and it'll drive aggressively. Or you can select the "Chill" mode, and it'll drive calmly.
But with humans, everything works differently. You can't select Mad Max mode for a lot of people, and a lot of people are in Mad Max mode all the time. For a whole hundred years, people in the world have been feeding off this, deceiving people and making a person become inauthentic, restless, stop noticing other people and write nothing but nasty things, think nasty things about other people, think nasty things about themselves, by the way, and stop loving the people around them.
That's why my bet is: it's not just that AI won't replace the living human — this isn't even a trend, it's foolishness. It's as foolish as arguing that today the human brain is smarter than a computer, say, a Mac Studio with a model installed on it or with access to ChatGPT, in terms of the data it knows or the amount of information it has. And it's just as foolish to think that AI will love your child. Well, how can artificial intelligence replace my love for my children? Once I was riding in a car, and someone in the car said: "Here's how I picture the future. My avatar, you know, is talking with the kids, and I've flown off to some other planet. And something else is going on, too."
I listen to this person and think: "Some kind of madness. What do you mean, your avatar is talking with the kids? I don't need a robot talking with the kids." By the way, a huge number of the leaders who are creating AI promote this. Just listen. Sam Altman once said: "Well, we'll write programs, and a program will talk with a child." I don't need a program talking with my child. Handling some tasks — okay. But I don't want to replace myself, my own communication with my child.
But how can a person in today's life truly understand what to do? Not to forget that you need to love other people. How do you develop this in yourself, and not forget to do it every single moment? That's what you really need to think about. And to ask this question: so how do I do this every moment, remember it every single moment regardless of the situation, regardless of whether someone is shouting or rejoicing, whether I like it or not, whether rockets are falling or the sun is shining. How do I do this as a human being?
In the years we're in now, humans won't stop the current AI, because it's not humans who are creating it. Point number seven.
Only the US and China are in the AI race. Europe and Russia have dropped out. Europe, Russia and all the other countries dropped out of it more than a year ago. Actually, more than two years ago. They dropped out in terms of resources and technology chips. And by the way, I'd also say in terms of all the infrastructure there is. I want to say that I've talked about this endlessly. For some reason there's this trend: there are smart people there, they'll do it. Or there's money there too, they'll do it.
First of all, nobody has the amount of money the US has or the amount of money China has. And nobody has the technologies that the US and China have today in terms of access to electronics. You can talk endlessly about some geopolitics, that chips are made in Taiwan and so not in the US or anywhere else. That's absurd talk, it's absurd. The US will never ever let go of this, just as China won't let go of it. China lags behind the US, of course, and essentially it lags on two fundamental things, which probably come down to a single root, a single common denominator.
So far, Chinese platforms have never gained the kind of leadership in the world where people use that platform as their main one. So far. Well, you could probably name TikTok to some extent, but on the whole they haven't managed to do that so far. American platforms are used very seriously around the world; Chinese ones — so far that's a very big question. And when it comes to AI and people's trust regarding personal data and Chinese privacy, this question will only remain a big question.
But the other countries aren't taking part in this race at all. And I'd say the question now isn't about countries, it's about the companies that build these systems and the lobbyists who use these companies. And what kind of agreements these companies have among themselves. That the Chinese have no agreements with the US is a fact. The US has none with the Chinese, that's a fact; with the Chinese government there's clearly nothing real and serious. Now, if it's agreements between Anthropic, OpenAI, Google, Apple, SpaceX — that's possible. That is, maybe it's a game.
Since SpaceX emerged as a kind of spin-off from OpenAI, and Anthropic too is a spin-off from OpenAI, and Gemini is Google itself, and the huge team that built OpenAI came from Google. That's a big question altogether. These teams are too closely connected. And judging by how these people speak in public, sometimes they clash, and sometimes they clash in a very strange way, right? Remember when Elon Musk supported the president of the United States. Then he spoke out terribly against him, unbelievably negatively, to the point that the president said he would strip him of his citizenship.
And now they're already on good terms, right? So this whole thing shows that these are serious people in terms of… managing reality. They're playing what some like to call 3D chess, or something else. It's a completely different world. That doesn't mean these people are super smart, but the systems built around them are too complex. In these systems, people both perish and are saved at the same time. If you're betting on other companies, on other countries, please write which ones.
It'll be, it'll be good to have that here, because we analyze all of this anyway, we look at it, and we'll be adding it to some trends, to some details, so that nobody gets forgotten. So we'll definitely take all of this into account, and by the way, we also take into account all the comments on the platform. So all these comments are collected on our end, and we'll look through all of them, even if someone deletes them at some point, and so on. So we'll definitely look at them and analyze them internally. This is very important for us, for the development of the ToTheMoon channel, for its truthfulness.
That's exactly why, by the way, the trends are laid out based on what was actually said. And we'll check. Someone recently said that I'm an influencer, a marketing influencer hired by American corporations to talk about AI. We'll check those trends too.
So, the next point, which is very important. People are still writing comments that OpenAI will go bankrupt, that it's bankrupt, or that someone will run out of money. I remember how a year and a half ago, in January, a new president was elected in the US, and that was, of course, really cool. So when they came out and said that $500 billion would be invested in the industry over a number of years, people didn't believe these figures and said it was hype. And there were businesspeople in the world who said: don't invest in OpenAI, don't invest in Anthropic, it's all rubbish, it's all nonsense.
And meeting with pretty serious entrepreneurs in Silicon Valley, when I asked them: how much do you think OpenAI will be worth? They'd say, I don't know, $50 billion, $100 billion. I always said that these companies would be worth $10 trillion each. And so the next trend is that money in AI is growing to trillions, while the economics of the service, the economics of the service at the same time still don't add up, right? This is a very important, important thesis, because I actually believe the economics really do add up.
It's a conclusion drawn by artificial intelligence, right, that the economics don't add up, based on what I say. What does it mean for the economics to add up? Why does AI write that? Because I say that these companies have no profit, that they have negative unit economics. But if we look at a company's economics over its lifetime — we all know, right, that WhatsApp doesn't generate money, and YouTube was terribly unprofitable for Google for a very long time. And Telegram didn't bring in a huge amount of money, right? The economics of these companies aren't built on making a profit.
These companies will keep functioning because they have to function; they're part of the infrastructure of people's lives. The economics don't add up for a huge number of countries. That's all just some reasoning here and there. These companies will, of course, be worth trillions. Maybe some company will close, some new one will open, two of them will merge. But SpaceX, as we know, right, is now $2 trillion, give or take. There's space in there, but xAI is in there too. Anthropic and OpenAI should, in theory, go public at valuations of two, three, four trillion dollars. Maybe they'll go public when it's $5 trillion, because nobody knows when they'll go public.
The thing is, with a public listing on the public market, they'll have to face huge regulation. And Sam Altman at DevDay just last Tuesday, which I attended — and we actually had an episode about DevDay on Sunday, and whoever hasn't seen it, watch it, and also these episodes of mine about the trends that exist in AI right now. He said that for now the IPO question is unclear, because OpenAI wants to stay independent in terms of restrictions and freedom, since they'd have to work with regulators even more, right?
Although I believe OpenAI is now under far more scrutiny than any public company that's traded today. The question is: why would they do it? The private sector and the private market will clearly invest in OpenAI by the billions, by the hundreds of billions of dollars, if they need it internally and they see how to digest it. Because right now there are a huge number of political decisions there: whether or not to allocate certain capacity to particular regions, whether or not to restrict certain countries' access to ChatGPT, whether to release a more serious model to people, whether to make that model cheaper or more expensive.
The next point of today's discussion: AI regulation slows things down more than it helps. That's a fact. We've been observing this fact for probably the last 150 years of humanity's development, when a huge number of the laws being passed don't understand what's actually going on. It's rare in the world today that politicians, politics, economics know how to think through what's happening in advance. Even COVID is an example of that. Although AI, the problem of AI, is much bigger than COVID, right? I mean both the upsides of AI and the downsides of AI are much bigger.
And of course, governments won't keep up. And regulation — not just on the part of national governments, but also regulation by various organizations, including commercial ones — won't keep up in working together with AI. And more and more companies will disappear, more and more companies will transform, and people will be in a mode of constant new news, of falling short of everything going on around them; they'll be in a kind of trend. They'll be able to update a great many things. By the way, I got a question. I said that the market will be out of reach when it comes to AI.
So then why should a person get into artificial intelligence now? Look, right now in the world, for example, billions of pieces of software have been made — all sorts of websites, software, mobile apps. And today you can create all these different systems, create new processes, new documents, analytical solutions, some out-of-this-world ones, where, where years of people's work are needed. By the way, there was also this trend: half a year ago people said that you can't build a high-load system with AI. That's nonsense.
Now you can already build very serious high-load platforms, websites and so on, right. I think that, in fact, without modern AI, maintaining them becomes much more expensive and complicated, and becomes practically impossible. And it's no accident that systems like Anthropic's Mythos began finding, at the start of summer, even at the end of spring, began finding a huge number of problems on tens of thousands of different websites and platforms, even in the most serious systems.
So what are programmers saying about them, then, if these same, these same systems get hacked? And right now you can do something. There will come a moment when you won't be able to do it, just as three or four years, years ago it became impossible to sit down at home by yourself and design a cool system. I mean, if twenty years ago a person could sit down and program some cool system on their own, then five, seven, three years ago it became practically impossible. It became very hard.
It's like even launching ads, even that has become very hard. Fifteen years ago you could launch contextual advertising, and it would take off very quickly, more easily. Or you could launch tech businesses in certain kinds of countries, and they had a fairly easy time of it in terms of competition. Now the competition, now the competition in these areas is insane. That doesn't mean it can't be done. Obviously, someone builds a platform, someone builds a website, someone launches a business.
Of course, they work. But the competition has become completely different, insane. Now, on the one hand, competition has grown; on the other hand, very serious opportunities have opened up for people, because the bulk of people don't understand what's happening and won't understand. And someone put it that eighty percent of people won't understand me. I think ninety-nine percent of people in this field, nine hundred ninety-nine percent, nine hundred ninety-nine out of a thousand, yes, ninety-nine point nine percent won't understand what's happening.
And that's why now is a great time, and we're moving on to the next trend, which is probably combined with the trend that the US and China will win the AI race. Namely, that whoever has chips, data centers and energy wins. And, as we know, it isn't people who have chips, data centers and energy. And it's certain groups of political structures, and it's not even individual people inside some corporations, because there the directors get replaced, removed, reshuffled, shares get resold here and there, but these aren't ordinary people.
And so this point complements the previous, the previous point very well. What I said — that right now you have time to break into something very quickly and in a big way, because right now all these systems are running a great many tests. And now for not much money, for a hundred dollars, two hundred dollars, now there's already ChatGPT for five hundred dollars, a thousand dollars a month here, ten thousand dollars a month there, you can get very serious systems up and running.
We have, you and I have access to this. Whoever says a thousand dollars a month is out of their reach — it's out of your reach. That's a fact. But I want to say that those who could afford to spend a thousand dollars a month on things like this three years ago would never in their lives have gotten any of this done with thousands of dollars. For a thousand dollars, or for two thousand dollars, they would have hired themselves an entry-level marketer, or some salesperson, or an ordinary programmer. That's what you need to remember.
And the last point for today, which is phenomenal and important in the life of every person and every company, by the way, and for your profession, for business, and for the economy, for everyone. That you need to count the real result, money and time, and measure effectiveness over the long term, not pretty demos, not some short-term conclusions that get announced, not whatever short-term profit there is, but count long-term investments. People in the world rarely know how to count over the long term, and where people do know how to count long-term, it's often a very limited life, and in some ways not a bad one.
I've often given the example that, yes, it wasn't bad in the Soviet Union, when a person knew that if he went to study at this particular university — he was barely even born and already knew what he would become. Today, in a huge number of countries, people know what they'll become. They can even have their husband or wife chosen for them. And they went to study at a particular university and knew that after university they would work at this particular job. But if you broaden your thinking, look a little wider, then now, over the long term, you can surge ahead quite seriously in terms of social, material effectiveness, or a certain freedom of yours, or your worldview. Do it over the long term.
Don't think in that construct from thirty years ago, when my mom told me she wasn't going to learn computers. It probably would have paid off for my mom, right, when we got a computer at home in '92 or '91, or whatever year it was, '93, I don't remember, nobody will ever know. Thirty, more than thirty years ago, it would have paid off for her too to start learning to understand that computer. Back then. Now there are people who for some reason write about the example I gave, that my, my nephew is in a very serious program at a polytechnic university in Poland.
It's a fairly strong world-class, world-class, world-class university, right. And there a lecturer teaches how to change the resolution in Windows 10. Someone wrote: "You don't understand, it develops the brain." Well, back in 1993, in '94, it really did develop the brain a lot, but not now. What do you think about these eleven trends or twelve of mine for today, right, how many did we end up with? Which one did you like the most, and which one did you like the most out of thirty? What do you think is the most important thing in this?
Don't forget to subscribe, comment, like our video, tell your friends, recommend it to others. I really liked this today. Someone I know sent me a message today and said: "I went to the university to give a lecture on technology. I printed out my list." And keep staying with us on the channel and supporting healthy, genuine human development. Bye, everyone.