There's a world of artificial intelligence, uh, that's a deception of ours. We've got fifty companies, a variety of beautiful logo, new products, new founders of start-ups, unicorns. But if you look at real money, the picture is very hard, yeah. Almost eighty per cent of all capital in the new Forbes list. Artificial intelligence, there's fifty companies. And, uh, eighty percent of OpenAI Anthropic. And the question is, if the models took the money, what happens to work, business processes, professionalism? Because on this list... It's not a chat-bot list, either, yeah. There are lawyers, doctors, programmers, bankers. It's not a Hype list. It's a map of where artificial intelligence is starting to get between a man and his work. And today I'll show you what companies are really important, who can win, and, for whom it's gonna be a big problem, yes, and who's gonna be, uh, part of the code or part of some system, or just a button in that system.
ToTheMoon Canal. Watch. What's interesting? That there is a new Forbes list for the 50 most promising private companies in the world of artificial intelligence. And it's important that it's not a list of public, but, uh, shares or companies that are trading somewhere. And it's not a classic rating that does, uh, Forbes. Rather, it is a cut in which private companies in the artificial intelligence of Forbes consider key to the next phase of the market, yes. And the official page forbes describes it as a list of the most promising businesses, and, uh, artificial intelligence. And the project itself, this is AI50, is being done in partnership with Forbes, together with the very important investment funds, like Sequoia and Merit Tech Capital.
Oh, and of course, it's worth, uh, considering and studying very seriously and looking at what there are, uh, areas there. There's no point in reading the list completely. I was studying this list and I wanted to share it with you, it doesn't make sense, okay, read alphabetical. It's important to see what there is, uh, those blocks, yes, which, for example, are models of research labs. Well, logically, yes, we'll remember that there are companies like OpenAI Anthropic, uh, Mistral, and so on, yes. But there are separate blocks, such as code and sophth, medicine, science and law, there, creativity, copper, operating agents, infrastructure, data, security, there, physical peace, etc. There's a search, knowledge, education, working tools.
I wonder what education has always been put somewhere, yes, like, together with the search and with the working tools. So, formally again on the list, uh, fifty companies, but the money picture looks like a market for two big giants and forty-eight catching companies. It's very important to understand everything. We're talking about this a lot in the ToTheMoon episodes, we're talking about this. There's someone who thinks that, uh, that doesn't mean that OpenAI Anthropic is the same company as everybody. Uh, I think these are companies that are very different from the whole market and, uh, leaders who are around, others, like Gemini from Google or, uh, xAI, which is inside SpaceX, They're just, like, in the middle of a two-way drive, yeah. But it's important to remember that these are companies that have already attracted $300 billion. And that sum, it must all be fundamental, because still a large number of people think that artificial intelligence is a question, uh, some ideas or brain, or whatever it is, it's absolutely absolutely. A copycat, just like China, for example, can completely copy everything. By the way, it is important to understand that the OpenAI was the first to be the total leader, and in particular companies that attracted money. So if you see how attracted, OpenAI attracted $180,2 billion and Anthropic attracted $66 billion, yes. It's a $27 billion together, and, uh, eighty percent of this list, which was three hundred and five billion. So that doesn't mean that the rest of the companies are weak. Uh, that means these companies, which, I guess, they're a very capital-intensive base model. Oh, we all know that this is, uh, energy, infrastructure, and, uh, defence industry, all this is at the same time, yes, that is, crazy, huge infrastructure.
And when you see, uh, not just fifty companies, even on this Forbes list, and when you see tens of thousands of companies in the world that create artificial intelligence, you'll notice that these fifty are also Separate companies, yes, the market money shows very much where the center of gravity is, yes. And it's clear that the next war, it's definitely not for a beautiful chat, there, or some microfunction, is more about who's gonna take the whole whole job. Forbes is a man like Rush Srivasta, and he makes the list meaning that the race is not just about the most powerful model, but about who turns artificial intelligence into products and sustainable businesses, yes. My personal opinion, uh, I'll continue to report that there's no question now who A is turning into some basic first product. Now, the question is, who can create an artificial intelligence that will unbelievably jump, ahead of all the others, and in fact, can't repeat it? Or who can create an artificial intelligence that starts, uh, to fertilize and to create self-sustaining businesses, right? Not even, but, uh, the company itself as OpenAI or Anthropic will be sustainable, but because I'm not sure they need to be super sustainable. They are unlikely to be lost in coming years or decades. The question is, is it, or is these companies able to create a real sustainable business? That's interesting, of course. So it seemed to have been the first time that companies like Anthropic and OpenAI, that they were reinforcing, and, uh, creating models. And other companies that are around, they build apps, yes, and they build applications that use, like, Claude Code or Codex, for example. Now Forbes has clearly indicated in his report that AI Fifty is in this report that AI Fifty, that Aicic and OpenAI, they are incredibly climbing, putting them on the programming market, yes.
And where, for example, companies like Cursor should prove their autonomy. And I mean, it seemed that the programming market, in fact, or even the software development, was supposed to grow. But there's a feeling that as programmers disappear. But the fact that you get very cool tools to develop, you're starting to get the most. But if we look at how much the market believes in companies that create software today, we'll see what's going on. I was just about a few days ago opening Epam's accounts, and I had a tyket in my tyket. It is a company from Belarus that is one, well, that is quite serious, known for the design of the software for the order. And their assessment just dropped rapidly. I think in the last, there's a pick they had five years there during the rug back, he's now reached that minimum. They're already standing there, I think there's less than $4 billion. They're still considered a fairly stable business. Although is a big question than stability if such swings go there in terms of assessing them. So, of course, the market stops believing in such companies. So the market stops believing in someone who has programmers who can develop something. I mean, it's important that when systems like Codex or Claude Code have emerged, there's, like, some things like Cursor, it still seemed that people who were professional in the programming field should be First to win, right? And we see, even from the point of view of Epam, the market doesn't really believe in them. I'm saying that I believe more in infrastructure companies like Nebius, but I'm not talking about OpenAI and Anthropic than in the company, like Booking.com. I mean, here, of course, those people who are making soup, they're going to a far-reaching plan at all, and the market is already showing a disastrous view. And basically, the fact that OpenAI and Anthropic take this pressure, even on companies like Codex, even companies like Cursor, in the idea where the whole code is being developed, yeah, that's that, that's that, that's that. History, she's the main conflict in all this Forbes list. Artificial intelligence Fifty.
If the laboratory itself takes up the business processes
takes the work process very hard, so strong, and then build it that is built, for example, in the courseors, and, if it is, have a very strong product, there, the user's habit should be very strong, There must be very deep data, industry depth and so on. And my favourite term, of course, is that a large number of sophths that are used in automation, in sales aid, in CRM, are those different agents that are now doing, Salesforce, including ERP and MRP. systems are different, they're in big trouble, just a lot of trouble. And I see everything as a market, it's not-- well, how many people don't understand it, because people, me, here, people in the ToTheMoon project and, in general, where I am, because of my own location, Some, there, connections to the Kremnival Valley, there, outside Russia, etc. People send a lot of projects. These projects, mostly we've made another agent who's helping something out, there, helping out on sale, something thinks they're watching, doing something. And you-- I don't even care if I'm honest with you, these systems are open. It's not interesting to open up because I don't believe people can do something cool, but it doesn't solve the fundamental, I think, challenge the future. I mean, like, I have a real estate business, and I don't do an O.R. there. But I was so impressed by the story that I raised some infrastructure and, uh, replaced the internal CRM systems, and there are two CRM systems in the internal are the professional system, there, Bonza and Arif. market. Put them in some of his data-collection infrastructure and linked them to each other with help, there, Codex and Claude, yes. And what I finally got and the reports I got, the data I got, uh, you know, you could start by thinking it's just analyst. I mean, I can easily see what kind of leather channels are going through, how fast they're processing, like, uh, sales managers reacting fast, as they call, the quality of talking about these lids, how much this is. The leather is qualified or not qualified, and so on. What was in fact impossible to do, by the way, manually. Why? Because you need to analyze all the calls between you, sync them up and so on. And, well, someone's gonna say, "We're doing it on the side," but still. And Imagine, these reports were received, and these reports started to go to different places. So they're in Telegram, they can easily get into WhatsApp, they're visible on the portal, so people have access to them from different systems. So, like, portal, and as such, there's Bonza or Arif or CRM systems don't need, right? But not enough, the person who sits in an analytical environment, for example, I don't know, is directly in Claude, Code or Codex, it's clear that he can look at any data at all. There's a question. The sales manager comes to work, and there's a lot of mass-markets, he's very relative. And there, well, you get some sort of leather on the day of the manager, and he's processing these leathers. And I understand that two hours of work in Claude, Code or Codex will allow the manager to be in any massager or his information system to come up with the necessary leaflets for his own information system. I'll call you, yeah. And he can through this system, there, leave comments or votes, or text. And the question is, why the CRM system? The CRM system, why? And you realize you're not tied to the system. You're not tied to Telegram, you're not tied to whatsApp. You can do it through some portal. You can probably send it through the text. The manager can talk fast, speak the text. Anyway, when he calls corporate phone through, uh, his phone, it automatically records everything, is integrated, and everything is automatically recognized. So why the CRM system? We'll keep talking about-- it's a little case, yeah. I'll probably take a further edition of this case where I'll tell you more about it. I think he's very specialized, he's not on the wide audience we have in ToTheMoon. At the same time, it shows very clearly the whole market, the medium to which it is in the market. It's clear that even a case like this is so that I'm not a voice anymore, it's clear that if you have a thousand sales managers, it's not just a finger-checked job and it's a big deal. But we're still in a place where ChatGPT 5.5 Pro works better than a month and a half ago. It's a lot better, yeah. Or where Claude Opus 4.8 works with his dynamic agents incredibly cool ways to cycling sessions, conditionalities, sessions or internal design agents between itself and to create a fairly good, complete one. systems. Maybe not super-crunch, but pretty good. I mean, by the way, that ChatGPT 5.5 Pro, I don't know who, but I have access to the U.S., I don't understand what his context window is. What I see is a context window that's been a long time ago, and, uh, a million tokens. And most importantly, it seemed that more than a million thokane context windows should be scattered, that it should not take a crazy amount of data. But I'm the one with the last one of the briefcases now, which is an analysis of two million records, notes, uh, work, uh, commercial structure of the corporation. I needed to be in my system of management of their records. I'll take a look at it. It's about eight hundred and nine hundred thousand conversations, living, real conversations, real transcribations. And ChatGPT 5.5 Pro was able to re-evaluate all this in a single session and continue to work in it. To be honest, it's really awkward, just a fundamentally hostile. What's going on? So what else is on this Forbes list? Artificial intelligence. Although I think I've been carrying the basic thoughts, but I want you to show me these other things. So the subject on the list is that of lawyers, there are doctors, client support, real estate.
That's what I said, financials. This whole story, the list shows that the best applications companies are building into a particular, expensive job. Yeah? And there are companies like Harvey, Legora for the right, right? Or Abridge, Open Evidence for Medicine, there's Poro for finance. They point out, by the way, there's a service or an artificial intelligence for real estate. I think, to be honest, that these services, they're carrying some abstraction and they're just chosen by the Forbes list on some strange principle. But sauce-- what does Forbes say? That the vertical market is this, these systems, they're easier to coinate. And there's a clear pain, there's some expensive professionals inside, there's a lot of documents and a Wednesday that exists. That is how, for example, it is seen in medicine in particular. There's a repetitive operation and a measurable result, right? And Forbes is doing this story. I'll remind you again that this issue is still for you, including a little bit of a show of Forbes, not only my opinion that Forbes is saying that it's not what the wrong person says, "We have artificial intelligence for us." "and whoever closes some expensive specific operation is better than a man or sharply speeds a man. So artificial-- right now. There's also a subject here, by the way, yeah, that artificial intelligence doesn't have to replace a lawyer, right? It's enough to take some of his work from the lawyer, yes, which the clients used to pay. That's a very good point, which exists in the world. And of course, it's from the perspective of the legal market. I mean, whoever said anything, it's clear that some kind of lawyer in the world will win. But the fact that the market will be reviewed is redistributed is a fact. Well, that is, the number of legally meaningful treaties I made through artificial intelligence or made a large number of my friends, or acquaintances, or colleagues, he's just a phenomenon. I will not trust a lawyer to date, I will trust the lawyer in the job, rather to comment on some of the paper, or I will need a lawyer where it is required from the point of view. - the law, right? Where it is very important for me to sign a document by a certain person. What do you think about this job? You'll have to say it. It's gonna be very interesting. It's important that the list contains Mistral. I'll remind you that Mistral is a French artificial intelligence operator and in the Ram-Europe is considering this Mistral on this list of fifty, right?
I guess if I were doing fifty-four artificial intelligence, I'd probably have Mistral.
I'd probably not have put Mistral. They put it in. And ho-- and I think I want to make this comment, yes, that Mistral AI is a big story not because it's a very cool model or it's a need to compare it together, Like, with Chinese systems, right? Well, it's clear she'll never override the OpenAI Anthropic on quality, even in terms of European law, she won't be able to do it. It's like a different story. What Europe says they have control, uh, their own, has their own value at all of this creature, has our weights, which we know, open. And there is a so-called European technological souvenir, yes, which is extremely important. I've seen, uh, saying one thing, uh, Russian one or two, whether a scientist, I don't want to lie. And he said there were three leading operators in artificial intelligence in the world. Oh, I get it. I read about the law on Russian artificial intelligence. And he said there were three leaders in the world. So it was, uh, the US, China and Russia. Yeah, I was laughing, I thought, "Well, if Russia is the leader, well, at least, uh, then I haven't forgotten about Europe in terms of their Mistral. Because Mistral, even as opposed to Russian models, is still well described and involved in a large number of different experiments. And I think she has much more-- more independence. Although again, you can express your views here and show us something else. I think all these systems are already uneven. So there's still some way to compare the serious operators from China. Well, it's clear that OpenAI and Anthropic have serious competitors, Gemini from Google and xAI from SpaceX. But all the rest is, of course, very relative.
So, uh, physically artificial intelligence, and there are a few companies on the list that work not with-- not just the chat room text, but with the physical world. Work with robots, autonomous transport, spatial understanding. And here again, I think there are some misunderstandings, because OpenAI and Anthropic, uh, are guaranteed to work with the physical world. And the leader, I think, who combines everything together, is Ilona Mask's company, right? It's SpaceX plus Tesla. So they're like no other job-- how no one else works with physically artificial intelligence and, uh, artificial intelligence that is supposed to be physical, right? And I will recall that a year ago my closest comrade, dating Elon Musk, he made a very clear conclusion after the meeting. I came, told him after talking to him, Auntie, that Ilon Mack doesn't even make a chat-bot stand, and that clearly and logically, yes. So to conclude that OpenAI and Anthropic are making a chat-bott stake, like or how Forbes describes it, I feel it's not correct. And so I love talking about when OpenAI and Anthropic learn to program for micro-controllers, then the world will change very much. There's a physical intelligence, Skilled AI, Oral Labs. We'll show these companies specifically to make many of the... maybe many of you will be interested in this Forbes list. There's Applied Intuition. It's just a bet, all about robots, cars, space, real world action. Well, the real world's action is clear that there's a lot of other operators working, I think, including robots. Meta works the same way, yeah. I mean, given that Meta literally, uh, today released a release, there, her latest new points, and it's possible-- it's clear that this is not a very interesting thing, there, on the market for artificial intelligence, or I'm in the process market. That, I think, is a pretty settled matter. He's everywhere. And Snap recently released glasses, there, 3D reality and everything that's related to it. They're actually-- they're actually stockings down, but they're still part of this market, yeah. And it's clear there's a lot of people who'll try to make these devices. I'll remind you that the OpenAI united last year with Joni Ive.
It's the most important designer of Apple, former and just like that, who developed the first electrical Ferrari from the point of view of style and design. And they're trying to build their device. And obviously, the OpenAI is very serious about this, and we'll see it at some point. It's just that the question is, if they approach this device as they approach their interface, there's nothing good there. But if they do it as fundamental as the ChatGPT came in, it could be very serious. I'm certainly waiting for a fundamental device on the market for artificial intelligence, which is totally unbelievable. There may be operators who will make very cool integration, and companies will be united, yes. OpenAI at some point or Anthropic will buy major robot production operators, electric vehicles, various spatial management. Because everything, well, I think, is conglomerate and X-- SpaceX plus Tesla, and they have a pretty cool, serious line. And there's no question of the linear, by the way, cars. I'm talking about the line of robots and robots, not those robots that just walk around, which Tesla shows, right? And robots, like some, like in a bunch of steams, robots, that's all Robots, right? From some micro-smart thing that can help in surgery to sew a man's hole to, uh, industrial robots. From robots that can even be a little like a man at home, although it doesn't make sense, does it? To the robot pilos. I mean, of course, this world will be unbelievablely developed. And the question of companies like OpenAI, they'll be buying companies like Roborock, that make robots or do something they want. We'll see that. If we're gonna see, Hyundai, for example, you-you--you--you--- bought all the Boston Dynamics. He bought them from SoftBank. It's quite recent information and very interesting, I think the bet, uh, is a big player in the world in the automobile, automobile industry, like Hyundai, a world-wide, very serious operator, A highly developed, technologically advanced machine-producing, electrical machines, generally capable of introducing very quickly different things and having such a serious company as Boston Dynamics, huh?
Boston Dynamics have different lines of robots. And would Hyundai be a company in this physically artificial intelligence? It's gonna be very interesting, yeah. There is a specific subject big, which concerns what is called a creative action-- creative artificial intelligence. What companies are there? It's clear that there's a Midjourney, all of them are absolutely famous, yes. By the way, we see how Midjourney is now in the construction of a completely different device, there, and a scanner, yes, in the offline. Companies like Runway. Everything about the video, for example, is it? Companies like ElevenLabs. That's all about audio and audio processing. Very, by the way, a big company. Again, if she can resist companies like Google and OpenAI, I have a big question. Although I recently in one of my infrastructure projects, in addition to the models of text recognition, text transcrimination, the re-cashing of the GPT read, um, GPT, yes, OpenAI GPT has connected the models. ElevenLabs. I had a comrade at dinner, and he said that ElevenLabs would recognize the Russian-- better Russian language than anyone. And it really did. That was very interesting. It's clear there's a Suno company, Synthesia. We all know HeyGen. A lot of talking about her. And while I think she's on some hype in the market, I remember when things could be generated and she was about to stay. See if she'll get an incredible development. They all have companies like CLIA, Black Forest Labs. Of course, they have a lot of money on the one hand. Not everyone has a lot of cash, but many of them have a lot of cash. Now, in the Silicon Valley and investors are looking very hard at the positive cash flow. I believe that the positive cash flow is the right measure of success of these companies. My personal history, I believe that these companies can properly integrate with giant operators like OpenAI, like Anthropic, like Mistral and so on. I mean, how much these companies can really be in parallel with them. Because when the OpenAI adds to yourself the incredible coolest opportunity to work with audio, you think at some point, why do you need ElevenLabs if you have a big infrastructure at OpenAI?
I mean, it's getting a big question.
I mean, it's getting a big question. That is, as long as the world allows for the use of Codex and Claude Code in parallel. And I think it's a underestimated story, but even ChatGPT 5.5-pro and, in fact, some Claude Anthropic, they recommend working parallel to Claude--Claude Code and Codex. They often make such a recommendation. Well, the frequent recommendation of ChatGPT that everything should do Codex, and Claude Code, it should audit, and Claude Code makes some other recommendation. In fact, I would like to express my concern at this recommendation, that this recommendation, which is based on low-cost models of Codex, low-cost ChatGPT models and low-cost Claude Code models, and Anthropic. Because when you're in a session like ChatGPT 5.5, you say I'm gonna use a expensive, uh, expensive Claude Code, like Opus 4.8 with a million toxins in ultracode or, uh, Codex 5.5 high, at that point. The system is starting to differentiate you to create a number and propose a stage of development, a number of sessions and a different recommendation. And that, I think, is an incredibly important omission, yes, which is on the market.
So the Forbes and PSAT list is providing an invisible infrastructure. That's all about dates, starts, chip safety, markings. It's obvious that this looks less interesting to you and us. Here. And there are companies all Databricks, Cruza, there, SambaNova, Together AI, et cetera. Yeah, I'm not gonna list these companies, but, uh, it's more like a story that everybody's looking at magic, okay, now in the interface, and a real artificial intelligence economy, it's in data, electricity, calculations, Introducing. And as I said, a year and a half ago, that's why, for example, I'm very investing. Everything about infrastructure and infrastructure, right? What you often say is invest in someone who needs to give a shovel. And when I got a lot of people here after SpaceX was there asking if I bought SpaceX, I said, "I didn't buy SpaceX, I'd buy someone SpaceX gave money. Well, at least that's my strategy. So Forbes first identified a list of early companies, a separate list of early companies.
It's not the winners of the current war-wave, but the next cycle. And the threshold list of early companies includes twenty companies at the so-called planting stage. Well, it's on the invest-- in the inquest of terminology, right? Well, or there's a first stage. Well, at least it's a start-up company. Together, they attracted, according to Forbes, more than three and a half billion dollars in the early stages. And it makes sense that there are companies that are engaged in accounting and various scientific experiments, and mathematics, and treatment, and finance, and even things like voice entry. Though it seems like, yes, that is, the OpenAI, and that is, the ElevenLabs. What a voice-injection, huh? Anyway, this whole thing is very alive, very alive. There are companies that are involved in various automation tasks, and this shows that the market is not over in the current 50 companies, and the next wave is being formed and will be formed. Although, of course, the investment market will be, it will be. Money is a lot, they have to be invested, they have to be spent. Big money is now coming with the exit and the installation of OpenAI Aic. And whatever happens, it's clear there's gonna be some streams, micro, macro. But the point is, the market will still develop, and they'll need to, they'll need contractors, yeah. They won't need contractors on one condition, on one condition, if they're going to grow incredibly much. And if we're looking at you, then OpenAI Anthropic, they don't fire people, and they fire companies with others who say that they're artificially inspired, embedded and fired. So, this is a very strange statement, right? Because OpenAI Anthropic, they are as good as anyone else in the arts and as no one else has developed artificial intelligence. And they don't have any dismissals, and we'll see how they grow up. And I think that a very serious effect will be when a lot of people are in these companies. There's 100,000 plus man. Could there be a situation where they won't have people, no people? Well, in theory, she's capable. But then, it really has to be this great artificial intelligence, super-court intelligence, high-level artificial intelligence. I mean, something fundamentally incredible, something fundamentally cool and very interesting. I'd be happy to see your opinion on these companies. I'd be happy to see those companies you're highlighting as the most promising and the ideas you really have. Send interesting ideas, not a little bit. I think it's very useful, including our entire audience. Suggest topics for these episodes. And don't forget to support our channel, to slide. Leave the comments. And we're waiting for you on the new editions. Bye, everybody.