Hello, everyone! We're on ToTheMoon. Technology news and sites from the Silicon Valley, the USA and the world. We're going out every Sunday. We're always four people.
We're at a stage to win. Well, he's objectively looking into the truth that America didn't win China.
You'll find out what you've said about GPT-5.
For America, competition in AI is a matter of economic and military superiority.
There's no idea of a genetic artificial intelligence, no LLM.
To get, I don't know, a Trump tie.
How's 100 million people going? How do you get a billion-dollar man?
I don't need a reasoning for my inquiries, do I? Yeah, now, it's a billion in 30. Yeah. That's a very precise word. Not alive. The American Government is more, more resolutely, more anarchical than modern entrepreneurs. By the way, it was usually the other way, wasn't it?
Today, I think it's such a very important, interesting issue. And he'll have different topics as we meet today, which will give, if everyone together see, a fairly serious picture of what is happening now in terms of artificial art. Intellect and technology services and technological and technological development. Very relevant, actually, for a man, right? What, you can use it for yourself? And before we go to the baths, to the Grok Ilona Mask, the OpenAI Batalias, there, together with them, to JD Vance's speech in Europe in the forum, I think it's in the European forum that matters.
I want to tell you a quick story from the point of view of the world of technology we live in. I often say that we have a whole automation on the planet, it's so early when we look at some super-sized, developed artificial intelligence. As we all know, Elon Musk is now in the state administration and public administration database in the USA. And then he recently put it out, not recently, but here, yesterday, he posted the post where he published the following. He has, therefore, downloaded information from the American Social Insurance database on all other living pensioners and uploaded the number of people in the range of age, how many people are. For example, between 50 and 50 years, there are 40, five million people living there, between 60 and 60 years, there are 40, six million people. There are very interesting numbers. So there's a very interesting figure. I'll sound them. Well, first of all, there's one person who lives in the range of 300 years. There are thousands of people who live in a range of two hundred and twenty-nine years. There are a hundred and twenty thousand people aged 60 plus. What is he doing? And so on. Yeah, there are three million people over the age of 40 plus years. What he said, aside from the joke, he assumed that some part of the U.S. population was vampires. Here. The only truth that explains it. It's an amazing story, actually, that shows the general insufficiency of many systems. And, in fact, we talk about hallucinations of different models. But the question, including what data are available everywhere, how does the information be maintained, how is it correct? Of course he does. Well, they're naturally turning some false information out now or then checking it out, turns out it's not the same, but it's a little bit of a lack of credibility, a lack of data. Yeah? And I want to ask you, and these people were being spent or were not being spent, right? Or how did you expect it? Because there are a lot of them. I mean, there's a million, 40,500 people in this range, there's a hundred and fifty years, so that's why. Here. Elnar, you wanted to comment?
Yeah. Data Science has a principle of garbage in garbage out. Yeah, I mean, if the trash is in the entrance, it's trash out. And most of the work, the date scientist, among other things, is just data cleansing. And what you're driving now is the emissions you need to work with. We need to figure out where they came from. It's, uh, the bug in the system, it's not recorded that someone died there, and that's where they got out of the system. Either something happened when you started it. I mean, it's, uh, this big routine job is the date scientist. But you're right to notice that it's interesting that, well, like this database, she lives with that kind of data. And it seems that this data has not been particularly addressed before. If there are people there who live three hundred and sixty plus years, almost from the base of America.
Yeah, I thought they were getting their money first, yeah. And it's interesting. Well, let's move on. Good comment, you said about Data Science, Data Science. Oh, by the way, I'm telling you now that there are millions in the US who are over 100 years old. What have you encountered in terms of uncorrect data? Not just technological, not just artificial intelligence, but also technological systems. Please write examples. It's always a very cool story to watch and read. And support the channel, of course, support the channel, lykait, write comments. It helps us move so that more people can watch us. Don't forget to do that, please. So before we go to Grok 3 and to the new model of even updating the models that happen, we'd like to touch on the subject of JD Vance in Europe.
I think that for the world, it is of great importance, for all, for all people. And, in fact, if we look at the agenda of last year, our episodes say that we live in peace and era when everything changes forever. So one person can completely change everything that has happened before or change a certain agenda. So he was in Europe, his main-- he was at the Arts Summit, and the first you-- the first one he started, he said, "You usually have an agenda dedicated to the safety of an artificial man." Intellect. I mean, we should not be concerned about security, but about prospects. And if the world is concerned about the problems that have created artificial intelligence, Europe creates this huge amount of obstacles and suggests that information is being compiled, it says that people will lose their jobs, then we will. "Really, artificial intelligence will help to develop humanity, help to increase people ' s productivity and increase, short, the number of places, not reduce their number of jobs." And he says, "I think it's our entire agenda. And most importantly, Europe, you are parasites from the point of view... And that was, by the way, very interesting to hear. In terms of limitations on artificial intelligence, that you all block completely that your GDPR is European, European laws to restrict work with personal data, it's just non-working. Absolutely. And if you block not just the major American corporations, but you block even a small business in America, you'll have problems. America will dominate this system. We'll divide, there, a trillion dollars in the next five years, there, or four years to develop artificial intelligence, and we'll be the coolest, all willing to work and partner, especially with those who are ready to work. working with us. But no restrictions will be imposed on anyone in the world. I mean, the basic message, I think, was that the artificial intelligence needs to be given a green light. And we had a lot of episodes last year where we discussed the laws where people had a lot of discussions, yeah, where to limit where to resolve. And it can be seen how many companies have lifted restrictions in recent weeks. In particular, Google has rewritten its agenda and said that their artificial intelligence has now disappeared the paragraph that it will not be used in the subject of weapons, yes, and in the subject of war, there, military, military, military, in all of them. military. They've already taken this agenda off before he speaks. He certainly says that America, American companies, corporations, small businesses, medium business for the next four years will unbelievably develop their hands in terms of development. technology and there will be no specific restrictions. We'll look at it. What do you think about this part? I think for the American market, it's a kind of technologically positive, right? If globally, of course, to see what happens to humanity, it's a friend-- it's another question. But we have a technology issue and a quality of these technologies.
I think that--
Yeah, well, I think it's super, Sasha. I mean, yes, I mean, it's super, which actually says, like, that we're not in a state of regulation, but we're in a phase of winning. Well, we don't mean, well, he's objectively looking into the truth that America didn't win China. And he's talking straight about, well, Europe's really behind us, well, right behind us. And it's about how America is concerned about how America beats-- wins. Well, because everyone knows that, for example, in the drones, America's gone unreliably, yes. Oh, well, I mean, by the way, military advantage, but many people think I'm not an expert, many people think that, for example, military is exactly what it would be like to be so physically wary, if, There will be a clash, China will win because America can't produce drones, technology is no for drones, short, nothing for drones, right? And so at least in the world, like a software like cybersecurity, cyberdominating something, right? I mean, not in the physical world, but as virtual, rude. Of course America needs to be, given America.
And I-- and what I heard from J.D. Vance, that he said that if you wanted, you could partner with us. And it seems to me very, uh, correctly that, um, it is very clear that there are very clear priorities that, for America, competition in AI is a question of what it is, economic and military superiority. There will be no endless summit to jointly develop, and, security policies, an implementation policy. I think it's very important in the context of the fact that, at this point, the best open source of the model is Chinese. In all domains, now the best open source of the model is Chinese, right? Maybe, except for pictures, which, like, a sca-- it's a argument that, like, a Flux model and a picture-making model, it's open source, and it's, uh, American, but she still doesn't have that super-star. adoption. And if we take more, there, useful housers, like videos, yeah, we'll get audio, we'll go into the voiceover, that's, uh, the dypsies, right? We'll take a video understanding, we'll take a reasoning, we'll take the generosity of the text, the ordinary LLM. So, Chinese people outnumber them everywhere, huh? So I think that-- and understand that, well, it's easier for business to always introduce an open source model. Oh, and open-- because in open source, open source model, they live like a perimeter, yeah, they're more controlled. The safety of the enterprise itself can be controlled, yes. And so far, it is that they win, win open source, long term in enterprise, Chinese models, respectively. And by close source, we don't have objective benchmarks again, because we don't baccalaureate in Chinese. Again, I just want to remind you that the baccalaureates we speak publicly, where OpenAI dominates, are still Baccalaureate in English, which is written in America. I-- I mean, it's just, you know, it's not like that. Maybe there are other baccalaureates in China, I never asked. I just wanted to emphasize here that in the open source China is full dominance, and that's right, America's paying great attention. Again, a lot of mems online about what, uh, you know, they're asking, like, the blue sky? The sky is blue. The tree is green? The tree is green. Taiwan is an independent country? Taiwan is part of China. Well, this is, like, a psychic dialogue, right? Which, as open source of the model, still has some kind of pinetuning processes, yes, RLHF and more advanced equipment that lay down certain behavioural patterns. And anyway, this is being introduced by enterprise, and it's, well, eventually, these Chinese open source model, they'll be spinning in Fortune 500 companies.
If you don't show up on the map of China in Taiwan's regions, you've got huge fines, you know, or some kind of punishment for that, right? That's for the word. In general, you said a good thing. What JD Vance said to "partner" and he said he was giving-- what would give $800 billion or $3illion. That suggests that, of course, America invites the world to participate in technological development and expects a large number of sophths to be produced in their own world. It is important, however, that he will not be restricted because what they will do to Europe, I do not know, because it will hold their GDPR and hold their rules and their laws very hard. I think he did it, he did it. Europe's probably not gonna turn around on this. At least she didn't turn around. They had no pre-conditions in the past decades. They only drowned in personal data restrictions. Here. But it's really cool for the U.S. If we're talking about laws, what JD Vance is talking about freedom of artificial intelligence. You as audiences, what are the limitations and problems that can be encountered in view of if they take down this policy of artificial intelligence security in the next years? Or you don't see it at all and count it... and you agree with the position of the U.S. government that you have to give it all green light and completely let it go, and then you can figure it out after the time. This will create such a huge technological, strong movement, and at the end it will be normal to regulate it, knowing that it is regulated. Given that there are many legislators who are not clear at all. Please write in the comments now under video.
It's, uh, I think it's a very important thing to know.
There is still some news that the bill has been put forward for consideration, I think, to prohibit the use of deep seek and the fines imposed on the U.S. territory. I mean, on the one hand, like we're gonna take all the restrictions away, but it's for our companies, and for the outside companies, we're just--
Well, this bill was thrown out, thrown out some Republicans who throw out, threw out a senator from New York yesterday, that the birth of Trampa should be made a public holiday for the United States and so on. I mean, I think it's--
It's important that the capital not be renamed.
Yeah. I think it's just bills that, you know, like structures. But what deep seek, by the way, banned on my own, went through, and went through, the details that DeepSeek would be banned in the U.S. State. It's a fact. But there's probably nothing wrong with that, right? It's kind of...
Yeah, yeah, that makes sense.
Yeah, that probably doesn't hurt, because every country obviously has to use outside SaaS services that learn data and so on, it probably doesn't want it to be. Well, I guess that's-- that's okay.
Yeah, Sas, there's a little better to figure out what DeepSeek meant. Because if they're not allowed to use this site, which is DeepSeek, that is, sending data, to China, on a par with the answer--
You're right. Or they've banned models, yes.
Yeah. And there are, for example, the different services are big, like AI together, there's at least some other stuff that's on its servers that's gonna turn off--
We're done.
Models. Yeah. And this company can be in the US, servers can be in the US. The same model is just, well, there in the U.S. perimeter.
It's a matter of limitations, it's a fact. I'd like to know if we'd moved to some limitations, but today we're gonna have enough information. There's a Toms-- Thomson Reuters. She won the first major copyright claim for AI training. What does that mean? That the court found that Ross Intelligence illegally used a secure content to train models. It appears that Thomson Reuters is a very strong provider of legal financial information, which is based on all courts and corporations. And now AI companies using this without a license are not entitled to use it. I mean, remember, we always said that now, in fact, ChatGPT or any AI assistant, he's not worse than a very cool lawyer, and he can make quite serious contracts with him. I myself have a lot of paperwork or documents to be made by external models. I mean, it's really cool. The question is how they'll prove who they have information and who doesn't, right? Because in some of the episodes, I once said that these systems were, I also gave an example of the Russian-speaking system, Consultant Plus, they're supposed to die, right? Because there are some assistants who keep all this Clondike information. And here's the case, it's quite indicative that they said, "No, you've learned our data, so you can't use it." How this is how copyright is distributed and who owns it, we'll see. The most important thing is that these systems are being built. We, I think, as usual users, are the most important thing that systems come in. And those people who look at us, if you do some kind of start-ups, write, by the way, in the commentaries, what kind of systems you work on, projects, what you think you can earn money from the start-ups perspective. Or you do something from the point of view of the start-ups. It takes into account certain aspects, yes, how some law or system can stop your start-up or, on the contrary, give the whole market an enormous amount of market. I mean, there are some laws or some consequences that give. If there were restrictions in the US last year, Gidi Van is now a scasal, there's a lot of unlocking to create different systems.
Now, that green light is open in many things. That's a very interesting subject. So? Grok 3, right? Grok 3. I am impressed, of course, with the quality of the pre-presentation that OpenAI, Grok. I think I'm just saying that, well, Americans are making presentations like that, doing it. Well, it's Apple that comes back. And you think they have a good presentation. And of course, Ilona Mask's cut face at the presentation. They're sitting there, so they're sitting next to this company, and they're periodically switching on the screen. It looks really unprofessional. Here. Well, it is. What do you think, huh? I'll give you some opinion. I'm just wondering, I think it's been a week. Now, now, you're gonna find out what you've been saying about GPT-5 or some stuff, yeah, there's something they've suspended from doing and why
They stopped doing three and why? Oh, what do you think--
Yeah. I'll probably tell you right here, yeah. I mean, I-I'm really having a hard time knowing what kind of skies. M, that is, models have become very similar to each other in quality. And I don't have a fraimvor to make decisions on the use of a model. Sorry, I'm talking about myself. I just had this idea quite recently. I'll try to explain. Oh, there are news skies. So there's a lot of tweet noise going on, right? And the current autlets, there, New York Times, other things, they, well, everyone has a very strong bike, right? And of course, we need an independent model that, as a community note in Twitter, is, like, working, that is, that actually allows us to understand what's going on. Like, Sas, you were saying that Dodge found out there are people who live for over three hundred years, right? And I want to, again, so I can, for example, explain this to me, well, it's like a five-year-old, right? I mean, what data are the findings and so on. Because Twitter has a big argument about this and there's no time to figure it out. So, of course, social security is a very sensitive subject for America. Of course, that is an important topic for discussion. But I'll see who's right, who's responsible for numbers is not yet possible. I'd like to use LLM for that. No, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no,
Yeah.
Then...
Oh, there's a question. Well, we've been comparing different ones there, schools, colleges, and, uh, family. I mean, well, there's a detailed level of analysis, and unfortunately, they don't give them systems either. And then there's, uh, some, uh, travel planning skies with the purchases. It's not like models can't do it, can they? I mean, they still have any comparative table of products specific, manually made on Amazon, which will be better than, there, analysis. It'll be more actionable than an analysis. So here, it's actually a little hard to get a different kind of signature. So X also shows different tyres at the OpenAI. And OpenAI says they'll make GPT 5, which will be all connected. But it's hard to understand. So far, yes, I think it's very difficult for a concimer to understand, actually, what it's all about is what these models are. I think it's very hard to understand.
Well, they're all doing peddata now, right? And everybody says they're getting a little pedgy. Well, we see four-Os always pinging. Elon Mask said yesterday that every day would be Grok 3, and there would be a change in it all the time. From the end of the line, I think all users, they don't even understand the difference. That's a lot of talk. I even see her wife's example, she's a big user of ChatGPT, but she doesn't know exactly what the buttons are. Now, this Deep Research, there, or o1, Pro, oh1, Mini, considering another version of her $200, she doesn't even know what it is for the volume that it's inside for the size of the model, yes, different. And it's very difficult to choose between them what better use is what better answers to a request. Well, I guess it's almost impossible. On the other hand, the quality of these things. What did Grok show you yesterday, I think? He showed a copy, a copy, you know, like, uh, collected, uh, from all, all the systems, all the chips. So when they're looking at it, they're showing up this reasoning, and the reasoning is detailed. DeepSeek showed it, in fact, first, and they wrote it down very detail yesterday, how they were doing this research. And it looked like fun, you could come in and study how they were, so they'd all analyzed it. At the same time, like we've seen it in other systems. What's the difference? They show a comparison table that Grok 3 is the coolest of all the others, and there's a kind of about one Pro.
Yeah, well, there's some systems cooler. On what? There are systems, and they're getting very similar. And this is the statement you're saying now that OpenAI says that we're not gonna let out at all, that we, uh, concentrate on possibly release GPT 5 and GPT 5 will consolidate. All of them. So for ordinary users what does that mean? Which, in the idea, would be a new instrument. Because, look, we all work in the old chat room, right? The chat that was invented there a few years ago, and he's alone, on par with the same one. That there's something that's gonna happen to determine how to communicate with a man. I think it should, and he must determine that where a person can send a voice mail where a person can send a picture where a man can find a deep search where a person can analyze where a person can. I'll ask you questions. Well, that's some kind of working tool. I think it's the right thing to follow, and it's probably all gonna come to that. Here. And I want to remember Lekuna, who said that in five years, we'd have other names.
There's no idea of a genetic artificial intelligence, there's no LLM, and there's something else going on. Illnar, what do you feel about Grok 3?
It's exactly the same. Just another model. There's no breakthrough yet. That's what you've said, and that all of Sasha's said last time, they've got low-weight fruit. Everything that could be done without, in there, super expensive, without any breakthrough in the engineering component, they've introduced. That was funny. Well, that's on the one hand, right on the other hand, discredited DeepSeek. It was funny that he was bored, that he was spent there so many times more computers on, you know, computers to teach Grok 3 compared to Grok 2. And DeepSeek, on the contrary, recently destroyed the market with his statements that they spent less than a computer to learn their model. I mean, this is some kind of funny dissonance. Well, well, just another model. I wonder if this model also raised the cost of the subscription award in X, yes, there was $22 in my opinion, it was $50,000. So they've already raised the price so much. Not two hundred more, of course, like OpenAI, but also the product, to be honest, a little bit of cheese. Well, I wonder if they're Grok two want to put them in a open source soon. But I don't know, it feels like DeepSeek Grok is alive, and it's not clear what license he's gonna be, and DeepSeek MIT-check is a license, that is, a practically full freedom to use. I don't understand why. And the statement is so slightly derogated, in my view. But it's great that there's another player here for--
Yeah.
I mean, you know, you know, you're gonna be out, but there's a lot of stuff going on here, right?
Yeah, it's a little silly, yeah, it's a little silly. I mean, I don't think anyone can explain why? Well, I mean, I'm kind of surprised that the modern-- well, the current American government, it's as much as less, it's more direct and less bizarre than, unfortunately, Ilone Mac and Sam Althman. So neither Elon Musk nor Sam Altman can tell you what they're doing and why. Well, just, I mean, basically, why would Grok have two, yeah, open source? So who's gonna make it easier? I mean, just to say that getting, uh, I don't know, a Trump tie that they've released something in an open source, competing with Chinese. I don't understand that, to be honest.
I think that's very little, very little point.
Well, that's very good, very much like the quality of the presentation that's being done. I mean, in fact, they could have imagined that even in terms of the growth of all shares or their portfolio or their own name. The general introduction was, well, some points, very clear and we are developing what we have to do. Here, Sasha, what you do all the time is a very good thing. I can see here, okay, they let out, the gicams look nice. Someone who hates OpenAI or DeepSeek or something, and amateurs, so X said, "Good, so I'm gonna use their model." They said, "We've got a lot of money." What does it matter to $50,000 and $200 for the expensive version? Well, people who want to pay money for it don't even care. Well, they paid and paid. It's not clear what quality inside the model is. It's like DeepSeek to use a lot of things free. And what's next? If they showed, that's because we made Grok three, we'll be able to release this in six months. Or because we've done Grok three now, we'll be able to reach that final result in a year. Well, just show them they can collect tetrics, it's just a profanation. It's like the OpenAI show that we have in Insta Cart a store, a store to buy food like sushi cooking. So, you guys, you guys are a case? You're really doing this case, aren't you? You really attract these people. And this tether is the subject, she's very... They showed up in the example, I don't know which one of you was watching. Elnar, you saw, right? Well, it's with some kind of thing, yeah! What are you doing? Well, that's all written down. There are little improvements in terms of, for example, programming, the programming arrangements that these directly deal with. From the point of view of the simple system of the dialogue window you're showing. I mean, a full duplicate of all other systems that exist on the market. Okay, DeepSeek released the same buttons.
- You let the same buttons out there, and the same-- you're supposed to have something else to show or show some kind of integration in terms of X systems, like the same, right? Or processing some information, maybe inside the X, I don't know how one of the most downloadable applications ever, and I guess it's a volume-leading one. There's a lot of discussion going on. In terms of showing this paradigm of credibility. That information may be checked, well, on some point, even from the point of view of the pyara, to give it up. Pierre is very strange. Here. I'm sorry, we can't see the market's reaction to this from the point of view of the X, can we? Because he's not trading, and you can look at the reaction, I don't know, Tesla and so on, but it kind of doesn't matter. Ia will see what's going on. But anyway, it's a pleasure that Grok has been updating. Again, the indicator is these buttons, the fact that they have searches, they have deep research, they have deep reasoning, and so on. It's still a figure, I think, that, uh, that's what's gonna happen, that's all the big players that's gonna get there, right?
All the big players will get to it. That's very good.
Sasha, low-weight fruit. Perplexity deep research has already been released.
Yeah.
I mean, it's not something super innovative, engineeringly complicated. I mean, yeah, that's something interesting. But that's how it went for a week, a half from the anon. Perplexity says, "We already have and--
Well, because it's a model pump.
Dead and fast.
You're just gonna betting your necklace. It's just some kind of a little bit of extra work, it's not something dark or super smart. And this deep research, the only problem in deep research, the fact that I see OpenAI now, he's been lying even more. Well, of course. And when he comes up, he started to make some data, right? And I think one of the episodes that is missing in the solution is when Chat4o is now in dependence. Here's four o'c's been permanently jealous in the last four months. What does that mean, obsession? You asked me to answer something. Next, you try to get another answer, and he's not a dick, he's just giving you the last answer, and you have to make a new chat like that, right? I think that these things, the solution of these things, they would probably be fundamental in the sense of the skies. Sometimes they don't understand their strategy for mass use. So, Sasha and I are talking a lot, and how are we gonna put 100 million people in, how about a billion-dollar man? Well, I mean, speaking of implementation, like yesterday's Grok performance, it's not clear how they get it from those users. Just like the release of some product. Now, OpenAI has once again, for fuck's sake, released this product again. Fuck you know how his ordinary people will use and what they'll do to him and how they'll work with him. Especially by letting out model o3, in one and a half months, saying that we're here now, we're not gonna be out of the whole thing. Why do you torture people with all your models, man, and all your stories?
Yeah, and with this model, not with the model, with deep research, I'm like the user's feeling that OpenAI, well, is not doing my job. That's what Perplexity on top has brought some extra stuff on existing models, and now she's been doing this story, it's like it's okay. And from OpenAI and other companies, you're waiting for the cool fundamental models that are some technological breakthrough. They have a lot of talents inside, super engineers, science, everything. And they, uh, these, uh, simple things do that any company can do practically.
Well, now, you know, yeah, I mean, that's exactly what you have right now... Just if everyone says that, well, low hanging fruit is generally over, that is, actually, further, that is, that further improvements will be as good as reasoning, and that there is a man, well, that's more than that. Some interactive espionage, but models are no longer going to be more intact. So people are on this same-- a lot of people are on it.
Yeah, yeah, yeah. Well, on this basis, Skewed and raises huge budgets for research in this field to make some other principled approaches to understand how I understand that, at least.
Yeah, I get it, too. So it's a billion-dollar Green Oaks, which is exactly what it's been, in a conditional way, that all the markets have failed, some yuzkeis, for which it's been compromised, that there are 10 billion. I can make money with GenAI. Well, as it may be, you can make it, like, a hundred and more in the future, right? And I think that's what he says, yes, that's just the current models, everything stopped. We're working on the next generation of models. And assuming that they invest in it, there's a billion in evaluation, I don't know what they say, like thirty.
He seemed to have the first billion in five, if I remember correctly.
Yeah. Now, now, a billion is on the 30th grade. Yeah. Well, I mean, it's clear that if he takes over half the current GenAI market with a new model, well, it's gonna jest that assessment.
I think he has some kind of thing, that's his way of looking at things, that's his point.
Tanya, how did you see any changes in the past six months in terms of simple use, like ChatGPT or some basic-bearing systems? We're talking about something, maybe specific, there's technological expansions. How simple things do you see changes? Or, for example, from the last six months, it's just... when you learned more of the boxes to find interesting.
Well, you know, I'm actually listening to you more like a simple user, yeah, which, uh, I don't see a lot of change, I guess, like Pauline, your wife, there. First, I don't have a professional version, common ChatGPT. The only thing I think I don't know is, maybe I have a special subject, I use it more to-- work there's a pretty straightforward response, that's, like, some size, some kind of stuff. technology, see something like that. Oh, so the data are pretty certain, and how much it would be easy to find in Google, but ChatGPT is accelerating it. I mean, he's like a little helper, right? I mean, I'm asking him a question by the vote, but he's answering me, but just so that there's a real change that I'm listening to every single issue of ours. There's a model like that, a model that's coming out, they added this reasoning there, so I don't need a reasoning for my inquiries, do I? I kind of, uh, I know, in principle, where I come from, and I just need to remind myself of some data, yeah, there, to clarify something. Oh, I mean, as a regular user, I don't see any crazy progress.
Well, yes, there might be a more tangible version of the story, but I don't.
And you've been making use of some tools, like I remember when we were just starting to discuss AI at all, I was, you know, telling you that you could, for example, have some sort of three or three tools there. Models, like, build a code to automatically write, there, and so on. Is that what you started to use, make use of, or not?
Codes. Codes, you know I don't use them. I've been trying some codes on the car cad a couple times, but I realized it was all too much. I don't need to go that far. Everything I can do is enough. The concept could be used to produce pictures, but not in GPT chat rooms, of course, but they're all getting some more specialized, too, like a Pinterest picture. I don't know, some, I don't even know how to explain. Well, that's not enough. Conceptually, we can get some superficial ideas and develop them. But just so I can describe it, I want this, this, this, this interior, that color, that size. And I have three of my visualizations-- no, I can't see that yet. Even, I'm not a professional at this, but I know people who are artists, and they create some paintings, interers.
Still, I think it's still at a very basic level.
I wonder if this is always, I think, very indicative of the real introduction, the real introduction. You can tell a lot of things, some kind of stuff, right? And in fact, we're really making a real move on how we start to use. I mean, I thought it was fun to be a thing when I could use a camera visually through the camera like I could use my eyes. I never found such briefcases. So we were in the Black mirror watching, yeah, when they had something in their pupils, it's like a script working, that's through the pupils, when you interact or when you automatically have something in your headphone. It's either in the head, right? And from the point of view here through the phone, when I'm videotaped, he's starting to answer me, like, some kind of unlived scenarios, and you're still just going to pee or sound like you're saying, You're going to some extra design.
It's a very precise word "unliving." I mean, even these pictures that go there in others, they're not alive, so they're gonna have to be very refined. You can use as a source of ideas, but so, dosed.
Well, that's an interesting story. I think that artificial intel--
That's a good story. I think that the artificial intelligence that is being discussed is being discussed ahead of us, it is precisely the possibility of having some systems that are easy for people to do easily. I mean, even, for example, the creation of cover pictures in YouTube. There's no normal simple services, simple, I'm just saying simple things now, when anyone who generates cover, they or anyone who wants to make cover, they're starting to use these services. It's very easy, you know, accessible, how you use an iPhone camera. Very accessible, it must be easy. It's complicated now. Intuitive, yes. To be able to write very cool proms, to be able to enter certain systems, and to be, to be, to turn around. - And then the designer, yeah, being a designer, there, turning everything around. Well, that's a serious story, right? And I'm, like, one of the big tasks that's worth, and I'm on it, and I don't know how to get to her. It's a challenge where I have a huge amount of certain data to teach it to make this model perfect with these data through simple means. I think we can do that. Once you get involved, there's always some difficulty, restrictions, details. Anyway, it's complicated. You should be a date-sientist or an engineer very strong, straight-up scientist.
That's what they say, there, like, Harvard, was taught that a student. Well, well, that's another class of people, right? That's their job. They sat with these data, doing it completely. That's it. It's not what I came in, it's a volume of data and got some information. I mean, getting this information is so specific. I want to touch the news that is the last thing I think is incredibly important today in the final edition. We support this line, we have supported this entire twenty-fourth year. This is the history of the interaction in general of the presence of Elon Musk and, in particular, of OpenAI Sam Altman. And you and I have seen some conflict that has been happening, and we have been up again for the past two weeks. In particular, we know that the President, the new President-elect of the United States, convened the conference on the first day. This conference was attended by Sam Althman to open a new Stargate company, which, uh, will be Stargate right, which they're planning to allocate half a trillion dollars there. There was no Ilona Mask. And here, so Ilon Mack was a week ago or when he sent a proposal that he was willing to buy a cash, well, for a cash, it's from the calculation account, for a probable, for a list. It's their conditional cash, of course. I think he doesn't have that cash, right? For a hundred billion dollars. It's the same thing the media writes kash. It feels like you know that Ilon Mack has opened and received $100 billion. In the suitcase. Yes. Or he's got a hundred billion dollars on his account. Well, where is he, where are they?
Apple, Apple can still buy cash in theory because Apple has 100 or 20 or so, or there's $100,50 billion, but like Ilona Mask's structure doesn't have it. So some consortium is ready to buy $100 billion. And it's very interesting, I think, the precedent that is now on the market, because they've released it. It looked like I'd even read a lot of news, it looked like the OpenAI type couldn't give up. And if they refuse, they lose all their way, all their moves forward, and they're supposed to be out of business. They're more like, and there were even some interesting descriptions from analysts, which is what it's done to make someone else want to invest in the OpenAI to create this competition. Some kind of thing. It's a very strange offer, because OpenAI is already worth more than three hundred billion. And, well, as far as the descriptions are in circles. And of course, there was a further rumour and Sam Altman Elon Musk responded at the beginning. He wrote, "You know, we might not agree with your offer, but we're ready to buy Twitter there for $9 billion, or ten." He said Twitter, not X. And then Elon Musk commented, and, accordingly, Sam Althman called Elon Musk's undetected and unhappy. For what? What Elon Musk said was that he was a jerk on Sam Altman. There was another rumour. The board of directors again refused, and the board of directors of the OpenAI refused the deal. To be honest, it looks superrealistic, doesn't it? That people who operate such companies with such volumes and influence the global development of technology generally have access to human surveillance and a lot more than the rest of them are generally speaking. That's all the stitches there, like a jerk, idiot, there's something, something like that, and it doesn't look like it, and it doesn't even look like a piano. No, well, Sam Althman's answering, too. Doesn't even look like a piano? Well, I mean, there's a pyara that adds to each other, and then when... But he's losing it. Look, he let go, Mask, he let out, you're right. But first, he offered to buy a company that costs a lot more money. And I think it's unprofessional, it's unprofessional. I mean, it's just stupid, that's what they used to say, they were fun, and now it's like a losing game. You know, when you're from the fifteenth place, you're yelling again, you know, why, right? And then you release Grok 3, who, well, is somewhere a little going on and doing something. Well, you'd have released something revolutionary, said you had a huge amount of money and offered to buy it.
Why would you suggest buying OpenAI for 100 billions when they just talked to the President of the United States about half a trillion dollars they're only investing with SoftBank and Oracle, Yes, and Arm. I mean, it looks very specific. They're still building projects, by the way. Arm said together that they would make their chip in partnership, I think with someone, with Microsoft or with someone, right? What Arm is an English company, and they're gonna do, release their chip. Uh. Interesting, by the way, movement on the securities market, too. They don't move much of their stock. Ia will see what's going on. Because this game is in the chips now, of course, it's important. The Trump, by the way, has again said that we have to win the electricity race, that they are trying to unlock the total amount of resources, because the current electricity in the country is only required for AI. By the way, I think the U.S. President's team is on-going, and it's the, uh, one that's on the artificial intelligence. I'm not talking about, uh, JD Vance, Vans, because he's like the right arm of the Trumpa, he's actually a... very strong IT and AI agenda. But Trump is very serious about it, always talking about it. This agenda has even now emerged in the discussions between the delegation and Israel, and with the delegation of China, and with the delegation of Russia. So it sounds like it all the time, and with Europe, it always sounds like an artificial intelligenceual agenda. Still, the U.S. team is looking really hard on the target. Anybody, Sasha, what do you think? The supplement is the squeeze of Ilona Mask and Sam Althman. Or is it just a standard topic?
I think it's like an offtopics, right? I mean, you can fight as much as you want, but it doesn't matter, I guess. I mean, at the end, F-- I don't think it's important. So they can create this drama. And Perplexity can create a drama, but I can't say that I can't say that it's productive and that solves some problems. Again, yes, I say that, unfortunately, now that the American government is stronger, it is better to anarchulate than modern entrepreneurs. Unfortunately, we're in this situation. Although it was usually the opposite. Usually, the government needs a lot of michuras to do some kind of thing, and now the other way around. I mean, yes, but maybe they might be, but maybe they're just wondering what they're gonna be, uh, that Trump, JD Vance can speed up Vance and OpenAI.
Yeah, it's kind of cool, especially if you're in the U.S. And in general, I think it's great for the world, because, as in DeepSeek, still, uh, more and more possibilities are emerging. So the world is crambling, so much of it is stuck together. Well, that's great, that's great! We meet up on the To Moon Channel in exactly a week. Technology news and sites from the Silicon Valley, AI around the world. We're always four people. We'll wait till we meet again. Bye.