Finally, something new came up.
AI servers are less accessible today than just accessible, massagers and some standard servers.
I remember their first Bart demonstration when Bart lied to his demonstrations right at the Bart demonstration.
It's just $6 billion-- uh, it's a very sick-large-- that's a very big hug for a company that's developing models.
So hallucinations are going well.
I was gonna say that there's a little other point of view, right? It can't be negative anymore, can it? Hello, everybody! We're on ToTheMoon. Let's go. Hey, guys.
Hi.
Hi.
I want to start.
I just drove, took the kids to school and I was back, and I had to send some information to my mom, a classmate--- a school mom and my son. I wanted to recheck the English language through ChatGPT. And then ChatGPT hangs. Here, uh, the news is, at this morning, ChatGPT hangs, Claude hangs. And I felt, to be honest, a little helplessness, because I'm used to being easy, um, voice, so I'm gonna talk, get a very quick answer and a certain kind of answer, yeah. And I've taught my ChatGPT how to tell me to answer. What I'm interested in. It's a super fast-tracking language. And of course, there are so many different cases that you're starting to use in the day. I'm using a lot of different case files myself, and you're at some point in time like you've got a phone missing like, uh, you've been off the grid. So, some, uh, this isn't the dependence, but it's basically all that's happening now in the development of artificial intelligence, uh, leads to this... Well, we're still in there, we're not controlled by satellites, cell signals, a lot of different sophths, there, the availability of managers, the access of the Messengers, the rest. But it still looks, uh, that the AI servers are available, but it's less controlled to date than just the availability of massagers and some standard servers. I wonder how this is gonna be all going to develop.
TechCrunch has already been given headlines. AI Apocalypse's on.
Yeah? AI Apocalypse.
Because Claude, OpenAI, Perplexity. Well, Perplexity is clear. They simply suffer indirectly from the accessibility of everything else. And that's all inaccessible.
Someone wrote in the comments, in the past ' s comments, that NVIDIA is becoming the most expensive company in the world. He said, "No shit will happen. So, Taiwan's gonna be taken now, and then, uh, everything NVIDIA's production is gonna die. Which, in general, probably always is.
Come on, I'm just a little bit interested in news about Elon Musk and his, uh, some projects. Uh, first of all, the hiAI raised $6 billion and investments from various funds., X is Twitter, if anyone doesn't know, Twitter, not everybody uses X, - it's very fun, it's seen that there's different possibilities, different possibilities, like a series of liquors. Now, a couple of days ago, I think yesterday was the news that they're about to start posting and letting him be housed. What is this? Is Elon Mask in the innovations or is he testing, uh, behavioural characteristics of the world, how does the world respond to this? So that's where he's going? And I wonder if people are so many innova-- different things he did in X, all of them are talking about it, and then they forget it completely. So he has a very interesting strategy, it's like testing people's behavioural characteristics very large and very fast and very easy.
And I'm just $6 billion-- uh, that's really sick-- that's a very big hug for a company that's developing models. I mean, I'm, uh, checking it out. OpenAI has raised everything all over 11.3.
Yeah. So, here's Anthropic 7.6 billion total, I've been checking, and that's, uh, absolution--- well, and that's, uh, abso-- more than anyone, I think it's more than almost any start-up. So Mistral at the-- that's where they're raising another round, and they want to reach a billion. And if we look at Cohere, they've all raised 430 million. And I think that this investment in the xAI, this investment will end with the creation, er, of competitive space for those who can actually produce models. I think that the xAI, Meta, OpenAI, Google and Anthropic, they're very much ahead of the
others.
Huh?
Yeah.
No, I mean, it's gonna be a development, a development, like Meta does. In the interv-- considering that Meta is in development, uh, his, uh, his, uh, starting, his chat or LLM, it's injecting them, in whatsApp, like, yes, injecting in or, uh, Facebook is injecting them. So, uh, is there a similar focus or not?
In consumers, the introduction is much faster in consumers, especially if it's a private company. The mask can do anything. So, xAI has capitalization as much as the capitalization of Twitter, right? So X is directly. But there's a ogre-- there's a very strong synergy. What he has is a chance to irritate faster than anyone else, I think.
The last private investment they've attracted was estimated at 20 miles to 20 billion. That's what I know. They're so aware that X has, like, a p-p-per-p-p-p-p-p-p-p-t-practical, more than, uh, ChatGPT, and they can go Grok to automate everything, all, all, all. processes and a complete user experience.
It's interesting, of course, that with such a crowd, company is worth a little bit less than that company that is built like, uh, a owner, but still a little different in AI.
Yeah, but I think it's like they're going together, and that X is not a smack now, I think Mask has banned everyone from swamping. I think it's the synergistic one, she's giving, uh, how capitalizing is huge. I mean, the fact that these two companies work in the tandem is a huge
advantage.
And yet, literally recently, Mask has been putting the results of the Twitter ballot back on his Twitter, yes, the same X. And they chose the best model. And, strangely, the best-selling company that Grok is out there. And there was the comment that the first most slaline was that, I think, uh, the audience that was voted on was a little preconceived. Well, it's obvious, right? Twitter users are on Twitter and, there, companies associated with Twitter are fairly well. And that's a huge plus Mask to spread his models, because by integrating the services that millions of people use, he's getting a big audience, right? And as far as it goes, through the bonus, uh, access to X is developing or something, it's already, uh, the following questions, but it's coming straight to the ground with a huge, ready audience that's quite loyal to the company.
Why does Google have one of those if Google and search are immediately integrated? He has much more opportunities, for example, to test AI than X, with the way people are in it, what they read, what they do there.
I remember their first Bart demonstration when Bart was lying right at Bart's demonstration, right? So he was asked some questions, I don't remember how to do it right now, but he's been giving false data in there, right? So he's hallucinated. And in this sense, when it's directly integrated into search, you know, or then, when it's integrated, it's very unpredictable results that are hard to control. And now, I'm not in my hand right now, but, uh, there was news that, uh, Google, there, in some regions, AII.'s already been integrated into my search and there's absolutely, uh, weird advice, right? I mean, for example, there's a pizza recipe, like, there's a glue, yeah, there, or something.
Yeah, I saw that.
So hallucinations are going well. Here, here, hallucinations are going well. And that's why it's a Google introduction, too, technically complicated, isn't it? And that's how you keep your reputation, you don't lose the advertisers, and you're basically the one who's left the service.
I, uh, make a clue. I'm watching her now, because I still use Google, I'm just looking at her as an extra story. Sometimes there's answers, sometimes I don't. Well, I mean, I think I've seen 40 or 50 of these different answers in the last week. But just in terms of the involvement, I think it's a very cool channel, uh, real involvement. Just like Bing when they introduced ChatGPT. I think it's a real mass-used use of this, because I used AI a little bit in whatsApp. For example, when you were on a plane, only a massager was available. There was no-- no Internet. I used it there. It's convenient, right? Oh, then I don't use it anyway, because whatsApp I use for people, people. And yet, if I want some information separately, I open, for example, ChatGPT, or open Google. I'm still using a searcher, and there's all their chips that are worth extra. I don't expect them to, first of all, supersequent because I'm used to being a searcher, uh, giving a lame, and I need to find, try to find, find, directly spend time in it. And as much as I, I, uh-- this is the question I asked about the xAI, my just a little bit of a feeling. So how much is Musk going to test anything new in terms of introduction and use or not? I wonder if they'll have any such exits or outfits like when they're acting on a protest that we don't even expect, yes, somewhere.
I'd be a little step back, if I could comment. Uh, still on time, well, maybe we can watch the stats, I'm not sure, but it feels like the time of a X-ray service is more than Google. Yeah, so you asked Google, you did, you got the references, you went over and you're using other services. And now the time, well, the usual time of use of this service, it's fairly limited and small, well, I think it feels like it. And if we're talking about social media, the same X, for example, people get used to spending enough time there. And if some useful, convenient service is integrated, it is, in fact, likely that people will be actively using it, in my opinion, it will be higher. Mm-hmm, specifically about some new stuff. But, uh, Musk is used to surprise. Yeah, well, we were just doing some of the Tesla episodes back there talking a lot, yeah, there's a reason why these innovations are going through.
I think something like that will happen here, too. Oh, what's interesting is, uh, I don't know, there, Sas, maybe you know Grok, if the HAI has any innovative technical advantage with regard to, uh, other companies already in existence, What are the basic models developing? Or is this money that they've raised now, it's purely on, uh, uh, Mask, in his name, that there are many successful businesses with him?
Yeah, there's a bike about that, which, uh, sec-- thanks to Sequoia, this round got here. A-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the-the- I mean, there, over 20 years ago. And that's why this whole deal went down. I think Grok has no clear advantage now. I mean, it seems that in the text generation, um, how all these models would have reached about one level. Google-- Google has a lot of hallucinations, but Google has a lot better video--- in the Anderstening video. A-a-m, and GPT-4o, they were the ones who were surprised by the speed of work and the generation of voices, and at-- like all-time recognition. I mean, it seems to me that it is from the point of view of the quality of the text generation now, uh, well, as a little stitch, that everyone is at about the same level.
Back to history, uh, where we started, yes, AI-Apocalypse, which, first of all, is some level of generosity of texts that, uh, models have grown up, and that's probably the ceiling we've been in so far, or a time-step we're going through.
But there's another story, like, uh, a little less pessimistic-- a little less optimistic, uh, Jan Lecuna. I'm, uh, so, say, duty pays a little Twitter to read, and Twitter is the kind of thing that Mashk's been having with Jan Lekun. Jan Lekun is a very, very, uh, famous man in the AI area. He's a professor at New York University, the author, there's a lot of books and work on this machine learning. And if I remember correctly, he's on Facebook now, I think, uh, he's running their office. And that's when Jan Lekun says that, uh, it's very similar to the dotcom bubble. All this Haip, connected to, uh, Generic AI, that he's gonna get as close as he used to be with the dots, that's the first. And the second thing that AI in this case is just, um, fashion and hip-name for linear algebra. I mean, basically, there's just simple operations, there, with the matrices over-renewed, with vectors. And just people who, in math, don't get so well, they call it I., and they wonder how good this thing, uh, can answer some simple questions. Yeah, that's why it's like, uh-oh, it's a stew. On the one hand, I want to believe that this will make a difference to our lives. Well, it's really cheating, right? Here, uh, everything, uh, even the b-banal Whisper that makes the transcribation, is a useful thing that changes, there, quick answer to the question, something else. But on the other hand, and the position of one of the, uh-e-e-h-goods of those AIs, Ian Lecuna, is also quite interesting, makes it a little more thought: What's going on?
Look, it's not gonna be a big deal or there's no way a story like, uh, it was with the dot. It's because there's no such thing as AI company, uh, a few trillion dollars, at least one. I mean, it doesn't exist. So, in the big operators, they're up, uh, serious, these-- these operators do, do something serious. What difference does it make when billions of dollars were involved? For these companies, these billions of dollars are just a schmuck. How much money does it have to be cash, cash balance? They have a much larger balance in their accounts than they attract. Rather, the question, um, all rounds, the correct reflection, the right shape and the rest. So Apple can buy all these start-ups with the balance sheets on the accounts, then buy them all and invest everything. The point is, not the money, but the point is, there's some possibility and that we've still got a pretty serious, big tool. Right? Why does he say that? Uh, and, uh, probably, uh, that many people overestimate AI, as they say it's a spaceship, but it's not a spaceship, as you said. Aah, autopilot now is such a relative autopilot, yes, very relative. Bye, for now.
I think there's another understanding coming. We'll see. I'm just gonna go to such technical stuff. Uh, there's a lot of interesting work coming up, and they've presented a new function in Perplexity. Ah, and these different things are fun. And I think these things, uh, cool people, uh, through, through themselves. It's like a new, new user experience, right?
I really liked it crazy. Oh, well, short, yeah, for those who don't know the news. Look at this. There's a presentation that's interesting enough, how it looks, what it can be used for. But if in two words, Perplexity is a search, so, say, combined with, uh, a chat-bot, which the search results analyze and give you some information on the basis of what he collected. And what did they do, huh? They added that you could collect, uh, perplexity pages, that is, the page, the data you were looking for, the information you were processing, the processing, the processing. And you have a note that you have, yes, a little weirdo that you can use. You can keep it there somehow, and you can go back to her. So, uh, what did I like most? Oh, well, let's do this., if short, it was for Perplexity to add Notion, yes, and now you don't need the results of a search to collect one of your own, formalize, they're like automatic, um, beautiful, such a little weirdo that you're showing up, You can keep it. What did I like here? There's finally something new. We were so excited, uh, first, how much longer? Two, two and a half years ago, when ChatGPT came in, first, there's 3.5 versions, then four. And now we have a new experience like this when we're talking to chat, and there's something going on. And since then, there's nothing that changes, right? In fact, the interface is the same, uh, the same way to interact. But yes, Perplexity came along, but in fact, it's like the search results were added inside some LLM, someone they've processed. That's where you got something, right? I mean, it wasn't a big deal new. Now at least it's some interesting experience, user experience, as it is called when it's not necessarily in this company, but it can be something we're used to. We'll use a lot of things.
If, for example, Google or some other company does, I think it's a fairly fresh, interesting story that might change our experience of using the Internet, searchers.
Did you see the final product look like this? Is it like a spectacle, some schematics or what?
Yeah, there's a prezi, there's a prezi, it's really fun out there. I like this stuff very much. I think that, uh, sometimes it's a shock to how much corporations don't--
How real corporations can't produce classy, user-friendly things. And-or shows that they are not interested, not people, but interested in their own interests, their own power, right? And it's obvious that the OpenAI is not interested in ordinary users, but to find this one for itself... to get into the AGI market and find AGI, right? Not some kind of prank. Perplexity, but it's probably more of a user job. It's like Amazon still missing-- it's basically, they've made beautiful, uh, pages, like, "Lings for-- "extradition." It's like Amazon, if there were-- uh-uh, there were pages about-- done-- some kind of thing, uh, not about the merchandise, the brand or the group. About the brand, huh? It's usually about brand or salesman. That's the stuff, just the stitch, right? And in Alibaba, in Alibaba, these pages are cool, very beautiful, and, uh, eight years ago. Very cool, huh? And they took half a million dollars for their promotion, and they took half a million dollars for... Because it's an opportunity when you come in and you have a very cool brandy story, a great, well-trained user, a probable, chaotic, with a lot of different possibilities, right? And we see how many companies are growing, and they just don't. And now, in Perplexity, when they released these pages, it's very-- it's such a compassionate experience, very serious. I mean, it's even closer to e-commerce.
You, Sash, what are you smiling about?
And here, I was gonna argue that a little different view, yes, a little more negative, huh? The search, it's a very broad task, and there's a field of invisible evidence. There's a trend that the thinners use TikTok more and more to look for. I think there's statistics that more than a third of the United States' TikyTok is used to look more than Google. Iia, that is, there's a lot of things, there's a great opportunity to make new research. And I think that Perplexity is not a successful company to do a lot of different products at the same time. I mean, you can't say they've won the search yet. I mean, for example, I've been doing a little lately, uh, recently, like, a little reserver. Play--- , it was, uh, the finale of the Western NBA conference, and it was, uh, Dallas versus Minnesota. I wanted to, uh-uh, re-check the statistics and use Google and Perplexity for that. In fact, both and both companies have very strong knowledgel-knowledge base, and in such a way as small-sized-- in purely numerical indicators, they can be quite mistaken, and they can be very strong. I'm sure they're wrong enough. So it's just important there, where often the player throws three-way eyebrows, for example, is it? I'm here, I'm gonna meet them, so I've been testing them for five minutes and I'm very disappointed both. And I think there's a huge job band, like, to improve the search, and to take over the market. And what-- and probably what Perplexity does parallel to new products is not from the position-- it's not from the point of view of power, to be mild. I mean, for the start-up, it's okay. So the start-up should be looking for something that's gonna blow up. But it's more like a test for me that they have some stagnation in their search.
I'm a user here, yes, I'm, uh, I'm not gonna start using Perplexity. I'm not using it now, and I'm not gonna start with this thing. But as a user, I am very happy that something new has finally come up because if it isn't, some other big company has this idea, if it looks interesting, it's gonna get her, it's gonna work, and then it's that. We'll be able to use it. It's just, remember, there, I don't know how many, 18-20 years ago, uh, there was a huge number of phone de factos, right? There were phones without keyboards. The numberkia was like a pen, just with a turning machine. And as a total delusion, there were a lot of them. The market was looking at different sides, yeah, there, designers, uh, interfaces could use it all. Since 2008, the iPhone has actually been the same, just a screen and nothing else. And, uh, on the market, the phone if something comes up, I'll be happy. And here's finally, uh, a company we've been talking about, and there's enough Perplexity, too, and something that I think is different from what's on the market, right? It's not just like they're taking something to look for, added LLM to it, and they're being processed. I'm very trivial now to explain how it works, uh, Perplexity. I'm probably wrong about that, but it looks like it's for me as for the user. And there's some new use format that I hope, uh, that other big companies will catch.
And the point is, everyone seemed to think it was infinitely expensive to do something, uh, new yea, maybe it was all about, yeah, in corporations, that's to do innovation, because it could be said, "No, no, It will be expensive. " Google, uh, Google spent a lot of money to do it. And now, uh, you can see, according to Perplexity, that this is possible to reproduce. Of course, there's a critic here again, which is possible as Perplexity actually uses Google results. And the viro-- and they're kind of officially partnership-- they're having such a formal partnership. But, but still they, they spent an , not so much to do-- do, uh, the first product. And a lot of companies could afford it. But now, Sasha, now, we're all going back to the previous top, and we're all going back to the big companies.
I wanted to agree that the company has a huge-mm, tens of billions of dollars in Apple, Microsoft, and so on, that they should invest. I agree that, at the very heart of these companies, the entire industry is developing. And I'm here to note that recently, there was a problem with Salesforce. That must be a m-mm... I mean, salesforce recently lost $50 billion a day. For a company that's worth-- that's 20% of their market cap, so, roughly. Uh, uh, s-- and that's, uh, just if we look at the level as below, the Apple, below the NVIDIA, below Microsoft, yeah, let's see the Salesforce is the lead company, that is, uh, uh, that's, uh, in marketing, sales, customer service, this is analyst, a leading company in these areas. And they have an outstanding CEO. And the main thing is that Salesforce is indeed one of the companies that invested in AI. We know that when the OpenAI was the mut, Mark Benionff, their famous CEO and faunder, called personally and wrote on Twitter, recherches, like they want to move into Salesforce. I mean, uh, that's no joke, that's not what-- that's not a joke. So they took a serious look at AI, and they had problems like this., and I understand that the crypt of Salesforce, Salesforce is that maybe they won't be the beneficiaries of AI-- from AI in that it might be expensive for them, even for them, AI, and perhaps for them. On the contrary, to copy some con-- some specific product line that now has a Salesforce, it's easier and cheaper. A-a-a-a-a-a-a-a-a-a-a-a-a--generative AI-a-a-a-a-a-a--a--a--a-a--a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a--a-a----a-a-a-a-a-a-a-a------------a-a-a------------a-a-a-a-----a-a-a-a-a-a-a-a-a---a-a-a-a-----a-a-a-a-a-a-a-a-a-a-a--a-a-a-a-a-a-a-a-----a-a-----a--- I'm only gonna have to pay my pocket, but only units can afford it. So, salesforce cannot allocate 6 billion to its models, as Mask has done, in blatant words. I mean, I think they're gonna be very hard.
I mean, if they, uh, in the last year they generated only 6 billion, say that all of these 6 billion we're gonna invest in AI, I think it's very difficult for a shareholder.
In fact, a year ago, they were worth as much as they were standing now, a little less. And their fall and loss of $50 billion is more, uh, growth, last year, and then a little fall. I mean, if we look at the overall curve, it's gonna be like this anyway. But it's interesting to develop their systems, if they need to do their systems, it's a question. I don't know if these companies have to do their systems. And I think the salesforce could have been attracted, uh... I have a feeling that Salesforce might have attracted billions of dollars of investment if they wanted to create separate AI-decisions because they have access to sales and marketing, uh, as well as anyone, customers, and they can just be an incredible number of things, start moving. And they have a huge number of ML, and AI, uh, operated inside the system. The other thing is, they're not a mass market player, if you're talking about it, right? So they don't go out to markets with their own search, and they don't have their own search. But I'm in the view of the end result and the introduction of this general outcome.
The results of this general introduction, they have many opportunities, in fact, more. We'll see what happens to them. We're on ToTheMoon. I'll see you in a week. We're on Sundays now. We were in a format, we tested Fridays, Sunday, actually, once a week. We're always four people. Sign on the channel and wait for your comments. We read all the comments. Ah, in terms of lighting different topics. Here we are discussing technological news from the Silicon Valley, different sites and the world as a whole. See you again!