Elon Musks Chilling Warning: "Grok 3 Is SCARY Smart!"

Grok3 has very, very powerful reasoning capabilities. So in the tests that we've done thus far, Grok3 is outperforming anything that's been released that we're aware of. So that's, that's a, that's a good sign.

It's.

In fact, at Times I think Grok3 is kind of scary smart. You're like, wow, this thing's smart.

So that was Elon Musk actually talking about the latest chatbot from his company, X AI, and he's actually referring to Grok3 as a scarily smart chatbot model. Now this is something that I don't think I've heard any CEO of any AI company say when talking about their AI model. Now, some could argue that Elon Musk is doing this for hype. Elon Musk has been known to drum up the hype before any major releases. As many of you know, this guy has been doing that for Tesla many different years right now. But I do think that potentially this time could be different because with AI, every single time there's been a major AI release, usually we are quite stunned, surprised by whatever these companies manage to pull out of the magical hat that they call their AI labs. Now, I think this talk by Elon Musk at this conference is super insightful because a lot of the times when new AI models are going to be released, we don't really get that much insights with as to how powerful the models are going to be before their immediate release. And considering this is going to be an AI model we get in about a week or two, depending on the entire landscape, because I don't think Elon Musk is going to be the only person releasing his AI model. I think it's super interesting to say that these models are now getting to the stage where they are scary smart. Now I do hope that that statement does hold up. Maybe he's giving himself some very big shoes to fill. But considering that we are in a new paradigm of a, this could potentially be quite true.

You know, it comes up with solutions that you didn't even think were like, you wouldn't even anticipate, you know, not obvious solution. So Grok3 was trained with the most amount of compute and I think very efficiently trained. Also notably, Grok 3 was trained on a lot of synthetic data. So, and then it goes back and forth through the data and tries to achieve logical consistency. So, so when if it's got data that is wrong, it'll, it'll actually reflect upon that and remove the data that is that is wrong. It does not concord with reality. So its base reasoning is very good. In fact, the. Even without fine tuning Grok 3, the base model is better than Grok 2. So with. So we're really in the final stages of polishing Grok 3, probably it gets released in about a week or two.

So with that being said, you can clearly see that Elon Musk has looked at this model and clearly thinks that this is going to be the best model that currently exists by far. Now that is of course a very big statement, considering one of the things that is probably the hardest thing to do right now for any AI company is to be the market leader. And I think probably it might not even be hard to actually be the market leader. And the reason I say that is because oftentimes we've seen every single company on this leaderboard right now, whether it be Deepseek, Anthropic, Alibaba, OpenAI, Google, whatever company we've seen. It's very. I don't want to say it's actually easy. It's not easy. But even if, let's say anthropic, I mean, X AI managed to actually take the top spot from ChatGPT on this chatbot arena, do we think that they will hold that leaderboard spot for a very long time? It's one thing to take that top spot, but I cannot see any company and this isn't just a dig at X AI but. But considering the AI space moves so quickly, that number one top spot changes almost every two weeks. So I do wonder, even if X AI is able to grasp the top spot, just how long they're going to hold that spot. Considering now we've got new entrants like Deep seek and Quen 2.5 max and a variety of other models that are going to be coming very soon that look to take the top spot as well. It is a very, very tough environment even for an accomplished company like X AI. Now if you're actually wondering, have there been any leaks with regards to these models and just how capable they are when it comes to Grok 3, I did come across this tweet on Twitter where someone quote tweeted someone that is an X AI employee showing early hints of Grok3's coding abilities by demonstrating this tree demo. Now, I gotta be honest, I'm not too familiar with this tree demo, but I'm guessing that this is a tree demo that is coded and it shows the different seasons. Now I'm guessing also that this is a pretty complicated task for a computer to do. And coding is one of the benchmarks that many people do like to use to kind of gauge and see just how relatively good the model is compared to other systems. The reason many people do this is because the very best model that many people use for coding is called 3.6 Sonnet. And that model has gained a variety of different users attention because it's just so coherent and so high quality in terms of anything that you ask it to do. So with this short demo, we're basically seeing that this model might actually be rather effective. Now there's also some more information. We also got this tweet as well. We got the fact that there is a model currently in the Chatbot arena called Chocolate, and we're currently seeing that this model is a bit of a personality. You can see that there is a quote tweet from this tweet that says the Chocolate model in Chatbot arena is likely Grok 3. Nothing else can be this cringe. When the model was asked, what model are you? The model responded, the curious mind wants to know, doesn't it? Let's just say I'm a mysterious stranger and the digital realm here to help and provide assistance without the need for labels or badges. Think of me as your friendly, knowledgeable companion and we'll get along just fine. Now, shall we focus on something more interesting? Yada yada yada. And this is kind of reminiscent of how Grok talks to you because it probably has a large personality compared to other AI models. There was also this post from Reddit that spoke about how Chocolate and kiwi on the LMSYS arena are a likely variant of Grok 3 and that they're very good at web design. Claude 3.6 Sonnet apparently finally has got some competition and they actually show a result of the prompt write full HTML, CSS and JavaScript for a very realistic internal Reddit intranet logon page with both models and it showed that the Grok model was really good. Now at this time there is also some crazy Grok 3 drama going on that I wanted to include in this video because it actually states otherwise. And it's going to be super interesting because this is coming from someone that works at the company. So we had this guy, Benjamin D. Kraiker, who was an ex human Data team member at X AI. This is a guy who was working at X AI and then he actually tweeted something. Okay, so he tweeted this a few days ago. So last week he tweeted that this is the current ranking in AI for code, he puts ChatGPT01Pro at the top, 01 second, and 03Mini at third, saying that they're all kind of tied. And then he puts Grok3 is expected to be underneath ChatGPT01 Pro and O1 and O3 mini. So this is surprising because he puts it below other companies and this actually caused quite a stir. Now, some would argue that this is the person being honest because maybe they've internally tested the model, maybe they've realized that the model just isn't good as O1 Pro or O3 Mini, and they're just being honest about the internal performance. You can see this actually got 5.2 million views. But we can also see that There is above Claude 3.5 Sonnet and these other two models. Now, some people would say that if you're, you know, working at a company, you should never publicly admit that your software is going to be worse than other people's. It's just bad press. But crazily you can see that this actually led him to resign from the company over talking about the performance. So he speaks and he says that I actually resigned from X AI tonight. It makes me very sad that this was the right thing to do. And here's why. Xai told me either I had to delete the post quoted below or face being fired. After reviewing everything and thinking a lot, I've decided I'm not going to delete the post, which was very clearly a harmless personal opinion. Why did they tell me to remove this opinion? Well, according to them, the reason is that I acknowledge that Grok3 exists and he says, I wish I was joking. I'm not. That's the reason the fact that I wrote Grok3 to be decided is grounds for being fired. But wait, hasn't Grok 3 officially been announced by X AI? Yes, it has. And basically what he's saying here is that they wanted him to remove that post. And I think it's quite obvious why they wanted him to remove that post is because Grok3 isn't at the top. But they actually said, you know, the people working@x AI said that, look, we want you to remove the post because it speaks about Grok 3 and that's confidential information. However, I would say that it's not confidential information because Elon Musk has been talking about Grok 3 pretty publicly. But I'm guessing that probably they just wanted him to remove the post because discussing the model's performance beforehand is probably a bit confidential and stating that it's not as good as others is clearly negative pr. I mean, when we think about the scale of that kind of pr, you have to understand that X AI is a company that's raising billions of dollars for their clusters. And if they can't raise billions of dollars because maybe you've got insiders saying that look, these models aren't as good as ChatGPT, that seriously isn't going to do very well. Now you can see right here that he says to be clear, the post they wanted me to remove is 100% just my personal opinion. I don't know where Grok 3 will stack up against other models. Hopefully it does well, I don't know. That's why it says my opinion. And to be determined, it'll probably be pretty good at some things and perfect at others. And I didn't think this was a particularly wild opinion. Again, the official demand that they said is that even writing Grok 3 to be decided is somehow confidential information. And this is absolutely absurd since it's repeatedly been acknowledged by the company and its famous CEO. Are they mad that my clearly labeled opinion didn't guess that the still unreleased Croc 3 will be higher? Maybe, probably. And again, maybe it is the top. I genuinely don't know. That's why it says to be determined. But what's also interesting about this is that, you know, amidst the drama he talks about how there is a specific feature of Grock that he spent the majority of his time working on with a really hard working team that is very cool and I hope it works extremely well for everyone. I won't say what it is because that would be actual confidential information, but maybe after it comes out. So this is a potential feature. I genuinely don't know what this is. There haven't been any leaks or any suggestions with regards to what this unique feature for Grok is going to be. Maybe it's agents, maybe it's computer use. I genuinely have no idea just yet, but I will be intrigued to see what what that is from X AI. But you know, seeing this person not forced to resign, but choosing to resign over this, of course it doesn't look good for X AI because if you knew that the model was good anyways, I guess you wouldn't really care if this post was up. But this is one of those things where it's like trying to remove attention from a post just grants it even more attention. Like this post I'm pretty sure wouldn't have had 5.2 million views if this person wasn't they weren't forced to resign. But if this whole debacle didn't happen and they didn't choose to resign, I'm pretty sure now people wouldn't be expecting Grok3's performance to be underneath. ChatGpts I mean, it's going to be super, super interesting. Elon Musk has already said that this model is going to be absolutely incredible. It's going to be the best thing currently released, and we already know just how competitive the entire AI space is, so overall, it is going to be a super interesting time. I'm pretty excited to see how these models perform, and I can't wait for Grok 3 to be released because it's probably going to ramp up competition once again. That being said, if you guys enjoyed the video, I'll see you in the next video.