e doesn't tell you what it was trained on. The data that it was trained on is not something we have access to. So what if they just trained it on a bunch of pro-China propaganda, you know, plus all the generic stuff in the world, so that it was essentially a brainwashing tool? How do we know? Well, the only way we'd know is by asking it questions that would get to that very point. And it turns out people are asking that question. And sure enough, so Joshua Harley was asking it some questions. And if you ask it some questions about China that it doesn't want to answer, it actually starts to answer them and then goes into racism while you watch. I can't talk about that. So it does know some bad stuff about China, such as the Uyghurs, let's say. But it will start to write it, and then it will, whoops, which means that they've got some kind of safety valve on there that makes it change its mind. I don't know if you could find that safety valve in the code. Where else would it be? So that's an open question.

So we've got the issue. And then Joshua Harley also asked it if it used any or how it was created. And it said that it was created with these Nvidia GPUs. Now that's DeepSeek itself saying it was created using these chips that it should not have legal access to. So the question I have is, is that real? Did China really allow it to say that it uses stolen technology?

Here's what I think. This is just a presumption. What I think is that it could be that DeepSeek hallucinated the use of those chips, because that would be a good hallucination. Because if it knows that there's no way it could be AI unless it had had some of those chips involved, wouldn't it just make the connection on its own? So let me say it more to the point. If what DeepSeek knows is only what it was trained on, why in the world would China train it on secret data that says they stole some GPUs from Nvidia or got them in some illicit way? How in the world would DeepSeek even have access to that to train on it? So I wonder if it found that in its training and China didn't know how to, that it had to get rid of that. Or did it just assume that there's no other way it could exist unless some Nvidia GPUs were involved? So you really can't tell if it really thought that or it just assumed that. So that's interesting.

So we don't know if it's a propaganda machine. We can see evidence that it is hiding what we would think is the truth. We don't know if it's hallucinating. And it's also should be noted that most people who use it in the long run will use the version that's connected to China so they can collect your personal information again. Now a lot of the news about DeepSeek was that so exciting you could run it on a local computer. But the local one will never be as good as the one that's connected all the time. So we're going to get all excited about the local one, which basically will be a toy, you know, just maybe another user interface. But the real one, the one people would use if they just go to their browser because they don't want to figure out how to load it on their personal computer, will be connected. So again, it's a way for China to collect all kinds of information on its users.

So that, and then we find out that there's this huge cyber attack on DeepSeek. Now who would do that? Who would do a major cyber attack on DeepSeek? Well, first of all, they'd have to have a lot of skill. Let's say, could it be China? Well, no, it's China's product, so they wouldn't attack it. Could it be Russia? No, it doesn't seem like something Russia would do. Why would they? Iran? No. Israel? No. Europe? No, they don't, they're barely in AI. So who would do such a thing? I can only think of us. If it's not the United States, it should be. Because if it's true that our biggest AI companies are, and it looks like it's confirmed, are going to be in bed with the CIA, the CIA can't let China destroy the AI industry in the United States. There's probably no limit to what the CIA would do to make sure that the US is dominant in AI. So the first thing the CIA would do if they're doing their job and if they're in bed with AI is they would look to destroy DeepSeek. So if I had to guess who's doing a major cyber attack, probably us. I mean, can you come up with a better idea? Do you think the Sri Lankans just decided to go to war? Like who else would it be? Africa? What, Greenland? Greenland decided to kind of assert itself? I mean, honestly, who else would it be?

No