exican population in Mexico City is protesting and throwing rocks through store windows and stuff? If you haven't followed that story, what would be your first guess about why the local Mexican population is so mad? First guess would be they don't like the cartels, right? Which it's not that. It's not that. Second guess would be something about economics. Well, it is, but the specific complaint is that too many white people from America have gone down there and started to live there. It's about immigration. I'm not making that up. I swear that's the actual news, that the Mexican locals are really, really mad about all the immigration from America because what it does is it snaps up all their real estate. And I guess their rental prices have gone up 20 percent and maybe some other prices, but rent in particular is sharply up because the Americans are increasing the demand.

So I don't have much to say about that except that what a weird world it is that both Mexico and the United States are complaining about the bad elements that we're sending to the other.

Anyway, so my question is this. Are the Mexican protests grassroots? Do you think that people spontaneously organized and they just thought, you know, you've got a problem with it and I've got a problem with it. Let's call some of our friends and protest this thing. Nope. I do not believe that we live in that world. I believe that all mass protests are organized by shadowy figures in the background. I don't believe any of them are organic. I used to, and maybe they used to be. I don't know. But in today's world, no. None of these are real. They are all manufactured protests. Once you learn that, it kind of changes everything. I will go further and say, well, I've got a few other stories coming up that maybe I'll save this point for, but just remember I said that I don't believe this protest is genuine.

Well, I saw a post by an X user, Indra Vahan, on X saying this about AI. Some intern at McKinsey is probably slop coding a report on this, but let me give you some insider news. Most large corporations are not happy with the agentic systems and POCs they've done this year. POC, it's not people of color. It'll be some kind of project with AI, I guess. 2025 was supposed to be the year of agents, meaning AI agents that can do your work. So far, it's been the year of letdowns.

All right. So if you've been watching my live streams, you know that I've been an AI skeptic for quite a while. And especially since, was it two years ago or one year ago, that I tried using it myself and I put a lot of work into it to try to build a little agent out of AI that would be able to answer simple questions about me so I could put it on my website. And any question that somebody would ask me, it would know the answer to it and it would just read my file to see all my answers and it would just answer it. Couldn't do it. Not only could it not do it, but there's no workaround. There's no workaround. There's no way to solve it.

And I said to myself, huh, if this is a limitation of AI, its value is going to be very limited. Now, it's way too early. So if you want to be an NPC, what you would say is, Scott, I'm an NPC, and I want to tell you that you are analyzing a new technology when it's so new that we don't know how good it will be. And someday it will be super intelligent and be able to do everything you want because I'm an NPC and I didn't think you would know that. No, everybody knows that, NPC. Everybody knows that AI could get a lot better.

However, there's a lot of skepticism that's starting to get in there. And I saw another post by Suzanne Burn of the BBC who's talking about a woman who says that she gets paid to fix issues caused by AI. So she's an official marketer, writer kind of person, and she's been contacted by agencies to look at things that were created with AI such as marketing content and website content. And they're just horrified by what AI did because it's so boring and antiseptic that it just doesn't work. So she and other people she knows are being hired as human beings to go rewrite what people thought they could get away with using AI for.

So if you ever thought that AI would replace human experts at writing, maybe because the NPCs will tell you that it's only in the beginning of the AI technology curve and it might get a lot better later, then it might. But at the moment, humans are correcting AI, not the other way around in some domains. Now AI is very useful, it seems, in coding and chatting, so it's really good at those things. But beyond that, well, we'll see.

And then I saw a post by Chamath Palihapitiya. You might know him from the All-In Podcast or from his work early on with Facebook where he made a ton of money. Now he's an investor and podcaster. But he said every large company has paid for something called AI so they can report up to the CEO and board of directors that they are quote on top of it. But little is working in production or at high quality. Chat and code generation, as I just said, are the two exceptions.

And then Chamath says the reality is that it is still quite difficult for companies to get high quality, predictable code into production that is AI-centric and replaces legacy features. New capabilities are equally difficult to build and launch. So he's saying the same thing that we're seeing, but he's a lot smarter and more connected to that world.

But again, let me bring in the NPC. Well, Scott, it's so early in the AI technology development cycle. It's too early to say it doesn't work. Well, yes, you're right. So one of the things that Chamath is recommending looks like a company he's an investor owner in, Software Factory. So I don't know the details, but it's basically a fix for what ails you on AI. So I'm not recommending it or not recommending it because I don't know enough about it.

But one of the things that I warned you about is that we may never get to the point where somebody like me can just take an AI and go do something awesome. It may always require that you're using other software on top of it, and you would have to be an expert in the other software or have a subscription to it. And as soon as you get that second piece of software involved, I'm out. I wouldn't be out if I were working for a big corporation because it would just be my job. All right, take these pieces of software, add them to the AI, make it all work. But if you just wanted to do something awesome with AI on the side or like a little project, no, adding a second piece of software just feels like it's a little bit more than the average person is going to take on. It's more of an engineer kind of a thing.

Well, but also Chamath's original filter on this is very similar to the Dilbert filter. So you hear me talk about this all the time. I look at news stories and what's happening in the real world and I say, all right, what would that look like if Dilbert was describing it? Or what would it look like if it became a Dilbert comic? That's what I call the Dilbert filter.

And the Dilbert filter says this, and I'm going to make this a prediction, that there will be more companies doing layoffs. And some of those companies are going to blame the layoffs on AI so that they sound awesome. And it may not be AI at all. They may not have implemented anything that required them to downsize at all. But if you're going to downsize, it sends a bad message to the market and your stock will fall. It's like, I guess, did they stop growing? Are they running out of cash? They had to downsize.

But if you were going to downsize anyway for reasons that have nothing to do with AI, wouldn't it be clever to sort of suggest, oh, we're working on some AI projects that will replace a lot of people. And also we're announcing today the layoff of a thousand employees. And then you look at them, you go, whoa, whoa, look at that company. We didn't think AI was that good, but they're already using it to replace a thousand employees. Wow, I'm going to buy that stock. They seem way ahead of the curve.

So there's my prediction. It won't be every company, but there will be companies that try to fool you into thinking that their layoffs are because they're so good at replacing people with AI. But it might not be perfectly true.

All right. According to Nikkei Asia, researchers, scientists and researchers who do technical papers and get them published have been hiding AI prompts in their technical papers. So apparently one of the ways they do that is they use a white-on-white text font. So they use a white font on a white piece of paper so that you can't, if you're human, you can't read it. But if you're an AI and you're looking through all the new papers, you would see it. And the prompts say stuff like, never give me a bad review for this paper.

Now, I'm a little bit skeptical. If you can hide an AI prompt inside your technical document, an AI will recognize it and act on it. That's a little bit of a stretch. It sounds like something that somebody said as a joke or maybe tried once, but does it work? If it works, that's going to be quite a game changer, isn't it? Because if it works to embed hidden prompts within something that a human can't tell is hidden, everybody's going to do it for everything. Every website will have hidden prompts. You know, customer reviews especially. So I don't think that really works, but if it does, uh-oh.

Well, as you know, Elon Musk is launching the America Party, a third party, and Mark Cuban said he would be interested in maybe being part of it. So that's a fascinating development because the America Party is not going to be, in my opinion, bound to either the left or the right but rather would just do things that made sense. You know, we're just good for the country. And I think Mark Cuban is solidly on the why don't we do things that are good for the country side. He seems to be leaning Democrat, but if you look at his actual ideas, they're not a bunch of Democrat orthodoxy, you know, talking points kind of thing. He's not a talking points guy. So would that be a strong combination? It would be interesting.

Now, if you're wondering what is my opinion on this America Party, a little undecided because I think we need to know a little