right? Remember, this is not medical advice. So there's no medical advice that's going to follow. But in my opinion, as a patient, I am now about to embark on the two most promising ways to treat my specific situation. Some people complained and they said, "Wait a minute, why is this rich guy getting this special Trump administration treatment and would regular people get this treatment?" And the answer is I'm doing this for everyone. Now obviously it's mostly to keep myself alive but you don't think if I fix this problem, the problem being fixing the distribution of this promising drug, you don't think if I fix at least the communication with the patients and raise the awareness of this drug, you don't think that helps other people? The whole theory here is that if I can fix it for myself then it gets fixed. It's not just for me. It would be primarily for anybody who had the same problem and didn't have the good fortune to have apparently some of the best friends in the world. Some of them I didn't even know about. But boy am I appreciative. And I promise you that if I get a good result, everybody's going to know. That's part of the play. Part of the play is that first I escape from the jail, but then I go back and I free the other prisoners. In this case, the prisoners would be people who have cancer, the kind I have. And if I can, I'll burn down the prison and take the warden as a hostage. So this is always a bigger play. It's not about me specifically, but I understand the criticism. I could understand why people would see that.

Yesterday even Elon Musk weighed in, used Grok to show me that there were some cancer treatment alternatives if the ones I'm trying don't work. So yes, my medical treatment involved Trump, the administration, Elon Musk, Kaiser, and by the way, Kaiser is doing a great job at the moment. They're doing a great job of communicating and getting me in where I need to get. So A+ for Kaiser for making the adjustment. You know how I judge people, right? I've told you this is a reframe as well. The best reframe for judging people or processes is not what they did, although it seems obvious that that should be the way, right? It's how they respond to what they did. How they're responding is excellent. And that's how I will evaluate them. I'll evaluate them based on the response. So A+.

You might remember I brag about this too often that I am the only non-AI expert, I think. No that's not true. There must be lots of others. But I'm one of the public figures who's been saying since the early days of AI that, hey, I don't think this large language model thing that keeps hallucinating could possibly be useful for anything except fun little chats. Like, you'd never be able to use it for anything. Because when AI was new, you knew that I tried to use it for something. And what I tried to use it for was what I thought was literally the easiest thing it could do, which is look at a file I'd created and tell me what's in the file. Like, what could be easier than that? If you're AI, it can't do that. And if it can't look at a file and accurately tell you what's in it, and I know you think it can, and you think, "Oh, I build this special file. It's called a rag. Then it does." No, it can't. No.

But here's what the New Yorker says. There's an MIT study that found that 95% of the companies that invested in AI tools, these are not the companies producing AI, but the ones using them, were seeing zero returns. And they say it jives with the emerging idea that generative AI, quote, in its current incarnation simply isn't all it's cracked up to be. John Cassidy is writing about that in the New Yorker. Now, does that sound like me two years ago? It does, right? Was I not two years ahead of that? If you used it for five minutes, you could see that it just didn't have the right tool. Just wasn't ready. And it didn't look like it could possibly be ready, which is what I think is different in my case. A lot of people said it's not ready, but other people said if you just keep feeding it words, it'll become smarter. No, I said if you keep feeding it words, it'll become more like people. It won't get smarter, if you know what I mean.

So Axios is writing also that the layoffs might be going up and that companies are only using AI as an excuse for their public explanation of why they're laying off people. Who was the first one to tell you that the companies would lie that AI was the reason they were laying off people? Because then they could get a twofer. The twofer is, oh, you reduced expenses by laying off people. Yay. Oh, you're also a pioneer in AI and you've made it work so quickly that you could lay off people. God, you're amazing. I told you that the most likely Gilbert future was that companies would lie and say the AI is why they were laying off people. And here it is. Axios is reporting companies are lying. They're calling it the layoff boomerang. Meaning that they lay them off, but you're going to have to hire them back eventually when the AI doesn't work. So that's a pretty big deal.

And one last thing on that same point, actually two last things. ChatGPT has announced that ChatGPT will no longer give health or legal advice. What do you use AI for? Mostly health and legal advice. Those are the two categories I use it the most. Now I was aware that I would still have to check my work, but it is what I use it the most for. I mean there are all kinds of legal if you count tax and insurance and all that within the legal domain all the time.

Now let me ask you this for those of you who've been watching me. Did I or did I not tell you at the birth of this AI bubble? Did I not tell you that AI would be limited by these special interest human groups who didn't want to be replaced? Is that what's happening? Or is ChatGPT just independently thinking they're going to get in trouble if they accidentally give bad legal advice or accidentally give bad health advice? Both of which are guaranteed if you have a hallucinating AI right now. How did anybody else tell you that humans will block AI from doing what AI does? Even if it could do perfect legal advice, even if it could do perfect health advice, I told you that humans would block it because they don't want to lose the power of being the gatekeeper to what is true about your health or what is true about your legal situation. Now, that was a pretty damn good prediction, wasn't it? I mean, I feel like I can take credit for that.

And then it gets better. There's a new study according to Medium, Lewis Call is writing about this, that finds that AI models write code. Oh, okay. Well, here's the one thing that AI can do well, right? The one thing that people say, well, AI can help you write code faster. But 18 to 50% of the time it writes code with security flaws. Do you think the human is going to catch all the security flaws by looking carefully at every line of code written by the AI? Or do you think that a normal human being would say, "Oh, AI, write this part of code," slap it in their program, and then write the part that they write, and then slap in some more AI code? Which do you think sounds more reasonable? That the human would in great detail check every line of code the AI wrote just to make sure it didn't have these security flaws? No, no, no. Well, they're just going to put them in the program unless they're gaping and obvious, I guess. So let's see. It can't do coding. It can't do legal. It can't do health. And it can't help you in any productivity way by doing tasks. It's called AI people. It might be a bubble.

But I will give some comfort to those of you who are complaining in your head right now. I do understand that we're at the beginning of AI, not the end. Can you give me that? I do understand that somebody might figure out how to solve all these problems. I understand. But at the moment, it's right on my prediction. Which doesn't mean it will always be so. So I do accept the inevitability of a superior AI intellect, but we're just not close. It would be some entirely different technology. And there are people working on entirely different technologies. So it's not like it's not going to happen. It just isn't happening yet. That's my only point.

All right. Apparently there's some new news about Comey. So you know Comey is in trouble. You know, I hate all these legal stories, but as best I remember, Comey had his friend leak some stuff and then did he lie to Congress about leaking stuff and now the lie is the issue that he might be jailed for the lie? Well, apparently some more documents were discovered from Comey and that time. And he said among other things, "Well done, my friend. Who knew this would be so much f