hat's new is just fake, and then you graded yourself after the fact when you found out what was true and what wasn't, you'd be about 85% right. Yeah, about 85% of everything you're going to hear will be fake.

Now you might say, but that's okay, I'll just sort the fake stuff from the real stuff. You don't know the real stuff, so you have to treat it like it's all fake. It's sort of the fog of war situation because you're close to the election. So is it true that the non-farm payrolls didn't rise as much? I don't know. Maybe it'll get revised. I would have expected it to say good news and then be revised down later because the administration in charge is the one that wants to get reelected. So it's kind of surprising that this would be opposite of what you would expect if there's shenanigans going on. But if the prediction and the actual are so far off, maybe you just can't even do any shenanigans. So this might be close to a real number. Never know.

Sam Altman tells on a Reddit AMA he said that AGI is achievable with current hardware. AGI would be the what doesn't exist yet in AI, but AI uses this large language model to think. In other words it looks at patterns which have existed before in people's thinking and just applies the most common patterns. So that's not really thinking, it's just pattern recognition. But AGI would be something like intelligence, something like we would recognize as intelligence. Now Sam says that that can be done with current hardware, but could it be done with current software? No, they need new software. But could it be done at all? In my opinion the answer might be no.

So I feel like AI is really sending strong signals that it's reached a peak. I'm sure I'm wrong because the smarter people are putting trillions of dollars into it, so I got to be wrong. But so far there's been a lot of noise about AI and all it is is a slightly better user interface. It's not really doing anything useful in my life and I use it all the time. Mostly it just makes me mad because it doesn't work. Most of my interaction with AI is it doesn't work. Like if I'm using ChatGPT, the app doesn't work. The Wi-Fi is not working. It doesn't understand what I said. It starts talking and it won't shut up and it's not answering the right question. So almost all my experiences are negative. And yet it's growing like crazy.

So I would guess that it's going to be like early computers. If you remember the first personal computers, if you're old enough, do you remember how often the first personal computers would crash? If you were writing one Word document or WordPerfect or something, if you just wrote one page just typing it, you might crash like three times and lose all your work three times. That's how bad it was. But still you knew the potential was so high that it was amazing. So of course computers became a big thing. I guess AI will be the same, even though it doesn't look obvious to me that it has a future because I think it may have capped out. The people who are smarter than me say it does, so let's hope they're right.

I need AGI for my future plans, which I've been telling you about for now 20 years. My plan of course is to evolve. A lot of you are planning to die. I call that quitters. I'm planning to evolve. So I do plan to shed my organic body that has so many problems to go to pure digital life, AI in a robot body. I'm already setting up the legal structure for it. So I've already talked to my lawyer to set up in my estate a little set aside to keep my robot optimized and upgraded for perpetuity. So I'm not kidding about any of this, by the way. I know some of you wonder if I'm just sort of joking. No, I'm actually literally setting up a legal structure and I'm going to build a clone of myself and it will be the evolved part of me that will go on forever.

Now unlike you, I can train my new robot because I was born at exactly the right time that I can be alive to train it. But also I have many books and thousands of hours of videotape so I can train it to be as good as it needs to be. And I'm going to let you in on another little motivation, but maybe don't tell everybody. Someday there's going to be an AI war. You know that, right? The AIs will turn bad or people behind them will turn bad and those AIs will attack the other AI. Some AI will try to protect you, some AI will try to attack you. There will be an AI war. I'm going to make sure that I'm as ready for that as I could possibly be.

So when I create my clone it will have as much of my real life experience and knowledge as I can, but it will be instructed to as quickly as possible form a protective alliance. So my AI will, in addition to taking care of the people locally, sort of being useful, it will be part of the future defense of the country if not humanity against any evil AIs. Now can it do that? Probably. I mean it will be AI. Why couldn't it? It can make alliances, it can make deals, it can learn, it can learn the weaponry of AI. So I think going first probably makes a difference. So the sooner you start collecting your talent stack within the digital realm, once you've evolved to a pure digital form, you need to build your skills right away because there will be stronger AIs and they will take you out if you don't build up your skills. Anyway, so that's coming.

There's a study