t's an open casket you're like, what are we going to do about that? Because I think the rigor mortis sets in about the same time as the erectile dysfunction. So you could die with a wall saluting and then they probably just have to make some adjustments. Now depending on your size I guess they'd make some adjustments to the casket top, make one a little extra deep sort of thing. I don't know. I'm just thinking through all the possibilities. I don't think it's a big risk but you know it's good to make sure you've covered all your bases.

The question I ask is why did it take so long for the FDA to approve this? I think we've known for a long time this was a good play. I knew it. I'm not in the FDA but yeah I mean I suppose there were some testing and they were waiting for some data. I don't know. It feels like too much of an emergency situation to have not done something about it earlier. So I can't be happy about the speed of that but at least it happened.

And by the way I would like to congratulate all of you because I do think that public opinion has some influence on this, don't you? And certainly you know my audience has been pushing for it as I have for two years. How long have we been pushing for this? Two years. And you know I'm not going to say it wouldn't have happened anyway because it's just a good idea. But I do feel like if you're putting pressure on something that ends up happening you can claim that you were part of the productive process. I think we were. I don't know that it made a difference but we were certainly pushing in the right direction and you should be proud of that. Because I think that proves you were on the right path and you were advocating for the right stuff. So good for you.

Elon Musk and Steve Wozniak and a bunch of other tech leaders have signed a petition I guess or a letter asking for AI development to pause for six months. Pause immediately for at least six months. The training of AI to not make anything more powerful than the current version. And it's because of the risks to society. What do you think of that? Do you think it's a good idea to pause AI development until we have a robust monitoring, you know some kind of a human system for maintaining it? Yeah, yeah I'm gonna say yes unambiguously to that. There are very few things which I can say I agree 100 percent. This is one of them. This is one of them.

And here's the argument. The argument goes like this. Pausing for six months probably won't hurt us. Would you agree? Pausing for six months probably won't hurt us but it could save civilization. Now if you say to yourself, Scott it's going to get out anyway, there's nothing you can do, you could try to control it but it's too late. I agree. I agree that there's a really good risk or high risk that it's too late. Some people said well I can already download it onto my local device. I could keep training it. There's nothing you can do.

But I would point out, do you remember when the internet was new and all the people who were in charge of copyrights and books and stuff, they all said oh no, publishing will go away because everybody will just steal it and put it on the internet. So there's no way to have a business model when the moment you make something it's just stolen and on the internet. Now that's the current situation. The current situation is all of my work gets stolen and put on the internet immediately, almost the moment it's created. But it never hurts because the law was still robust and it's still illegal to routinely put my work on the internet. Lawyers would come get you.

So I would say you should not underestimate how comprehensive the law could be if the law tried to make AI illegal. I think it could do it. I think it could. That doesn't mean people wouldn't cheat. But if you're cheating sort of locally, if you're not Google and you're not some giant AI company, you're not going to do too much in your bedroom with your local computer. I think you only have to worry about the big entities and the big entities can be completely controlled. Google is not going to violate the law. They have too much at risk.

So I think you underestimate how effectively the law could shut down AI. I think you're seeing the beginning of a movement that's going to be chipping at AI from every direction until it's almost completely neutered. So my guess is that AI is far too dangerous for the right reasons and the wrong reasons. The right reasons that AI is dangerous is it might teach us what's real. If America ever found out what was real the whole system would collapse because our system depends on people believing stuff that isn't true. AI could change that. So our whole system could be at risk just from the truth. Then that's just one risk. If you imagine all the other worst case risks they're pretty amazing.

Now I am opposed to stopping AI permanently. I think it's necessary. I think you don't want China to have it before we do. But I do think it makes sense to be really, really serious about safety. I think the net effect of that is AI might have a military use and almost nothing else because we'll neuter it in so many ways.

Now what happens when AI starts taking jobs? Don't you think there will be legislation to make it illegal to replace a human with a machine? Now we've been replacing humans with machines for 100 years or more. But once they can replace all the humans I think we're gonna have to rethink it.

So here's what I think is going to happen in the medium term. I think there are too many humans who are incapable of taking care of themselves, more so than ever probably. Because in the old days you could at least be a farmer or work on a farm. But today there's a huge segment of the population that really isn't capable of doing anything in the modern world. They don't have any capabilities whatsoever. They're a liability. So I think we're going to have to create two civilizations. One who lives on some kind of UBI and maybe the robots do the hard stuff for them. And they just take them away from everybody else because there'll be crime-ridden and low bad mindsets and it'll be every problem in the world. But they won't