g to New Atlas, there's a company that's asking for some kind of government approval that I believe they will get to take the type of nuclear reactors that are already in naval ships and have been operating for 70 years without trouble and to use that design for domestic energy production.

Now I don't know if you remember this, maybe I started 10 years ago talking about how nuclear should be bigger and should be more of a focus. And one of the things that Mark Schneider taught me at the time was that we already had a design that was used in the military, the Navy especially, and it didn't have problems and you could build them small and they would be powering battleships and stuff like that.

Now I think the first part of the request, and it's really sort of a two-parter, is that the company wants to take the existing nuclear reactors that are on ships but only the ones that are being decommissioned. So if they're being decommissioned anyway, you know, you don't want to waste a perfectly good nuclear reactor, right? So they want to take those and presumably modify them and stuff but use them. They would take them off the ship so they wouldn't be using them on the ship. They would take them off the ship and repurpose them.

But here's the good part. They also want to build new ones because there wouldn't be enough decommissioned ships to make much of a dent. So they want to take the design that's been proven over 70 years and it would cost about somewhere in the $2 billion range to create a modular reactor versus what we see with the big reactors which could be tens of billions of dollars. So it's smart, it's well proven, and it's economical. And I think they only need approval from the Trump administration to do this sort of thing. So again, optimistic.

Speaking of optimism, one of the products at the Consumer Electronics Show is a leaf blower, an aerospace-powered quiet leaf blower that cuts noise by 70%. Do you know what a plague the leaf blowers are in high-end neighborhoods? I hate to be like a rich person complaining, but there's at least one to two days every week where it becomes impossible to take a nap or anything because either your own gardener is right outside or your neighbor's gardener is right outside and it's so freaking loud.

Now you might say, "I'll bet that's expensive and I'll bet your gardener is not going to want to pay for that." And then I thought to myself, I'll buy it. If my gardener was willing to take this product and it worked, yeah, I don't know. I'm not sure it's for sale. It might be just announcing that it will be. I would immediately go to my gardener and I would say, I will buy this for you if you'll use it. I assume he'd say yes because it doesn't give up anything in performance. Then I would figure out who the other gardeners were, like my neighbors' gardeners, and I would say, you know, you should do the same thing. Just buy one for your gardener and you'll make all the neighbors happy. If they said no, then I would say, at least the immediate neighbors, I would say, well, I'll buy it. You know, I have some extra cash, so let me buy your gardener one of these.

So I wonder if the way it will spread is that the homeowners will buy it for their gardeners even though that's not their responsibility really.

Well, according to science, Ashish Gupta is writing that Stanford's doing a new approach to AI that solves the following problem. Have you ever wondered why AI can create a video of somebody doing something that looks exactly like a person doing something, but if you try to tell the AI to do exactly what the AI is showing in a picture, it can't do it? And the reason is that the videos are created by just predicting where pixels should be on the screen. But what they'd like to do is use the AI's imagination, they call it, where it can create a picture of somebody doing something and tie that to what the AI learns by having it learn by its own dreams.

So before it tries to do something, let's say you want your AI robot, so before it does anything, it first imagines it in pixels and then the part they're working on that I guess they're getting close is to figure out how the AI can learn from its own pixel pattern. So if it created a dream basically, which would just be an AI video, if it created one that was folding a certain kind of laundry, can you say, "All right, learn from your own picture how to do this." So the physics got modeled right. So the idea is to add the physics to what it can already imagine. But most of the AI has passed the six-finger problem. Now the AI is so much better. So maybe that's a big deal.

Well, also talking about my silly state, according to U-Haul, more people are leaving California than any other state for the sixth year in a row. Holy shoot. We have the highest state income tax in America and lots of other problems.

Let's talk about Greenland. Things are changing in Greenland. So Steve Miller was on a show