Now let's talk about the bubble problem. Tracy, you guys have talked about this a lot with different people. Make the best case you can for me against bubble ideas and beyond. Oh, hey. So the idea of bubbles is very simple. This idea that I talked about earlier is basically a winner-take-all strategy, and if everyone develops the product they say they're going to build, and they develop an AI model or system that magically solves every business and personal problem in the world, then maybe they can justify some of those valuations. When magic happens, it's not a bubble. That's correct. That's correct. And that's what a lot of these companies are dealing with. They promise magic. That's the way they talk. So I think there's a concern that as AI becomes a more dominant force in the U.S. economy, it will potentially have an economic impact that feeds on itself if the bubble bursts, or even if the promised income and savings don't materialize on the scale that people think. This is similar to what we saw, again, less pessimistic, but similar to what we saw just before the Great Financial Crisis. Housing has become a very important driver of U.S. economic growth. Everyone was buying houses and building houses. We have seen the share of home construction in the U.S. economy rise, but eventually home construction became too large to become a source of broader problems in the U.S. economy. It wasn't always like that. In the past, it was common for housing to take a hit when the U.S. economy was in trouble. What happened was that houses became so large that they became the direct cause of problems in the entire American economy. And the fear now is that we could be heading down the same path as AI. So I showed you a graph of the cyclicality of many of these businesses. I always think about the “It's Sunny in Philadelphia” meme of a guy standing in front of a board with a red thread connecting everyone. Check this out. Look at this. That's exactly what it feels like when you start unraveling these relationships. But another concern is the current lack of transparency around how AI is actually financed. There's a lot going on in the private credit market. We don't know — tell us what the private credit market is. True, in private credit markets, companies sometimes receive loans from banks, but most often they receive loans from other types of investors. And these loans and bonds are not publicly issued or publicly traded. So typically when something like IBM or Microsoft issues a bond, it comes with a prospectus and there's a lot of information about it available online. Once people know the terms, they will trade. Anyone can purchase. You will be replacing it later. Private credit is even more bespoke. A customized loan between a company and an investor. For obvious reasons, it's very difficult to gain much insight into that particular market. The clue is in the name. It's all private. So when it comes to funding, I think it's very difficult not only to grasp the scale of what's going on, but also to grasp who's actually funding what. Sometimes you hear big investors, like big private credit investors like Apollo, saying things like, “Oh, we're crazy about data centers right now, but it's hard to figure out how much of an impact it is.''
