Jevons paradox is propping up the entire global economy

So the entire global economy is being propped up by Jevons Paradox. You may have heard of it as this obscure kind of economic paradox, but it's having very material implications right now for all of us, for countries, for entire economies, and for the world of AI. This video is part of an AI Fundamentals series where we're covering a lot of just the base ideas and dynamics that matter to this world and this evolution of AI and the associated technology. So in this video, we're going to cover what exactly I mean by that statement. And that will help you hopefully understand AI a little bit better and make better decisions in terms of your career and investments. That's going to be the video going to be a fun one. Let's go. Speaking of tools that actually save you time, This one is different from anything we've covered on this channel because it's not an app or a website. It's a pair of glasses. Now, this part of the video is sponsored by Rayneo, and these are the Rayneo IO glasses. And the first thing that I think is important to know about these is that they actually look like glasses. And a lot of smart glasses still look like these tech prototypes. They're something you would demo once, but you wouldn't actually wear on a day-to-day basis. These, meanwhile, are 33 grams with a 50-50 weight balance. And for sake of comparison, 33 grams is the weight of six quarters or 11 pennies, so very light. A magnesium aluminum frame and a titanium template. So if I didn't tell you that these things were smart glasses, you probably wouldn't be able to tell. And trust me, that is a good thing. And the concept with them is that the average person checks their phone over 100 times a day, and each time you do that, you're distracting yourself from what you're actually doing. And with these, instead, you just look up. There's an ultra-transparent waveguide display that's built into the lens, powered by Rayneo's Firefly nano-optical engine, and it projects a personal dashboard directly into your line of sight. That means time, calendar, reminders, notification, all of that information comes to you and your phone stays in your pocket. And then there's the AI layer on top of that. The voice memo records, transcribes, and summarizes conversations. It uses a hybrid mic system, a microphone array, plus bone conduction. And so it isolates your voice and keeps transcriptions accurate. So you walk out of a meeting, your notes and everything is already handled by default. And then live captions gives you real life captions for conversations. And then for my personal favorite as a creator, they have a teleprompter that's built in with AI. So it follows your speaking pace automatically. And if you want to record something like this, I can just be looking, wearing the glasses at the screen and know exactly what to say. There's also what's called LifeBlog running in the background. and that captures your ideas and builds this kind of daily summary of your day and then for privacy there is an indicator light that turns on whenever something is recording or translation is active so people around you do know now as far as the practical details battery gets you 63 hours of power on one charge it's IP54 water and dust resistant so as far as that would you actually wear them past this does check that box so if you want all the key information in your digital world in your line of sight instead of on your phone, you can check out the Rayneo IO glasses at the link in the description or on this website on the screen. So check them out. And that said, let's get back. So Jevons Paradox is this economic concept you can observe sometimes which says, hey, if cost goes down, maybe demand will actually go up. And typically, economically, it works the opposite way of that, right? If there's more demand for something, what that's going to do is make the price go up if we assume that supply is constant. That makes a lot of logical sense, right? If the line is longer for something, if more people want it, they're going to be willing to pay more for it. But what this is saying is the opposite. It's saying, okay, this thing is becoming a lot cheaper, and yet we're also seeing a lot more demand for it. And you might already start to be seeing the parallels here to the world of AI. But first, I want to explain how I get from AI to affecting the global economy. And so the first connection there is the United States economy as affecting the world. And basically, the US economy is what the entire world hinges on, right? You can look at what happened to Europe following the 2000s and the 08 crisis and so on in the US. You can look at the dependence that China has for its export market on the US in a lot of ways, and just in terms of, for example, the stock market capitalization globally and the distribution of that allocated to the U.S., the world needs the U.S. If the U.S. enters a period of intense financial pain, it is going to cascade across the world. And so then we ask, okay, well, what is the U.S. economy reliant on if the global economy is in many ways reliant on the U.S.? And what the U.S. is reliant upon currently is artificial intelligence as a sector. There's all these stats coming out now that basically we would be in a recession without AI. And there's all these stats coming out now that we would be in a recession without AI. And if you strip out AI companies specifically in the U.S. stock market or the indexes versus the rest of the market, it's a whole different world. Basically, the whole U.S. economy is being propped up by artificial intelligence and materially by a small number of very valuable companies, so things like NVIDIA, your AMD, your OpenAI, your Anthropic, or by the spending of companies that are invested in something related to AI, namely the hyperscalers and something like half a trillion dollars plus a year of CapEx being invested into physical infrastructure. for AI, which at this point is the largest infrastructure investment ever adjusted for GDP in the history of the United States. So more than the railroads, the World War II, etc. So we're making a few connections here, right? We're saying, okay, the world economy in some ways is reliant or hinges upon the United States economy. The United States economy hinges upon the performance of the artificial intelligence sector, at least over the next few years. So then what is the underlying bet in that space, right? You've heard all this talk of, oh, is AI a bubble, right? So what's the bet that determines whether all of that investment and performance for the artificial intelligence sector and the market cap of these companies is justified? Well, what it is at a very high kind of fundamental level is Jevons paradox. And that's how all of this ties back to this concept, which is demand is going to keep going up even as costs fall and let me give you an example of what happens if we assume this paradox is false and then how that whole system could unwind. So let's say open source models actually become incredibly good they're just as good as the frontier layer so the closed source models and they're basically free to produce and you don't need all this expensive hardware GPUs to run them you can just run them on the CPUs on your computer. What we're basically saying there is, hey, your data centers and your GPUs, all of that doesn't really matter to have really good AI. If that happens, what would happen pretty much overnight, if that was a certainty, and it wouldn't actually play out like this, but just for the sake of the thought experiment, NVIDIA's market cap would crash, AMD's market cap would crash, the multi-trillions of dollars of OpenAI and Anthropic equity would just disappear effectively overnight. That's not completely true because they have distribution, but just a lot of it would go away. And so you'd have to unwind this trillion dollar plus a year in spending, this 10 plus trillion dollars worth of equity values. What that would do overnight is crush the U.S. economy, would put us in a recessionary state. And if the U.S. enters into a recessionary state, then that's how you get this kind of cascade effect across the world. That might not be a long term thing, but just for, you know, three months, six months, a year, two years, however long, you'd be in this period of fiscal pain around the world. And all of that would be a result of just violating this assumption of, okay, if costs come down, will demand continue to increase? Because what you're saying there is as cost comes down, demand doesn't increase from the perspective of the traditional AI companies. So the closed source labs and the companies which produce those closed source models. So to say that again, then the bets that Anthropic, OpenAI, NVIDIA, the hyperscalers, so Google, Amazon, Microsoft, et cetera, are making is that even though we know that all the inputs to AI are getting cheaper, right? The models are becoming more efficient. The distillation is becoming better. GPUs are becoming better, more powerful across the board. even though all of those things are true, you still think that the demand for frontier models is going to massively increase. So even as the costs drop, demand is still going to increase. That is the paradox, and that is the bet of this group of companies. And I saw this, I think, tweet recently, and it was like, either Anthropic and OpenAI are either going to be worth $20 trillion dollars or zero, right? And that's a pretty fair, not literally, but a fair materialization of this concept, which is either you do still get demand going through the roof for these models and for things like GPUs, in which case the market caps are more than justified, right? These companies should all be worth a lot more. They'll be doing trillions of dollars a year in revenue. Or, hey, open models are actually incredibly good. That's where all of the demand goes. And then these companies are extremely overvalued and AI may not be a bubble, but the valuation and earnings of those companies were a bubble. That's going to be a very important dynamic that plays out over the next five, 10 plus years. And it is concept, which is affecting in large part how the market cap and earnings of these companies move right now. So we're going to take this idea that we're discussing here which is very abstract and that's going to be what moves the valuation of these companies up or down by hundreds of billions of dollars if not trillions there you go that's a little spiel there and this is part of that AI fundamentals uh series there are a lot of other videos in the series so if this interests you then you can check out the remainder of the videos in the series with that in mind appreciate you being here hope you found it interesting