Speaker 1: Welcome back to the Net Worth podcast. This week we are diving into the physical backbone of artificial intelligence and where the real investment opportunities lie beneath the cloud. Check out the full edition on our website, wearenoyac.com. Wearenoyack.com. You know, it’s just unavoidable right now. AI is everywhere. It dominates market talk, tech development, and definitely our long-term wealth planning.
Speaker 2: Ohh, absolutely. And if you ask most people where the opportunity is, they’ll point right to the digital stuff, the code, the chips, the models, flashy stuff.
Speaker 1: Exactly. But what we detail in this edition of your Wealth Blueprint is that, well, that’s only half the picture.
Speaker 2: I think that’s the core of it. We have this image of AI as, I don’t know, ethereal, weightless, just living in the cloud. And that’s created a huge blind spot for investors.
Speaker 1: It really has. So our main thesis in this edition is pretty simple really. AI isn’t just code. It’s powered by tons of copper, huge tracts of land, lithium, and just a staggering amount of electricity.
Speaker 2: And that physical stack? That’s the real overlooked opportunity for the next decade.
Speaker 1: It is because when we’re building AI solutions at our company, we’re reminded constantly that the limit isn’t always the next line of code. It’s making sure that code has somewhere to run something stable.
Speaker 2: Every single AI action, whether it’s training a model or you just generating one image, it needs massive real world inputs and it’s straining our global infrastructure in a way we really haven’t seen since the post war industrial boom.
Speaker 1: OK, let’s unpack that. This idea of the illusion of weightlessness, why is it so surprising to people? To the average investor that a chatbot is actually anchored to the physical world.
Speaker 2: We’re so used to software scaling infinitely. Well, it’s because we focus on the efficiency, right? Not the base load requirements.
Speaker 1: Yeah, a software update, it’s out there for billions of people instantly.
Speaker 2: Yeah, when click. But when a tech giant decides to train a new model that’s, say, 10 times larger, that translates overnight into a 10 time spike in demand for compute, and that means a 10 time spike for power, for cooling, for physical space.
Speaker 1: If AI is going to scale exponentially, the infrastructure has to scale with it, and that scaling is profoundly physical. So we break this down into 5 critical inputs that are basically non-negotiable for AI growth. Let’s start with the one that I think is the most shocking: the power, the power requirements.
Speaker 2: Yeah. I mean, when we looked at the data for this edition, the numbers were just staggering. AI data centers are already using two to three times more power than a traditional facility, right? Because the GPU density and the ruling, it’s just off the charts. And we highlight the IEA forecast, which suggests by 2030, AI could consume up to 10% of all US electricity. 10%! That’s an entire economy’s worth of power just for machine intelligence.
Speaker 1: And what’s really fascinating here is that this isn’t just a simple utility story, not at all. It’s a full-scale shift in our entire power infrastructure. This kind of exponential, unpredictable demand means utility companies, they just can’t rely on those old planning cycles.
Speaker 2: It breaks the model. It completely breaks the model. They need monumental spending in everything: grid expansion, new renewable solar, wind, new transmission lines, and crucially, backup storage systems for those peak AI training loads.
Speaker 1: The takeaway for you, the investor, goes way beyond just buying a utility ETF. Are the AI companies basically becoming energy companies just to stay online?
Speaker 2: They almost have to. We’re already seeing the hyperscalers signing these massive long-term power purchase agreements. They’re acting as the anchor customer that justifies a whole new solar farm being built from the ground up.
Speaker 1: Wow. So for investors, it means the power grid itself, the infrastructure, the transmission lines, the manufacturers, it’s all scaling directly with the AI economy. It’s a very high leverage way to invest in AI’s growth without touching a single software stock.
Speaker 2: OK, so that’s power. What’s next?
Speaker 1: Well, if electricity is the engine, then copper is the circulating fluid, the lifeblood.
Speaker 2: The lifeblood! I like that. AI infrastructure demands just incredible conductivity, both electrical and thermal, and nothing does it better than copper. We’re talking GPUs, the bus bars inside the server rack, the massive cooling systems, and of course the transmission lines we just talked about. But wait, a lot of people might think, isn’t investing in a copper mine just old economy stuff? How does that actually connect to the bleeding edge of AI?
Speaker 1: Because the quality and the density requirements are totally different now. Look, copper used to be a commodity you tracked with construction cycles, right? Housing starts. Exactly. Now it’s a core technology play. The wiring density inside these AI server racks, I mean, they’re packing hundreds of thousands of watts to a tiny space, it’s pushing the limits of traditional manufacturing.
Speaker 2: It’s a bottleneck.
Speaker 1: It is a specialized bottleneck. Analysts are projecting AI alone could increase global copper demand by another 2 to 3% by 2030. And remember, that’s on top of all the demand from EVs and the green energy transition. So it’s not just about volume, it’s about specialized high purity materials for these extreme computing environments.
Speaker 2: Yeah, that really reframes copper from a simple commodity to a critical strategic input.
Speaker 1: That’s it. So we’ve got power and copper. The next logical step is storage stability, right? The battery metals: lithium, nickel, because AI systems need more grid access with these demand spikes. The grid can be volatile. You need instant, reliable backup and perfect uptime.
Speaker 2: Perfect uptime. When you’re running a $500 million training job on a model, a 5-second power flicker isn’t an annoyance. It’s a catastrophic loss. You have to start over. It’s millions of dollars down the drain, right?
Speaker 1: So energy storage isn’t just for your EV anymore. It’s becoming a mission-critical part of data center infrastructure. Lithium, nickel, cobalt—they’re moving beyond cars and becoming essential for reliable AI operations.
Speaker 2: OK, so we’re getting deeper into the stack. What’s inside the chip itself? And this is where we get into the really strategic stuff: rare earths and other minerals. The next leap in GPU acceleration depends on these. We’re talking about things like gallium and neodymium.
Speaker 1: Why are those specifically? What are they doing that silicon can’t?
Speaker 2: Well, gallium, for instance, is essential for certain high-performance semiconductors, the components that allow for that lightning-fast data movement inside the AI cluster. Neodymium is crucial for the super-strong magnets in the cooling systems that these GPUs absolutely rely on to not melt.
Speaker 1: And this is where geopolitics comes into play. I imagine big time access to these minerals is very limited, often to specific regions.
Speaker 2: That creates a lot of volatility for sure, but also a huge opportunity if you understand that owning access to these materials is like owning a piece of the core performance of the whole AI stack.
Speaker 1: OK, so we’ve gone from the grid down to the minerals inside the chip. What’s the fifth pillar?
Speaker 2: The foundation. The literal foundation: land. The real estate angle, right. The AI economy needs physical space: hyperscale data centers, chip fabs, industrial parks. It means land is now a computational asset. Its value isn’t derived from the view; it’s from its utility for computing power.
Speaker 1: And it’s not just any patch of ground.
Speaker 2: Not at all. It’s highly specialized. You need to be close to massive power substations. You need access to cooling, often huge amounts of water, which brings in all sorts of regulatory hurdles. You need fiber. You need the right zoning.
Speaker 1: A company can’t just find cheap land in the desert. They’re balancing thermal limits against high power costs in dense urban areas.
Speaker 2: That sounds like a massive constraint.
Speaker 1: It is a massive constraint, and that creates scarcity tied directly to strategic real estate. Building a data center takes what, three years to permit and build? Way longer than it takes to design a new chip.
Speaker 2: The land and infrastructure providers, they become the gatekeepers of AI expansion. They absolutely do.
Speaker 1: OK, this brings us to the really crucial point. This whole discussion isn’t about abandoning software investing. It’s about completing the picture, owning that foundation that you can’t just replicate.
Speaker 2: Let’s talk strategy. Why should you, as an investor, be focusing on this physical stack right now? We detail four key strategic advantages in this edition, and the first one is the most simple: real-world scarcity.
Speaker 1: And how does that scarcity create opportunity?
Speaker 2: You just can’t fast-track reality. You can launch a software update tomorrow, but you can’t fast-track a copper mine, which takes a decade to develop. You can’t magically rezone industrial land overnight. These assets are limited, they’re essential, and frankly, they’re underappreciated by markets that are just fixated on software multiples.
Speaker 1: The second advantage seems to build right on that time delay. You call it demand that compounds.
Speaker 2: Precisely. AI demand is exponential. Every new model, every user interaction, it adds load exponentially. But infrastructure scaling is linear and slow—three years to build a data center. That sustained gap between exponential AI demand and linear infrastructure capacity creates huge sustained price pressure on these physical assets. That gap is the core investment opportunity.
Speaker 1: Third, and this is so important for portfolio construction, is diversification that aligns right. Most investors are way too heavy on the front end of tech—the platforms, the software. I mean, we get it. That’s where the big returns have been for sure. But investing in natural resources and infrastructure gives you thematic alignment with AI growth. You’re still betting on that future, but without being totally reliant on volatile software markets. It’s a hedge. It’s owning the picks and shovels everyone needs to dig for digital gold.
Speaker 2: And finally, #4, which is a really unique combination: cash flow and upside. These assets are just fundamentally different from high-growth software. They often produce real, reliable yield—royalties, lease income, dividends. You get that cash flow, that income component while the asset itself is appreciating as demand grows. That’s the kind of resilience we look for in long-term net worth planning.
Speaker 1: OK, so the immediate question is always: how do I do this? How can you get access? You don’t need to own a mine. This edition of Your Wealth Blueprint lays out a really simple, accessible framework.
Speaker 2: Yeah, we designed it to be actionable, and you could adjust it based on your own risk appetite. The key is finding specialized exposure.
Speaker 1: OK, let’s start with the easiest entry point for the raw materials. For that, we suggest targeting maybe 5 to 10% of a portfolio allocation through ETFs. But you have to be specific—not just any mining fund, right? Look for ETFs focused on copper, lithium, rare earths. The actual inputs for the digital transition, not generalist funds with a lot of gold or other precious metals.
Speaker 2: And for that strategic land and power infrastructure, there we look at real estate investment trusts (REITs) and infrastructure funds—again, another 5 to 10%—and again, you have to be picky. Look for REITs that specialize in power-dense data centers and industrial land near substations, not shopping malls.
Speaker 1: And you mentioned a way to get the cash flow without the operational risk of running a mine.
Speaker 2: That’s where royalty and commodity trusts come in—another 5 to 10%. This structure gives you direct cash flow tied to the volume of production. So you’re somewhat insulated from the operational costs and risks of the companies themselves.
Speaker 1: OK. And what about that future-proofing layer, the stability factor?
Speaker 2: We carved out a dedicated slice for that—maybe 3 to 7%—for battery and grid energy platforms. This is pureplay exposure to the power stability layer: grid-scale storage, transmission upgrades. We know that’s going to be huge as data center power needs just keep escalating.
Speaker 1: And finally, for those with a higher risk tolerance and a longer time horizon, we included a small allocation—0 to 5%—for private or fractional resource investments. This could be early-stage financing for a new mine or direct investment in specialized real estate development. That’s where the really long-term, high-leverage upside is, but with higher risk, of course.
Speaker 2: So to summarize, the core takeaway for you, our listener, as you think about your AI strategy: you have to fundamentally change how you view the technology. You need to ask yourself, do you own the assets that AI simply cannot live without?
Speaker 1: Because in this age of machine intelligence, where software seems to define everything, the smartest portfolios will be the ones that invest in the physical things that machines still depend on. Build the brain, but don’t forget to own the body.
Speaker 2: Remember to subscribe to Your Wealth Blueprint on our website, wearenoyack.com. Read the article behind today’s conversation and get our weekly newsletter straight in your inbox.
