NVIDIA Can't Fix This Problem

During several interviews in 2024 and 2025, Sam Altman said something that should have stopped every investor in their tracks.

He said: "The ultimate bottleneck for AI compute isn't chips, it's electrons."

The man running the most important AI company on the planet just told you the constraint isn't software. It isn't semiconductors. It isn't data.

It's electricity. The wires, transformers, and substations that carry it from where it's made to where it's needed.

While the world chased the next AI ticker, fought over who makes the best chip, and argued about which large language model would "win", the real war was being fought underground.

In copper cables. In transformer yards. In substations built before my parents were born.

The grid, that invisible, taken-for-granted skeleton of modern civilization, is breaking.

And the companies building, upgrading, and reinforcing it are quietly printing money while everyone else stares at their screens waiting for the next NVIDIA earnings call.

Let me show you what I mean.

Every AI Query Has a Power Bill

Here's what the AI hype machine doesn't want you to think about:

Every query you send to ChatGPT, every image Midjourney generates, every autonomous vehicle calculating its next turn… all of it requires electricity.

Not a little. A staggering amount.

A single hyperscale data center consumes as much power as a small city.

Right now, Microsoft, Amazon, Google, and Meta are racing to build hundreds of them.

But over 3,000 gigawatts of power projects sit in a queue worldwide, waiting on hold to connect to the grid because the grid can't absorb them fast enough.

The time to connect a new power source to the grid now takes four to ten years in most countries. You can actually build a power plant faster than you can plug it in.

Six U.S. states have officially proposed pausing the construction of AI data centers.

The reason? Grid strain.

Even President Trump acknowledged it publicly: the U.S. grid is "old and tired." He's not wrong.

Over 50% of America's power transformers were installed more than 40 years ago and are past their intended useful life.

Seventy percent of transmission lines are over 25 years old. Most of the developed world's grid infrastructure runs on designs from the 1950s.

This isn't a future problem. It's a right-now problem. And it's the single biggest chokepoint standing between the AI revolution and reality.

I've spent the last several weeks going deep on this theme, and I mean deep.

What I've found has convinced me to add two new positions to the Moonshot Minute Portfolio. More on that below.

$15.8 Trillion. That's the Bill.

Bloomberg New Energy Finance recently updated its estimate for total global grid investment needed by 2050.

The number: $15.8 trillion. Up from $13.2 trillion just a year earlier.

The International Energy Agency says grid investment needs to increase by roughly 50% annually through 2030 just to keep pace with rising demand.

But in 2025, global grid spending hit only $483 billion. A fraction of what was spent on EVs ($893 billion) or renewables ($690 billion).

Everyone's funding the shiny stuff. Almost no one's funding the thing that makes the shiny stuff actually work.

By 2050, approximately two-thirds of the world's power grid will need to be replaced. Not upgraded. Replaced.

Developed economies need to expand grid capacity by at least 50%. Developing economies? By 150%.

The total length of transmission and distribution infrastructure worldwide needs to double, from 80 million kilometers to 166 million kilometers.

This is the largest non-discretionary infrastructure buildout in human history.

Governments don't get to choose whether to fund it. They have to. The alternative is economic stagnation, technological irrelevance, and national security failure.

I grew up watching my parents, immigrants who came here with nothing, build something from scratch. Not with fancy tools. With grit and whatever was available.

That's what this grid story reminds me of. The foundation isn't glamorous. Nobody brags about it at dinner. But without it, nothing else stands.

The Bottleneck Behind the Bottleneck

It gets worse.

The equipment needed to modernize the grid is in critically short supply. Demand for grid components in the U.S. has surged 35% to 274% since 2019, according to Wood Mackenzie.

Lead times for large power transformers, the backbone of grid infrastructure, now stretch two to three years.

And here's the kicker: China controls the supply chain.

China produces roughly 75% of the world's capacitor-grade dielectric film, a material inside every substation, every renewable-energy inverter, every battery storage system, and every data center backup unit.

No U.S. manufacturer currently makes the equipment to produce it.

China dominates the supply of grain-oriented electrical steel, which accounts for 25% to 50% of the cost of a large power transformer.

China controls the supply of refined gallium, an essential material for power semiconductors used in smart grid systems.

China is building dominance in silicon carbide wafers, which reduce power losses by 50% to 70% in high-voltage applications.

And China accounts for over 55% of global primary aluminum production. The material that makes up the overhead conductors carrying electricity across every transmission line you've ever driven under.

Not only are we behind on building the grid, but we’re also dependent on our biggest strategic competitor for the materials to build it.

When I tell you the two companies I'm adding to the portfolio this coming Wednesday have direct exposure to this exact dynamic, the collision of surging demand, constrained supply, and multi-decade government spending, I hope you understand why I'm this excited.

I don't add positions lightly. These earned their way in.

The Stocks Nobody's Watching Are Crushing Everything

Here's where it gets interesting for us, the people actually trying to build wealth in the real world.

While the S&P 500 has returned a modest 1.5% year-to-date, a popular grid infrastructure index, a basket of global companies that manufacture grid equipment, high-capacity transmission cables, and smart-grid software, has returned over 23% YTD.

Grid stocks are outperforming the S&P 500 by more than 15x this year. Since its inception in 2021, the index has returned 256.1% versus 74.1% for the S&P 500.

This shows that capital is flowing into companies, addressing a problem that cannot be ignored. Governments worldwide, from Washington to Beijing to Brussels, are being forced to spend.

I don’t need to tell you what happens when governments are forced to spend.

The U.S.-China tech arms race makes grid reliability a matter of national security. The AI buildout makes it a matter of economic survival.

The companies building this infrastructure, the ones making transformers, cables, smart-grid sensors, digital twins, and grid-edge intelligence systems, sit at the center of a multi-decade, multi-trillion-dollar spending cycle.

The world is looking left. The money is moving right.

The Smart Money Already Sees It

Institutions aren't waiting for CNBC to run a segment on transformer lead times. They're already positioned. Here's what they see:

Digital twins, virtual replicas of entire grid systems, reduce infrastructure planning time by 70%.

Utilities already use them to simulate AI-driven load surges before they happen.

The New York Power Authority built a digital twin of the entire state grid to test cyberattack scenarios and validate new technologies before a single dollar is spent in the field.

Virtual power plants coordinate thousands of distributed energy assets via software, serving as dispatchable resources that respond within minutes. In California, aggregators delivered roughly 539 megawatts during evening peaks by coordinating over 100,000 residential batteries.

That's the future of grid management, and it’s happening now.

Grid digitization alone is expected to account for $2.2 trillion in investment by 2050, 13% of total grid capex.

Technologies like the TRUSense Gateway convert legacy meters into grid sensors, enabling real-time monitoring of EV and data-center surges without ripping out existing infrastructure.

The companies enabling all of this?

Most still fly under the radar of mainstream financial media. No household names. No Super Bowl commercials.

But they're building the backbone of the next century of economic growth, and the performance numbers already prove it.

Two of those companies will enter the Moonshot Minute Portfolio this coming Wednesday.

One Question You Should Ask Today

I'm not going to pretend this is simple. Grid infrastructure isn't as sexy as buying NVIDIA or betting on the next AI chatbot.

It doesn't make for great cocktail party conversation. Though if your cocktail parties are anything like mine, the bar is already pretty low.

But here's what I've learned after years of doing this: the biggest gains come from the things nobody wants to talk about.

Look at your portfolio and ask yourself… do I own any of the physical infrastructure that makes AI possible?

Not the software. Not the apps. Not the models. The infrastructure. The wires. The transformers. The cables. The metals. The grid.

If the answer is no, you're betting on a revolution while ignoring the foundation it's built on. That's like investing in the internet in 1998 without owning a single telecom stock.

The grid upgrade isn't an optional, “nice-to-have” thing. It's inevitable. It’s a must. And the market is only beginning to price it in.

The electrons have to flow. The question is whether you'll be positioned when the world finally figures that out.

This Coming Wednesday, I'm Dropping Two New Grid Plays Into the Portfolio

I don't often tease new positions. When I do, it's because I believe the setup is exceptional.

After weeks of research, digging through earnings calls, supply chain data, backlog numbers, and government spending commitments…

I've identified two companies that sit directly in the path of this $15.8 trillion wave.

One of them has a backlog so large it has visibility stretching years into the future.

The other is solving a supply bottleneck so severe that customers are paying upfront just to reserve a spot in the production line.

These aren't speculative moonshots.

These are companies with real revenue, real earnings growth, and a structural tailwind that doesn't depend on whether the next AI model is better or worse than the last one.

The grid has to be built regardless. These two are building it.

Next Wednesday, paid subscribers to the Moonshot Minute will get the full breakdown — the tickers, my entry strategy, the price levels I'm watching, and exactly how these fit into the broader portfolio alongside our metals and uranium positions.

If you're already a paid subscriber, you don't need to do anything. Just watch your inbox on Wednesday.

If you're not, this is the week to fix that. I'm not going to pretend there will never be another good idea. There will.

But I can tell you that the amount of work behind these two picks is substantial, the thesis is airtight, and I want you in before the names go out, not after, when the edge is already smaller.

Join Moonshot Minute Premium before Wednesday, and you'll have the recommendations the moment they drop.

The foundation always gets built before the building goes up.

Most people only notice the building. The money is in the foundation.

Double D

🔓 Premium Content Begins Here 🔒