Yesterday, I introduced what I believe are the five layers of the AI revolution that all investors must learn.
Most investors think of AI as a single trend. I see it as an entire ecosystem, with opportunities stretching far beyond the companies making headlines.
Today, I want to start with what I believe is the most important layer of all… The foundation.
Because before AI can write a sentence, diagnose a disease, drive a car, or power a humanoid robot, it needs one thing above everything else:
Infrastructure.
The Biggest Buildout Since the Internet
Think of AI as building the world’s largest city.
Everyone gets excited about the skyscrapers. But before those buildings can go up, someone has to pour the concrete, build the roads, connect the power grid, install the plumbing, and lay the fiber-optic cables.
That’s exactly where we are today with artificial intelligence.
The world is in the middle of the largest infrastructure buildout since the internet.
Every major technology company is spending at levels we’ve never seen before. Microsoft, Amazon, Meta, and Alphabet are committing hundreds of billions of dollars to AI infrastructure over the next several years. Those aren’t expenses - they’re investments in what they believe will become the backbone of the global economy.
The Chip Winners Beyond Nvidia
Every time someone asks ChatGPT a question, generates an image, or uses an AI agent, enormous amounts of computing power are required behind the scenes. AI workloads consume many times more processing power than traditional internet searches, which means demand for computing infrastructure continues to grow at an astonishing pace.
That creates opportunities throughout the supply chain.
Everyone knows Nvidia (NVDA) has become the poster child for AI chips. But GPUs are only one piece of the puzzle.
Companies like Broadcom (AVGO) are building the networking chips that move massive amounts of data between AI servers.
Marvell Technology (MRVL) is helping solve the high-speed connectivity bottlenecks inside next-generation data centers.
Arista Networks (ANET) supplies the networking equipment that allows thousands of AI servers to communicate with one another almost instantaneously.
Astera Labs (ALAB) develops the connectivity solutions that help all of these incredibly powerful chips work together efficiently.
AI Is an Electricity Story
Then comes the next challenge.
Power.
The average data center already consumes enormous amounts of electricity. AI data centers require dramatically more.
That’s why I’ve been pounding the table on what I call the Electricity of Everything.
Artificial intelligence isn’t just creating demand for better software.
It’s creating demand for more electricity than we’ve needed in decades.
That benefits companies like GE Vernova (GEV), which provides critical power generation equipment. It benefits electrical infrastructure leaders like Eaton (ETN), whose products help distribute and manage electricity safely and efficiently. It benefits companies involved in transformers, switchgear, grid modernization, and even nuclear energy.
In other words, AI isn’t just an AI story.
It’s an electricity story.
Cooling Is the Next Bottleneck
Then there’s heat.
As these processors become more powerful, they generate tremendous amounts of heat. Traditional air cooling is reaching its limits.
That’s creating another investment opportunity.
Companies like Vertiv (VRT) have become leaders in advanced cooling systems, liquid cooling technology, and data center infrastructure designed specifically for AI workloads.
Without efficient cooling, these multi-billion-dollar AI facilities simply cannot operate at scale.
That’s why I often remind investors that the greatest fortunes during technological revolutions aren’t always made by the companies everyone talks about.
Sometimes they’re made by the businesses quietly selling the picks and shovels.
During the California Gold Rush, most prospectors never struck it rich… The merchants selling the tools often did.
We’re seeing something remarkably similar unfold today.
Bottom Line
Every new AI model. Every AI assistant. Every robot. Every autonomous vehicle. Every medical breakthrough powered by artificial intelligence begins with this first layer.
The infrastructure.
This is why I remain incredibly bullish on AI - not because of one company, but because an entire industrial ecosystem is being built around it.
And here’s the part that excites me the most…
We’re still early.
The infrastructure being built today isn’t designed for the AI demand we have now.
It’s being built for the AI demand companies expect to have five and ten years from today.
That tells you just how big they believe this opportunity will become.
Tomorrow, we’ll move to the second layer of the AI revolution - the companies building the intelligence itself. From large language models to AI agents capable of performing increasingly complex tasks, we’re entering an era where software doesn’t just respond to commands…
It begins to think, reason, and work alongside us.
And I believe that may transform knowledge work forever.
Here’s to the future,
Matt McCall
Founder, NXT Wave Research
