Legendary investor Peter Lynch said, “The best way to find great investments is to look around and see what’s changing in the way people live and work.”

When running Fidelity’s Magellan Fund, he bought Walmart (WMT)—a 30-bagger—because he saw new stores sprouting up everywhere. He first invested in Dunkin’ Donuts because he read about it in a newspaper, liked its coffee, and saw stores being built around Boston.

What would Lynch see today if he were investing professionally?

Artificial intelligence (AI).

• What I love about AI is that, unlike most frontier technologies, you can get your hands dirty with it.

You don’t have to read analyst reports about ChatGPT. The world’s best AI systems are only a few clicks away and cost $20/month.

I use ChatGPT, Google’s (GOOGL) Gemini, and xAI’s Grok all day, every day.

They’re my editor, research assistant, and graphic designer all rolled into one. I laugh at the fact that I used to spend countless hours sifting through Google Search looking for answers.

I literally can’t imagine working without AI now. On a recent 11-hour flight to San Francisco, I was without my trusty AI sidekicks. For the first time ever, I thought about paying for internet on a flight.

I even have my six- and four-year-old kids using ChatGPT’s Advanced Voice Mode. They call it “the lady on the phone.” My daughter Aubrey asks it all kinds of random questions, like, “How far can mice see?”

• AI usage is going vertical.

Almost 800 million people use ChatGPT each month. Remarkable given it launched less than three years ago:

AI startups are growing faster than any other group in history. I’m incredibly impressed by this new generation of AI companies. They look more like the early days of Google than they do the last decade of tech companies.

Unlike the software-as-a-service cash incinerators, they’re extremely lean, profitable, and growing quickly. Take Cursor, for example. The AI startup is raking in $300 million/year with less than 50 employees.

The real beauty of AI is that it’s a general-purpose technology. You can use ChatGPT to plan a vacation or to homeschool your kids.

What sets AI apart is how fast it’s spreading compared to past general-purpose technologies. Electricity was discovered in the 1800s, but it took nearly a century before it powered most homes and factories. It took 150 years before steam power became widespread in factories and railroads! AI is being integrated into everyday tools in a matter of months.

• I think AI is the greatest productivity tool of our lifetimes.

It’s not just tech companies that are benefiting from AI. Industries you might not expect are seeing massive improvements, too.

Oil giant BP PLC (BP) boosted the output of its software developers by 70% over the past year using AI.

AI helped the Commonwealth Bank of Australia cut scam losses in half. How many tens of millions of dollars is that worth?

What most people miss is how fast AI is improving. I’ve never seen a technology move as fast. AI models released just six months ago are basically dinosaurs.

A year ago, AI was more frustrating than helpful. Now, I can't imagine working without it.

As these charts show, you now have access to a PhD-level scientist or world-class coder in your pocket!

Source: @DKThomp on X

AI’s improvement isn’t magic. It’s math. The two raw ingredients you need to make AI better are: data (books, essays, and web pages) and compute (GPUs).

Scaling laws are brutally simple: Crank up the data and computing power by 10X, and the AI’s quality roughly doubles.

This is why Nvidia’s (NVDA) revenues are going vertical…

It’s also why we’re building bigger and bigger GPU clusters inside data centers. If current trends continue, we’ll have 1,000,000 chip clusters in just two years:

Source: Created by ChatGPT

Total dollars spent on compute will exceed the spend on crude oil in the next decade.

I think AI scaling laws are going to be the “Moore’s Law” of the 21st century. The bedrock upon which so much innovation is built. They’re one of the most—if not the most—important trends in the world today.

So, what does this mean for us? Full steam ahead for the AI buildout.

America’s big tech giants reiterated their plans to spend a combined $350 billion+ on AI this year.

The recent rolling back of AI chip export restrictions is also extremely positive for US AI winners. Recent reports indicate:

• Abu Dhabi’s AI powerhouse G42 will buy up to 500,000 of Nvidia’s top-tier chips each year.
• Nvidia will sell more than 18,000 of its latest AI chips to Saudi company Humain.

Houston, we’re going to need a lot more AI chips, servers, data centers, and energy to power them.

In March in my Disruption_X advisory, I was cautious about AI spending, saying:

While spending on this physical AI buildout will continue… what matters to us as investors is the “rate of change.” The percent growth is slowing down, even as the absolute dollar amounts hit record highs.

Q1 earnings confirmed this. Big tech’s AI spending was lower than during the last three months of 2024:

Source: FactSet

Slowing growth could hurt big AI winners like Nvidia. But smaller firms can plough through any weakness.

The best analogy here is digital ad giant Google. In the wake of the 2008/09 financial crisis, global advertising spend plunged 13%. Meanwhile, Google—which was still a small player in ads back then—bucked the trend and grew 9% in 2009.

Little-known disruptors at the vanguard of the AI buildout are in a similar position today.

If you liked this, you might enjoy my Jolt investing letter, published every Monday and Friday. In it, I share my unfiltered investing insights and analysis. Specifically, I write about disruptive technologies (like AI) and how to invest in them. Go here to join.

______________

Stephen McBride is Chief Analyst, RiskHedge. To get more ideas like this sent straight to your inbox every Monday and Friday, make sure to sign up for The Jolt, a free investment letter focused on profiting from disruption.