Listen to the audio version of this article (generated by AI).
Hello, Reader.
Tom Yeung here with today’s Smart Money.
There is usually a “tuition cost” that comes with learning how to invest.
- That first stomach-churning loss…
- That first painful tax bill…
- That first accidental “buy” order instead of a “sell”…
Everyone remembers their early mistakes. It’s what makes you a better trader.
That’s because learning to invest by paper trading is like figuring out how to swim by reading a book. There is no substitute for diving in and trying not to drown.
Now, most of us pay that tuition a little at a time. Preferably very early on.
But one AI hedge fund appears to have paid a very expensive tuition bill in recent weeks.
I’m talking about Situational Awareness, an AI fund run by one of Wall Street’s brightest new stars, Leopold Aschenbrenner. The 24-year-old former OpenAI researcher first gained notice in 2024 after publishing a lengthy essay called “Situational Awareness: The Decade Ahead” that predicted the rise of artificial general intelligence (AGI).
Then Aschenbrenner put money behind that idea. He launched a hedge fund with the same name and built enormous positions around the AI boom.
For a while, the results looked almost supernatural. The fund gained 2,000% in 2025, and another 439% in the first half of 2026. Situational Awareness was worth $45 billion at its peak.
But then AI stocks hit a speed bump this summer.
Soon, the hedge fund began losing money. Then much more. So much, in fact, that it was forced to dump whatever it could. It ultimately sold its public-stock portfolio to a different hedge fund, Ken Griffin’s Citadel, notching an 80% loss.
In other words, Aschenbrenner had seen the AI future before almost everyone else…
But he had not built a portfolio that could survive the trip.
Fortunately, you don’t need a billion-dollar hedge fund – or a mountain of leverage – to profit from the AI Revolution.
Today, I’ll explain why simply being right about AI isn’t enough… lay out the two qualities I believe separate long-term winners from eventual blowups… and introduce you to one company I think fits that description.
Then, I’ll show you where you can find three more stocks that we’re watching…
The Right Thesis, The Wrong Trade
In fairness, I believe Aschenbrenner remains directionally correct about AI.
AI systems are becoming more capable. Businesses are spending hundreds of billions of dollars to build data centers and build better AI models. The AI Revolution will have many years of growth ahead.
However, a correct prediction is not automatically a good investment.
Imagine someone knowing in 1997 that the internet would transform the global economy. They would have been absolutely right. And if they had put their life savings into Amazon.com Inc. (AMZN), they would have made millions… if not billions of dollars. The stock is up more than 300,000% since its listing in 1997.
But what if that same investor had bought Pets.com instead? After all, Pets.com was also an early e-commerce player. Besides, it had a sock-puppet mascot that showed up in a Super Bowl ad and on Good Morning America. Jeff Bezos never thought of doing that!
Instead, Pets.com turned out to be a total disaster. The pet food delivery company could not figure out how to become profitable, and the stock went from an IPO price of $11 down to $0.19 before it was totally liquidated in 2001. Hundreds of employees lost their jobs, and investors were wiped out.
The same will be true of the AI Revolution. Not every AI firm will succeed, and some will fail spectacularly.
But that’s not really what doomed Situational Awareness. The fund’s biggest problem wasn’t that it believed in AI. It was that it used enormous leverage to amplify those bets. When AI stocks stumbled, even temporarily, those leveraged positions quickly became impossible to maintain.
That’s an important lesson for individual investors. You can be absolutely right about the future… and still lose money if you own the wrong companies, pay too much for them, or take on too much risk.
The Right Way to Play the AI Revolution
You’re probably now wondering how to separate the “Amazon.com successes” from the “Pets.com flops” of the AI Revolution.
I have some good news for you. In my experience, great businesses usually share a few common characteristics, even in industries moving as quickly as AI:
1. A wide business moat. Some AI companies own valuable technology and enjoy real pricing power. These are called “moats” because they protect those companies from competition. And they’re a key reason for a company’s long-term success.
2. The right price. The best investments are bought cheaply before everyone has discovered their worth. If a stock today is worth $1,000 per share, an investor would have made far greater profits if they had bought for $10… or $1… or better yet $0.10.
Those are two of the qualities that Eric looks for when researching AI investments.
He isn’t simply searching for “the next Nvidia” or “the next Amazon.” Even though there are some fantastic mega-cap AI companies out there, these stocks have already been discovered by just about every person on Earth with a working brokerage account.
In fact, if Amazon rose another 300,000% because of its AI business, it would be worth almost $9 quadrillion. If you spent $1 billion per day, it would take roughly 25,000 years to burn through that amount!
Nor is Eric trying to replicate the highly leveraged approach that helped Situational Awareness generate spectacular gains – and equally spectacular losses. Extraordinary returns are wonderful if you can keep them. But if your portfolio loses 80% every time the market hits a rough patch, you’re probably not going to stay in the game very long.
Instead, Eric focuses his search on a core group of companies that are building AI’s “Golden Rivets.” These are the specific, irreplaceable pieces needed to construct and power the AI buildout.
These Golden Rivet producers are not necessarily the companies receiving the loudest television coverage. Nor are they being bought up by the hottest AI hedge funds in town. In many cases, they are old-economy businesses that Wall Street overlooked while everyone chased chips and chatbots.
That is precisely what makes them interesting.
One example is Teradyne Inc. (TER). Rather than competing to build the next AI model, Teradyne supplies the sophisticated automated testing equipment and software that semiconductor manufacturers rely on to ensure increasingly complex AI chips actually work before they leave the factory.
Whether Nvidia Corp. (NVDA), Advanced Micro Devices Inc. (AMD), or another chipmaker wins the AI race, those chips still need to be tested. That’s exactly the kind of “Golden Rivets” business Eric likes to own.
These are the firms that will be fueling the big AI names. They will be building the chips… powering the data centers… and perhaps even running AI servers in space.
Teradyne is one of four semiconductor-related companies Eric discusses in his . There, he explains why he believes AI’s next phase could reward these overlooked “Golden Rivets” businesses far more than today’s crowded AI trades—and reveals the other three stocks currently on his radar.
Every investor pays tuition eventually. The trick is paying a few hundred dollars… instead of a few billion. Hopefully, today’s lesson saves you from the latter.
If you’d like to see the rest of Eric’s “Golden Rivets” framework, I think you’ll get a great deal out of his .
Regards,
Thomas Yeung, CFA
ÃÛÌÒ´«Ã½ Analyst, InvestorPlace
