Lessons From the AI Genius Who Tried to Outsmart the Market

Listen to the audio version of this article (generated by AI).

Wall Street history is littered with brilliant people who thought they had figured out the market.

Take Long-Term Capital Management (LTCM).

The hedge fund was run by some of the smartest minds in finance, including two Nobel Prize-winning economists. It used sophisticated mathematical models to exploit tiny pricing discrepancies in global bond markets.

For years, the strategy worked.

Then Russia defaulted on its debt in 1998. Markets moved in ways LTCM’s models had not anticipated, and the fund’s enormous leverage turned mounting losses into a crisis. It came so close to collapsing that the Federal Reserve helped organize a Wall Street rescue.

Then there was Archegos Capital Management in 2021.

Founder Bill Hwang built huge positions in a handful of stocks using borrowed money and derivatives. But there was more going on than aggressive investing. Hwang and his team lied to Wall Street banks about the size and concentration of Archegos’s portfolio and manipulated the prices of stocks it owned.

When those stocks started falling, Archegos faced massive margin calls it could not meet. Banks unloaded tens of billions of dollars in stock, and the firm imploded within days. Hwang was later convicted of fraud.

Now we have another spectacular hedge-fund blowup to add to the list.

This one involves Situational Awareness, an AI-focused hedge fund run by a 24-year-old former OpenAI researcher named Leopold Aschenbrenner.

And to be clear, there is no indication that Aschenbrenner did anything fraudulent.

His mistake appears to have been much simpler.

He got very smart about artificial intelligence, made a fortune betting on it and then used enough leverage that the market eventually took control of his portfolio away from him.

For a while, it looked brilliant.

Situational Awareness reportedly soared 439% in the first half of 2026.

Then July arrived – and one of the smartest young minds in AI got pummeled by the market.

Folks, there’s an important lesson here.

Being smart can give you an edge in the market. But believing you’re smart enough to outwit the market can get very expensive, very quickly.

And I think that lesson is especially important for AI investors right now.

Because after three years of enormous gains, the AI landscape is starting to change. The companies and technologies that led the first stage of this boom will not necessarily lead the next one.

In today’s Market 360, we’ll take a closer look at how Situational Awareness went from one of the hottest hedge funds on Wall Street to a forced liquidation in a matter of weeks. Then I’ll explain why its collapse offers a timely warning for AI investors as we approach what I believe could be the next major reset in the artificial intelligence market.

The “Nostradamus of AI”

Now, Aschenbrenner wasn’t some amateur who wandered into the AI trade at the top.

He entered Columbia at 15, graduated as valedictorian at 19 and later worked for the FTX-linked Future Fund before joining OpenAI. He was eventually fired from OpenAI after a dispute over the company’s information security practices.

His reputation really took off after he published an essay in 2024 called Situational Awareness. It laid out an aggressive vision for AI and the enormous amounts of chips, memory, electricity and data-center infrastructure it would require.

Then he put real money behind that thesis.

Aschenbrenner launched a hedge fund under the same name. The early results were extraordinary.

MarketWatch reported that Situational Awareness had gained more than 2,000% since its September 2024 launch. The fund was up roughly439% in the first half of 2026 alone.

Assets under management reached as much as $45 billion by early July.

At 24 years old, Aschenbrenner looked like a genius.

In fact, when the week began, he was preparing for a multiday wedding celebration in Carmel. By the time guests started arriving, his fund was unraveling and the vultures were circling.

Then the market turned.

When Leverage Takes Over

July was already shaping up to be a rough month for AI and data-center stocks.

The first wave hit when mean-reversion algorithms began attacking some of the market’s strongest momentum stocks. AI and data-center names were hit especially hard.

Then came another scare on July 17. Chinese AI startup Moonshot released a new large language model called Kimi K3, claiming it could rival leading Western models at a fraction of the cost.

Short sellers quickly seized on the news and started calling it a “DeepSeek 2.0 moment.” The argument was familiar: If Chinese companies could build cheaper AI models, perhaps the U.S. would not need as many expensive chips, data centers and power projects after all.

That narrative hit AI infrastructure stocks hard and pushed the NASDAQ close to correction territory.

Then DeepSeek added more pressure late in the month by releasing a new model at a fraction of the price charged by U.S. competitors.

OpenAI responded by cutting the price of one of its models, while Google rolled out new efficiency-focused offerings.

That sparked talk of an AI price war.

And unlike some of the rumors short sellers like to spread, there may be something behind this concern.

AI model costs are falling quickly. But I’ve consistently pointed out that cheaper AI does not necessarily mean less AI.

Lower costs can drive a lot more usage. And all that additional AI activity still requires chips, memory, electricity and data centers.

Unfortunately for Situational Awareness, that distinction didn’t matter in the middle of a bloody selloff.

The fund was heavily concentrated in many of the AI and data-center stocks already under pressure. And according to media reports, at times it was leveraged by up to four times its underlying capital.

As those stocks kept falling, prime brokers reportedly began demanding that Situational Awareness reduce risk. That forced the fund to sell into an already weak market, adding even more pressure to the same stocks it owned.

By the end of July, The Wall Street Journal reported that Situational Awareness had plunged 67% for the month. The fund eventually sold the bulk of its leveraged public-stock portfolio to Citadel in a massive block trade.

That appears to have been the final capitulation event.

And it is also where this story gets especially interesting for us…

Why Capitulation Is Good News…

What I find fascinating is that seven companies disclosed in Situational Awareness’s latest 13F also appear in our current Growth Investor portfolio.

Seven.

That includes names like Bloom Energy Corporation (BE), Micron Technology, Inc. (MU) and NVIDIA Corporation (NVDA).

So, I’m certainly not going to criticize Aschenbrenner for recognizing good AI opportunities. In several cases, we identified the exact same companies.

We first bought NVIDIA in May 2019. As of right now, we were sitting on a gain of 5,245%.

We added Bloom Energy in September 2025, and it is up about 212%.

And Micron, which we added at the end of January, had nearly doubled in about six months, giving us a gain of roughly 112%.

The difference wasn’t the stocks.

The difference was that we didn’t get greedy.

Situational Awareness used borrowed money to magnify its bets. That produced spectacular returns on the way up, but it also meant Aschenbrenner could not simply wait when the market temporarily moved against him.

His lenders eventually forced his hand. Meanwhile, over at Growth Investor, all we had to do was wait out the storm.

And what happened next helps explain why I believe the worst of July’s correction may be behind us.

For days, Wall Street had suspected that Situational Awareness was contributing to the relentless selling in AI stocks. Once news broke that Citadel had absorbed much of its public-equity portfolio, traders suddenly knew that a major forced seller was now out of the market.

Buyers rushed back in.

Bloom Energy surged about 26% the next day. Sandisk Corporation (SNDK) jumped roughly 26%. Several other recent Situational Awareness holdings posted huge one-day gains, too.

What that tells me is that Situational Awareness was dumping stock because it had to. Once that liquidation ran its course, buyers were willing to step back in.

That’s what we call capitulation, folks.

Don’t Outsmart the Market

There is a simple lesson in all of this.

Aschenbrenner knew AI. He identified several major winners. And for a while, he made an extraordinary amount of money.

Frankly, I wouldn’t be surprised if we hear from Aschenbrenner again. He’s only 24 years old, he clearly understands AI, and several of his stock ideas were very good. He just learned an extraordinarily expensive lesson.

But even the smartest investor can get pummeled by trying to outsmart the market.

That’s worth remembering now because AI is entering another major transition.

I remain extremely bullish on AI. But I don’t expect the companies and technologies that dominated the first phase of this boom to automatically dominate the next one.

My research team and I have spent months studying a massive new effort taking shape across America’s national laboratories.

President Trump has compared it to a new Manhattan Project for AI. At its center is a new network of government supercomputers and AI infrastructure that I call Golden Dawn.

The goal is to use AI to accelerate scientific breakthroughs in areas ranging from energy and medicine to advanced materials and quantum computing. And I believe the companies helping build that infrastructure could represent the next major group of AI winners.

That’s why I recently put together a special presentation called The AI Reset of 2026.

In it, I explain what Golden Dawn is, reveal the stocks I believe are positioned to profit from it and identify several well-known stocks I think investors should approach with caution.

Click here to watch my full AI Reset of 2026 presentation and see how I’m preparing for what comes next.

Sincerely,

Louis Navellier

Editor, Market 360

The Editor hereby discloses that as of the date of this email, the Editor, directly or indirectly, owns the following securities that are the subject of the commentary, analysis, opinions, advice, or recommendations in, or which are otherwise mentioned in, the essay set forth below:

Bloom Energy Corporation (BE), Micron Technology, Inc. (MU), NVIDIA Corporation (NVDA) and Sandisk Corporation (SNDK)