Just came across something pretty wild in the AI investment space. There's this 24-year-old named Leopold Aschenbrenner who's quietly become one of the most interesting fund managers to watch right now—and his track record is honestly insane.

So here's the thing: his fund, Situational Awareness LP, went from managing what looked like $255 million in publicly disclosed holdings at the end of 2024 to sitting on $5.5 billion by Q4 2025. That's not a typo. In less than a year.

Most people don't realize Leopold Aschenbrenner actually has roots in the crypto world. He was part of the FTX Future Fund team back in 2022—worked there until the whole thing collapsed. Before that, he was at OpenAI on the Superalignment team, the one focused on keeping AI systems in check. Got fired in 2024 over alleged information leaks (though he disputed the details), and interestingly, the entire Superalignment team disbanded shortly after. His fiancée, Avital Balwit, now runs operations at Anthropic, which is basically OpenAI's biggest competitor.

But the real story is what happened next. In June 2024, Leopold Aschenbrenner dropped a 165-page paper called "Situational Awareness: The Decade Ahead" that basically became required reading in Silicon Valley. The thesis was bold—AGI is coming sooner than most people think, and there are specific infrastructure plays that will dominate. Three months later, he launched his fund based on exactly that thesis.

Here's where it gets interesting: Leopold Aschenbrenner's portfolio doesn't chase the obvious mega-cap AI plays. You won't find Nvidia, Microsoft, Amazon, or Google in his top holdings. Instead, he's betting on the unglamorous stuff—energy infrastructure, computing power, optical communications, storage. His top 10 holdings alone represent 86% of the fund's positions. Bloom Energy (solid oxide fuel cells for data centers) has absolutely crushed it. Stock's more than doubled since end of 2024.

There's one detail that caught my eye though: Situational Awareness LP holds positions in several Bitcoin mining companies—Core Scientific, Cipher Mining, Iren, Bitdeer. But here's the thing—Leopold Aschenbrenner isn't betting on crypto mining itself. He's betting on these companies pivoting into AI computing centers. Makes sense when you think about it: mining operations already have land, power infrastructure, and cooling systems dialed in. As the crypto market stagnates and AI demand explodes, it's a natural transition.

Looking back at Leopold Aschenbrenner's arc, there's something almost poetic about it. He and SBF both came through FTX, both were true believers in effective altruism, both generated serious returns. But SBF ended up in prison, while Leopold Aschenbrenner basically got kicked off one path and discovered something better suited to him. The crypto industry itself is at that same inflection point—old narratives are dying, new ones are emerging, and the ones who can see the shift early tend to do pretty well.

Funny how that works. When the old story breaks down, the real opportunity is in building the next one.