In a recent LinkedIn post, Linas Beliūnas discusses the remarkable trajectory of Leopold Aschenbrenner, a former OpenAI employee who transitioned from AI safety research to founding a rapidly growing hedge fund focused on the physical infrastructure underpinning artificial intelligence. Beliūnas frames Aschenbrenner’s story as a compelling case study in identifying and capitalizing on emergent market needs.
Beliūnas highlights Aschenbrenner’s early career, noting his departure from OpenAI’s Superalignment team at the age of 22. According to the post, Aschenbrenner was dismissed after circulating internal memos concerning AI safety. Rather than fading from the scene, Aschenbrenner published an extensive manifesto forecasting the advent of Artificial General Intelligence (AGI) and a subsequent surge in demand for AI infrastructure.
“Then he was fired after circulating internal AI safety memos. Most people would quietly disappear. Instead, he published a 165-page manifesto predicting AGI by ~2027 and a massive explosion in AI infrastructure spending.”
A Bold Investment Thesis Emerges
Linas Beliūnas details how Aschenbrenner leveraged this foresight to launch Situational Awareness LP. The fund quickly amassed significant capital, attracting backing from prominent figures such as the Collison brothers of Stripe. Beliūnas points out the impressive performance metrics of the fund, emphasizing its rapid growth and substantial returns.
As Beliūnas explains, the fund’s strategy is particularly noteworthy for its investment focus. Instead of concentrating on AI applications or the large model developers themselves, Aschenbrenner’s fund has heavily invested in the foundational physical elements required for AI’s expansion.
“Not AI apps. Not even the biggest model companies. Instead, his fund is all-in on the physical infrastructure behind AI: GPU cloud provider CoreWeave; Power infrastructure like Bloom Energy… Data-center optics companies; Bitcoin miners pivoting to AI hosting like Core Scientific…”
The Critical Role of Physical Infrastructure
Beliūnas elaborates on the core of Aschenbrenner’s investment thesis, which Beliūnas presents as a contrarian yet increasingly validated view. The argument posits that as AI technology continues to scale, the primary bottlenecks will not be the algorithms or the processing chips, but rather the availability and management of power, data centers, and the connectivity between them.
Power as the Deciding Factor
In Beliūnas’s telling, Aschenbrenner’s fund is making a significant bet that control over essential physical resources, particularly electricity, will be a deciding factor in the future of the AI boom. This perspective is supported by the fund’s substantial holdings in companies involved in power generation and data center operations.
“The bet is simple: As AI scales, the real constraint won’t be models or chips. It will be power, data centers, and connectivity. In other words, the AI boom might ultimately be decided not by algorithms but by who controls the electricity.”
Linas Beliūnas concludes by framing Leopold Aschenbrenner as a visionary, highlighting how a young fund manager, despite an unconventional start, has positioned his fund to benefit significantly from this anticipated infrastructure demand. The success of Situational Awareness LP, as outlined by Beliūnas, underscores the importance of looking beyond the immediate software and hardware innovations to the fundamental physical realities that enable them.
📝 About This Content
This article is based on insights shared by Linas Beliūnas on LinkedIn.
📅 Originally posted on April 16, 2026 | View original post on LinkedIn →