Leopold Aschenbrenner, the former OpenAI researcher turned hedge fund manager, has indicated a massive shift in how institutional capital views the AI trade. His firm, Situational Awareness LP, disclosed an $8.46 billion short position against the semiconductor sector in its latest 13F filing. The fund is betting heavily against industry titans including Nvidia, Broadcom, and AMD, while simultaneously pivoting into long positions in crypto mining stocks like CleanSpark, Riot Platforms, Applied Digital, and IREN.
The strategy rests on a thesis of infrastructure arbitrage. Aschenbrenner argues that the market has overvalued the hardware manufacturers–the companies selling the shovels–while undervaluing the entities that already possess the physical power and data center infrastructure required to run large-scale AI models. By shorting the chipmakers, he is effectively hedging against a potential cooling in hardware demand, while his long exposure to miners serves as a play on the scarcity of high-capacity energy and compute assets.
Aschenbrenner’s background adds a layer of intrigue to the trade. After his departure from OpenAI in 2024, he authored a widely circulated essay predicting AGI arrival by 2030, framing the physical infrastructure buildout as the defining investment of the decade. His fund, which raised $13.7 billion in under two years, is now putting that theory to the test with a portfolio that explicitly favors energy-heavy crypto miners over the silicon giants that have dominated the market for the past two years.
The scale of the put options is staggering. With $2 billion in exposure to the VanEck Semiconductor ETF and $1.6 billion against Nvidia alone, the fund is positioned for a significant correction in the chip sector. If the AI hardware cycle hits a plateau or if supply chain bottlenecks ease, these short positions could yield massive returns. Conversely, the success of his long positions in miners depends on their ability to pivot their existing power infrastructure toward AI data center operations.
Traders should monitor the upcoming quarterly earnings for Nvidia and the broader semiconductor index for signs of margin compression. Any indication that capital expenditure on AI hardware is slowing will likely serve as the primary catalyst for Aschenbrenner’s short thesis to play out. Watch the price action in the mining sector closely, as these stocks are now effectively trading as proxies for AI infrastructure capacity rather than just Bitcoin price derivatives.
Leopold Aschenbrenner’s $13.7B Fund Bets Against Chips, Longs Miners
A $13.7 billion AI-focused hedge fund has initiated a significant position in crypto mining companies while simultaneously shorting semiconductor stocks. The fund's strategy relies on the thesis that crypto miners possess the physical infrastructure necessary for AI development.