Leopold Aschenbrenner, the former OpenAI researcher turned hedge fund manager who has argued that Artificial General Intelligence (AGI) deserves greater focus than AI chips, has backed up that theory through a major portfolio repositioning.
This followed Aschenbrenner's Q1 13F disclosure with the U.S. Securities and Exchange Commission on 15 May, accepted on May 18, which showed his disclosed equity exposure rising from $5.5 billion to $13.67 billion, centred on energy infrastructure, computing capacity, and crypto-linked data centres. These sectors are increasingly seen as crucial for supporting the infrastructure needed to develop and deploy AGI at scale.
Aschenbrenner's Eye-Catching Portfolio
The figures show that Aschenbrenner's 165-page 'Situational Awareness' essay continues to influence his investment approach. In that thesis, he argued against overlooking AGI, believing it was closer than most market participants expected. That perspective is now reflected in his portfolio.
Aschenbrenner bet against semiconductor companies like Nvidia, VanEck Semiconductor ETF, Oracle, and Broadcom. The scale of these positions is notable. Below are the latest disclosed put positions for these companies.
- NVIDIA ($1.57 billion)
- VanEck Semiconductor ETF ($1.07 billion)
- Oracle Corp ($1.07 billion)
- Broadcom Inc ($1.01 billion)
These bets indicate a strategic move based on his belief that the current enthusiasm for AI chips may be premature or overvalued.
The Big Pivot for Aschenbrenner
For those unfamiliar with Aschenbrenner, he's the former OpenAI researcher dismissed in 2024 over an alleged information leak. After that, the German native shifted to public markets and published a 165-page essay titled 'Situational Awareness: The Decade Ahead.'
The essay argued that artificial general intelligence was likely to arrive faster than markets expected, but progress would be limited by physical infrastructure rather than algorithms. This framework now guides his current positioning.
Aschenbrenner believes AI development will eventually hit a bottleneck, not due to chips, but the power and land needed to support them. This reinforces his thesis that AGI progress depends more on computer infrastructure than on model development.
This is evident in 'Situational Awareness's disclosed major holdings in Bitcoin miners. The list includes IREN, Core Scientific, Riot Platforms, CleanSpark, Bitfarms, Bitdeer, and Hive Digital—all sitting alongside energy and compute-related names like Bloom Energy, SanDisk, and CoreWeave.
The focus is on power and capacity. It was noted that Bitcoin miners already have high-density power sites and grid access, which AI companies may not be able to replicate immediately.
This shift is also reflected in the reported performance of some companies. In the first quarter of 2026, TeraWulf reported that its AI and high-performance computing hosting revenue, roughly $21 million, surpassed its Bitcoin mining revenue ($13 million) for the first time.
There is also Core Scientific, one of the fund's disclosed holdings. It revealed plans to convert its Pecos site into a 1.5-gigawatt AI data-center campus, repurposing 300 megawatts of existing mining capacity.
Markets Remain Skeptical
Despite these reports, the market remains unswayed. Confidence in Nvidia and other AI chip leaders stays intact, indicating that investors are not yet ready to accept a near-term AI unwind.
A key test will be Nvidia's earnings report for the first quarter of fiscal year 2027 on 20 May. Market expectations remain high, which analysts continue to incorporate into their outlook. However, that could change if the numbers show a significant downward shift.
Disclaimer: Our digital media content is for informational purposes only and does not constitute investment advice. Please conduct your own analysis or seek professional guidance before investing. Remember, investments are subject to market risks, and past performance does not guarantee future results.
