前OpenAI研究員對沖基金在SEC文件中大舉押注比特幣礦工

Leopold Aschenbrenner, a former OpenAI researcher who departed the lab’s superalignment cadre to launch the San Francisco‑based hedge fund Situational Awareness LP, has steered his portfolio toward the AI compute backbone. The latest 13F filing for Q4 2025 reveals a dramatic scale‑up: the fund reports about $5.52 billion in US equity exposures across 29 positions, up from a few hundred million dollars at the start of 2025. Rather than chasing consumer AI software, the strategy bets on the infrastructure that powers the AI boom—power plants, data centers, and the hardware that underpins high‑end computation. The concentration is clear: a small cadre of AI infrastructure names and energy plays that the fund believes will capture the surge in demand for AI workloads.

Key takeaways

The Q4 2025 13F shows Situational Awareness with roughly $5.52 billion in US equity holdings across 29 positions, signaling a deliberate tilt toward AI infrastructure and energy‑intensive compute.

Top holdings include CoreWeave, Bloom Energy, Intel, Lumentum, and Core Scientific, reflecting a strategy anchored in data center capacity and related hardware ecosystems.

The amended Schedule 13D reveals a 9.4% stake in Core Scientific, amounting to 28,756,478 shares with shared voting and disposition power, indicating a levered view on the company’s expansion into AI hosting and HPC environments.

Beyond pure mining, the fund has increased exposure to Bitcoin miners and energy players such as IREN, Cipher Mining, Riot Platforms, Bitdeer, and Applied Digital, signaling a broader bets on AI compute throughput via specialized energy infrastructure.

Aschenbrenner’s strategy also includes a noted short in Infosys, reflecting a view that large‑language model adoption and AI coding tools could pressure traditional outsourcing software services models.

Tickers mentioned: $BTC

Market context: The shift underscores a growing convergence between crypto mining and AI compute ecosystems, where megawatt‑dense sites and long‑term data‑center arrangements are increasingly treated as scarce, high‑value assets in the new compute economy.

Sentiment: Neutral

Price impact: Neutral. The moves reflect strategic positioning in a sector undergoing structural changes rather than immediate price catalysts.

Trading idea (Not Financial Advice): Hold. The cross‑section of AI infrastructure and mining assets suggests exposure to broader compute demand, but the concentration in a handful of names warrants careful risk management.

Market context: The AI compute narrative is evolving from a focus on chip supply and software models to ownership of the physical and energy assets that enable massive data‑center deployments. The post‑halving environment has encouraged miners to pivot toward hosting AI workloads, recasting megawatts and data‑center capacity as strategic assets rather than mere hash rate capacity.

Why it matters

The portfolio strategy signals a shift in how investors view AI reverberations across sectors. By placing heavy bets on AI infrastructure players like CoreWeave and Bloom Energy, the fund aligns with the premise that the next era of AI growth will be defined by the reliability and scalability of compute foundations. CoreWeave, a major AI cloud firm, has pursued long‑term HPC hosting contracts, reinforcing the idea that enterprise‑grade compute capacity will anchor AI deployment for years to come. That dynamic is echoed in the fund’s positioning around Core Scientific and other miners‑turned‑infrastructure operators, highlighting a broader trend where mining assets are repurposed as high‑density compute farms capable of supporting AI workloads.

Moreover, the mix of energy‑oriented firms with traditional chip and optics players points to a convergence of energy efficiency, power reliability, and advanced hardware as the bedrock of AI scalability. The emphasis on Bloom Energy and similar energy infrastructure names acknowledges that the economics of AI compute increasingly hinge on dependable, low‑cost power and resilient facilities. In this context, the bitcoin ecosystem—often used as a proxy for large‑scale, independent energy demand—appears intertwined with broader infrastructure plays, rather than living in its own isolated corner of markets. Bitcoin (CRYPTO: BTC) remains a barometer for how much compute demand miners can leverage, particularly as large data centers seek to optimize energy intensity and uptime amid rising competition for grid capacity.

The presence of a substantial Core Scientific stake via an amended Schedule 13D demonstrates the degree to which the fund leverages governance and ownership rights to influence a company’s expansion into AI hosting and HPC. This move aligns with a broader industry pattern where miners diversify beyond hashing to become multipurpose data‑center operators that can monetize surplus capacity across AI workloads, rendering traditional hash rate metrics less decisive in evaluating value creation.

Finally, the Infosys short reflects an acceleration of AI coding tools and large‑language models that, in the hedge fund’s view, could erode the traditional outsourcing model long relied upon by software services giants. If AI tools increasingly curtail demand for routine outsourcing tasks, equities tied to that segment may face new headwinds, even as AI infrastructure assets benefit from expanding compute demand. The net effect is a nuanced stance: bets that the core compute economy—powered by energy, data centers, and HPC—will drive durable value, tempered by selective shorts on areas perceived as vulnerable to AI displacement.

What to watch next

Next 13F filing cycles (early 2026) to reveal whether the $5.5B positioning is sustained or expanded across additional AI infrastructure names.

Any new or amended Schedule 13D/13G disclosures around Core Scientific or other holdings, signaling shifts in control or strategy.

Updates on long‑term HPC contracts and data‑center expansions tied to CoreWeave and similar operators, which would validate the thesis of AI hosting as a growth engine.

Further moves in mining‑to‑infrastructure transitions, including additional energy‑asset investments from the broader field, and how such moves interact with regulatory and grid‑capacity constraints.

Regulatory or policy developments affecting large‑scale AI compute deployments and crypto mining operations, which could influence capital flows into AI infrastructure equities.

Sources & verification

Situational Awareness 13F Filing, Q4 2025 — 13f.info

Amended Schedule 13D for Core Scientific — filing PDF

Fortune profile on Leopold Aschenbrenner and the fund’s size — fortune.com

Hut 8 research/coverage on AI data center pivot and compute revenue — cointelegraph.com

CoreWeave and AI data center partnerships — cointelegraph.com

Market reaction and key details

Situational Awareness has built a narrative around a recalibration of AI investment risk, moving from a focus on peak‑AI software potential to the tangible, capital‑intensive backbone that makes AI feasible at scale. The 13F results highlight how a single fund can tilt an entire sub‑sector toward a handful of strategic names, elevating the importance of long‑term contracts, energy reliability, and data‑center capacity in determining which players benefit most from the AI era. While the broader market continues to wrestle with volatility and regulatory questions, the fund’s emphasis on compute infrastructure—paired with a measured portfolio tilt toward miners pivoting to AI hosting—illustrates a disciplined approach to navigating the evolving landscape of AI, crypto, and high‑performance computing.

What to watch next

This article was originally published as Ex-OpenAI Researcher Hedge Fund Bets Big on BTC Miners in SEC Filing on Crypto Breaking News – your trusted source for crypto news, Bitcoin news, and blockchain updates.