Capitalize on the artificial intelligence boom by investing in physical infrastructure plays like SK Hynix, Nebius, IREN, and Bloom Energy, which benefit from surging long-term demand by hyperscalers.

Expect high volatility in these hardware names, as demonstrated by recent single-day surges of 16% to 27% following severe market sell-offs.

Avoid shorting traditional software applications like Microsoft, because strong enterprise earnings and resilient cloud revenues create significant near-term headwinds for that bearish trade.

Consider targeting private frontier AI leaders, such as Anthropic, which currently trades at a secondary market valuation near $1.2 trillion, to capture high-growth private asset exposure.

Use these high-value private AI holdings as strategic portfolio buffers to help cushion against potential drawdowns in public equities.

Anthropic

• Held as a venture investment within the Situational Awareness (SALP) fund, surviving the wipeout of the fund's public equity book.
• The fund holds about $5 billion in Anthropic shares, which originally stemmed from an early-stage investment in Series B when the company was valued at $60 billion.
• Secondary markets currently value Anthropic near $1.2 trillion.
• The value of these Anthropic shares cushioned the blow of massive public market losses, allowing the fund to report an 80% gain year-to-date despite a 67% monthly drawdown.

Takeaways

• Early-stage venture investments in frontier AI labs can act as massive portfolio buffers during public market liquidations.
• Diversifying away from purely public, leveraged books into high-growth private AI assets saved the fund from total destruction.

AI Physical Infrastructure (Memory, Compute, and GPUs)

• The core thesis of the fund was that physical AI infrastructure (compute, GPUs, and memory) represents one of the best investment opportunities available due to massive, ongoing demand.
• Specific memory and hardware stocks mentioned in the portfolio include SK Hynix, Nebius, IREN, and Bloom Energy.
• These infrastructure plays experienced a heavy market sell-off and short pressure in July, but surged between 16% and 27% in a single day immediately after the forced liquidation of the fund's leveraged positions.
• Long-term demand drivers remain strong as hyperscalers (like Google, Microsoft, and Amazon) continue to ramp up capital expenditures and cloud infrastructure spending.

Takeaways

• Physical infrastructure remains a high-conviction thematic play for the artificial intelligence boom, though timing and risk management are critical.
• Extreme leverage on cyclical or volatile hardware trades can lead to cascading liquidations, even when the underlying investment thesis is ultimately correct.

Software Applications

• The fund held a bearish stance (short position) on traditional software applications, believing that foundational AI model labs like Anthropic and OpenAI would eventually absorb and replace standard enterprise software (such as Microsoft's offerings).
• Counter-argument discussed in the episode: Major software providers like Microsoft recently reported record-breaking earnings, and various cybersecurity and software stocks have continued to trade positively, suggesting the short thesis on software applications faces strong near-term headwinds.

Takeaways

• Shorting software applications during a period of record enterprise cloud and software revenue carries significant execution risk.
• Investors should carefully weigh the speed at which AI models can realistically cannibalize entrenched enterprise software against current cash flows and earnings realities.