The discussion opens with the fallout from Leopold Aschenbrenner's hedge fund, which was forced to unwind its leveraged positions after a -67% monthly loss, despite being up 80% year-to-date. The hosts frame this as a cautionary tale about overconfidence, contrasting his effective altruism-inspired, expected-value mindset with traditional Silicon Valley optimism. They argue his long-term bets on AI infrastructure are sound, even if Wall Street's algorithmic trading misunderstands the sector. The conversation then shifts to the sustainability of massive AI infrastructure spending, comparing it to past boom-bust cycles. They analyze OpenAI's aggressive price cuts, which pressure model providers but benefit GPU suppliers like Nvidia, and argue that falling prices will spur increased usage, making a zero-sum outcome unlikely. The hosts also examine Nvidia's $5 billion investment in Ilya Sutskever's SSI, framing it as a low-risk bet since the computing infrastructure retains value. Finally, they review the latest earnings from Microsoft, Amazon, Google, and Meta, highlighting how the market rewarded companies with clear AI monetization strategies while punishing those with high costs and unclear plans, and touch on Apple's long-term challenges in AI.