David George discusses SpaceX's IPO as a milestone enabling retail participation. He emphasizes that late-stage venture is about founders, not capital markets, and that the AI wave will create even more value than previous tech cycles. He sees orbital data centers as inevitable once Starship achieves rapid reusability.

Selene Woo argues memory demand is structurally driven by AI inference and agentic AI, not just cyclical supply bottlenecks. Supply constraints (manufacturing complexity, capex intensity, technology migration) will keep the balance in favor of memory stocks. She sees multi-year outperformance for memory and AI-infrastructure equities, with physical AI as a long-term multi-trillion opportunity.

OpenAI unveiled its first custom AI chip 'Jalapeno' developed with Broadcom, offering 50% lower cost vs typical AI GPUs. This is part of OpenAI's strategy to own more of the infrastructure stack and reduce reliance on Nvidia. Broadcom CEO Hock Tan sees every frontier lab eventually having custom chips, with no limit to demand for this infrastructure.

Andrew Feldman explains Cerebras' margin contraction as a strategic decision to rent back gear to customers to meet extraordinary demand. He highlights Cerebras' architectural advantages: no HBM dependency, no CoWoS, and use of 5nm node instead of 3nm, insulating them from key supply chain constraints. He sees a heterogeneous AI chip market with room for GPUs, ASICs, and pioneering architectures.