| Ticker | Direction | Speaker | Thesis | Time |
|---|---|---|---|---|
| LONG |
David Friedberg
The Production Board / CEO |
Friedberg highlights that "Central banks have decided they no longer want to hold US treasuries... Gold is now a larger share of holdings." Chamath notes "Copper is up 26% in a month." The US fiscal situation (printing money to pay debt interest) forces dollar devaluation. In this environment, fiat purchasing power drops, but nominal asset prices (Gold, Commodities, Real Estate) rise. LONG. This is a hedge against the "debt spiral" and M2 money supply expansion. Fed hawkishness or a deflationary crash (recession) temporarily strengthening the dollar. | 73:13 | |
| LONG |
Chamath Palihapitiya
Host, All-In Podcast / CEO, Social Capital |
Chamath states, "Yesterday, Elon shut down the Model S and Model X production lines in Fremont... to make Optimus." This confirms Tesla is transitioning from a pure EV car company to a robotics/labor substitution company. If Optimus scales ("put two in your garage"), the TAM expands from transportation to the entire labor market. LONG. The market likely still prices TSLA as an auto manufacturer, not a labor supplier. Execution risk on Optimus; loss of revenue from high-margin S/X models in the interim. | 54:39 | |
| SHORT/AVOID |
Jason Calacanis
Host / Angel Investor |
Jason describes his AI agent: "It vibe coded a CRM for itself... it built its own SaaS tools." If AI agents can instantaneously build bespoke software tools and perform SDR (Sales Development Rep) functions, the value proposition of expensive, seat-based B2B SaaS (like Salesforce) collapses. Companies will stop paying $100/seat for software that an agent builds for free. SHORT/AVOID. Structural deflationary pressure on the entire SaaS business model. SaaS companies successfully embedding these agents to retain moats. | 48:28 | |
| LONG |
David Sacks
Craft Ventures / General Partner |
Sacks argues that for AI agents (like "Claudebot") to be useful, they need access to personal data. "Google has all my email, my calendar, my documents." While open-source models are rising, the "Context Window" advantage belongs to whoever holds the user's life data. Google is the only major player with the full suite (G-Suite) to integrate an agent seamlessly without security friction. LONG. Google captures the "Personal AI Assistant" market due to data incumbency. Trust/Privacy concerns; Open-source tools finding ways to bypass Google's ecosystem. | — | |
| LONG |
Jason Calacanis
Host / Angel Investor |
The group discusses running open-source models (Kimmy K2.5) locally to avoid API costs and data privacy issues. Jason notes they are "ordering Mac Studios... and stacking them" to run these models. The shift to "Local Inference" requires high-performance local compute with unified memory. Apple Silicon (M-series chips in Mac Studio) is currently the hardware of choice for running heavy open-source models locally. LONG. Apple benefits from the "de-clouding" of AI inference. PC competitors releasing NPU-heavy chips that undercut Mac pricing. | — | |
| LONG |
David Sacks
Craft Ventures / General Partner |
Sacks notes that Trump is forcing NATO countries to increase spending from 3% to 5%. "They're going to have to start investing in those things, buying weapons." European domestic defense production is insufficient. To meet these targets and "keep the US in Europe," EU nations will be forced to buy American hardware. LONG. Direct revenue transfer from EU budgets to US defense primes. Geopolitical de-escalation (unlikely per speaker sentiment). | — |