Trade Ideas
Jason notes that Anthropic's announcements (Claude for Legal, Claude for Code) caused immediate 10%+ drops in stocks like Thomson Reuters, LegalZoom, CrowdStrike, and IBM. Chamath explains a structural shift in valuation: The market has moved from asking "when" growth slows to "if" these businesses will exist at all. This forces investors to increase the Weighted Average Cost of Capital (WACC) and compress PE multiples to create a massive margin of safety. SHORT. The "growth annuity" model of SaaS is broken; terminal value is now questionable due to AI displacement. AI integration might allow incumbents to cut costs faster than they lose revenue, temporarily boosting margins.
Chamath cites data showing 40% of data center projects face local opposition ("Bananas" movement), causing the industry to lose ~5GW of capacity and potentially $130B in revenue over two years. The demand for compute is inelastic, but supply is artificially constrained by regulation/NIMBYism. Trump's administration is pushing for "behind the meter" solutions where tech companies build their own power plants. This favors companies that can secure land/power or provide independent power generation. LONG. Scarcity of power + high demand = pricing power for energy infrastructure owners. Continued regulatory gridlock or successful "Bananas" lawsuits halting construction entirely.
Life Biosciences (private) reached an agreement with the FDA to treat humans with Yamanaka factors to reverse aging in the eye. This is the first human application of cellular rejuvenation technology. Success here validates the entire longevity/epigenetic reprogramming sector, likely sparking a boom in related public biotech stocks. WATCH. While the specific company is private, positive Phase 1 results will lift the entire genomics/longevity sector. Clinical trials fail or show adverse effects (e.g., cancer risk associated with cellular reprogramming).
Rumors suggest the new Mac Studio will feature an M5 chip specifically designed to be "language model ready." As AI moves from cloud-only to local inference (privacy/speed), hardware that can run models locally becomes essential. Apple is positioning to own the "local AI" hardware stack. LONG. A massive upgrade cycle for high-end compute hardware. The M5 chip may not deliver the expected performance jump, or local models may remain inferior to cloud models.
This All-In Podcast video, published February 28, 2026,
features Chamath Palihapitiya, David Friedberg, Jason Calacanis
discussing IBM, TRI, CRWD, OKTA, EQIX, XLE, XBI, ARKG, AAPL.
4 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
Chamath Palihapitiya,
David Friedberg,
Jason Calacanis
· Tickers:
IBM,
TRI,
CRWD,
OKTA,
EQIX,
XLE,
XBI,
ARKG,
AAPL