Back to News
Market Impact: 0.3

At Davos, CEOs said AI isn’t coming for jobs as fast as Anthropic CEO Dario Amodei thinks

NVDACRWVCRMGOOGLPLTRMETAMSFTNYT
Artificial IntelligenceTechnology & InnovationPrivate Markets & VentureRegulation & LegislationCybersecurity & Data PrivacyInfrastructure & DefenseElections & Domestic PoliticsManagement & Governance

Nvidia invested $2.0 billion in CoreWeave at $87.20 per share to lift its stake to over 11%, implying a CoreWeave valuation of $52 billion and accelerating AI data-center construction through 2030 while creating a circular demand dynamic for Nvidia hardware. Concurrent industry developments include Anthropic CEO Dario Amodei’s stark essay on AI risks (including a 50% displacement claim for entry-level white‑collar jobs), Anthropic research showing an “activation capping” technique that cut harmful responses by 50%, Cognizant and EY outlining large addressable opportunities (a cited $4.5 trillion U.S. productivity upside and a new $6 billion tax market), and rising regulatory and privacy moves as the U.S. DOT pilots Google’s Gemini for rule drafting and the U.K. scales live facial recognition—factors that could increase regulatory scrutiny and policy risk for AI infrastructure investments.

Analysis

Market structure: The Nvidia–CoreWeave deal structurally props up NVDA’s GPU pricing power and locks a large, dedicated demand stream for high-end accelerators through 2030; direct winners are NVDA, specialized cloud infra providers (CRWV) and enterprise AI software vendors (CRM, GOOGL, MSFT) that monetize agents. Losers are commodity cloud resellers and surveillance/data-mining vendors facing regulatory backlash (PLTR) and legacy labor-intensive services whose margins compress as AI automates workflows. Expect tighter GPU supply vs. demand for next 12–36 months, supporting ASPs and capex cycles; that raises corporate borrowing for data-center builds and marginally pushes industrial commodity demand for copper/natural gas and utility capex. Risk assessment: Tail risks include aggressive antitrust/regulatory scrutiny of circular financing, high-profile AI safety incidents (Amodei narrative) triggering moratoria, and power-grid bottlenecks that delay builds; any of these can halve TAM growth in a 6–18 month window. Immediate catalysts (days–weeks) are DOT/Gemini news and UK facial recognition rollouts; near-term (weeks–months) catalysts are Nvidia GTC (Mar), company earnings and any regulatory inquiries into the NVDA–CRWV structure; long-term (years) risks are geopolitical export controls and China open-source competition compressing margins. Trade implications: Tactical portfolio tilt: overweight AI infra and enterprise SaaS (NVDA, CRWV, CRM/GOOGL) and underweight surveillance/legacy IT (PLTR). Use 9–24 month LEAP call spreads on NVDA to capture secular GPU demand while limiting capital; initiate small core positions in CRWV (1–2%) sized for idiosyncratic governance risk. Pair trade idea: long NVDA (or CRWV) vs short PLTR (or buy PLTR 3–6 month puts) to express infrastructure winners vs reputational/regulatory losers; set stop-losses ~12–15% and take-profit at +30%. Contrarian angles: Consensus overstates immediate job destruction and underestimates service expansion (EY example) — invest in mid‑market enabling software (CRM) where AI unlocks new $bn TAMs over 12–24 months. The market may underprice CoreWeave’s revenue acceleration despite governance optics; conversely, sentiment-driven shorts on all AI names are overdone if GPU scarcity persists. Hidden outcome: circular funding invites regulatory review — hedge with 6–12 month protective puts on NVDA/CRWV if regulatory filings appear within 90 days.