OpenAI’s former head of sales is entering VC. She still calls herself an ‘AGI sherpa’

OpenAI veteran Aliisa Rosenthal left to join Acrew Capital as a general partner after building a >300-person go-to-market organization at OpenAI, signaling continued VC interest in AI-native commercial expertise. Apple is reportedly developing an always-on AI wearable pin that could target manufacturing of up to 20 million units at launch (potentially as soon as 2027), while Google DeepMind has acqui-hired Hume AI’s CEO and senior engineers; Hume has raised $74M and projects $100M revenue in 2026. OpenAI is reorganizing around a general-manager structure for enterprise and consumer products with Barret Zoph leading enterprise, and Deloitte data shows 60% of workers now have employer-sanctioned AI tools—underscoring accelerating enterprise AI adoption and competitive pressure across Big Tech.

Analysis

Market structure: Apple (AAPL) and Google (GOOGL) are the primary beneficiaries of the announced product/talent moves—Apple if the wearable pin ships (20M units target implies ~$2–4B revenue run-rate at $100–200 ASP) and Google via voice/IP accretion from Hume improving assistant stickiness. Nvidia (NVDA) retains pricing power on datacenter GPUs as enterprises scale agent deployments (Deloitte: 74% will use agents within 2 years), tightening supply vs. demand for high-end compute through 2026. Incumbent consumer platforms (META) face mid-term share pressure in AR/voice but retain installed-base advantages in social/ads monetization.

Risk assessment: Tail risks include rapid privacy/regulatory intervention (EU/US fines or limits on always-on sensors) and a hardware recall or data-breach forcing product delays—each could compress multiples by 10–25% within 90 days. Operational dependency on component supply (cameras/mics/sensors) and datacenter power/energy costs (electricity/copper) are second-order constraints that could push manufacturing timelines from 2027 to 2028. Catalysts: Apple supply-chain confirmations (next 6–12 months), OpenAI enterprise GM announcements (next 3 months), and NeurIPS/academic credibility events.

Trade implications: Expect NVDA to keep premium TAM capture—favor delta-positive exposure into March GTC and FY earnings; AAPL is a conditional asymmetric long if Apple signals supplier orders or a 2026 pilot (establish 2–3% position if supplier bookings >10M units by mid-2026). Conversely, short-duration short ideas include META on execution risk for consumer AR (10–15% downside if monthly active usage fails to grow Q/Q). Use options to express convexity: buy 6–12 month NVDA calls (target 20–30% move) and sell short-dated META covered calls to harvest premium.

Contrarian angles: The market assumes Apple’s pin will be a 20M hit—this is binary; a 50% miss (10M) would reduce the implied revenue runway by ~$1–2B and disproportionally penalize suppliers. OpenAI’s enterprise GM push could concentrate revenue with a few large contracts, increasing client-concentration risk and bargaining power that could pressure per-seat pricing over 12–24 months. Historical parallel: early smartphone accessory winners were small and fragmented—expect many startup winners rather than a single dominant application-layer victor.

Key Decisions for Investors

• Establish a 2.5% long position in AAPL (size relative to portfolio) IF Apple confirms >5M initial manufacturing orders by end-2026 or supplier bookings indicate 10M+ units; trim to 1% if preorders/filings disappoint within 90 days.
• Add a 3% long position in NVDA now and size up to 5% on any pullback >12% from current price; complement with 6–9 month ATM call options (target 20–30% upside) to capture continued GPU demand into Nvidia GTC and FY earnings.
• Initiate a relative-value pair: long GOOGL 2% / short META 2% to express conviction that Google’s voice/IP gains and enterprise positioning will outcompete Meta’s consumer AR execution over 12 months; rebalance if META reports MAU growth >5% quarter-over-quarter.
• Sell covered calls on META (1–2% notional of portfolio) with 3-month expiries at strikes ~8–12% above spot to monetize elevated uncertainty while retaining upside; if implied vol spikes >30%, close within 10 trading days.
• Monitor regulatory and supply-chain catalysts actively: if EU/US regulatory proposal limits always-on sensors or if key sensor suppliers report production shortfalls >15% vs. plan, reduce AAPL exposure by half within 30 days and rotate 1–2% into enterprise software names (GOOGL, GS replacement) focused on voice/agents.