Billionaire Ken Griffin Buys 2 AI Stocks Chasing a $1 Trillion Market Opportunity in Robotaxis (Hint: Not Tesla)

Robotaxis represent a $1T+ U.S. opportunity and Nvidia and Amazon are positioned as likely major beneficiaries; Ken Griffin held both as his largest Q4 positions. Nvidia is presented as indispensable to robotaxi stacks (GPUs plus Omniverse/Cosmos/Alpamayo and Hyperion), with Wall Street modeling ~38% EPS CAGR over three years and the shares trading at ~35x earnings; management says robotaxis could generate ‘hundreds of billions’ over the next decade. Amazon’s Zoox has delivered >350,000 autonomous rides, is testing in Austin and Miami, operates under an NHTSA demonstration exemption but has applied to run up to 2,500 commercial robotaxis with a decision due in early April, and Morgan Stanley projects Zoox could capture ~12% of autonomous rides by 2032; Amazon’s EPS is forecast to grow ~19% annually and the stock trades near ~29x earnings.

Analysis

Nvidia’s integrated stack (simulation + foundation models + optimized silicon) creates a structurally different cost curve for AV development: teams that adopt an integrated sim-to-deploy pipeline can cut validation time measured in real-world miles by an estimated 50-70% versus a mostly on-road approach, which translates to meaningful OPEX savings for fleet operators and accelerates time-to-scale. That creates durable vendor lock‑in: once a fleet operator standardizes on a single sim/model/hardware chain, switching costs (re‑label, re‑validate, and re‑certify) will be measured in years and tens to hundreds of millions of dollars per large operator.

Second‑order winners include TSMC/advanced foundry capacity and a narrowed cohort of Tier‑1 sensor integrators; demand for high‑end GPUs and co-packaged optics will lift a handful of suppliers and tighten component lead times, compressing margins for legacy Tier‑2s that can’t match integrated supply. Amazon’s combination of cloud, logistics and experiment-in-market economics lets it subsidize early rider acquisition and gather proprietary route/usage data, raising the bar for pure-play ride-hail competitors and forcing them into niche or commodity roles.

Key near‑term catalysts are binary regulatory approvals for commercial operations and quarterly cadence showing enterprise GPU deployment; both can re-rate winners quickly. Tail risks that would reverse the trade include a high‑visibility safety incident, a new accelerator delivering >2x perf/watt at lower cost (driving rapid share loss), or a prolonged datacenter capacity shortage that spikes GPU pricing; these risks play out on 0–24 month horizons, while material robotaxi cashflows are a 3–7 year story.

Key Decisions for Investors

• Long NVDA via 9–15 month call spreads (size 2–4% portfolio): buy deep ITM calls and sell higher strike calls to finance exposure. Rationale: captures continued datacenter upgrade + robotaxi premium while capping theta; target 30–50% gross return if NVDA sustains current growth; max loss = premium paid.
• Buy AMZN 18–36 month LEAP calls (size 2–3% portfolio) or use financed call structures: objective is asymmetric exposure to Zoox optionality plus cloud/ads upside. Risk/reward: expect 2–3x upside if fleet commercialization paths clear and AWS growth continues; downside limited to premium (~100% loss of option premium).
• Pair trade — long NVDA / short INTC equal notional (size 1–2% portfolio): play secular share shift in AI compute. Timeframe 6–24 months; reward: capture relative outperformance if NVDA maintains architecture lock; risk: INTC recovers via alternative roadmap or Intel wins OEM integrations.
• Tactical short UBER (or buy puts, size 1%): hedge exposure to ride‑hailing margin compression from robotaxi rollouts and Amazon cross‑subsidization over 12–36 months. Risk/reward: potential 30–60% downside if autonomous supply reduces per‑mile price materially; downside capped by regulatory/geographic rollout delays.