Novel now, big scale shortly; AV speed reinforced by Bay Area trip

Strategic Context & Market Status

• Autonomous driving is positioned as the "first large-scale commercial application" for physical AI, analogous to chatbots for generative AI.
• Waymo currently hosts over 500,000 paid autonomous rides per week in the U.S., representing a doubling of volume compared to the prior year.
• Global commercial robotaxi fleet grew from 1,000 vehicles in 2024 to roughly 7,000 by end of 2025, with current estimates of 8,000–10,000 globally (4,000–5,000 in the U.S.).
• Despite rapid growth, robotaxis represent less than 0.1% of the global taxi/ride-hail market (20 million vehicles) and under 1% of the U.S. market (2 million vehicles).
• China-specific projections suggest a potential 400,000 robotaxis by 2030, with global forecasts pointing toward millions within the next decade.

Technological Shifts & Scaling

• Industry focus has shifted from "can it work?" to "how quickly can it scale," with skepticism regarding basic functionality largely replaced by concerns over manufacturing, safety validation, and fleet operations.
• The sector is transitioning from "AV 1.0" (rule-based, hand-coded systems) to "AV 2.0" (AI-native architecture using end-to-end AI, foundation models, and simulation).
• AI models now allow for generalization across new locations with minimal training, reducing reliance on city-specific mapping and enabling expansion beyond geofenced urban areas.
• Expansion is expected to move beyond dense city centers to include airport routes, suburban corridors, and autonomous freight lanes.

Economics & Pricing

• AV ride pricing is declining; while some markets currently command a premium, Tesla's RoboTaxi is already priced approximately 30% lower than Uber/Lyft to build consumer base.
• Target cost for Tesla's purpose-built CyberCab is projected at $0.30 per mile at scale, significantly undercutting human-driven ride-share economics.
• Vehicle costs are dropping, with the CyberCab under $30,000 and converted Model Ys used in fleets well under $50,000, aided by vision-only sensor approaches.

Business Models & Competitive Dynamics

• No single business model has emerged; multiple approaches are coexisting in parallel:
  • Vertically Integrated: Companies like Waymo, Tesla, and Zoox build tech, operate fleets, and manage customer experience.
  • Licensing/Software-First: Firms like Waymo partner with OEMs to provide the intelligence stack while partners handle vehicle manufacturing and operations (e.g., Lucid/Neuro).
  • Hybrid Partnerships: Mixed models combining vehicle, tech stack, and network providers (e.g., Uber, Lucid, Neuro).
• Debate is emerging regarding "personally owned autonomous vehicles" that can join ride-hail networks when not in use, blurring lines between private ownership and fleet services.
• Winners will likely be determined by execution in scaling manufacturing, safety validation, and fleet operations rather than solely by technology stacks or demo performance.

Impact on Traditional Auto & Ownership

• In the U.S., personal vehicle ownership is not expected to decline significantly in the near term; autonomous features will become a primary purchase criteria for OEMs.
• Point-to-point autonomy (L2+) is expected to become a commoditized feature within 5–10 years, potentially shifting revenue from high-margin features to subscription models or base vehicle pricing.
• Supplier benefits are expected to extend beyond OEMs to full-stack tech providers (e.g., Waymo, Neuro) and traditional forward-thinking suppliers (e.g., Aptiv).

Autonomous Trucking & Logistics

• Autonomous trucking is further along in commercialization than consumer robotics, with three providers already operating driverless freight services on highways.
• Highways provide a more structured environment than dense urban centers, accelerating technological generalization.
• Economic drivers include a 2.5x increase in utilization (20 operational hours vs. 8–9 for humans) and mitigation of acute driver shortages.
• Potential to compress multi-day supply chain routes to under 24 hours, reshaping logistics networks and distribution centers within the $4 trillion global freight market.

Physical AI & Broader Implications

• Autonomous vehicles serve as a massive real-world training ground for AI, developing capabilities in perception, prediction, and decision-making applicable to robotics and drones.
• Key concepts driving the "Physical AI" narrative include foundation models, world models, and embodied AI.
• Companies are explicitly linking autonomy infrastructure to robotics development (e.g., Tesla leveraging autonomy data for Optimus robots; Waymo establishing dedicated embodied AI labs).
• Robotics and humanoids are expected to follow the adoption curve of autonomous vehicles from 5–10 years ago.

Forward-Looking Statements

• Most new consumer vehicles are expected to include at least point-to-point autonomous features within a couple of years.
• Robotaxi fleets are projected to grow from single-digit thousands to tens of thousands in the next few years.
• Regulatory approval for operations beyond dense urban geofences is identified as the primary constraint, with technology capable of scaling before regulators authorize broader deployment.
• Suburban expansion and autonomous freight corridors are expected to influence real estate development, work locations, and the concept of vehicles as "mobile office spaces."

Event Context

• Insights derived from a field trip to the Bay Area on September 16, 2026, featuring meetings with 15 corporations, including Tesla, Waymo, Zoox, and AVride.