The robotaxi sector is no longer a laboratory curiosity; it is a live demonstration of competing principles operating at scale. Alphabet’s Waymo has built a dense, sensor-laden circuit of operations, while Tesla propagates a leaner, vision-driven field across multiple cities. A dense accumulation of recent reports 1,17,20,24,51 allows us to map these lines of force—technological approach, geographic reach, operational scale, and economic pressure—and to identify where the current flows smoothly and where resistance appears.
Waymo’s Established Circuit of Operations
Let us begin with the observed phenomena. Waymo’s domestic fleet comprises some 3,000 to 4,000 vehicles 1,17,20,24,51, though one outlier measurement claims a mere 700 37. These vehicles, predominantly Jaguar I-Pace and Zeekr vans equipped with lidar, radar, and cameras 21,22,34, are the product of 17 years of development 18 and over $20 billion in investment 40. The scale of their operations is substantial: approximately 500,000 paid rides are delivered each week 2,6,7,8,9,10,11,20,37,39,40,41,51, with the aim of reaching one million per week by year-end 3,12,13,20,24. Total autonomous miles have surpassed 220 million 5,18,29,35, and rider-only miles alone reached 49.9 million in the first quarter of 2026, averaging 3.8 million miles per week 35,41. More than 20 million total rides have been completed to date 20,24, across a service area spanning over 1,400 square miles in 11 U.S. metropolitan areas 4,14,20—a geographic footprint that noticeably exceeds Tesla’s present reach 41.
From an economic standpoint, Waymo is not yet a profitable whole 34,40, but it has achieved operational profitability in mature markets like San Francisco 37. This suggests that as the network densifies, unit economics may improve: cost-per-mile is estimated between $0.30 and $1.50 37, while revenue-per-mile can reach $4.00 37, with pricing akin to Uber Black 34 and dynamic demand-based adjustments 37.
Tesla’s Propagating Field of Unsupervised Operation
Tesla, by contrast, has introduced a different experimental apparatus. Its robotaxi service has been set in motion across six metropolitan areas: Austin, Dallas, Houston, Miami, Orlando, and Tampa 26,43,44,45,46,48. Austin, in particular, has been mapped in detail, covering its 245-square-mile metro area 25,36. Preparations are also evident in Phoenix and Las Vegas 15. In the San Francisco Bay Area, however, the field is not yet fully autonomous—a safety driver remains in the vehicle 26,44,47,49.
The fleet currently consists of lightly modified Model Y SUVs 16,22 and purpose-built Cybercabs 22,31,33, all relying solely on Tesla’s camera-only Full Self-Driving software 23,52. Tesla reports zero notable incidents during operations 23,50, and its pricing undercuts Waymo significantly: typical one-way trips cost $13–$14 compared to Waymo’s $22–$25 37. The company manages charging and cleaning in-house 37, pursuing vertical integration 19.
Points of Resistance: Ambiguities in Autonomy and Scale
When we probe the experimental record more closely, certain contradictions emerge. While many claims assert fully unsupervised operations in the six cities 42,52, others indicate that a safety monitor remains present in Austin 36. This discrepancy may reflect varying phases of the rollout or differing definitions of “unsupervised,” reminding us that terminology can obscure the true nature of the apparatus. Similarly, fleet size estimates for Waymo diverge: the more corroborated and recent figures point toward 4,000 vehicles 17,24, while the 700-vehicle claim stands as an outlier 37. Profitability likewise resists simple characterization—overall unprofitability 40 coexists with operational profitability in specific markets 37. These nuances are not flaws in the data but rather invitations to more rigorous inquiry.
Economic Currents and Market Disruption
The comparative economics tell a story of two competing philosophies. Waymo’s sensor-heavy, capital-intensive model is pitted against Tesla’s lean, vision-based approach. Tesla’s lower price point positions it as a mass-market alternative, which could accelerate adoption but also raises questions about its own path to profitability. Meanwhile, Uber finds itself a weakening node in the network. It is years behind in autonomy 38 and is resorting to legislative tactics to block independent robotaxis 38, even while placing a 10,000-unit order for electric vehicles from Rivian 14. Waymo’s plan to launch its own app in January 2028 and to terminate its Uber contract in Atlanta and Austin by May 2028 16 could further isolate the incumbent, leaving it disconnected from the primary current of innovation.
The Path Ahead: Implications for Tesla
What can we deduce from these observations? First, Waymo’s lead in cumulative miles (over 220 million) and weekly rides (500,000) establishes a high bar for safety validation. Its accident rate is substantially lower than that of human drivers 16,30,32, a testament to its experimental rigor. Tesla, with far fewer miles, must demonstrate that its camera-only system can achieve comparable safety at scale. Second, Tesla’s aggressive pricing and rapid city expansion suggest early operational momentum, but the small fleet size 17,28 and reliance on consumer-grade vehicles may constrain growth relative to Waymo’s purpose-built fleet and its planned Ojai vehicle with sixth-generation hardware 20,30.
The broader field is fragmenting into multiple closed ecosystems—ride-hailing incumbents like Uber face disintermediation, while Waymo eyes international expansion into London 24 and Germany 27. For Tesla, the critical experiment is to close the gap in operational experience and regulatory trust while preserving its cost advantage. Investors would do well to monitor regulatory developments and safety data as key differentiators. Early mover advantage in geography currently favors Waymo, but if Tesla’s lean approach proves reliable at scale, it could reconfigure the entire field. As always, the burden of proof lies with the demonstrator.