TechCrunch reporters Sean O’Kane and Kirsten Korosec reported exclusively on June 3 that Uber plans to deploy 500sensor-equipped vehicles this year to collect real-world driving data for autonomous vehicle partners including Waymo, WeRide, and Avride. The vehicles are modified Hyundai Ioniq 5s fitted with 14 cameras, eight solid-state lidar sensors, and nine radars, all processed through Nvidia’s Dual Drive Thor autonomous vehicle computer. The hardware is not the story. The ownership question underneath it is.
Every mile driven by these vehicles generates information. Traffic patterns, construction zones, lane markings, pedestrian behavior, weather conditions, road quality — all of it becomes training material for autonomous systems. Modern artificial intelligence depends on massive datasets to improve performance, and real-world driving data is among the most valuable inputs in the industry because it cannot be fully simulated. Uber told TechCrunch its aim is to develop the world’s most geographically diverse set of autonomous driving training data, capable of collecting 2 million miles of high-fidelity data per month. The quality of future self-driving technology will depend largely on who accumulated the most accurate information, across the most diverse conditions, over the most years. Uber is positioning itself to be that source.
This creates a compounding advantage that is easy to underestimate. Companies that control proprietary datasets gain a structural edge because rivals cannot replicate years of accumulated information simply by entering the market later. The result is a race not for customers but for intelligence. Every additional mile strengthens systems that may eventually underpin transportation networks worth billions of dollars — and the communities whose streets, intersections, and daily movement patterns produced that intelligence will not share in the returns.
That dynamic reflects a pattern SSC has tracked across the digital economy. Social media companies derive value from communication. Search engines derive value from curiosity. Streaming platforms derive value from attention. Autonomous transportation extends that model into physical space. Public movement becomes a source of private value creation. The city itself becomes an information asset, and the gap between who generates that value and who captures it is the same gap SSC identified this week in Tesla‘s Austin expansion — a market position built on public infrastructure, announced as innovation, and owned entirely by the platform. Read that piece here: Tesla’s Austin Robotaxi Launch Is a Headline. The Fleet Size Is the Story.
The debate is no longer simply about autonomous vehicles. It is about who owns the intelligence generated through collective public life — and whether the communities producing that intelligence have any claim on what it becomes worth.