Introduction: The Problem Is Larger Than the Vehicle The development of autonomous vehicles has largely focused on making the individual vehicle more intelligent. Better cameras, better radar, better lidar, larger neural networks, better maps, more training data, and increasingly sophisticated systems for interpreting the environment around the vehicle have collectively pushed autonomous driving substantially forward. There is, however, a more fundamental architectural problem that additional compute alone cannot solve: the modern autonomous vehicle is still, for the most part, attempting to understand a shared environment from an individual perspective. A vehicle observes the world around it, attempts to identify what those observations represent, predicts what the objects it can observe are likely to do, and then determines how it should respond. In many ways, we have recreated the fundamental limitations of a human driver inside a substantially more capable computational system. The machine may have cameras facing in every direction. It may have radar, lidar, ultrasonic sensors, GPS, high-definition maps, and neural networks capable of processing enormous quantities of information in real time. It may react considerably faster than a person. It is nevertheless primarily reasoning from what that particular vehicle can perceive. This creates an interesting contradiction. Roads are inherently shared environments. Every vehicle, pedestrian, cyclist, traffic signal, construction zone, emergency vehicle, and piece of transportation infrastructure exists within the same physical system. Their actions affect one another continuously. Yet we are attempting to automate this shared environment by creating thousands or millions of independently intelligent participants, each reconstructing approximately the same environment for itself. The alternative is to stop treating autonomous driving exclusively as a vehicle problem. It is an ecosystem problem. How Autonomous Vehicles Generally Work Today Although implementations differ considerably between manufacturers, autonomous driving systems generally follow a recognizable architectural pattern. The vehicle begins with sensors.