
The car that drives itself — and the trillion-dollar question of who owns the road
The autonomous vehicle is not a single technology. It is a stack — sensors, compute, mapping, prediction, planning, and actuation — all required to function simultaneously and flawlessly at highway speed, in rain, in construction zones, behind unpredictable human drivers.
Progress has been uneven across that stack. Sensor technology matured rapidly. Perception models achieved superhuman performance on benchmark datasets. But the edge cases — the scenarios that appear once per million miles — proved harder than the core problem. Self-driving development has consumed more capital and more time than the most aggressive early projections imagined.
What has arrived, however, is commercially real. Robotaxi networks operate without safety drivers in Phoenix, San Francisco, and Austin. Long-haul autonomous trucking has logged millions of miles commercially. The technology works at the center of its distribution. The question is how fast it expands to the tails.
The business model transformation is as significant as the technology itself. If vehicles drive autonomously, they can be owned by fleets rather than individuals. Transportation becomes a service. Parking infrastructure becomes obsolete. Insurance reprices entirely based on fleet data. Urban planning rewrites itself around throughput rather than storage.

Synthetic emotional connection at scale — and the questions it forces about loneliness, intimacy, and what relationship means

The self-optimizing factory — where machines monitor, predict, and adapt without human intervention

The internal combustion engine's century-long dominance ends — and the grid absorbs the transition