Confronting the Weirdness of a Waymo Future | Interesting Times with Ross Douthat

Eliminating 40,000 annual deaths

With driver error causing the vast majority of road fatalities and injuries, Miller argues that scaled automated driving would prevent nearly all deadly incidents currently accepted as normal.

Mid-2030s inflection point

Waymo has expanded from a handful of cities to over 15, with growth accelerating such that hailing a robotaxi will be completely normal by 2035.

Liberation of attention

Automation would return hundreds of millions of hours annually to Americans currently required to focus entirely on the road.

The sensor cost trade-off

Waymo uses expensive lidar, radar, and cameras for reliable bad-weather performance, while Tesla bets solely on cheap cameras to achieve manufacturing scale faster despite potential safety compromises.

Frequent interventions in Austin

Tesla's current robotaxi fleet requires safety drivers to intervene constantly, unlike Waymo's driverless operations, though Tesla's safety record remains inferior to Waymo's at comparable development stages.

Remote assistance reality

Waymo vehicles rely on human operators abroad and domestically to provide "digital breadcrumbs" when confused by unusual scenarios, rather than being fully autonomous.

Suburban opportunity vs rural exclusion

Robotaxis will likely serve suburbs through municipal subsidies but remain economically unviable for rural America, similar to current Uber limitations.

Rejection of efficient ride-sharing

Despite transport planners advocating shared rides to maximize road capacity, Americans overwhelmingly prefer solo robotaxi trips for privacy, threatening efficient traffic utilization.

Liability as the ownership barrier

Unclear legal frameworks for accident responsibility represent the single greatest obstacle to private self-driving vehicle ownership, surpassing cost or technological limitations.