一百年前，许多城市已经围绕汽车重新设计道路和土地用途。文中引述Henry Grabar的观点，指出美国每辆车配置的停车空间已超过每个人分配到的住房空间，形成明显的空间不平衡。Robotaxi与robobus已在多个中国与美国城市常态化，现正扩展到如Zagreb与London等欧洲城市，Waymo是代表案例之一。文章认为自动驾驶车的到来是历史分岔点：若管理得当，城市可扭转以车辆为中心的布局；若缺乏规制，既有问题可能恶化。

在正向情境中，政府及早修法，优先放行共享型自动化运输（robotaxi、robobus与自动配送车），并抑制私人持有。需求从持有私车转向叫车：城市中几乎所有停车位都可能变得冗余，转作公园、脚踏车道和住宅用途。较小的单人车厢、紧凑车距行驶，以及不间断营运与夜间停置于城外，可在施加车队数量上限时提升通行效率并减少塞车。无人车队若全面电动化，将降低城市排放并减少事故。根据American Automobile Association，美国新车年度拥有与营运成本为11,577美元。政府还可对空转行驶（deadheading）课税并要求提供交通流量资料，用于补助失去职位的巴士与计程车司机转向远端监控、车队调度与维修维护工作。

在负向情境中，若监管跟不上，私人自动驾驶车大量放行，起初主要由富裕阶层购买，停车需求仍在，且因可「不必亲自驾驶」而推升行驶里程与车辆使用。道路更拥塞，公共运输因搭乘率下降与财务萎缩而更弱，进一步推高持车需求，通勤距离被延长，导致郊区蔓延加剧并侵蚀高密度城市基础设施。最终受害最大的是无法负担车辆的低收入人群，等于科技不是自动让城市更好，而是让既有的不平等与低效扩大。

About a century ago, cities reshaped roads and land use around the car. The text cites Henry Grabar’s point that in the United States the parking space allocated per car exceeds that allocated for housing each person, showing a severe spatial imbalance. Robotaxis and robobuses are already operating in many Chinese and American cities and are now spreading to European cities such as Zagreb and London, with Waymo a major example. The article treats autonomous vehicles as a historical fork: if managed well, cities can reverse car-first planning; if poorly regulated, existing urban problems are likely to worsen.

In a positive scenario, governments rewrite laws early, prioritize shared autonomous modes (robotaxis, robobuses, and autonomous delivery), and discourage private ownership. Demand shifts from owning cars to calling rides; nearly all urban parking could become redundant and be repurposed for parks, bike lanes, and housing. Smaller one-person vehicles, close-headway operation, and round-the-clock service with overnight parking outside city centers can raise throughput and reduce congestion, especially if fleet caps are enforced. Electric autonomous fleets would cut emissions and accidents. The American Automobile Association estimated annual ownership-and-operation cost of a new U.S. car at 11,577 USD. Authorities can tax deadheading and require traffic-flow data sharing to fund retraining of displaced bus and taxi workers into remote support, fleet coordination, and maintenance roles.

In a negative scenario, regulators lag and private driverless cars proliferate, initially purchased mainly by wealthy households. Parking remains necessary, vehicle usage rises because driving becomes painless, and roads get clogged. Public transit, already vulnerable, loses passengers and funding, pushing more people toward car purchases and longer commutes; suburban sprawl then expands, weakening dense-city infrastructure. The worst losers are low-income households unable to afford vehicles. Thus, the technology can either help redistribute scarce urban space toward people or amplify inequality and inefficiency if allowed to spread without strong policy constraints.