As Waymo expands its fully driverless ride-hailing service to more US cities and Chinese firms WeRide and AVOMO, together with Uber, secure Spain's first national permits for L4 autonomous passenger vehicles on public roads, the robotaxi is no longer a showroom demo but a commercial operation fighting for road rights, orders and cost efficiency, according to a commentary published by Chinese tech outlet Huxiu.
The article, written by an investor under the pen name Lingxiao, argues that the more consequential question for ordinary people is not which company enters which city, but whether falling per-kilometre costs of autonomous driving will make them sell their private cars. "When driverless travel becomes cheap enough, the residual value and ownership logic of private cars will be reassessed," the author writes.
The cost curve
The piece notes that capital has shifted from funding algorithm demos to demanding commercial validation. Pony.ai, for example, has set a target to bring the total cost of its seventh-generation robotaxi, including the vehicle and autonomous-driving kit, to below 230,000 yuan (about $32,400) by mid-2027. That is a company goal, not industry reality, but it signals the direction: when hardware, driver and operating costs cross a threshold, machines become cheaper mobility.
Early L4 robotaxis struggled to scale not just because algorithms were immature but because they were too expensive. Lidar, compute boards and redundant systems pushed up vehicle costs, making it hard to recoup investment through fares alone. Now solid-state sensors, mass-produced chips and platform-based vehicle architectures are lowering bill-of-materials costs, while end-to-end models reduce reliance on high-definition maps and hand-written rules, the article says.
From asset to expense
The investor's central argument is that most private cars sit parked for most of the day while still generating depreciation, insurance, parking and maintenance costs. A robotaxi, by contrast, is designed to run longer hours and spread depreciation over more paid kilometres. It may not replace private cars soon, but it will first make "buying a car versus buying a service" a real arithmetic question in dense cities and airport transfers.
"Buying a car used to be a symbol of middle-class life upgrade. In the future, it will increasingly become a pure consumption expense," the author writes. "You can pay for convenience and driving pleasure, but don't default to treating it as an asset."
The piece also sketches a future division of labour in the industry: technology companies supply the "virtual driver," platforms control users and dispatch, and asset operators handle vehicle purchase, charging, cleaning, maintenance and insurance. Software capabilities may become standardised, but physical-world assets such as licences, fleets, stations, energy and operational data will not get cheaper automatically, the author argues.
For individuals, the advice is to avoid leveraging up on rapidly depreciating consumer goods and to build skills that are hard to replace by algorithms. "I don't think private cars will disappear, but their financial attributes are being stripped away, leaving only use value and emotional value," the article concludes.
