At the Web Summit in Lisbon , David Risher , CEO of Lyft, delivered a position that contrasts sharply with the prevailing technological euphoria.
Far from the scenarios where autonomous taxis crisscross our cities, he believes that reality will significantly slow their deployment. And according to him, platforms like his will continue to rely on human drivers for much longer than predicted.
Delivering a very clear speech, David Risher detailed the technical , economic , and societal limitations that still significantly hinder the adoption of autonomous driving. A surprising message in a sector often presented as being on the verge of a revolution…
A technology far from being ready for real life
On stage, in front of an audience largely convinced that AI can solve everything, the head of Lyft reminded everyone of a point often overlooked: autonomous vehicles still perform very poorly in the face of rain, snow, or heavy traffic.
According to him, manufacturers are not entirely ready , nor are regulatory authorities, who are proceeding cautiously, so much so that several US states are still closing the door to widespread adoption, due to a lack of robust guarantees on safety and liability.
But the obstacle could also come from the users themselves. In his presentation, David Risher believes that the majority of passengers remain wary , and he states that he would be surprised if autonomous cars represented even 10% of Lyft’s volume by 2030 .
An economic equation that is far from favorable
Beyond the technical constraints, it is primarily the economic model of total robotization that raises questions. While, on paper, replacing drivers seems to reduce costs, in practice, the logic appears to be quite different.
Autonomous vehicles today cost between $250,000 and $300,000 per vehicle , compared to $30,000 to $40,000 for a conventional sedan driven by an independent contractor. Maintenance, cleaning, insurance, and energy costs are additional and would be entirely borne by the platform .
David Risher points out that these fleets would ” sleep ” at night, when demand is at its lowest, while simultaneously losing value on extremely expensive assets. This is an unappealing prospect for a company whose business model relies on flexibility and cost control.
In his view, it would be much more efficient to continue relying on a network of drivers , who own their vehicles and rent them according to the platform’s needs, but for how much longer?