Unlock free Digest editor
Roula Khalaf, editor of FT, chooses her favorite story in this weekly newsletter.
The path to independent cars is paved with promises to save people from themselves (and each other). Waymo, Google’s Samovoze Project, he says that “status quo safety on roads is unacceptable” and that autonomous driving “can save lives.” Elon Musk, no one would ever surpass, said In October, when he discovered Teslin Cybercab that autonomous cars would become “10 times safer than a man” and “save lives – like, a lot of life.”
It is a worthy goal and one that sounds achievable. After all, people are terrible drivers, aren’t they? 2022, the last year for which there is detailed information, 42,514 people were killed in traffic accidents on US road motor vehicles, according to National Road Traffic Safety Administration. Of these deaths, 29 percent were killed in collisions related to the ride too fast, and 32 percent were classified as “fatal driving cases with alcohol damage”.
Fully independent cars, in contrast, do not drink and drive. They do not accelerate because they are impatient or late suffering. They do not become sleepy. They don’t interfere with phones. On top of all, they have a vision of 360 degrees. “You have to have eyes in the back of my head,” I remember the truck driver once told me. Autonomous vehicles actually work.
But for companies that make vehicles, there are two problems with their promises to save lives. The first thing is that human drivers have set a tape more than you might think. Once you think about how many kilometers we drive, deaths are actually very rare. The mortality rate per 100 mn kilometers of vehicles left to the US was only 1.33 in 2022. She was close to 1 about a decade.
The latest information of Waymo shows By the end of September last year, his vehicles transferred only 33 million miles without a human driver. This means that he simply did not drive anywhere near the miles to make a statistical comparison with human drivers when it comes to mortality rate.
Accordingly, Waym’s record of less serious accidents looks encouraging. Company data analysis He suggests that his vehicles so far had fewer collisions that cause injuries and collisions reported to police compared to people driving in the same cities. Waymo told me that “it already significantly reduces the number of serious collisions compared to human drivers in the places we operate.” But Phil Koopman, a vehicle safety expert, warns about Saucešće. “We know the computers fail differently from people,” he told me. “Human drivers fail individually but [with] Computers, every car has the same driver, so something can happen that everyone does not succeed. “Besides, he says, computers may not get drunk, but they don’t even have” common sense. “
This brings us to another problem. Even if the vehicle’s self -management companies collect enough information to show that their technology saves lives, there are still widely published examples with each time that cars just drive make bizarre mistakes that most people just wouldn’t make.
There was a Waymo taxi that was driving lap For example, in circles in the parking lot with a confused passenger. Or time not to not one but two Waymos ran into The back of the same truck that dragged itself down the road at a slightly unusual angle, confusing software. The most serious, there were times when a pedestrian was thrown Another vehicle on the car’s path for independent guidance General Motors, who then pulled her by 20 feet below. In this case, a subsequent review ordered by Cruise establish That “a waking and careful human driver is aware that there was some impact and would not continue to drive without further exploring the situation.”
That could be unfair, Koopman says, but it is human nature that stories like these stages in the mind more than statistics. If your point of sale is “our cars are safer drivers than people”, every time one of your cars does something strange that a person would not do, your public case is eroded. In contrast, accidents that did not occur thanks to your technology rarely bring news.
That could be a reason Annual surveys The American car association shows that people are increasingly afraid of technology over time. The proportion of American drivers who say they believe in self -driven vehicles dropped from 14 percent to 2021. At 9 percent in 2024, while a share that is said to be “scared” from 54 to 66 percent.
Self -driving companies could say that they want to save lives, but it will be harder than they think they convince people to let them try.
