Driverless cars may kill off the world’s deadliest invention

We often make lists of the best inventions of all time. But which are the worst?

The nuclear bomb would be up there. The AK-47 Kalashnikov assault rifle also has a particular claim to infamy. The World Bank has estimated that there are about 75m such guns in circulation, causing tens of thousands of deaths a year.

But one other invention has killed many, many times more people than both of those weapons put together: the motor car.

About 1.25m people are killed in road traffic accidents a year, accounting for 2.2 per cent of all deaths globally, according to the World Health Organization. Accurate data on how many people in total have been killed by cars in the past century are hard to find. But 50m seems a fair guess. That compares with the 123m in all wars in the 20th century.

If that level of carnage were not bad enough, cars have also contributed massively to environmental pollution and adverse climate change. As an additional malus, the car’s thirst for oil has handed billions of dollars in revenues to some of the world’s most regressive regimes: Saudi Arabia, Russia and Venezuela.

Fortunately, the end of the human-driven, petrol-fuelled car may be in sight. Some computer scientists describe driverless car technology as essentially a “solved problem”, even if huge challenges remain before car companies are confident enough to launch fully autonomous passenger vehicles on to the road. Protocols still need to be worked out to determine how one driverless car will interact with another and how to prevent hacking.

But human, rather than technological, resistance seems likelier to be the biggest brake on the driverless car revolution. Whether we are behind the wheel or thinking about new technologies, humans remain inherently erratic.

For the moment, governments are putting in place regulations to encourage the adoption of driverless, electric cars and manufacturers are experimenting fast.

Last week, Philip Hammond, the British chancellor, said he wanted the UK to be one of the first countries to allow “genuine driverless cars” on its roads by 2021 (although he did not want to be photographed in one for fear of embarrassing political metaphors about the state of the British government).

Car companies are gearing up to deliver. Uber announced last week that it would buy up to 24,000 Volvo cars by 2021 to prepare a fleet of fully autonomous, on-demand passenger vehicles.

One of the biggest risks is that governments and car companies will rush into allowing poorly tested autonomous vehicles on to our roads prematurely, triggering a public backlash. No matter how good the technology becomes, autonomous cars are still going to kill people. “The assumption that you can go from 1m [road deaths] to zero is very naive,” says one tech executive.

Proving that autonomous cars are demonstrably safer than human-driven cars is therefore going to be a tough battle of statistics and public perception. The first death caused by a driverless car is certain to turn the likes of Jeremy Clarkson into a beetroot-faced, “told-you-so” motormouth.

The TV car show presenter has already shouted his doubts about the reliability of autonomous driving technology, challenging its promoters to sit in one of their cars as it drives down Death Road in Bolivia.

But he is right that motorists have an extraordinary attachment to their cars. The car has brought enormous mobility, liberty and enjoyment to millions of people. A huge number of jobs, investments and corporate interests are also dependent on the current model of car transport, no matter how dangerous it may be. For those reasons, some transport experts doubt whether the driverless car revolution will ever happen.

At the other extreme are those, like the technology think-tank RethinkX, who predict that “we are on the cusp of one of the fastest . . . most consequential disruptions of transportation in history”.

They forecast that 95 per cent of US passenger miles will be made by autonomous, electric, on-demand vehicle fleets within 10 years of receiving widespread regulatory approval as we move towards a “transport-as-a-service” model. By 2030, the number of passenger vehicles on American roads will drop from 247m to 44m, destroying millions of driving jobs and the oil industry.

Which scenario unfolds in the real world will be a matter of societal choice as much as technological possibility. We should triple-test the technology, weigh the evidence carefully and choose wisely. We have had enough of bad inventions.

john.thornhill@ft.com

Letters in response to this article:

Uber exemplifies the Tragedy of the Commons / From Tad Borek, San Francisco, CA, US

Driverless cars face long and winding road ahead / From Rod Stephens, Standon, Herts, UK

Can you think of a worse invention than the car? Leave your suggestions below.