Counterpunch on the Dangers of the Driverless Car

Ralph Nader in an article posted on Tuesday’s Counterpunch took to task the current hype about driverless cars following a day long conference on them at Washington University’s law school.

Driverless cars are being promoted because sales are cars are expected to flatten out due to car-sharing, or even fall as the younger generation are less inclined to buy them. Rather than actually investing in public transport, the car industry is promoting driverless automobiles as a way of stimulating sales again.

Nader is rightly sceptical about how well such vehicles will perform in the real world. There are 250 million motor vehicles in the US. This means that real driving conditions are way more complicated than the simple routes on which these vehicles are developed and tested. And while the car industry claims that they will be safer than human-driven vehicles, the reality is most people won’t want a car that they can’t control, whose guiding computer can malfunction or which could be hacked into, whether by the manufacturers or others. Along with such vehicles come increased pressure from the manufacturers for less protective regulation for their drivers, passengers and the general American public.

As for developing a driverless car, which can be taken over by its human occupant in an emergency, researchers at MIT have already shown that this won’t work. Instead of producing driverless cars, we’d be better of concentrating on creating vehicles that are safer, more efficient and less harmful to the environment.

He concludes

The driverless car is bursting forth without a legal, ethical and priorities framework. Already asking for public subsidies, companies can drain much-needed funds for available mass transit services and the industry’s own vehicle safety upgrades in favor of a technological will-o’-the-wisp.

He also links to a report by Harvey Rosenfeld into the dangers posed by driverless cars. It’s quite long – 36 pages. This makes it very clear, however, that driverless cars are disastrous. They’re literally a car crash waiting to happen. The report also claims that much of promotion of such vehicles comes from the insurance. Although driverless cars are likely to be much less safe than ordinary cars, the claims that they will be less liable to accidents will allow insurance companies to charge higher premiums for those driving ordinary vehicles.

Driverless cars have been under development since the 1980s, but I can’t see them becoming a viable reality any time soon. Last year the industry proudly announced two types of driverless car, one of which was called the Tesla, after the great Serbian physicist and inventor of Alternative Current. These were withdrawn after accidents in which people were killed.

I have to say, I don’t know anybody who wants one. The various pieces I’ve read about them say that for their owners, using them will actually be quite boring. This is despite the claims that people will be able to read, work or relax instead of driving. But you can do all of that anyway by simply travelling by bus, rail or tube.

Then there’s the threat of unemployment. Two-thirds of all jobs are expected to be lost to automation in the next decade. There are about 40,000 truckers in Britain. That’s 40,000 people, who may lose their jobs if driverless lorries are every produced. And they have been trying to develop them. I can remember Clarkson nearly bursting with excitement while riding in one during an episode of Top Gear a few years ago.

My guess is that the reasons behind their development is far more sinister, and almost certainly connected to the military. For years the military has been trying to develop autonomous, robotic weaponry. I’ve blogged about some of the war robots that have been created and which were featured a few years ago in the popular science magazine, Frontiers. These included various types of jeep, which had been altered so that they carried guns. Such machines have been under development for a very long time. Kevin Warwick, a robotics scientist at Reading University, describes how the US army created a robotic jeep equipped with a machine gun way back in the 1950s. This looks like another step along the way to producing the type of autonomous war robots, which Warwick and some other robotics scientists fear may pose a very real threat to the human race as they become more advanced and their intelligence greater. We are creating war machines very close to the Daleks of Dr. Who or The Terminator franchise.

It also seems to me that the navigation software and computer hardware needed for driverless cars will also find a major general applications in other types of robot. Despite claims by some neuroscientists that the human brain is a load of inefficient ‘kluge’ created by blind evolutionary forces that select for survival, rather than particular skill, cyberneticists have found it very difficult in practice to replicate the way living things, from insects all the way up to humans, actually navigate their way around the world.

Think about the way robots have to work their way around objects. They have to estimate exactly how far away the obstacle is, then work out a path around it, all done using maths. A human, meanwhile, rather than estimating how many steps it takes to the object, and then planning a path of some many steps, precisely laid out, are needed to walk around it, simply does it.

An example of how difficult robots actually find such navigation in practise was given by Warwick when he appeared at the Cheltenham Festival of Science over a decade ago. There’s been a kind of robot Olympics held in Reading. The various competing teams had tried to produce robots that could navigate their own way around the town. Warwick’s team had created a robot with an infrared detector, which would simply follow the light source planted on the back of the human running in front of it. Which to me sounds very much like cheating.

All went swimmingly, until suddenly the robot veered off the road and started shooting off somewhere else entirely, before it collided with something, fell over and stopped. Warwick and his team wondered what happened until it finally occurred to them that the robot had fixed on that big, infrared light source in the sky, the Sun, and ran off trying to pursue that.

This was a decade or so ago. I’ve no doubt that things have improved since, but I doubt that they’ve improved quite so much that driverless cars, or completely autonomous robots, are going to be appearing in the next few years.

And until they do, I shall be very suspicious of the hype.

Tags: Buses, Cars, Cheltenham Festival of Science, Dr. Who, Harvey Rosenfeld, Insurance Companies, Kevin Warwick, MIT, Neuroscience, Nikola Tesla, Pollution, Public Transport, Railways, Ralph Nader, Reading, Reading University, Robots, Washington University