Driverless cars are about more than driverless cars

DRIVERLESS CARS ARE ABOUT MORE THAN DRIVERLESS CARS.

Ever since DARPA issued a challenge to roboticists to get vehicles to drive themselves across miles of desert terrain and machines like Google’s self-driving cars have taken to the highways of the USA and other countries, there have been speculations about what this technology could mean if and when it is perfected. Mostly, the speculation has focused on what it could mean for those of us whose profession entails driving a vehicle of some sort or other. If trucks can drive themselves, what does that mean for the jobs of all those truck drivers? Will we one day be summoning a driverless Uber taxi and, if so, what will become of the taxi-drivers who currently provide this service?

But, really, I don’t think these concerns go far enough. The problem is, they are always confined to jobs that explicitly involve driving a vehicle, as if this is the only sort of job that driverless vehicles could replace. I don’t think the technology will be restricted to just this area, and to see why one needs to understand what problem driverless cars are attempting to solve.

The challenge that driverless car technology is trying to solve is as follows: It is attempting to give robots the ability to successfully navigate complex, dynamic environments. In order to really achieve this goal, robots would have to have the capability to generate detailed, three-dimensional maps of their surroundings. They would also have to be capable of interpreting such data and responding appropriately.

Another way to put this would be to say that the challenge of driverless cars lies in making robots that can ‘see’. I put that in quotation marks because it is not necessarily the case that robots would have to see just like we do. After all, when bats use echolocation to pinpoint their prey they are not ‘seeing’ quite like we do, but they are building maps of their complex, dynamic environment and responding appropriately, so in a sense they do ‘see’ their world. So robots may come to be able to ‘see’ their environments in ways we do not. But, however they ‘see’ their world, they need to be as good as we are at negotiating their environments if driverless tech is to really take off.

In many ways, robots are already as good as humans at certain abilities. Videos demonstrating DARPA’s robots shows a very agile wheeled robot that can quickly move around its environment. You can also find footage of artificial hands that are just as dexterous as human hands. Indeed, some video clips show robot manipulators performing feats that would be beyond the abilities of people. But all these demonstrations only work under specific laboratory conditions, and often there is actually a human operator remotely controlling the robot (though if it is a bipedal robot it would have to take care of certain tasks autonomously if it is to avoid falling over). The reason why we can’t just attach super-dexterous hands to super-manoeuvrable wheeled robots and let them loose upon the world is because they lack vision.

But what if they had the ability to ‘see’? Why, then they could operate successfully in any environment. We would see robots on farms, robots in shops, robots in manufacturing, robots everywhere. And it would not just be driverless tractors and combine harvesters but robot fruit pickers, robots shelf-stackers, robot cleaners, robot brick-layers. In fact, once driverless technology is perfected and added to what we have already achieved with robot mobility and dexterity, there is not a single productive job that could not be automated away.

So that’s why I say driverless cars are about more than just driverless vehicles. Potentially, it is about automating away every productive job in existence. Does that mean all jobs will disappear? No. Work that entails caring for people is incredibly hard to quantify which means it is incredibly difficult for computers to handle. Care-work may be the last jobs left for working-class people to do. This may not be a bad thing by any means, as it can be argued that caring for others is the most socially-rewarding work anyone can do. Unfortunately, our society does not place much monetary value on care work and it’s easy to imagine the value of such work will go down even more when there are tens of millions of people thrown out of other kinds of work and seeking new employment in the only sector left.

Then again I don’t want to be too pessimistic so maybe things will turn out much better?