On holiday in Venice recently, I found myself thinking a lot about driverless trucks. This may seem strange, because there are no trucks in Venice. Instead, everything is delivered by boat. At 6am every morning, for example, a large, broad-bottomed boat turned up and moored outside the small supermarket across the bridge from our hotel. The driver (or should I say captain?) then used the onboard crane to lift out heavy pallets of bottled water, boxes of provisions etc on to the quayside. Delivering the necessary supplies in this way took at least 30 minutes, after which he gunned the engine of the boat and headed off to make his next delivery.
Later that particular day, walking along the Grand Canal, we came across another large boat, this time emblazoned with the logo of DHL, the international couriers. The boat was piled high with cardboard boxes and two guys with clipboards were trying to sort out what needed to be delivered where. It was clearly a slow and painstaking process.
In both cases, the boats are the Venetian equivalent of the white vans that, in the UK, deliver stuff from Amazon et al. So, on my return, I whiled away an hour or two watching white-van drivers at work. It turns out that the driving is the easy bit: the real tasks are things like finding a place to park on a busy street, dodging traffic wardens, finding a house number or the right apartment in a block of flats, deciding what to do when there’s no one at home, getting a signature from a customer, and so on.
All of which helps to put the hype about driverless trucks into perspective. The narrative about autonomous vehicles that is being fostered by the tech industry and its media acolytes is that the technology is unstoppable and that it’s coming real soon now to a road near you. This is a subgenre of a bigger narrative about the inevitability of AI, robotics and automation generally. It’s basically technological determinism on steroids. And what it boils down to is this: if something can be done, then it will be done. And there’s nothing we can do to stop it, so anyone who is sceptical about the narrative is – wait for it – a luddite!
That this Manichaean, yah-boo discourse is idiotic, misleading and unproductive goes without saying. Technology has always been one of the forces that drive society, and technologies that historically went mainstream have usually done so because they offered great benefits to society. Think of public sanitation, vaccination, antibiotics, electricity, the internet – and, yes, maybe autonomous vehicles. Self-driving cars could make urban life safer, greener and more enjoyable, for example. And given that our economies have become totally dependent on logistics systems based on trucks, anything that makes trucks safer and more economical sounds like a good idea. Such vehicles could run 24/7 (no need for driver rest breaks), and at lower speeds (thereby saving fuel). And they would probably be significantly safer (because many accidents are a product of human error). And so on.
But while the potential benefits of new technology are great, that doesn’t mean it will automatically be adopted. In some cases, the social costs might greatly outweigh the benefits. But usually the biggest brakes on adoption come from economic and practical considerations. Consumers may not be the rational creatures of economists’ dreams, but most corporations are. They will invest in technology only if it makes economic sense to do so. One of the reasons the automation of jobs is happening so slowly at the moment is that, in important market segments, low wages make human employees cheap. Why invest in expensive, dextrous robots if you can have people on zero-hours contracts and pay them the minimum wage?
But the biggest impediments to automation are the practical difficulties that tech evangelists tend to ignore. Some of them have already sussed that self-driving cars are a distant prospect because their regulatory and infrastructural requirements are so complex. That’s why much of the excitement in the industry is now focused on trucks. It’s easy to see how autonomous “truck trains” could work on motorways, and indeed there have already been trials of such convoys.
The trouble starts when the vehicle has to leave the motorway in order to reach its final destination. Suddenly the truck faces the same obstacles as the self-driving car. So maybe it will be necessary to have human pilots to take it that last mile safely, just as ships have pilots to guide them into harbour. That’s also why we are unlikely to see autonomous white vans any time soon: their drivers do much more than simply drive – just like those DHL guys in Venice. So perhaps tech determinists need to revise their mantra: if something can be done, then it may be done – provided the economics and the practicalities are right.
