The driverless car and the fall of man
The quest for robotic cars is underwritten by a suspicion of humanity.
One of the most interesting things in the UK chancellor’s Budget in March was the announcement of an extra £100million of funding for research into driverless cars. Coming on the back of several similar initiatives globally (like Intel’s $100million Connected Car Fund and the well-known Google self-driving car programme), it seems like the driverless car is well on its way to becoming a reality in the 21st century. But what are the implications of driverless cars?
One putative implication became clear in the aftermath of the Budget, when leading taxi app Hailo launched a campaign called ‘Face to Faceless’, which consists of a series of portraits of taxi drivers projected on to London landmarks. It is meant to highlight the role of the driver and ‘remind the city that cabbies are part of its DNA’. Now, whatever your attitude towards London cabbies (and, yes, we all know about those ones that deliberately choose congested routes in order to up the journey fare and thus deserve every bit of the kicking a service like Uber is giving them), this reaction is important because it shows that one implication of driverless cars is their potential to disrupt – disrupt the labour market; disrupt the insurance industry; and disrupt our urban landscapes.
New technologies and technological progress always disorient us and pose a threat to existing practices. New technologies introduce uncertainty. We know from historical experience that the benefits and perceived deficits of a new technology are never distributed equally and that it is impossible to anticipate all the outcomes of the technology’s introduction. Some lose, some gain. The blacksmiths sung the praises of the coming of the automobile, but not many survived its ascendancy. Today’s cabbies might very well be the equivalent of nineteenth-century horse-and-cart men, blacksmiths and all the other related trades that were swept aside as the automobile conquered the world. The arrival of the driverless car does appear to represent such a moment, if not for society as a whole, then certainly for those parts of society that rely upon present-day transport arrangements.
Cabbies are not the only ones to raise concerns about driverless cars. Academics, philosophers, sociologists and policymakers have also raised important questions about the ethical and moral dimensions of autonomous cars: will these vehicles be programmed to act ethically, and whose ethics will they have? Indeed, some have asked if machines are capable of ‘having’ moral agency. Such concerns have led the UK’s leading robotics labs to get together and, with £1.4million of public funding from the Engineering and Physical Sciences Research Council, embark on a four-year project to lay down ground rules to ensure autonomous systems act reliably, safely and ethically when they interact with people. And the motorcar industry is teaming up with philosophy departments in universities across the world to bring an ethical framework to artificial-intelligence systems.
Enabling a complex machine like a car to act autonomously, and navigate a complicated infrastructure without human intervention, is an incredible feat. The convergence of sensor technologies with connected-vehicle communications for collision avoidance and traffic management, underpinned by faster and faster computing power, is a shining testimony to human ingenuity and ambition.
But just because such technology exists, or appears to exist, that doesn’t mean we should stop questioning the driverless car and its implications for the future. A legitimate starting point is to ask: what problem is it trying to solve?
A ‘solution for all seasons’?
Academics, the motor industry, technology companies like Google and Intel, and professional services businesses have all celebrated this new technology. As they see it, driverless cars will solve a huge number of intractable social problems: they will reduce the high cost of traffic crashes, which not only result in needless and meaningless deaths but also put pressure on creaking health services; they would decrease the need for the building and maintenance of transportation infrastructure, as we would no longer need existing traffic controls or lighting; they would reduce the millions of hours wasted in traffic jams, increasing productivity, and reduce stress and petrol consumption; and they would allow us to cut significantly the waste of urban space given over to parking lots – this would allow us to design our emerging megacities to be more efficient and pleasant environments in which to live and work.
There is no denying that these are real issues. According to research from the American Automobile Association (AAA), traffic crashes cost the US $300 billion annually, while it spends $75 billion annually on roads, highways, bridges, and other infrastructure designed to accommodate the imprecise and often unpredictable behaviour of human-driven vehicles. Self-driving vehicles, with the ability to ‘platoon’ – perhaps in special express lanes – would, it is suggested, bring an end to battles over the need for (and cost of) high-speed trains like HS2. Traffic congestion costs Americans 4.8 billion hours of travel delay each year – the equivalent of more than $100 billion annually in delays and needless fuel consumption, $23 billion of which can apparently be attributed to congestion caused by truck journeys.
If all of these claims are to be believed, the only problems the driverless car is not going to solve are the common cold, cancer and world peace. It appears that this ‘technology for all seasons’ has the potential to help solve some of the more intractable social and economic problems of the 21st century. But is this really the case?
Historical experience tells us to be sceptical about these claims. Technologies have an unerring tendency never to perform as their initial architects envisaged. But there is a more fundamental reason to stop and think: when all these claims are examined, they reveal one worrying underlying common assumption – namely, that human error and weakness are the real problems that need fixing.
People are seen as highly unpredictable and potentially dangerous. They drink and drive, they lose concentration, they show off, they compete, they get stressed, and, as a result of all these failings, and more, they are massively error-prone. To make the case for the driverless car, its advocates point to the fact that of the six million crashes each year, 93 per cent are attributable to human error. The costly infrastructure developed to accommodate these far-from-perfect drivers, they contend, results in high costs, wasteful petrol consumption and congestion. What they are really saying is that if we can solve the problem of human fallibility (that is, replace human drivers in cars with computers), then the environment will flourish, productivity will rise, the economy will prosper, and, therefore, society as a whole will benefit.
When discussed in these terms, the quest for the driverless car stops appearing quite so fantastically ambitious. In fact, it shows itself to be very much in step with contemporary misanthropy, a product of the culture of limits. It shows that far from representing a bold and progressive future-oriented vision, the driverless car is very much stuck in the present, as an attempt to automate what exists. This represents a diminishing of what it means to be human, a closing down, rather than a liberation, of society’s technological imagination.
Presentism and going back to the future
A major issue with driverless cars, which few advocates interrogate, is that the category is an historical misnomer. The evolution of the car is the story of the remarkable interaction between human agency and technical wizardry. The car was never built nor envisaged without its human driver. Both man and car evolved and, in so doing, created modern society as we know it. Just as it was not possible to see what impact the car would have a century later, we should not project our present-day knowledge and imaginations on to the future, as if we already know what the future will be like. The uncritical acceptance of the category of the ‘driverless car’ shows that our imaginations remain within the limits of contemporary society.
The question that needs posing is: if we remove the driver from the car, is it still a car? Yes, we will have a ‘carriage’ of some sort, which will be mobile. But this is no longer a car in the historical sense of the term. It avoids the history of the car and society and, more importantly, blinds us to what could be possible in the future. By projecting the category of a ‘driverless car’ into the future, we freeze the present and assume that all we can ever have is a slightly altered form of what we have today. With fantastic communications and sensor technology at our command, surely we can envisage a transport system, a mobile on-demand system of movement, that goes way beyond the car and how we travel today?
A critique of presentism is crucial to the discussion of the driverless car. The loss of an historical perspective is important because without it, without grasping what the car used to mean, we have no way of judging the meaning of the driverless car today. Presentism prevents us from posing the right questions. Surely, for instance, by removing human beings from the driver’s seat, we are changing not just the car but transport and moral agency, too? And surely this represents an unprecedented questioning of what it means to be human?
The speed of progress
The car exemplified everything that was progressive and flawed about industrialisation and free-market capitalism. A marvel of the coalescence of numerous inventions and technological developments, it literally placed mankind in the driving seat. The car, which was always conceived of as an extension of human agency, became a mobile monument to human autonomy, invention and freedom. It reshaped the city and urbanisation. Arterial motorway systems became the superhighways of free will, unfettered commerce and enterprise. Just as today’s younger generations embrace information and communication technologies to gain greater freedom, autonomy and identity, so the postwar generation embraced the car. They customised them, creating identities and forging their escape from the tyranny of smalltown parochialism and the gaze of their parents.
But the car was more than simply a means of escape or identity. Society was shaped by it. Driving became a means to express the moral agency of humanity. Although separated in individual cars, the act of driving affirmed every other driver as an autonomous being with the same capacity for moral agency as oneself. This is why passing a driving test was more than a tick-box exercise: it was also a symbolic moment that marked the passage from adolescence into adulthood.
Today’s ‘digital natives’ are not rushing to get driver’s licenses the way baby boomers did. In 1978, nearly half of all 16-year-olds and 75 per cent of all 17-year-olds in the US had driving licenses; by 2008, those numbers had dropped to 31 per cent and 49 per cent, respectively. This ought to be a concerning development. Not because it indicates that this generation is falling out of love with the car, but because it demonstrates an even more worrying trend: namely, the infantilisation of adulthood, a development that will have a severe impact not only on mobility, but on our whole approach to the future.
This is not an argument born of nostalgia, nor is it an attempt to resist technical progress. Rather, I am arguing that the car was not simply a technical achievement. It was also a social achievement, which, at its heart, expressed some of the key elements of what it means to be a human in possession of agency.
Today, we are increasingly sceptical about human agency, seeing it as a source of problems rather than solutions. Replacing human agency with technologies driven by ‘if then’ algorithms, which react through sensors to external stimuli and promise rigorous predictability, represents a highly problematic proposition. It raises questions as to the possibility of developing a machine that can replicate human free will. At the same time, it suggests that human fallibility, based as it is upon that same free will, can be regimented or diminished to the point where it no longer determines actions.
New technologies have always competed with old ones, initially for time and attention but increasingly for money, and then for dominance. The printing press attacked the illuminated manuscript; the photograph attacked the art of painting; TV attacked radio and the printed word; smartphones attacked TV. But what we are talking about in relation to the driverless car is an attack on human agency – the source of invention and the conscious application of knowledge to solve natural and manmade problems. The quest for predictability is in danger of robbing humanity of one of its most precious capabilities – the freedom to make errors and learn from them.
The implications of the driverless car are, therefore, immense. Every stage of technological change has provoked a degree of social anxiety. But what we are talking about here is something very different. The revolutionising of the textile industry at the start of the Industrial Revolution, which replaced skilled workers with workers who merely kept the machine operating, possessed a very different dynamic to the driverless-car revolution. The industrialisation of society during the nineteenth century was seen as a leap forward into a better future where invention, science and rationality represented progress. Yes, it meant huge social dislocation, from the replacement of skilled men by machines to the conviction that technical calculation was superior to human judgement. But technocratic capitalism’s deskilling of workers, reducing men to mere adjuncts of machinery, was driven by the capitalistic need to commodify labour, to turn it into just another factor of production, another market for capital to consume in its inexorable drive to accumulate. As unpleasant as this was for those directly affected, it was revolutionary because the new means of production and communications tore down old ties and traditions, and generated radical ideas about equality and human freedom. In reducing human agency to the adjunct of machines, industrial capitalism gave birth to a clash of political interests, a contest informed by competing ideas of human, moral agency. Ironically, technocratic capitalism placed autonomy and moral agency at the conflicted heart of modern society.
But now, we appear to be hoping for technology to supplant human moral agency altogether. This is highly problematic because it diminishes mankind and technology. It fixes in time what in essence cannot be fixed in time, and it therefore underestimates what might be possible in the future.
An open highway
Human knowledge and ingenuity are never fixed or closed. There would have been no progress if this were the case. The idea that it is possible to develop ethical frameworks for machines that will take into account all future eventualities is simply farcical. Human ethics are always a work in progress and remain historically contingent. We still experience events today that we don’t yet have well-developed moral and social codes for (assisted suicide is one case in point). And it does not take a great historical scholar to point out that certain moral and social codes that were once deemed acceptable and immutable are fervently disavowed today: slavery and segregation are two obvious examples. The assumption that it is desirable to diminish human agency by petrifying our current moral choices displays a deep distrust of mankind and our ability to develop socially and morally.
Transforming the sphere of moral judgement into a pre-programmed scientific accomplishment empties the autonomy of the individual of any meaning. Human agency is open-ended. Each human being is, in the words of Raymond Tallis, ‘an independent point of departure’. Every one of us is a new beginning, able to contribute to shaping the future. We are not fated to act out a preordained script. We are not driverless machines programmed to act according to if-then algorithmic logic. We follow the rules of the road – in the UK, the Highway Code – but we have the moral ability to make decisions, to act illegally if circumstances demand it. We can exercise judgement in a wide range of situations that don’t appear in a standard 40-minute driving test. We may break the speed limit, for example, when in danger, or cross a solid white line when our lane is partially blocked. Any car journey involves actions that entail hundreds of choices and decisions, many of which will be new and unfamiliar, but which we trust ourselves (and others) to make ethically and wisely.
But the misanthropy underpinning the idea of the driverless car aims to close this capacity down, and fix our choices like some pre-programmed algorithm. Why seek such limits? Autonomous cars aren’t simply replacing human drivers, just as human drivers in the first cars weren’t simply replacing horses. These new cars are seeking to automate the limits of contemporary society and strap mankind into a safety-first, predictable, fixed and static universe. Advocates of driverless cars are like those who assume that just because the modern railway gauge was reputedly determined by the width of a cart horse, so all railways will continue to work like this in the future. The biggest danger is not the lack of categories and concepts that could allow us to envision the mobility-on-demand system of the future. No, the biggest danger is that we sleepwalk into a future where the source of all this ingenuity – mankind in all its technological ingenuity – has itself been curtailed.
Norman Lewis works on innovation networks and is a co-author of Big Potatoes: The London Manifesto for Innovation.