In 2016, democracy made something of a comeback. First, in the UK, the British public voted to leave the European Union and overturn 40 years of integration with the European continent in a tightly fought referendum. Then, in the United States, a populist reality TV host stunned all pollsters to become president. Both results shocked the public and establishment alike. Those who believed that there was no alternative to the international liberal consensus were beaten on both counts by the oldest of democratic instruments: the vote.
While policymakers and political pundits across the world have since rediscovered the power of democracy, the same cannot be said of the tech industry. Everywhere, the narrative of unstoppable technological progress continues.
However, there are good reasons to expect that Silicon Valley will also face a reckoning over the next decades, if not in quite so sudden a way. The driverless car industry, for example, faces important roadblocks in the political sphere which could quite easily derail the efforts of the entire industry.
“The problem isn’t technology, it’s legislation” Renault-Nissan CEO Carlos Ghosn, June 2014.
Despite the rhetoric of innovators, the automotive industry is structurally different to other consumer industries, where politicians interfere only to tackle rare excesses: corruption, cartels and unethical practices.
The automotive industry is highly regulated, meaning that the boundaries between business and the state are especially weak. Where this is the case, voters can and do expect to influence corporate behaviour, even if traditional economics suggests that this would be counterproductive. A number of industries fall into this category.
In recent weeks, for example, the British government has announced that it will cap energy prices to prevent ‘excessive’ charges from energy providers. Historically, it has been impossible to replicate this kind of policy in the tech industry (imagine the price of a Macbook capped at under £1000). But, as entrepreneurs have turned towards public services and regulated industries, the scope for democratic involvement in business has expanded.
An Apathetic Revolution?
So how likely is it that driverless cars will feel the force of political discontent? According to some public polling: not very likely at all. Research by Auto Trader in the UK shows that some 49% of people have little to no knowledge of driverless cars – so little, in fact, that they have formed no opinion on the subject. Only 21% of those questioned knew what a fully autonomous car was.
Although this suggests that manufacturers need not worry about political opposition, apathy is a problem in its own right. Like smartwatches and smart meters, driverless cars are yet to capture the imagination of most consumers, a fact which might derail the project in an entirely different way. It is quite conceivable that industry big-wigs might have to engineer demand for their new products, as the energy sector is currently doing for smart meters.
However, politics is often as far behind technological change as regulation. To borrow from Gartner’s ‘hype cycle,’ new technologies only become more widely adopted after they have passed the peak of inflated expectations and traversed the quagmire of general disillusionment.
Driverless technology is fast travelling from the former to the latter, but wider adoption – what Gartner calls ‘the slope of enlightenment’ – remains some way off. When driverless cars reach this stage, however, three democratic challenges will undermine further progress.
Money, Money, Money
The first of these is money. For all the attention given to money in popular culture, the cost of implementing driverless cars has largely slipped under the radar. Unlike normal vehicles, driverless cars require a wholly different road infrastructure.
To ensure that they stay in lane, governments will have to oversee the transformation of road surfaces or, at the very least, deploy magnets along roadways to relay the position of other cars to each moving vehicle. The cost to introduce this new system could be enormous: starting at $10bn in the USA for the most basic solution.
Other costs abound. In order to properly navigate complex urban environments, municipal governments will have to invest heavily in mapping, a cost which is currently unquantified.
Finally, national governments will have to overhaul their systems for planning, driver registration and vehicle safety. None of these issues has received due attention, largely because they lack the inherent sexiness of new technology. But each of them will consume billions of dollars of taxpayers money.
In a time of squeezed living standards where increased government borrowing and privately-funded infrastructure have fallen out of favour, it is likely that revolutionary changes to global infrastructure will be opposed from all quarters. Driverless cars could be run off the road.
In Whose Interest?
Secondly, driverless cars pose existential questions for several important interest groups. By some accounts, there are 10m Americans employed in industries related to driving.
These include truck and taxi drivers, but also those who work for companies in the vast driving ecosystem: trucking motels, car washes, auto repair, the list goes on. As the election of Donald Trump has made clear, working-class Americans are deeply concerned by job losses – whether from outsourcing to the third world or technological change.
For many, trucking remains the last bastion of respectable, well-paid employment for those without a college degree. Any politician brave enough to introduce driverless cars would risk alienating this increasing vocal segment of society.
Two other interest groups will shape the adoption of driverless cars. Established players in industries like car insurance (worth $200bn a year), the car aftermarket ($300bn a year) and parking ($100bn a year) have little to gain from a wholesale shift to autonomous driving as they will struggle to adapt their business models to the new environment.
Their combined influence alone might be sufficient to disrupt the driverless revolution. Auto-manufacturers who sell cars to driving enthusiasts will be hit even harder, as will the drivers themselves.
Perhaps surprisingly, 45% of Britons enjoy driving ‘too much’ to consider buying an autonomous vehicle. While there is some evidence that millennials are less inclined to feel this way, the pace of demographic change suggests that older drivers will continue to influence the shape of automotive policy for decades to come.
Solving this “Clarkson Conundrum” will depend on satiating the love of driving in other, as yet undefined, ways.
If interest groups don’t derail driverless vehicles, the conservatism of the wider electorate might well.
It is quite remarkable how little of this natural conservatism filters through into the media-sphere, where pundits have spent the past five years competing to see who can deliver the most hyperbolic remarks about the potential of driverless cars.
Polling suggests that the public is far more cautious: 75% of Americans describe themselves as ‘afraid’ of autonomous vehicles, a pattern which is surprisingly similar for millennials – those normally the most enthusiastic about technological change.
The traditional counter-strategy of innovators has been to persuade through familiarity – the steady infiltration of new technology into everyday life. This approach will not be possible on this occasion. Until recently, there were two narratives about the future of driverless vehicles.
In one narrative, they would be slowly introduced to the roads until, many years off, all human drivers were replaced. This might be an impossibility. Through testing prototype cars, Google and other manufacturers have discovered that human supervisors lose concentration when not actively driving the vehicle; to remain safe, driverless vehicles will have to be fully autonomous from the very beginning.
But the interaction between human and computer-driven cars is a more fundamental problem. The latter are programmed to follow the highway code to the letter; the former will break the code where necessary – driving, after all, is an imprecise sport.
These two approaches to driving cannot co-exist without risk to drivers and passengers alike, as has already become apparent in states like California. The more driverless cars hit the roads, the greater the frequency of misunderstandings, accidents and fatalities. To avoid this, there will have to be a revolution.
Herein lies the problem. A revolution might sound good on paper – even exciting. But in reality, it will be messy, and messiness creates losers.
Consider the introduction of autonomous school buses. It would not be practical to roll out autonomous buses across the board instantly – even in the fastest of revolutions, things can be staggered. So imagine the fear and outrage of those parents whose children go first into the driverless death machine. With nothing less than a well-orchestrated campaign, moments like this could easily spiral out of control, halting the entire process.
Democracy and Driverless Cars
It is precisely the complexity of this kind of event that makes the future so difficult to predict. With excellent public relations, appropriate management of special interests and a careful eye on the public purse, the driverless revolution might go the way of Amazon, rather than GM crops. But the inverse could easily be true. Silicon Valley is well aware of this – of course – having traded on unpredictability and disruption for some 50 years.
But governments seem rather less alert to the roadblocks in the way. The UK government, for example, has adopted a strangely fatalist approach to driverless cars, intimating in various policy papers that their arrival is a formality – the only questions left to be decided are technical. This could hardly be further from the truth. More worryingly, it is precisely this anti-democratic attitude which angers so many Western voters. If policy makers are to avoid a re-run of 2016, they must engage with the real issues – and real people – now.