24 June 2019
The Case For a ‘People’s Uber’
by Aaron Bastani
I’ve previously written about how local bus services should be a universal basic service – which is to say free at the point of consumption and available to all citizens. Far from being an outlandish demand, such a policy has already been adopted by the likes of Estonia and Luxembourg. Closer to home Scottish Labour committed to it earlier this year. For an additional £5 billion, according to research undertaken by UCL, we could pay for free bus services for everyone in Britain while accommodating a 260% increase in use. And because of extreme supply – as I outline in Fully Automated Luxury Communism – the costs will only fall, with the inputs into running bus services: from the manufacture of the vehicles to the energy, and ultimately even much of the labour too (specifically with deep learning and robotics) declining every year. The question of the extent we choose to automate such a service – and in whose benefit – is a political one.
The dividend from free bus services would be immediate and impressive: it would relieve congestion and parking issues – the bread and butter of local politics; it would address issues of air pollution, which presently accounts for as many as 30,000 deaths a year in the UK; it would be a boon to isolated and low-income households; and would help revive high streets (along with a raft of measures I write about in the book and will blog about another time).
But while such a policy would be transformational, and easier to rapidly introduce than low-cost, frequent trains (also to be pursued), it shouldn’t stop there. That’s because free, local transport should also eventually include cars – which will remain an important mode of transport. Societal ageing means we’ll need transport systems capable of meeting the needs of the elderly and disabled; if we look to the coming decades ever more of us will suffer from age-related health conditions. Then there is the fact certain people prefer to travel alone, even if it is slightly less convenient – women later at night being one example. And have you have ever tried moving house using anything smaller than a Luton van? On top of all that is traditional logistics and distribution. Simply put, private vehicles aren’t going anywhere. We need them.
But more than accepting cars as an essential feature of modern life, integrating them into the broader UBS of transport could serve to undermine car ownership as a social presumption. Right now the average car is stationary and unused for around 95% of the time – a huge waste when you consider the successor vehicles of today will have the ability to autonomously pilot themselves. Why shouldn’t this surplus be deployed? It’s that question which makes me convinced that the argument for public ownership of cars – something equivalent to a municipal Uber – isn’t just ecologically preferable but common sense. It’s a superior alternative to a world of unfettered automotive ownership where we drive around in subterranean tunnels, spend hours in traffic jams and sometimes fail the MOT. You have the advantage of car ownership without the downsides.
Uber’s strategy regarding automation is well-documented. Until recently the company had been explicit in disclosing its longer term ambition: to gain a massive customer base before a shift to self-driving vehicles. In 2016 the company signed a $300 million deal with Volvo to develop an autonomous road-ready car by 2021. That August the company’s then CEO, Travis Kalanick, spoke candidly about the scale of the upstart’s ambitions, “It starts with understanding that the world is going to go self-driving and autonomous,” he told Business Insider, “so if that’s happening, what would happen if we weren’t a part of that future? If we weren’t part of the autonomy thing? Then the future passes us by basically”. Importantly for Uber, its platform would give the company access to massive amounts of data – a resource of critical importance for the development of ‘deep learning’ algorithms. If AI is the engine of the coming era, data is the oil. The larger Uber’s datasets are ahead of the ‘autonomy thing’ being technically indistinguishable from a human driver, and the more it invests in the relevant technology, the greater its first mover advantage. Which also explains the span of the company’s intentions, with Uber extending beyond just taxis and aiming to provide food delivery services and even electric bikes. It’s a plan which made a great deal of sense when Kalanick spilled the beans in 2016 – and still does.
But if it is the data which is increasingly the source of such value, and the ambition is to establish a monopoly and eliminate (or massively minimise) human labour, then why should that benefit Uber’s shareholders rather than the public? And what about the workers? Given approximately four million people drive for a living in the United States the consequences of such a shift will be profound. What is more, wholesale transition to electric cars will require gargantuan mineral resources – something recently highlighted in a letter authored by Professor Richard Herrington and fellow expert members of ‘SoS MinErals’. While that letter was irrationally pessimistic in some areas – such as the fact the UK would need to generate 20% more electricity if all vehicles were electric (that’s the point, the source just needs to be renewable) – it was entirely right to highlight the demand this would create for scarce resources like cobalt, lithium and neodymium when it comes to storage. As the letter itself makes clear:
“…to replace all UK-based vehicles today with electric vehicles (not including the LGV and HGV fleets), assuming they use the most resource-frugal next-generation NMC 811 batteries, would take 207,900 tonnes cobalt, 264,600 tonnes of lithium carbonate (LCE), at least 7,200 tonnes of neodymium and dysprosium, in addition to 2,362,500 tonnes copper. This represents, just under two times the total annual world cobalt production, nearly the entire world production of neodymium, three quarters the world’s lithium production and at least half of the world’s copper production during 2018. Even ensuring the annual supply of electric vehicles only, from 2035 as pledged, will require the UK to annually import the equivalent of the entire annual cobalt needs of European industry.”
Of course such a transition would take place over multiple decades, and further innovations will lead to greater energy efficiency and less waste. What is more, it’s important to highlight that known reserves of all the minerals listed above are sufficient to meet present global energy demand in the event of it doubling and, unlike extracted hydrocarbons, they can be recycled. But the critical point still stands: the ambition should not simply be a like-for-like replacement of renewables for fossil fuels, consumption patterns need to change as well.
Part of the answer is to massively improve existing transport infrastructure – including buses as a UBS – but another is to transform the relationship of the individual to the car. Here a presumption of acquisition and ownership needs to be replaced with ‘right of access’.
It’s been widely recognised that the emergence of the ‘sharing economy’ reflects a distorted form of collaboration which, if re-wired, could be the basis of a 21st century socialism. Here access matters more than ownership: if you could use a car whenever you liked, but had to wait a short while (maybe not that long once the algorithm had enough data) what would the point be of owning a vehicle? For a long time the answer was price: constantly using a taxi is more expensive than owning a car, principally because of the added labour costs. But once automation kicks in, alongside the constantly falling cost of energy and optimal efficiency as a result of big data, that becomes less of an issue. Much less.
One response might be that self-driving cars are unsafe and will never happen. As I outline in Fully Automated Luxury Communism, the truth is rather different. Although it’s fair to say the ride will be a bumpy one – Uber suspended its autonomous efforts when a pedestrian was killed in 2018 – it’s important to remember that around 30,000 people die every year in the US as a result of road accidents. Worldwide that figure stands at more than a million. What is more it is important to recall that previous disruptions in transport endured similar setbacks: in 1830 William Huskisson, at the time a leading British politician, died in an accident at the opening of the world’s first railway line between Manchester and Liverpool. That death, like Elaine Herzberg’s in 2018, was a tragedy, but I’d argue such logic will capsize once it becomes clear autonomous vehicles will mean a reduction in road-related deaths. By 2030 anyone who has lost a loved one in a traffic collision or suffers as a result of drink-driving could be asking why politicians haven’t addressed the issue by accelerating the shift to autonomous vehicles.
The next question is how far away this new world is: despite misgivings the technology has advanced rapidly. As recently as the 2004 DARPA Grand Challenge the best performing vehicle could only complete 11 kilometres of a 150 kilometre course. By 2010 Google announced its test vehicles had crossed the United States, with 99% of the journey performed with no human intervention. The U.S. Department of Transportation has six levels to classify autonomous vehicles, starting from entirely human operated, Level 0, to fully autonomous – Level 5. While there are presently no Level 5 cars on the market, nor the regulatory framework that would allow them to be used, a number of vehicles already operate at Level 3 autonomy – what is called ‘conditional autonomy’, where cars can pilot themselves but only under certain conditions. One example of this is the Audi A8. In certain contexts autonomous vehicles are already here, most likely is this will gradually creep until the technology is perfected, likely within a decade.
What would placing all this within the project of modern socialism and Universal Basic Services look like? A start would be the integration of ZipCar like services into local transport provision. One can see this initially operating most plausibly in London with Transport for London. That would provide a base to grow but also begin to give TFL that most important resource as they begin to include autonomous vehicles within their network: user data. Furthermore these vehicles would be electric, say from 2020, providing further incentives to create more charging stations around the city and offering Londoners a green option for the occasional car journey. These vehicles would be accessed using a mobile payment system, either that of a third party in cooperation with TFL or, preferably, a new mobile payment system connected to use of all UBS and connected to a state-controlled bank, something which could emerge from the launch of ‘PostBanks’ – something recently touted by the Labour party. The payment system being publicly owned is important, because the resource of value to deep learning as it is applied to Universal Basic Services is data – preferably as much as possible from consumer habits to driving scenarios. What is more, such data could be anonymised and subject to far greater levels of security than is presently the case with apps such as Uber. As a 2018 report by IPPR makes clear, public ownership of data is critical to building a ‘digital commonwealth’. Providing Universal Basic Services which build on innovations in AI, energy efficiency and the Internet of Things depend on it.
Such a payment system could be rolled out nationally alongside a variant of Oyster – not only capable of being used for buses across the country but train journeys too (although these would not be free, at least not for a while). As extreme supply began to kick in, and electric vehicles begin to first hit price parity with fossil fuel vehicles before getting cheaper, other municipalities could adopt similar measures by growing native electric fleets resembling the Zipcar model. As autonomous vehicles become the norm, probably by the late 2020s, such vehicles would be at the cutting edge of the self-driving revolution, and would only be available to those without cars of their own – creating further incentives to ditch private ownership. Alongside cars this would also apply to electric bikes, another area where the sharp fall in the cost of energy storage is meaning rapid change. Indeed Uber launched their ‘JUMP’ electric bikes in Islington this May. The reason? To keep Londoners on their app and keep collecting that invaluable data. But extreme supply in information and energy shouldn’t mean value capture for Uber, it should mean free electric bikes, and car journeys, for everyone.
While what is needed is a revolution in energy production and ownership, consumption patterns must also significantly change. At the heart of that is the notion that access to transport will matter more than ownership for a 21st century socialism built around Universal Basic Services.