Conference Synopsis

His Grace,  The Duke of Richmond opened Nucleus 2024 by welcoming the invited guests, and noting that the event he founded in 2015 was now in its tenth year.

His aim then was to accelerate the reinvention of mobility by bringing together those who lead, influence and govern that reinvention, he said. Mobility touches so many aspects of state and society – from the economy and technology to urban planning, ethics and the environment – that those who have the greatest influence on its future might never meet, and Nucleus was established to rectify that.

His Grace said that looking back now over a decade of discussions, it fascinated him that an event which set out to examine and encourage the future of sustainable mobility has now come to chart its history and its progress.

When Nucleus first met, the carmakers were trying to figure out how to make electric cars affordable and practical. We thought the big tech firms would disrupt the established carmakers. We thought that self-driving cars were just around the corner, and we knew that China would have a transformative impact on everything.

But none of those issues have played out quite as expected, and more have come to the fore. Over the past decade, the environmental, ethical and economic security implications of the mineral resources on which we are increasingly dependent have leapt up the agendas of the leaders of companies and countries alike, he noted.

And government has a greater role to play than ever: in regulating AI and autonomous driving, and in remaking supply chains and imposing tariffs with vast pieces of legislation, like the Inflation Reduction Act in the US.

Every one of those topics would be examined at Nucleus ’24, he said. But the future comes at us fast, and he wondered what issues Nucleus guests would be debating in another ten years’ time.

Opening the debate on ‘the great EV crisis’, the founder of a new EV Western manufacturer admitted that the geopolitical situation was ‘depressing’ and that electric vehicle adoption and sales were faced with a ‘profound challenge’. But he remained a relentless optimist, he said, determined to bring us all to a ‘state of enlightenment’ with EVs that cover more than five miles for every kilowatt-hour of electricity consumed.

When Nucleus first convened we might have seen a simple binary distinction between EVs and ICE cars, but it was time for that to end, he said. There were good and bad EVs too, dubbing those with poor efficiency ‘electron guzzlers’, and suggesting that it was impossible to make a practical, affordable electric pick-up using today’s technology.

His fellow panellist, an EV pioneer and now a senior leader with an established car company agreed that efficiency was key. Customers cared most about the amount of range they gained for a given time plugged in, which is a function not just of charging speed but of efficiency too.

He said that his business was ‘extraordinarily unhappy’ at its current heavy losses on its EV programs, but that profitability and greater affordability were coming, and he likened the EV adoption curve to that of other types of consumer technology: expensive and slow at first, followed by a rapid acceleration.

But you can’t just wait for consumer behaviour to change: the product has to change too. His business had been ‘rattled’ by the rate of change and by Tesla’s success and was willing to take very different approaches in its response, including vehicles which would move it into new market sectors, with smaller batteries and higher efficiency.

There were other causes for optimism too, he said. EV sales were not falling in most cases: the rate of growth has simply slowed. And his company’s ‘obsessive’ tracking of how people charge their EVs shows that EV users can now get close to matching the behaviour of those refilling with fossil fuels. Reliability of the charging network remains a key issue: you had a 99 per cent chance of leaving a Tesla Supercharger site with a charge, he said, but its rival US networks remain ‘terrible’.

For volume manufactures, IRA tax breaks were changing the situation ‘dramatically’. Chinese companies had long ago ‘run away’ with the technology, but partnering with Chinese battery firms to build in the US was the best option. Such incentives decline with time though, and he cautioned against building a business around them.

From the floor, a senior leader from an established European carmaker with long personal experience of working in China urged Nucleus to see all these issues – technology, infrastructure, state support and consumer demand – holistically, and reminded them that Chinese EV visionary Wan Gang set out a single national strategy for EVs almost twenty years ago. If Europe wants to succeed like China, it needs to plan like China. The second panellist agreed: for him, China is essentially one national carmaker with a bunch of subsidiaries.

Turning his attention first to broader geopolitics, the first speaker was as startlingly frank as the former central banker who addressed Nucleus two years earlier, but even gloomier in his predictions. The world was not the same following Russia’s invasion of Ukraine, he said, and would not be again. It was still in a process of transition and fragmentation, with free trade becoming increasingly challenging and resilience outweighing efficiency as national security concerns come to dominate political agendas. Europe would remain ‘war-ready’ for the foreseeable future, he said, but he was slightly more optimistic on the other likely geopolitical flashpoint. He did not believe that it was in China’s interests to take Taiwan by force: instead it will be done gradually by persuasion, pressure and harassment, rather than by siege or invasion.

He agreed with the proposition that the mobility industries – and electric vehicles specifically – are in the geopolitical crosshairs like never before, saying that they sit at the nexus of trade, technology, and both economic and national security. Questioned on the wisdom of the EU’s heavy new tariffs on Chinese-made EVs, he was supportive, saying that it was ‘the only way for the EU to go’; that the tariffs were actually quite mild by comparison with what yet may be introduced in the US, and were necessary to avoid the fate of other manufacturing sectors which have been lost entirely to China. The tariffs would not have a huge impact on prices, he believes, but will avoid a huge deviation of trade should – as seems almost certain – the United States increase its tariffs further. A show of hands among Nucleus delegates disagreed, with the majority opposing the EU’s new tariffs.

Questioned on fears of a trade war, he noted that this was the last thing China wants, and that its response to the EU’s tariffs had been measured and not escalatory. The real danger lay, he felt, with a possible incoming Trump administration which would be more isolationist, protectionist and nativist in outlook, and whose likely senior appointees he knew to be in favour of much higher tariffs on both Chinese and European imports. Although cognisant of actions the previous Trump administration had taken that were in America’s best interests, he feared the disruption that would be caused to global trade and diplomacy by a second Trump term, just when a steadier hand was most required.

The world has been convinced by artificial intelligence, the founder of a major AI enterprise told Nucleus, but the uses it has been put to thus far are only the tip of the iceberg. Once Nucleus has been meeting for twenty years rather than ten we will think of AI primarily for its physical, ’embodied’ uses, such as in robotics and in autonomous driving, he said. The three pioneers who assembled for the Nucleus panel discussing this new optimism around autonomy each represent a very different approach to implementing it. They may agree that it will work but they disagree on how, and mindful of autonomous driving’s history of over-promising, only one would put a date on when.

The first panellist’s approach is an end-to-end solution, training a single neural network to drive on any road and cope with any scenario without prior knowledge or mapping. Others have taken the same approach and faltered in the past, but the speaker said that the diversity of data now available, along with the constructive collision between tech and automotive roadmaps which puts the sensors in cars required to collect data has had a transformative effect. Two years ago he struggled to get meetings with carmakers: now they know that within a few years they will struggle to sell cars without hands- and eyes-off autonomy, and were keen to partner now on less advanced driver-assistance systems, sharing their data in order to move to greater autonomy faster.

The approach is working, he says: his tech has performed to the same standards in 15 cities, and development costs are lower.

The second panellist, a seasoned autonomy pioneer, struck several notes of caution on this approach. Broad but shallow ‘commuting’ data did not throw up enough difficult or rare scenarios to train AI sufficiently, so simulations were required. And rather than trusting the software to learn to drive anywhere, specific interjections should be coded-in, disallowing certain actions rather than trusting AI to learn that they are wrong. He also advocates tackling specific use cases first, where the benefits come earlier, more easily, and are of greater value. Trucking, for example, is a $4tn industry rather than the £160bn of ride-hailing. Displacing a truck driver saves $40 per hour rather than $15, and the time taken to get a truck from LA to Dallas can be cut from three days to 24 hours with autonomy.

The third speaker’s approach is different again. He simply accepts that soon all cars will be self-driving to some extent, and that the software that allows them to do so will simply be an enabler, rather than a differentiator. So in founding his own ride-hailing app he didn’t try to develop his own autonomous-driving software, but simply shopped those already in development, partnered with an existing self-driving software provider, and is instead focussed on delivering the best customer experience. It is still ‘super-hard’, according to the second panellist, to gauge the relative strengths of those working in this space. But Waymo and others were already operating well and safely, and there can be few greater signs of confidence than his fellow speaker’s decision to base a new global ride-hailing app around someone else’s autonomous-driving software, with the promise to have his first riders aboard by 2026.

The speaker described with some frankness the difficulties of running a lithium supplier in a market with fluctuating demand, dominated by China, subject to political interference via tariffs and incentives and beset by ethical and environmental concerns. What had been a steady-state industry with a relatively stable, low commodity price suddenly saw a spike in prices and a tenfold increase in predicted demand as the major carmakers pivoted to EVs, he said. But after going ‘hell for leather’ to find new resources in Africa and central America, the industry has quickly found itself in an oversupply situation, with the rate of growth in demand tailing off and prices back down sharply.

His business operates wholly-owned facilities in China, tries to operate as a good corporate citizen there and is currently welcomed, he said. But China has a natural resources strategy which is as considered, specific and long-established as its state policy on EVs, he said. China’s priority isn’t green energy but simple energy security: its aim is to be self-sufficient in critical natural resources, and he wondered for how long his business would be welcome in China after it reached that goal. There were parts of the world from which China could not source raw materials, such as the US, Canada, Australia and some parts of central and south America. But there were many other regions from which Western miners were locked out, either as a result of China’s ‘belt-and-road’ policy, or of ‘business practices which we don’t want to be a part of’. In those environments ‘we lose’, he said. ‘Western companies just have to walk away.’

The speaker also introduced the notion of ‘peak lithium’: the point at which the quantity of the metal being recycled from old batteries would be sufficient to reduce the mining of new material. Such recycling wouldn’t acquire real economies of scale for another decade, he thought, with ‘peak lithium’ not arriving until 2042, after a critical mass of EVs had both hit the road and then aged sufficiently to be recycled. Lithium-based batteries would continue to dominate for light-duty vehicles, he said, but there will come a time when old EVs will supply more of it than miners like him.

The speaker began by outlining the scale and scope of China’s espionage efforts against the West: it was a long-term, planned campaign, conducted daily, they said. China is a closed and secretive society and has a completely different set of values, particularly with regard to data and privacy, meaning that we do not compete on a level playing field. If data and intellectual property could not be stolen, China would buy them by employing staff, academics or companies, hoovering up databases or buying companies outright, sometimes through cut-outs to disguise the real owner.

The speaker said this was a question of pragmatic defence of our interests and values rather than being moralistic about espionage: it was their organisation’s job to steal secrets too, if by other means. It was the speaker’s firm belief that we should continue to engage with China. To do otherwise would be dangerous. By withdrawing, we would increase the risk of China misreading us. There should not be an economic ‘decoupling’: the benefits of continuing to trade with each other were too great. And we should not become overly focussed on the battle between the US and China for economic supremacy while the environment goes up in smoke.

It was heartening to hear of technology being developed for the right reasons, the speaker said, but it could also be used for ill intent, such as autonomous weaponry. Trying to understand what capabilities our potential enemies might have remained a core task for their service. In the past, we assumed that ‘the good guys’ would always have the best technology. That probably remains the case, they said, but it is easier to catch up if you’re prepared to steal. Western businesses, such as those represented at Nucleus, should not be naive about China’s willingness and ability to do so, and 90 per cent of the battle to prevent it was simple, effective digital ‘hygiene’.

The threats from organised terror and Russian espionage still lurked, they said, but had been much disrupted. Technology and data played a role in both preventing attacks at home from radicalized but isolated individuals, and with maintaining contact with clandestine sources in countries such as China, which is becoming increasingly difficult. The intelligence services might help protect tech firm’s IP and data, but tech and industry can help the intelligence world too through effective partnering.