Are You Gonna Go Huawei? 5G and the Rise of AI Nationalism

by James Meadway

1 May 2019

Pierre Metivier /Flickr

A full-scale Whitehall leak inquiry has been launched into reports from a meeting of the supposedly top-secret national security council that a number of Cabinet ministers expressed concerns about Huawei’s involvement in the provision of the UK’s next-generation 5G mobile networks.

The US has been attempting to enforce a hard line against the Chinese telecoms company, currently the world’s largest provider of network infrastructure, amongst the rest of the ‘Five Eyes’ group of English-speaking security collaborators. So far, only Australia has imposed an outright ban on Huawei equipment for 5G.

It’s fun to watch the Tories fighting like rats in a sack over who gets to succeed Theresa May’s miserable reign, and the sheer pettiness of the leak – relative to the matter at hand – is impressive in its own way. But step back from the self-centred squabbling and all this starts to tell us something interesting about where the economy is headed. Because, with slightly more meaningful consequences, this is the same week that talks are due to reopen between the US and Chinese trade representatives on the ongoing trade dispute between the world’s first and second largest economies.

Much of the reporting of the US-China dispute has focused on the classic trade squabble of goods tariffs, with steep increases in import duties now being applied by both countries across a range of traded goods. But, whilst attracting the bulk of press attention, tariffs on famous brands and familiar goods aren’t the driver of the row.

This is, as Foxconn’s CEO says, a dispute about technology. And at the centre of that is the great fear in the US that China’s Made In China 2025 programme, plus the extraordinary expenditures it is making in artificial intelligence (AI) and related technologies, could give the country an unsurpassable lead over the US in key areas in the very near future.

In other words, the principle at stake here is not whether a few tariff hikes can be applied a to a few goods, or even whether these tariffs are (as Donald Trump appears to think) a smart way to close the US-China trade deficit: it is about the control and use of knowledge and technology in a world where both are increasingly coming to dominate the conventional rules of economics.

Fundamentally, that means the control and use of data. Research produced by management consultants McKinsey suggests that international data transfers now “generate more value than the global goods trade”. Cross-border bandwidth is 45 times bigger today than it was 15 years ago, and this only forecast to increase. Today most trade of any sort comes with a digital shadow: 50% of services traded internationally are digitised, and virtually all goods trade, from coffee to cars, is monitored and flagged digitally.

The challenge for capitalism, to return to a previous theme, is in actually grabbing the value seemingly generated by the digital economy, when its outputs are increasingly not traded in the conventional sense – that is, through offering products for sale at a specific price. This is the classic model of economic competition, and the one taught in schools, in which many different firms will offer similar products and attempt to undercut each other to win sales.

The economics of 5G.

Much of the world economy still runs on roughly this price-competitive basis, of course, but the scale of the digital economy is starting to dwarf it, and even in manufacturing – where goods are most obviously still offered for sale at a specific price – data is coming to dominate, through the expanding use of digitised design and 3D printing, on one side, and the provision of in-service data after production. In aerospace, for example, Bombardier’s new C-series jetliner is expected to generate 844TB of data on a single 12-hour flight, roughly one-fifth of Facebook’s current daily global data usage. Aeroengine manufacturer Rolls Royce generates most of its revenue from post-production services, monitoring thousands of engines currently in use.

5G will further swell the deluge of data, with the proliferation of low-cost sensors in every electronic device, and the spread of massive, wireless communications capacity combining to further saturate both the economy and society. And it is the current models of artificial intelligence, reliant on machine learning from vast pools of data, that are poised to be exploit this, both to organise the enormous complexity of the ‘Internet of Things’, and to learn from it.

The economics of these two factors point, very strongly, to the reinforcement of existing tendencies in digital capitalism: that scale matters more than anything else, and that (as a result) it is control of access to infrastructure that determines who can grab any value generated. If you control access to the infrastructure, you can levy access charges.

But if AI reinforces these tendencies – and the kind of AI that we have, rooted in massive datasets, certainly does – and 5G reinforces them too, the logic of capitalist competition will be shifted further away from the textbook terrain of competition to offer the lowest price. The more the economy becomes digitised, the less it will be driven by price competition and the more competitive outcomes will depend on claims of ownership, access to revenue streams, and, ultimately, state power.

This is the logic spelled out in a thought-provoking essay by Songkick co-founder and angel investor Ian Hogarth, which he calls, rather bluntly, ‘AI nationalism’.  Because the economics of AI are dependent on the scale of its operations, and because the social impact of AI will be so great, Hogarth predicts that an “accelerated arms race will emerge between key countries and we will see increased protectionist state action to support national champions, block takeovers by foreign firms and attract talent.” This isn’t an unfamiliar pattern of development for capitalism; indeed, the high period of globalisation, from the 1990s to 2008, is historically weird in its rejection of this model, at least at the level of international governance. But since 2008, fusions of states and capital have proliferated.

Britain at the crossroads.

In the midst of these transformations, British capitalism is starting to look a little ‘also-ran’. Decades of neoliberal governments, fixated on using ‘competition policy’ to create ‘level playing fields’ in which capitalist enterprises would compete on equal terms has resulted in the decimation of domestically-owned manufacturing industries. Where Britain retains internationally significant capacity – in pharmaceuticals, aerospace, and motor manufacturing – the sectors are dominated by companies headquartered elsewhere in the world. An extraordinarily lax takeover regime allowed the sale of a genuinely globally-important tech company, ARM Holdings, to SoftBank, a Japanese conglomerate fund with close links to Saudi Arabia.

Ownership for larger companies in the UK is, in general, fragmented – according to the OECD, “in about 90% of companies listed on the [London Stock Exchange], there is no major shareholder owning 25% or more” – and significantly internationalised, with the majority of UK quoted shares now owned overseas. The result is a drive towards short-termism, chasing immediate returns at the expense of longer-term gains, and the fragmentation of liquidity of markets for ownership makes long-term relationships between government, owners and firms harder to maintain.

The point here isn’t to bang a big made-in-Britain drum. It’s to indicate that in a world where scale matters and price competition increasingly does not, it’s the role of the state and its relationship to major suppliers that matters most. Lose those relationships, and you are looking at losing overall. The market alone can’t deliver.

It’s in recognition of this that France and Germany are pushing to open up the EU’s notoriously strict state aid and competition rules, which could otherwise prevent national companies – or even EU-wide champions – operating at the scale and with the government support they need. Meanwhile, a recent European Commission report cited China as a “strategic competitor”, and industrial strategy – in the sense of deliberate government intervention to promote specific industries, firms or technologies – has returned to favour since 2008.

For a time when globalisation was in its high period, Britain could (and did) run an implicit industrial strategy for financial services, in which (whilst rounds of deregulation saw the disappearance of the older City of London firms) the combination of loose domestic regulation and heavy (if often implied) state support enabled a few UK-headquartered banks to grow to extraordinary size. RBS in particular ballooned itself from a still largely Scottish-focused domestic lender before 2000 to “the world’s largest bank” by asset value at the end of 2008, and had to be rescued by what the Treasury described at the time as “the world’s largest bailout”.

The collapse of global capital flows since the crisis, down around 65% on their 2008 peak, has been driven very substantially by the crumbling of the global bank model, with banks across the world retrenching to support their domestic markets. Britain remains a major money-dealing centre, London as the world’s largest foreign exchange market – and, notably, the world’s largest centre for renminbi trading outside of China. The appearance of fintech might, in theory, begin to provide a competitive edge of sorts, and both regulators and the Bank of England have rushed to embrace the potential. Britain’s public-sector research base contains major strengths in AI and Big Data technologies, but, even whilst there has been a flurry of start-up activity, its leading commercial AI research company, DeepMind, was sold to Google five years ago.

Which gets us back to Huawei. The US has alleged this officially employee-owned business is, in fact, controlled by the Chinese governmenta claim Huawei strongly denies – and has lodged charges against the company’s chief financial officer. These and other security concerns are the US’s key reason to insist the company is kept out of supplying 5G.

But it is also useful, for a US government deeply fearful about losing its technological edge to China, to cite these concerns. Perhaps in a sign of the times, even the usually loyal United Kingdom has only half-responded to them, with major telecoms companies citing the potential losses from a delayed 5G rollout. These points of tension are only going to become more apparent over time.

Build
 people-
  powered
   media.

Build people-powered media.

We’re up against huge power and influence. Our supporters keep us entirely free to access. We don’t have any ad partnerships or sponsored content.

Donate one hour’s wage per month—or whatever you can afford—today.

We’re up against huge power and influence. Our supporters keep us entirely free to access. We don’t have any ad partnerships or sponsored content.

Donate one hour’s wage per month—or whatever you can afford—today.