Unpredictable Patterns #96: Geopolitics and tech policy
Techno-containment, health tech as geopolitical advantage, decoupling myths and the question of the next Berlin wall (or stack)
It has become more or less a habit to mention geopolitical tensions when we discuss politics and policy these days - but what does that mean? And is our coarse-grained model of geopolitics helpful in understanding political pressures in any operational way? Are there things we should do differently, bets we should make or take, issues we should focus more on depending on what we believe about geopolitical futures? Or are statements about geopolitical trends more or less like statements about fashion?
In order to understand this we need to explore more in detail what we mean with geopolitics and how we can start to model geopolitical pressures in policy work.
What is geopolitics?
The idea of geopolitics starts out narrow - it starts out as the study of the impact of geography on politics and international relations. In early geopolitical models of medieval empires one key variable would have been the geographical distance between the capital where the decisions were made and the provinces - and even things like the time it took to get the decisions out to the provinces. This geopolitical lag where decisions and intelligence were made at the maddeningly slow rhythm of horseback riding created political challenges for any ruler at the time. The consequences - the importance of loyal local rulers with a sense of what the centre of the empire wanted - led to specific kinds of political structures and systems. In a sense geography presented a key constraint on possible political behaviours.
At one level geopolitics is really about geography.
Another example of how geography shapes politics is the distribution of natural resources. The distribution of oil and gas matters for political decisions - as has been made abundantly clear by the war in Ukraine - and as climate change forces us to shift away from fossil fuels we will see the emergence of new geographical distributions that matter. A simple example is the possible expanding network of hydrogen trade routes, plans and agreements that may create an entirely different dynamic from the fossil fuel network of today.
And it is not only energy that matters in terms of natural resources. The rare earth materials that are needed in semiconductor design is another example, as is the extraction of silicon.
Actual raw materials and natural resources, then, provide another constraint on political decision making that can be called geopolitical in a more traditional sense. But there is a difference between what a country chooses to focus on and the resources that only that country has. Not every country can meaningfully drill for oil, for example - but countries can decided what services and products to specialise in producing.
This brings us to a third component of geopolitics - trade. Trade is essentially the choices we have made to divide labour in the international economy, and these choices also have profound political consequences. Trade is a fairly recent basis for geopolitical considerations - both geography and natural resources have been around for some time, but world trade is a surprisingly recent phenomenon - and the kind of globalisation that the world saw from 1970 to 2010 was unparalleled. It also formed the basis of a new kind of geopolitics.
Trade requires access to markets and some basic sense of fairness. If one country subsidises their companies, sets tariffs and prevents foreign ownership and investment, that country can seek to craft a competitive advantage in world trade that can then be turned into relatively faster growth or expansion. The geopolitical question at the heart of trade, in many ways, is a question of what we believe are fair means of international competition and the distribution of global economic growth, rather than natural resources.
If trade, as it seems, creates a kind of economic convergence1 and a supply chain divergence - supply chains are globalised, and this trend is reinforce by very low interest rates in the later part of the period we are looking at (think of interest rates as the price of time: when the price of time is zero supply chains will grow in size and complexity2).
These new supply chains and the new economic convergence created by trade create a third, much more complex geographical constraint on politics: any movement to harm this system will not just create conflict between the participants in the system, it may also threaten the global economy as it increasingly has become integrated into a single complex system.
Trade is a strong force even as we discuss US-China decoupling today. It is easy to forget that Chinese exports to the US remain at close to record levels, and China remains the third largest trading partner for the US:
Geopolitics - then - consists of the constraints imposed by geography, natural resources and the complex economy produced by globalised trade, at least at a first approximation. This gives us a starting point to ask how these constraints then matter for tech policy.
Geopolitics and tech policy
Geopolitical constraints on tech policy sounds like a weird concept. Was technology not supposed to make geography irrelevant? This, indeed, was one of the early themes in tech commentary in the early 2000s. Frances Cairncross wrote, in the Death of Distance (1997) that time zones will matter more than miles and that geography will slowly fade as key constraint on, at least, businesses. And to be fair (it is oh-so-easy to deride early attempts to understand technological impact in hindsight) companies can think global first today if they think that is the right choice. But the overall tendency during the period 2000-2020 has been the opposite. We have lived through the careful re-territorialisation of the Internet: jurisdictions have been reasserted, digital sovereignty has emerged as an organising concept for the political will to have direct access to data and infrastructure, as well as a general accountability for Internet services and some countries have even developed alternative Internets, expecting to decouple from the Internet. It has become commonplace to describe the emerging phenomenon as the “splinternet” - a term coined by Clyde Wayne Crews in 2001 to describe a positive vision of multiple private Internets that could set their own rules (and hence enable a libertarian experiment), but now thought to describe balkanised web where the virtual world is closely mapped on the geography of the world.3
Natural resources matter, but they mostly matter in terms of trade. And that is where the majority of the tech policy issues connect with geopolitics. Trade wars, treaties and negotiations are increasingly the site of tech policy negotiations - and we should not be surprised by this. This debate has a multitude of dimensions that is helpful to disentangle.
Data. The flow of data - personal as well as non-personal - matters greatly for tech policy.4 Privacy issues have dominated this discussion, but as data increasingly becomes relevant in artificial intelligence and machine learning, we should expect that the focus will move from data protection to data governance across borders. Tech policy can create national advantages in trade and geopolitics in this space. One often undervalued example is health tech (and we will see that this is a pattern in Chinese investment). In the short run access to health data for machine learning will enable innovation in health in ways that can create geopolitical advantages. Assume - for the sake of argument - that one country creates an exemption for medical research and personal data, and in doing so spurs innovation that leads to significant progress in curing cancer: what will that then mean for the geopolitical balance if they can then control this capability? In the long run, however, the more important long term question is data governance. How a country sets out to record, curate, store and preserve data across public and private sector (and with a division of labour and responsibility that makes sense) will determine how well it can use machine learning to improve and adapt to a changing environment. Data governance, in that sense, is not just the care for databases that we sometimes envision, but the larger task of creating a new infrastructure that can power a new kind of collaborative learning for us as a society.
Compute. A lot has already been written about this - following the export controls put in place by the US on certain semiconductors - but the questions around compute governance - the way compute is tracked, monitored and distributed closely parallels the discussion about energy networks and supply chains and can, analogously, be seen to determine political power in the longer run.
Intellectual property rights and knowledge. The distribution, protection, enforcement and creation of intellectual property has always been a conservative force in globalisation, regionalising rights and licenses in different ways. This is likely to be a new avenue of protection and conflict geopolitically and can affect the basic flows of science across different networks. Overall, knowledge supply chains and networks are likely to become focal points in geopolitics, as country competition and technology races are re-introduced as political strategies.
And these are just three examples of areas where tech policy can affect geopolitical power balances and vice versa. We could certainly come up with many more if we wanted to - but the point is clear: the connection between tech policy and geopolitics is real.
What is the long game?
Geopolitical competition sounds innocuous enough: competition between different countries can be a good thing, right? But in order to understand geopolitics more in depth we need to also understand what the long game is. What is the geopolitical competition a competition about? International politics researcher Pak Nung Wong suggests that the ultimate goal here is much larger than we sometime admit:5
“Against such international dynamics, this chapter argues that, despite the U.S.-China trade war being apparently about trade disputes, it reflected a deeper dimension in the U.S.-China structural rivalry – technological competition for global economic supremacy.”
Now, this may seem obvious on the surface of it, but it does have some profound consequences. The difference between a trade conflict and competition for global supremacy is one of limits. If you are in a trade war you have limits to what you are willing to do and you will not be willing to translate the trade conflict into war or into a complex re-organisation of your own country’s economy. When you are in global contest for supremacy you do not operate under those limits. If Wong is right, there is no cost-deterrence operating as China evaluates options in the semiconductor trade wars, for example - they will do what it takes to reach semiconductor independency, since it is an existential issue.
Wong thinks this is true for the US as well, but in his analysis the US badly bungled what he sees as the start of a long conflict in China. He notes that: 6
COVID-19 also showed that the Trump government committed a serious strategic error in launching the trade war with China too early before the U.S. succeeded in relocating and restructuring the global supply chains.
If you are not just engaging in a trade conflict, but competing for global supremacy, well, then you need to act accordingly. We see more of this now, with the Biden administration and the re-territorialisation of the semiconductor supply chains, but China has responded forcefully in this direction as well - with plans to create technological independence across a number of different areas - from quantum computing to artificial intelligence.
Over time this long game will force tech policy issue not only into trade frameworks, but deep into national security contexts of different kinds. Decisions around data and privacy, intellectual property and even public access to government data may change fundamentally.
This shift can impact tech policy as well. Take the example of free expression. So far this issue has been debated in the context of democratic debate and censorship. Content moderation has been about removing illegal content and moderating what different countries considers harmful in one way or another. But very few security concerns have been present in the debate - but the recent advances in AI/ML technology that summarises and detects new patterns in information will mean an increasing in capability by several orders of magnitude for what is called open source intelligence. OSI is already much more powerful than most realise, but it is at the cusp of becoming much more powerful than covert intelligence, because we do not know what it is we reveal in aggregate.
Our public sector publishes enormous amounts of information, and individuals working for the government are generally thought to have the right to comment on government affairs - but what if that commentary and those documents can be analysed to discover key vulnerabilities and secrets that we do not want divulged? It is not hard to see how the question around public access to government documents and free speech can be cast in a national security context.
There is another aspect of the long game that is sometimes neglected - and that is the way tech policy can be used to further the stability of a political system. There is currently a giant natural experiment unfolding between democratic states and authoritarian ones, where the best is whether technology should be used by citizens to further democratic debate and deliberation or by the state to ensure compliance and stability. Technology, very clearly, can be used in both ways - and the experiment so far is inconclusive, even if the debate might make you think that the democratic countries feel they are losing: the rhetoric about social media as a threat to democracy is curious in the light of how technology is used as a real threat to democracy in authoritarian countries.
Now, of course, your conclusion could be that the authoritarian states will win this bet - that technology generally erodes democracy if to freely available and used without restrictions - and you would not be the first to think so. This is the conclusion that Plato has king Thamus draw when he denounces writing in The Phaedrus:
What you have discovered is a recipe not for memory, but for reminder. And it is no true wisdom that you offer your disciples, but only the semblance of wisdom, for by telling them of many things without teaching them you will make them seem to know much while for the most part they know nothing. And as men filled not with wisdom but with the conceit of wisdom they will be a burden to their fellows.
If this is indeed what you believe, then you should also advocate for a tech policy that can preserve and strengthen democracy - actively - rather than just argue for more content moderation. The argument that content moderation is the cure to democratic deficits is certainly possible, but seems absurdly myopic: is there nothing we should do about deliberation, citizen participation, government accountability, data governance, social learning and other subjects? And are these areas not infinitely more important than the barely concealed trade war with the US that some have argued that European tech policy risks devolving into?
The long geopolitical game is one with two dimensions at least: economic global supremacy and political system stability and endurance. It is, perhaps, ultimately a contest between democracy and authoritarianism.
The nature of that game is also important to explore closer. Wong, again, gives us an interesting model: one that upgrades Kennan’s containment strategy into a technological standards fight. Wong suggests that to understand the current trade wars and tensions, we need to look much more closely at early instances of similar conflicts - including the 5G-conflicts where China and Huawei sought to establish new standards for communication networks, and the battle for swing states in Europe and Africa was fought through public procurement, standards organisations and more. The first blow in the fight that is playing out today was actually dealt a long time ago, Wong suggests - as China tried to launch its own WIFI-standard to compete with WLAN. They lost that fight, but it taught the Chinese strategists that unless they have standard powers, they will not be able to defend against the containment strategies the US can employ technologically. The fight has since gone down into the stack to semiconductors, but China is contesting standards commanding heights even so - creating a long game that will be about communication infrastructure, operating systems and application layers.
After the Berlin wall we are building a technology stack that separates the two competing super powers.
In a way we are building a new Berlin wall - but this one is a stack - made up of protocols and infrastructure, and unfortunately may turn out to be much harder to dismantle.
So what?
We often pay respect to geopolitics, but as a backdrop against which local policy debates play out. If the trends in this note play out, however, we can make the following predictions.
Geopolitical concerns will shift tech policy from local policy to trade and then further into national security contexts. Increasing technological capabilities will affect the overall balance of power in our systems.
The race between the US and China may increase the pace of development in crucial technologies - but not as much as during the cold war, since technology is now primarily developed in the private sector in the West. One possibility is the creation of public-private partnerships around grand innovation projects, however, in the interest of geopolitical competition.
Technology will increasingly be used to further the stability of authoritarian political systems as a means to compete with democracies, and attacks on open technologies will worsen, as they are seen as a potential means to erode democracy. Defending against this by focusing on content moderation is likely to fail - if there are no simultaneous efforts to build out technological amplifications of democratic institutions.
Technological containment and counter-containment may create a separation of political and economic systems that will be much harder to dismantle than the Berlin Wall. Paying close attention to infrastructure, operating systems and application layer integrations across geographies will be a way to ascertain how techno-containment is playing out.
As always, thank you for reading and please keep the comments coming. There is so much interesting to think about here!
Nicklas
See Baldwin, Richard The Great Convergence: Information Technology and the New Globalization (2016) for a version of this argument.
See Chancellor, E The Price of Time (2022) - a magisterial review of the history of interest, natural rates of interest and the impact on policy and politics that interest has.
We should note, in fairness, that the virtual world has approximated the geographical more than we may expect, but because of other reasons: the trust needed to buy something online has, it seems, a geographical limit and language presents a strong constraint on how we consume information as well. It raises the interesting question of if commercial trust and language are underestimated geopolitical levers.
See Wong, P. N. (2021) Techno-Geopolitics. 1st edn. Taylor and Francis. Available at: https://www.perlego.com/book/2806927/technogeopolitics-uschina-tech-war-and-the-practice-of-digital-statecraft-pdf (Accessed: 5 November 2022).
Ibid.
Another fab piece. Lots of thought-provoking points here.
"Think of interest rates as the price of time" — love it.