Countries and Companies Part 1

At the AI Fringe last week, Peter Kyle MP noted that SpaceX had sent more into space, by weight, last year than 'any other country'. It was a slip of the tongue, but a revealing one. The day before Gina Neff had asked whether Big Tech companies were 'quasi-states'.

How different would the AI Summit - and the Sunak-Musk love-in - have been of we took this thought seriously? In this post I examine the plausibility of the claim. In Part 2 I look at the consequences.

What is a state?

I have always found Max Weber's definition of a state insightful:

a human community that (successfully) claims the monopoly of the legitimate use of physical force within a given territory

Let's examine the key terms, starting right to left. In Weber's definition the reference to a territory determines which people a given state has power over by their location in space and time. It also determines which natural and human resources that state can use its power to exploit for its own benefits. In both cases there will be some vagueness, some fuzziness around the edges when we apply these territorial boundaries in practice.

A non-territorial state would then need a monopoly of legitimate force over a population not identified geographically, and the potential to use that force to exploit certain natural and human resources, again not identified geographically. Our ability to imagine this is hindered by the fact that there is no longer any land which is not part of some territorial state, and thus that existing state already has a monopoly of the legitimate use of force on people who are there. It appears there is no population left for a non-territorial state to exercise its state powers over.

But what of 'international waters'? While there isn't a resident population¹, there is no distinction between legitimate and illegitimate use of force and no monopoly control of exploitation of natural and human resources within that extra-territorial region.² Does this allow us to imagine non-territorial forms of state power?

The James Bond film Tomorrow Never Dies involves someone manipulating GPS signals in order to trick a British warship into straying into Chinese territorial waters, triggering an international incident. This makes clear that those who control the technology used for navigation and communication in international waters have significant power over people in those places. While the film is based on the covert exercise of that power, one can easily imagine overt coercion through threats to withdraw services which are essential to survival on the high seas. While this is not quite 'physical force', there is the potential to use this power to control behaviour as effectively as the threat of arrest and incarceration controls behaviour in territorial states.

Quasi-States

We have seen how the provision of tech services remotely can give power over people insofar as those people depend upon those services to survive. This allows us to define a population - in a non-territorial manner - over which coercive force is exercised by their dependence upon those services. Monopoly provision, or at least huge barriers to switching provider, create powers similar in strength to the state. Starlink may not have police with guns, but in some circumstances they have as much coercive power as the threat created by the existence of armed forces.

However, what makes a state distinctive from other organisations which can use force is - on the Weberian definition - the distinction between legitimate and illegitimate use of force. Organised crime may have achieved a monopoly on the use of physical force in a specific town or neighbourhood, but that itself doesn't make the criminal organisation a quasi-state, because its use of force is not legitimate.

There is some circularity in Weber's definition here: what makes a state's use of force legitimate is that it is within the law, and yet it is the very state which creates the law. For example, in some states capital punishment, the ultimate use of physical force, is legal and in others it is not.

The situation is not that different with tech companies, which write, enforce and change their terms of service (and more detailed, often secret or obscure, rules built into how their service runs) themselves. They often import into those terms the laws of the territorial states where their users live, but they also often try to get around those laws, with quite striking degrees of success.

In order to look like a state, tech services also have to create an effective monopoly in their legitimate use of coercive power. Territorial states achieve this by using force to control anyone else who seeks to use force within their territory.

In tech services, this achieved through closed standards and lack of interoperability. It doesn't mean that there is only one service provider or that a user must choose just one (even if the service providers would like such situations). Rather, monopoly power here means that when someone does use a service, everything they do with that service is in the power of that service provider. Contrast email and messaging: I can use both WhatsApp and Telegram, but when I am using one of those, that service provider has control over everything, all my actions and interactions occur within their system, and they have a monopoly on power within that system; not so with email.³

Which brings us to the last element of the Weberian definition: states successfully claim a monopoly of legitimate force in their territories. The point here is that a state fails and ultimately collapses if it loses its monopoly, through revolution, civil war or invasion. Hence the need for armies to ensure 'security'. Other ways states can fail are through mass civil disobedience⁴ or emigration. Hence the need for large police forces and strong borders. In general states seek population growth, giving them more human resources to exploit and thus welcome immigrants and disincentivise emigration. Population growth is only constrained when over-crowding leads to civil unrest and thus threatens the success of the claim to a monopoly of legitimate use of force. Successfully claiming a monopoly on power is very clear for tech services: users keep using the services they provide (which give them that monopoly on legitimate power).

All this shows clear analogies in the relation between tech companies and their users. Their power depends upon their users continuing to use their service under its current terms of service and they use that power over their users to create an effective monopoly of the legitimate use of that power with each user.⁵

Tech Citizens

It seems then that insofar as you make Facebook / WhatsApp, Google or Apple fundamental to your life and wellbeing, you become a citizen of their quasi-state (and if your friends or colleagues force you to reluctantly use their services, you are like a tourist: still subject to the state power where you visit). Users may have multiple citizenships, but they cannot continue to use these services without being subject, in that use, to their monopoly on power.

Of course, you are 'free' to leave and use a different service or none at all. But they make that hard, make the transaction costs as high as possible, and use their existing powers to create coercive forces which you in their power. You are also free to leave the country where you live, work and see your family and friends, but it is a freedom it is hard to exercise and that increases the power of the state over its citizens.

Power vs Violence

Remote technologies give the providers of those services significant forms of power over populations which are not terrestrially defined. There exercise of that power can also acquire features similar to the monopoly and legitimacy of state power, Like most exercises of state power, this is primarily through coercion and manipulation rater than force. But the threat of force is there, though for tech service providers the exercise of force is limited to restriction or removal of services.

As Weber noted, a state's power and use of force is ultimately grounded in physical violence. While many states have eschewed the most obvious forms of violence, such as capital and corporal punishment, all retain incarceration and exile (aka deportation) which are forms of physical violence.

This might seem a clear difference between a genuine state and any company. That is historically false, as the infamous East India Company demonstrated, and there are reports of current companies whose business is the extraction of physical resources using physical violence. But tech companies are different in that physical violence, especially physical violence directed towards their users/citizens, would not serve their ends.

Countries are concerned with the physical behaviour of their citizens and when they do try to control their thoughts, that is to manipulate their physical behaviours. Good citizens go to work⁶, pay taxes, and don't emigrate.

Good 'citizens' of tech companies give their time and attention to that service, they use it. One can imagine a dystopia where use was mandated with the threat of physical violence, but so far that has not been necessary.