Mark Coeckelbergh, Green Leviathan or the Poetics of Political Liberty: Navigating Freedom in the Age of Climate Change and Artificial Intelligence , Routledge, London, 2021, 150pp., £39 (paperback), ISBN: 978-0-367-74779-4
It is well known that we are not perfectly rational and that our irrationality (or, more precisely, our stupidity) results in us making disastrously short-sighted decisions. This, at least in the case of the climate crisis, has led us to where we are now: on the edge of global catastrophe unless we do something urgently. What if one of the ways to mitigate this crisis is to have an authoritarian state, monitoring and controlling our every decision, to ensure that our actions align with global climate objectives? With the use of (big) data analytics, we could perhaps engineer the perfect political environment in which people would be structurally disabled from doing things that are bad for the biosphere. What about our leaders? Human leadership is also notoriously dreadful, especially during times of crisis. Perhaps we ought to rely on artificial intelligence (AI) systems to make decisions instead of humans? Such systems are not driven by human greed and hubris and could thus provide a further AI-based solution to climate change problems. To reap all these benefits, all we have to do is give up our freedom.
These are the central themes that Mark Coeckelbergh tackles in his latest book, Green Leviathan or the Poetics of Political Liberty. Essentially, he is concerned with the dilemma that AI and climate change present: Do we accept an authoritarian ‘green leviathan’ (fueled by AI and data) as a solution to the climate crisis, do we leave things as they are and become obsolete, or can we find a third way? Coeckelbergh believes there is indeed a third way, which forms the text’s central argument.
In the opening chapter, Coeckelbergh provides us with a road map for the rest of the book, and we are introduced to the fundamental concepts that form the bulk of the argument. While drawing, quite naturally, from standard political thinkers in the Western philosophical canon (such as Hobbes and Rousseau), Coeckelbergh refreshingly brings these thinkers to bear on questions concerned with nudging, the Anthropocene, and posthumanism. The guiding theme throughout is the question of liberty; that is, can a meaningful sense of political freedom be maintained in the face of the climate crisis and developments in AI.
Coeckelbergh makes explicit the key tension between individual freedom and the collective good at the heart of much of Western political philosophy. To get a proper handle on this tension, we need to see how it applies to the central concerns of the book – climate change and AI. The first thing Coeckelbergh points out is that the dominant mode of dealing with new technologies is broadly liberal-democratic: new technologies should respect freedom and not undermine democracy. However, some doubt whether such an approach, when it comes to AI, is feasible. These authors claim that to deal with the existential threats that AI might pose, we need to move beyond liberalism. These doubts stem from our ‘hackability’ as humans and the fact that AI systems trained on massive amounts of human-generated data will make it easier both to understand and to manipulate the actions of entire populations. Thus, we see the tension: AI threatens our liberty by virtue of its potential misuse by authoritarian regimes (and, as we shall see, this is still a problem even if the ends of such manipulation are ‘green’). But AI also encourages thoughts of radical libertarianism, where it is freedom at all costs, with no regulatory oversight, which comes with its own set of problems.
We can tell a similar story of the interaction between liberal democracy and climate change: Do we respect liberty to such an extent that we are crippled in our ability to put a system of checks and balances in place that demands accountability for those responsible for environmental destruction? Or, given both the nature of the threat and those who are most responsible for it (wealthy elites), do we give in to authoritarian impulses and force compliance? As Coeckelbergh puts it, ‘the choice seems to be between absolute freedom – which preserves liberty but at the cost of not dealing with the problem at hand – and absolute authoritarianism, which solves the problem but at the cost of liberty (pp.16–17). Coeckelbergh suggests that this situation, at the heart of debates surrounding the climate crisis and AI, resembles the ‘state of nature’ articulated by Thomas Hobbes in the seventeenth century. In each case, we have individual and corporate actors who aim to protect themselves through survival and competition, with little to no regulatory oversight. The result is a persistent conflict over resources and, in our case, the deterioration of our natural environment. For Hobbes, the solution to this problem was to create an all-powerful artificial body politic (his famous Leviathan), a kind of mortal god, capable of enforcing the law through fear, violence and intimidation. Coeckelbergh terms this kind of approach to climate change ‘green Leviathan’. This green Leviathan could counteract human stupidity by forcing us to agree on ways to reduce our carbon footprint (at both the individual and collective levels).
However, while this solution gives us an all-powerful dictator, it says nothing about whether (and how) this authority is meant to know what is best for us. Here Coeckelbergh draws on Plato and suggests that we could also imagine an aristocracy, which seems to be the Hobbesian solution with extra layers of fluff. Instead of relying on only the sword, we can now add wisdom and virtue to the equation. However, what unites both the Hobbesian and Platonic solutions is that they advance authoritarianism as a solution to our political problems, and thus neither proposal is democratic. If we care about democracy (and both Coeckelbergh and I think we should), then we need another way to deal with these issues. Luckily, there is a third way. This is where Coeckelbergh draws on one of my favourite stories in all of fiction, that of the Grand Inquisitor from Dostoevsky’s The Brothers Karamazov. The essential takeaway from the tale is that we cannot cope with freedom: free choice and autonomy are crippling for most people, and so making decisions on their behalf is, in fact, the kindest thing we can do for them. People must be deceived for their own good.
This bears on climate change and AI insofar as we can imagine an authoritarian super-state making use of sophisticated machine-learning systems that know us better than we know ourselves and implementing this information to ‘nudge’ us into making certain choices. Not only is the removal of our freedom personally and psychologically advantageous (given the take-home message of the Grand Inquisitor), but it also turns out to be politically advantageous, given how it increases our chances of successfully engaging with the climate crisis. People should not be left to make their own choices. This argument assumes a deeply mistrustful understanding of human nature: we cannot be trusted to act for the common good (or even for our own good). So we need to be kept in the dark about certain things, just in case they confuse what limited cognitive power we have at our disposal. Coeckelbergh then briefly discusses how a utilitarian style of reasoning might emerge from such a perspective. Given that human nature is in general quite awful, in conjunction with a political and moral theory that stresses the moral significance of consequences, utilitarians might feel fully justified in this kind of deception as it does indeed seem to maximize utility insofar as we do not become extinct. Again, we see the spectre of authoritarianism, this time in the form of calculating utilitarian rationality.
There is, however, a milder version of the story: drawing on work done in behavioural economics and social psychology, Coeckelbergh suggests that nudge theory might provide a means to preserve freedom without deceiving people. The basic idea behind nudging is that certain choice architectures are better or worse than others. Expressly, we should adopt and implement designs that ‘nudge’ our behaviour in desirable ways because of our cognitive limitations as human agents. From placing food at a buffet (first cheap carbohydrates, followed by more expensive proteins) to ‘opt-out’ models for retirement annuity contributions, nudges can be for better or worse. As noted above, if we cannot trust human nature to make the right choices, then perhaps we should trust nudgers (those experts who design these choice architectures). This ‘libertarian paternalism’, according to its defenders, does not infringe on the liberty of individuals. However, as Coeckelbergh notes, it is difficult to see how this is meant to play out. In practice, the nudgers decide for the nudgees: ordinary folk are not trusted to make their own judgments, and so again we see a general mistrust of human nature. Alternatively, one might even think that, given the existential threat of climate change, nudging leaves individuals with too much liberty and that what we need is more rather than less paternalism.
However, this is not the only way that we can understand human nature. For example, it is possible to take an egalitarian perspective, such as that championed by Jean Jacques Rousseau. In his view, the state of nature is a kind of Garden of Eden, and humans become corrupted only by society (specifically, property) and the inequalities it generates. Rousseau’s solution (to reconciling the freedom of the individual and the state’s authority) is that citizens subject themselves to ‘the general will’, which is a function of the collective wills of each individual. Thus, people obey their own wills, not the wills of any particular person, preserving individual liberty and guarding against authoritarianism. Coeckelbergh suggests that Rousseau’s view helpfully opens up a new direction for us to explore, that of positive liberty. If we replace the pessimistic (Hobbesian) view of human nature with a more positive one, we can shift our focus from negative liberty (‘freedom from’) to positive liberty (‘freedom to’). So, what other notions of liberty might be relevant to this positive conception?
When we frame freedom in purely negative terms, it is easy to think that regulations and constraints that aim to slow climate change also threaten individual liberty. However, conceived positively, it becomes possible to see these very same constraints as increasing our positive freedom by enabling or enhancing individual and collective wellbeing. This approach understands liberty as being concerned with certain capabilities (such as bodily integrity, to laugh and play, to have political choice and participation). In such an approach, there is still the risk of authoritarianism and paternalism. For example, someone might claim they know better than I do what my needs and capabilities are. This draws our attention back to negative liberty: How to protect individuals from this threat? Coeckelbergh proposes that we look towards a Deweyan understanding of democracy. Dewey argued that democracy is a means for communities to problem solve and viewed it as a kind of ‘organized intelligence’. Thus, when people engage in democratic decision making, we should trust them to decide on the best course of action. In the case of climate change, this could mean trusting them to do what is best both at the local and global levels. The point of politics, then, is to create ‘the conditions for people to enhance their capabilities and meet their needs’ (p.61). This goal should be pursued in a participatory way, where citizens and experts engage in a dynamic dialogue about what ought to be done. Moreover, this participatory approach emphasizes the virtues that we might expect from our politicians and policies. That is, they should reflect what we might think of as the ‘good life’ and aim to make this a reality.
Coeckelbergh reflects on just how we might get this done, specifically regarding climate change and AI. Here, we are introduced to various ways in which human agency influences the planet. More specifically, how, in the age of the Anthropocene, humans have become a ‘geological force’ dominating and capable of determining the future trajectory of our planet. Climate change and AI are essential parts of this story: we alter and pollute the environment, and we come up with new technologies to do so. As Coeckelbergh puts it: ‘AI is then a tool for the destruction of ancient natural environments and at the same time a tool for the management of the Anthropocenic planet’ (p.75). Moreover, just as AI can help us destroy, so it can help us create or preserve. By analyzing massive datasets, AI-powered systems might develop mitigation strategies that humans alone might never have dreamed of.
However, Coeckelbergh is critical of the usage of the Anthropocene concept, especially how it conceals that it is not just human beings in some general sense who have control, but rather a very select minority. This minority of powerful corporations and individuals is responsible for the current climate crisis, and to obscure the fact that some people are more culpable than others (as the Anthropocene concept seems to do) is problematic. However, when we take this seriously, the Hobbesian dilemma again rears its head: How to get people to comply? Force them. Furthermore, based on its mistrust of human nature, the Hobbesian solution would be not to trust people to do the right thing if left to their own devices. They need an authority with unilateral powers of enforcement. Coeckelbergh wants to go beyond such reasoning, and one way to do this is to investigate whether there are other political values besides survival.
For example, we might want to focus our attention on principles such as justice and equality. In the case of the climate crisis, we might say that instead of concentrating on liberty, we should turn our attention to how risks are distributed and ensure that this distribution is just. As Coeckelbergh puts it, ‘If we shift our concern to climate justice, understood as climate fairness, then we can and should talk about whether the distribution of risk and harm related to climate change within the collective is fair and then (if necessary at all) discuss what this means for liberty’ (pp.83–4). Following this, Coeckelbergh outlines various conceptions of distributive justice (sufficitarianism and prioritarianism), examines arguments concerning intergenerational ethics, and considers the likelihood of a ‘climate revolution’. To be appropriately successful, a revolution should preserve or enhance democracy, but this, of course, is not guaranteed. While borrowing from Marx and Engels to show how such a revolution might play out, he argues that our body politic is not some static structure that is temporally fixed. Instead, it is in a constant state of becoming, and so it is dynamic and open to change. One such way we might change it for the better is to advance a non-anthropocentric understanding of who or what comes to count, politically speaking.
Up to now, Coeckelbergh’s argument has been concerned only with human-centric perspectives on politics and liberty. Chapter 6 looks at various arguments for less anthropocentric perspectives, which seek to include both animals and the environment as part of our political world (p.101). That animals should be accorded some moral status is no longer controversial. What might be controversial, however, is including them as political subjects, as citizens. However, just because they lack political agency does not mean that their freedom does not matter. They have interests, and these interests ought to matter in our political theorizing.
Turning to nature itself, and borrowing from Bruno Latour, Coeckelbergh suggests that there is no neat separation between science and politics, which creates hybrids. Take COVID-19, it is both a scientific (epidemiological, virological) issue, but it is also a profoundly political one (lockdowns, restrictions on citizen freedoms, etc.). Consequently, Latour argues that political philosophy is not only about subjects (such as humans and animals), but also about actants: non-human actors that influence the world. Thus, we have an extension and expansion of what counts politically, which has implications for thinking through climate change and AI.
A consequence of this is that we can now see climate change and AI as political issues, as even climate change, which we think of as scientific, is also profoundly political (a kind of hybrid). We can also see that it is about the technology and science of AI. Even so, there are questions of power and politics: who develops these systems, what data they trained on, whether they respect citizen privacy, etc. Moreover, we can see that establishing a clear dividing line between that which is ‘natural’ and that which is ‘artificial’ is doomed to failure. Our existence is not something that can be separated from the environment in which we are embedded. We do not somehow exist outside nature, and thus the goal of our political projects should not be to go back to some original Garden of Eden. Instead, once we conceive of politics as being in a constant state of becoming, we understand that ‘we have to fine-tune and shape our ongoing, close relation to nature’ (p.115). The question of AI and climate change is therefore transformed into a relational one. How ought we to relate to the environment, given that we cannot ever escape from it or be outside it? Through this lens, our understanding of politics is reconfigured from a human-centred preoccupation to one that includes animals, the environment and even viruses.
In his final chapter, Coeckelbergh outlines some challenges that the (relational) liberal-democratic project may face and summarizes the golden thread of the book. The key take-home message is his attempt to forge a ‘third way’ that resists authoritarianism and radical libertarianism. He proposes that we understand politics as a poetic process: it is something we participate in but do not have complete control over. This also stresses the interdependent relations that we share with other creatures and ‘things’ and that understanding the complexity of these relations requires work. Politics is something we do and that we have to keep on doing. It is a process that has no definitive end and no definitive limits.
Overall, this is a wonderful book that ties together many seemingly disparate themes and unites them into an easily-digestible but also philosophically-rich story. Too often, specifically in the case of AI, attention is placed on specific technical solutions or problems that the technology might solve, and while these have their place, they also leave out a lot. Moreover, such technical approaches are often made at the expense of the potential moral and political implications these ‘solutions’ might have. Coeckelbergh’s book restores some balance to this debate, not necessarily by taking a specific stance towards technology, but – and perhaps more importantly – by providing us with a coherent framework within which we can make sense of developments in AI specifically, but also in technology more generally. Additionally, the book shows us that it is possible to make progress on issues such as climate change without necessarily infringing on people’s liberty. Once we accept that our freedom has both negative and positive components, it is easier to understand how policies that aim to secure an inhabitable planet for future generations can be justified.