This essay draws on Zero Ideas’ latest research paper, How AI will reshape the political opportunity for climate. Both that paper and this essay are collaborations with Mariia Shakh, a practitioner in the corporate implementation and governance of artificial intelligence and a senior director for AI transformation at one of Europe’s biggest banking and tech groups.

Artificial Intelligence stands to shape our politics in many ways. It may amplify issues that are already salient: concern about jobs, inequality, power (in both senses), energy prices, local planning. Its bots may disrupt the democratic process itself. These threats are real, well recognized, and much discussed.

More positively, perhaps we will each be ‘individually empowered’, in Mark Zuckerberg’s vision.1 But even without such a rosy view of what AI may bring, it is AI’s threats that will bring an urgently needed change to what is politically doable. Here’s why and how.

The biggest challenges we face in the world today are systemic. Climate change is the most prominent example. Both the risks we face and the responses required operate at the scale of economies, institutions and societies. But despite decades of trying, we have not made those systemic issues sufficiently or consistently politically salient. Political salience sits in the practical here and now, in the household, dominated in the UK, U.S. and elsewhere by the cost of living.

The structural shift in global geopolitics is another systemic challenge that our here-and-now politics isn’t connecting with. People can feel the gap. That feeling is breeding disillusionment with politics, but doesn’t help to close it.

Artificial Intelligence is clearly another big, systemic challenge. The difference, we argue, is that AI has what it takes to close the gap. AI has what climate and geopolitics have lacked. People experience it directly in their lives, while also asking the big questions about power, governance and purpose. Its mix of speed, scale and immediacy provides the paths we have been missing from personal experience to societal concern, as the past month’s discourse has demonstrated.

If concern about AI makes the big, systemic issues salient, that can open up the territory for climate change and geopolitics too. The significance of AI is not just in its own effects on society; it is also in how it may expand how we think politically about the future.

The key question is how to bring political discourse from the immediate issues in my life to the systemic issues shaping society. Figure 1 maps out the territories for what can be politically salient. As well as the distinction between my life and society, it also contrasts concern about what works with concern about what matters, in terms of values and meaning.

Figure 1. Schematic ‘salience map’

The default territory for political salience is in the bottom left: economic security and opportunity. This is what President Clinton’s campaign strategist meant when he said, ‘It’s the economy, stupid.’

Staying within my life and household, more personal and ‘softer’ issues can also become salient: the attachment to people, place, and community in the bottom right. This territory is concerned with belonging and meaning. It includes the connections I feel to friends and family, to where I live and the community (or communities) I am a part of, and to nature and the physical world.

Moving up to the societal level, the need for effective leadership in the top left is concerned with what is stopping the system working for people. It includes the distribution of wealth and power, how we manage the transition to whatever is next, and questions about the competence and effectiveness of the system of governance.

And finally in the top right, what it means to be human is concerned with the big questions of who we are and what we are here for: our identity and individuality; human dignity and purpose; the pursuit of the common good; and the agency and responsibility of society at large.

While the cost of living may be dominating politics today, this is not fixed. In our research paper2 we show that since the Second World War the UK, U.S., Germany and France have each had important national elections fought in every one of the four territories of the map in Figure 1. And what we see from those examples is not just that salience happens everywhere, but that it usually (though not always) follows from brokenness or loss:

Attachment to people, place, community typically becomes salient after the accumulation of a per­sonal sense of loss, occasionally acutely but more often chronically, building up over time and eroding social cohesion. As an example, the U.S. electing Donald Trump in 2016: the political recognition of a social loss (arising from an economic one) built up over time.

Need for effective leadership typically becomes sa­lient when confidence in an institution is lost, whether because of demonstrable failure, misbe­haviour, or a step change in the challenge. As an example, the UK electing Margaret Thatcher in 1979: responding to a sense of ungovernability after the Winter of Discontent, strikes, inflation, tax, and the growing exercise of union power.

What it means to be human typically becomes sa­lient when confidence in ourselves is lost, and also when a belief in ourselves can compensate for losses elsewhere. As an example, the U.S. in 1960, after the success of Sputnik challenged U.S. competence and leadership, setting the stage for the space race.

So far we have established that making systemic, societal issues politically salient is a realistic and achievable idea. But there are good reasons why it often doesn’t happen.

Climate change has failed to move up the salience map because most people in rich countries still experience it abstractly, while they worry about experiencing the interventions to address it concretely. The most se­rious effects of climate change remain a proba­bilistic future risk, mostly affecting future gen­erations and often people in other countries; whereas avoiding those effects is seen to impose costs and constraints on me and my family, here and now.

Both in the functional concerns about costs and jobs, and in the value-based con­cerns about ways of life, the immediate threats from climate actions are often more salient than the worst threats from climate change itself. And as Figure 2 shows, without more of an immediate sense of loss in the immediate (lower) half of the salience map, there is a weak foundation for growing sali­ence in the systemic (upper) half.

Figure 2. Climate change’s struggle for salience

In geopolitics, neoliberalism illustrates how a political theme can migrate across and up the map, from immediate to systemic—but also how slow the migration can be (Figure 3).

Figure 3. The evolving salience of neoliberalism

Workers in rich countries have experienced global competition, offshored jobs, and stagnating wages since the 1980s.3 Their economic secu­rity has been politically salient since then. But until the 2008 Global Financial Crisis, economic growth and abundant con­sumer credit helped contain the political con­sequences. Only since the crisis have the underlying is­sues become politically salient in terms of people’s attachment to people, place, and community, reflected in the rise of national­ist-populist parties and movements in the U.S., UK, Germany, France, and elsewhere.

What makes the issues systemic is that the sense of personal disconnection has now evolved beyond an immediate, personal issue, into a sense that society is not being run for me or people like me.

This is more than a frustration with individual leaders, governments, or policies. G7 govern­ments find themselves unable to respond effec­tively, because they can’t reconcile the expecta­tions of their people with what is needed to keep their economies competitive in the global mar­kets they helped to create. The social contracts they had made with their people during the Cold War, justified by the ideological competition from Communism and protected by borders, have proved unaffordable in an era of globalism.

As it becomes clear that this structural challenge can’t be solved by a change of government, people start to question who and what we are running society for. This is not (yet) at the level of what it means to be human, but there are signs that we are no longer only in the territory of what works, but increasingly of the bigger questions of what matters.

The journey across the map is still incomplete after more than 40 years.

Climate change hasn’t moved up the map. Neoliberalism has crept up slowly over a generation. But the dynamics of AI are different from both. What’s significant about AI in this context is not that it is also a systemic issue, but that it provides the paths from the immediate to the systemic that climate change and neoliberalism have failed to achieve.

People’s experience of AI is far from abstract. We are already reaching the point where most people in developed countries use generative AI directly, whether at work, home, or both. And those that don’t use it themselves experience it almost as directly in search results or customer service. It’s already a factor in individual career choices, even if the impacts are unknown.

The observation that salience usually follows from brokenness or loss is significant, because AI is positioned to drive various forms of loss in every territory on the salience map. Figure 4 summarizes the picture; our research paper cited earlier describes the details. The important point here is that, with all these losses being felt at once (whether as fear or as reality), there are multiple, direct paths from these immediate experiences to the salience of systemic, societal concerns. That is obviously not a desirable state, but the cloud may have a silver lining.

Figure 4. Potential salience-driving losses that may result from AI

Already today, opposition to the build-out of AI data centres has become a political issue in the U.S. midterm elections. The concern driving that opposition is not just about electricity water use. To quote a recent Brookings commentary, ‘AI has become an existential threat for many individuals, and data centres have become the means for people to express their worries about overall AI development. It is hard to fight AI rollouts, which are largely controlled by tech firms, but easy to gin up community opposition to data centres because they require local permits and environmental safeguards.’4

The functional, immediate concerns about job losses, career prospects and the future value of skills are already culturally salient. They are not yet established as electoral priorities, but they have several characteristics—personal exposure, potential loss and identifiable concentrations of power—that could make them so.

From there, what is remarkable about AI is that salience appears already to be spreading in all directions at once (Figure 5).

Figure 5. AI’s paths to systemic salience

On the functional (left) side of the map, the gap between how fast AI capabilities are growing and the sluggish capabilities of our institutions to respond is both vast and visible. There are increasingly noticeable signs that governments, business, and even the AI frontier labs themselves, are not in control. Leaders’ fear of missing out has been overriding people’s other important fears, as well as more purposeful ambitions. This comes at a time when populations are already questioning whether their governments are complicit in what’s wrong with the world, too dependent on the status quo to lead meaningful change.

At some point this institutional incapability may translate into illegitimacy, as people shift to the fundamental questions that most of our institutions aren’t asking, let alone answering, about what matters: Productivity to what end? What are we running society for? What constitutes human flourishing?

Meanwhile a more immediate, personal-scale loss of human relationships and interdependence, as people turn to AI for companionship and self-reliance, provides a parallel path to the same territory of what it means to be human.

Not all of these issues will translate into political salience. But the multiplicity of paths, and the short, direct routes to the upper (systemic) part of the map, suggest that AI may quickly shift people’s focus from the immediate to the systemic to a degree that challenges our assumptions about what people find salient.

When it does, AI stands to open up the salience of the systemic territories where climate change and geopolitics need to play, but which they currently can’t reach. Once those territories are politically salient, our big, systemic challenges can compete for political attention on their own terms, rather than hiding behind proxies like the cost of living. What’s most threatening about AI may create a critical political opportunity.

Let’s start big.

In all of human pre-history and history, we have had three great epochs, defined by three great breakthroughs in our command of energy: fire, farming, and fossil fuels. In time, these lasted, respectively, about 400,000 years, then 12,000 years, then less than 300 years. But each breakthrough unleashed such a jump in our population, that in human presence the three epochs were much the same size: about a trillion person-years.

The figure shows this schematically. (It’s not drawn to scale; if it were, it would be too extreme to see what is going on.)

Today, with the world’s population on track for ten billion people, and life expectancy in rich countries closing in on 100 years, the next trillion-person-years will last just one human lifetime.

This changes the context for how we might think about the next century. It won’t be just one more in a line of some 4,000 centuries of human existence. It will be the fourth great human epoch. This is the red column in the figure.

The danger of thinking on such big scales can be that it makes our own actions feel small. In the spirit of being a ‘good ancestor’ to future generations, The Long Now Foundation strives for a ‘world that operates at the timescale of civilization.’1 The idea is that:

Everyone in this world would find themselves in the middle of the story of civilization.

Everyone in this world would understand themselves as someone else’s ancestor.

Everyone in this world would remember that they are part of something larger than themselves.

To help to normalize this civilizational timescale, the Foundation adds a leading zero to its date formats; we are in the year 02026.

But that leading zero hints at our own insignificance. In the context of the whole of human civilization over tens of thousands of years, how much can it matter what you and I do? By suppressing our sense of agency, a long perspective of many generations gives us ‘a reason for passivity’.2 We can justify whatever we do today, because the causal impact on the long-term future seems tenuous, and we have plenty of time to change course.

For the sake of our planet, some would like us to extend our thinking to even longer time horizons—aligned with geology and not just human civilization. They offer this perspective, for example: if the lifetime of Earth is represented by one yard—the distance from your nose to the end of your finger—then one stroke of a nail file on that finger would erase the whole of human history.3 Shocking, but not empowering.

A shorter outlook emphasizes our agency. But if it’s too short, it can lack a long-term, sustained vision. Then it can provoke a precipitate response and consequent backlash. We are all familiar with the short-termism that comes from 4-5-year parliamentary terms or quarterly financial reporting, each designed to maximize our agency through accountability.

The trillion-person-year epoch resolves the tension. It is a grand epoch on a civilizational scale—the size of one third of humanity’s cumulative existence on Earth to date. And yet it stays within a relatable time horizon, on a human scale. My granddaughter, born this year, may get to live through all of it.

The idea of the ‘Anthropocene’ as a geological epoch was rejected in part because of just this accelerated time horizon: ‘with a span of less than a single human lifetime, it sits uncomfortably within the [Geological Time Scale] where the units span thousands or even millions of years.’4 On a scale of time you can’t equate the Holocene and the Anthropocene; but on a scale of human presence, you can.

By measuring civilizational progress not in time, but in human presence, we see how pivotal this moment is. This accelerated-epoch perspective should both give us perspective and stimulate our sense of agency. We get to make choices now that will shape what a whole epoch becomes.

The great human innovations that introduced the three previous epochs each allowed us to command substantially more energy than before:

• Fire let humans use the energy from biomass that they couldn’t otherwise exploit. We were no longer limited to the 60-100 Watts from our own internal metabolism; we could make use of an external energy source, for keeping warm, cooking meat, or clearing ground.

• Farming let humans redirect current photosynthesis into our own food, capturing for ourselves a far higher share of the solar energy arriving per hectare.

• Fossil fuels let humans release the vast accumulated energy stored in hydrocarbons from past photosynthesis.

With each of these breakthroughs came dramatic changes (even if not immediate) in human prosperity, human population and the planet’s ability to carry it, and also humanity’s impact on the environment.

Those three innovations set a high bar. We may yet meet that challenge, but the rationale for calling a new epoch is different this time. (And it’s not just because it’s a trillion person-years since the last one.) For the first time, the trigger for a new epoch is a need to move on from the old (fossil fuels and other extraction), rather than an opportunity to embrace the new. The three previous epochs were defined by expansiveness; our next one is defined by constraint.

But constraint need not mean worse. In fact, when we had no external constraints, we chose to create them internally in the form of laws and norms, both religious and societal. As Frank Lloyd Wright observed, ‘The human race built most nobly when limitations were greatest and therefore when most was required of imagination in order to build at all.’5

An epoch defined by constraint is not destined to be reductive. It can be as expansive as our previous epochs.

The constraint we face is fundamental: the scale of our collective impact on our planet has become so great that we have to restore and maintain the stability of the Earth system. Eventually that means limiting what we extract from the system, because the Earth’s material resources are finite. More urgently it means limiting what we put into the system, to repair and avoid the damage we have been doing to our soils, oceans, freshwater systems and atmosphere.6

Can such a constraint support a fertile future for humanity? The answer is yes, with a mix of new thinking and new capabilities. Three ideas suggest how.

First is the idea of ‘net zero’. Familiarly applied to greenhouse gas emissions, but applicable more broadly, ‘net zero’ is a target state in which our activities have no net impact on the Earth system. (We take out whatever we put in, and vice versa.) This is an expansive idea, because once you achieve that state, there is no limit to your scale. Twice nothing is nothing. If we achieved 100% collection and recycling of aluminium cans, we could drink from them as often as we liked; we just need to send the same aluminium around the circle of production and recycling more often. That holds for the use of materials; it doesn’t hold for energy, which is consumed every time the material goes around the loop. That is why we need the second idea.

Second is the idea of energy abundance. It has long been observed that you can’t have infinite growth on a finite planet. But when it comes to energy, our planet is not finite; it is constantly replenished. Our concerns about conserving energy are driven by the damaging emissions from burning fossil fuels, not from the energy itself. Mark Lynas, science advisor with the 74-country Climate Vulnerable Forum and head of policy at WePlanet, observes that ‘the availability of clean energy with renewables and nuclear is effectively unlimited… There is no reason why people (or businesses) should not be able to use as much clean energy as they like if they are willing to pay for it.’7

Third is the prospect of a step-change in efficiency. Getting from where we are today to a system of net-zero extraction and pollution and abundant clean energy will involve substantial disruption and up-front cost. So far, it has proved politically hard, bordering on impossible, to persuade societies to accept that disruption and cost, even though the collective returns from doing so are well established. But what if a co-incidental innovation could both jump-start the process of disruption and compensate for the cost, by delivering dramatic efficiency gains? Because, among its many other impacts, AI may be set to do exactly that.

These ideas don’t mean that we will have an expansive future. We are a long way from net zero. The UK (and some other countries) has moved beyond ‘peak stuff’—the economy, despite its growth, consumes less material each year than it did 20 years ago,8 and carbon emissions have halved9—but even the UK is still extracting and emitting, just at a lower rate. These ideas do, however, suggest that we can have an expansive future, if we respond to our constraint as Frank Lloyd Wright would like and expect us to. This is the sort of positive thinking behind Abundance.10 And it is why our sense of agency is so critical.

The point of thinking about the next trillion-person years is not to try to predict the future, and guess what will come after fossil fuels. Rather, it is to engage with the future, in order to influence it. This is an important corrective: we have become over-focused on and over-confident in predicting, and neglectful of our capacity to influence.

It’s not a trivial shift. Our bias towards predicting rather than influencing reflects what we expect of the world. Our experience has taught us that the context will be stable (and so predictable). And, at least in developed countries, we have become increasingly focused on ourselves as individuals rather than society at large. As long as our concern is about how we each better ourselves and our own family, then what’s relevant is to predict the future as best we can, so that we can navigate it; changing the context isn’t a realistic option. Jonathan White, a political scientist, illustrates the trend through the change in political manifestos. Once, these were ‘dramatic statements of principle intended to convey an enticing shared future’, but ‘by the late twentieth century, manifestos looked rather different—more like a shopping list than a statement of principle… A manifesto was conceived as an offer that should have something for everyone, focused on their satiable wants rather than any more general vision of the future.’11

As we transition to a new epoch, the context is no longer stable. Predicting becomes futile and influencing becomes critical. And individualized influencing doesn’t work.

Engaging with the dynamism of a new, accelerated epoch is going to be extremely challenging. In most developed countries, we have become accustomed to a period of unusual stability. That stability has favoured optimization over renewal, openness over resilience, today over tomorrow. This has left us poorly prepared for a more dynamic world. It’s no wonder that we cling to the present, rather than engage with an apparently apocalyptic future.12

Just because we haven’t, doesn’t mean we can’t. But it does mean that we won’t, unless we start to think and act differently. And this is where I hope the combination of impact and agency that is brought out by the idea of the next trillion person-years can contribute. Not to try to dictate the right aspiration for us to have; but to stimulate us to aspire to a fertile future in the first place.

The next century will not be just one more in a line of some 4,000 centuries of human existence. It will be the fourth great human epoch. Until now, our species has seen three great epochs shaped by our use of energy, each accounting for about one trillion person-years on Earth: Fire (a small number of people, lasting 400,000 years); Farming (many more people, for 14,000 years); and Fossil Fuels (billions of people, for just 300 years).

The next trillion-person-year epoch will take just one human lifetime. The choices we make now will shape what that epoch becomes. That mix of impact and urgency is the inspiration for this series of essays.

We enter this new epoch at a time of existential risks and a growing sense that we are losing our agency to act on them. In most developed countries we have become conditioned by a period of stability that favoured optimization over renewal, openness over resilience, today over tomorrow. This has left us poorly prepared for the world that climate change, AI and changing geopolitics are now creating. It’s no wonder that we cling to the present, rather than engage with an apparently apocalyptic future.

And yet, collectively, we are richer, healthier, and more capable than ever before. Have we really become incapable of shaping a future better than today—of making our new epoch a Fertile Future?