Intelligence is not what we lack.

Didier Aubourg

Writer | Literary critic | Engineer | Ancient civilizations | Poetry | AI | World memory | murmurescosmiques.fr

Based on a silent assumption of our time

COP30 was held in Belém, in the Amazon, in November 2025. Like the twenty-nine previous COPs, it brought together heads of state, ministers, scientists, negotiators, NGO representatives, and lobbyists. Like the twenty-nine previous COPs, it ended with a compromise text, partial progress, cautious wording, and significant omissions, notably the lack of a genuine roadmap for phasing out fossil fuels. A few months later, the situation remained almost unchanged: global greenhouse gas emissions had once again reached a record high in 2025.

The scenario is familiar. The climate is spiraling out of control despite agreements. Inequalities are worsening despite diagnoses. Wars are returning to Europe despite seventy-five years of supposedly learned lessons. With each new warning, the same cycle: a report, a brief media outcry, a promise, then oblivion. And the conclusion is inescapable: we know, and we don't act. Or rather, we act too little, too late, never commensurate with our knowledge.

Faced with this dissonance, a narrative is gradually taking hold in the public debate. It is sometimes explicit, more often implicit. If we fail to solve the problems we ourselves have created, it is because our collective intelligence is overwhelmed by the complexity of the world. The systems we have generated—climatic, economic, geopolitical—now exceed the capacity of the human mind to grasp and govern them. To cross this threshold, we would need something more powerful than ourselves. Augmented intelligence. Artificial general intelligence. A cognitive tool capable of modeling what our brains can no longer comprehend, of optimizing what our institutions can no longer coordinate.

This narrative is now circulating in scientific forums, technology conferences, and foresight essays. It has its figures, its formulas, its statistics. It is appealing because it offers a way out.

We must take it seriously. This is not to deny what artificial intelligence can contribute. It can accelerate calculations, model scenarios, optimize networks, and identify correlations invisible to the human eye. These capabilities are real and will continue to expand. But none of them answers the crucial question: what do we do with what we know?

This narrative rests on a presupposition that must be examined: the idea that our collective failures are primarily failures of intelligence.

We were not caught off guard. The warnings did not arrive suddenly, in a single decade. They have accumulated in successive layers, over more than a century.

In 1896, in Stockholm, the Swedish chemist Svante Arrhenius published a lengthy article quantifying the potential effect of changes in atmospheric carbon dioxide on Earth's temperature. His calculations, done by hand, were approximate; his conclusions, cautious. He didn't yet know that nascent industry would release into the atmosphere, in less than two centuries, a considerable portion of the carbon that the planet had taken hundreds of millions of years to bury. But the physical principle was established at that time: altering the concentration of CO₂ in the air means altering the Earth's thermal equilibrium.

In 1972, four researchers from the Massachusetts Institute of Technology published, for the Club of Rome, *The Limits to Growth*, better known as the Meadows Report. In it, they modeled the interactions between population, resources, pollution, and industrial production on a global scale. Their conclusion, hotly contested at the time, did not concern a precise date of collapse, but rather the physical impossibility of indefinite material growth in a finite system. Fifty years later, several reassessments have shown that the observed trajectories remain close to scenarios of continued trends, far from the scenario of stabilization.

In 1988, the Intergovernmental Panel on Climate Change (IPCC) was established under the auspices of the UN and the World Meteorological Organization. Its first report, in 1990, already established that human activities were increasing greenhouse gas concentrations and that this increase would lead to further warming of the Earth's surface. Five assessment reports have followed since then. Each has confirmed, refined, or exacerbated the previous findings. None has been refuted in its overall structure; on several points, observations have even confirmed the most alarming scenarios.

Climate change is just one issue among many. Biodiversity is collapsing, and we know it: the Intergovernmental Science-Policy Platform on Biodiversity (IPB) has been documenting an extinction crisis unparalleled in recent human history for over a decade. Agricultural soils are degrading, and we know it: FAO reports are available and precise. The oceans are acidifying, warming, losing oxygen and a part of their life, and we know it. Inequality has found its cartographers, from Thomas Piketty to Branko Milanović and Anthony Atkinson. The mechanisms of war have found their analysts, from the work of Raymond Aron to that of the contemporary international relations community.

The knowledge available today is considerable, accessible, translated into numerous languages, included in textbooks, debated in the media, and integrated into political programs, at least nominally.

We are not an ignorant civilization. We are an informed civilization.

If we know and do not act, it is because action does not depend solely on knowledge.

We confuse three things: intelligence, which understands; power, which enables; and will, which decides. Knowledge reveals what happens. Power determines what is possible. Will decides what will be done. These three aspects are not reducible to one another. A civilization can know without power. It can be able without will. It can even will without truly knowing what it wants.

It is at this intersection that our collective failures take on meaning. Large fossil fuel companies do not continue to extract because they ignore climate change: some of their own teams had documented it as early as the 1970s and 1980s, as revealed by internal archives that have since resurfaced. States do not rearm because they ignore the ravages of war: strategic libraries are full of their knowledge. Societies don't destroy their soils because they are unaware of their fragility: agronomists have been warning about this since Albert Howard and his Agricultural Testament, published in 1940. They do so because interests, dependencies, fears, power dynamics, and short-term electoral horizons prevail over available knowledge.

The obstacles are structural. The timescale of climate effects is measured in decades and centuries; that of political mandates in years. The benefits of an energy transition are distributed over the long term and across all of humanity; its costs are immediate and concentrated on identifiable sectors, which wield considerable influence. The international cooperation required for action proportionate to the stakes presupposes mutual trust between states engaged in power rivalries. The political economy of fossil fuels relies on massive infrastructure, jobs, state revenues, and balanced budgets. None of these obstacles are cognitive in nature. They will not disappear simply through increased intelligence.

An intelligence superior to our own could illuminate them, model them, make them more visible. It would not, however, abolish them, because they are not primarily based on a lack of understanding. They are based on trade-offs, interests, affiliations, and loyalties. And these dimensions cannot be dissolved by calculation.

An analogy helps to grasp this. A person who knows that smoking shortens their life doesn't necessarily stop smoking. Another person who knows that their meat consumption contributes to climate change doesn't immediately become a vegetarian. Knowledge doesn't dictate action. Between the two lies the full complexity of humanity: its habits, its pleasures, its social affiliations, its economic constraints, its trade-offs between the present and the future. What holds true for the individual holds true, even more so, for societies. A society is this complex entity where millions of individuals, institutions, businesses, and states are constantly negotiating what they are willing to lose, what they want to protect, and what they are willing to transform. This negotiation isn't a matter of calculation. It's a political issue in the deepest sense of the term: the question of what we want to be together.

No intelligence, however infinite, will decide or will do things for us.

Why, despite everything, does the idea of ​​a saving intelligence continue to hold such appeal?

It's appealing because it shifts responsibility. If our problems stem from a cognitive limitation, they can be solved by increasing capacity. There's no longer any need to transform economic structures, question lifestyles, or redistribute power. We simply wait for the tool that will think better than us, faster than us, and further than us. Collective responsibility is transformed into an engineering problem. And the present moment, which demands painful choices, can be suspended while awaiting a future solution that will render these choices obsolete.

This represents a new form of magical thinking: no longer the magic of the gods, but the magic of systems; no longer the expectation of a miracle, but the expectation of optimization. The gesture is ancient, the packaging recent. Societies have often sought intermediaries to whom they could entrust what they couldn't solve themselves. Oracles, sacred kings, and the great machines of the industrial age have each in turn occupied this function of external third party, charged with welcoming what humanity did not want to face head-on.

This story is instructive. Every technological revolution has carried a promise of liberation that far exceeded its actual capabilities. The steam engine was supposed to free humanity from arduous labor: it instead created working-class conditions whose traces remain in 19th-century writings. Electricity was meant to equalize living conditions and bring every household into modernity: it did indeed transform daily life, but it also created new fault lines, between connected and forgotten territories, between powered centers and relegated peripheries. The atom promised almost unlimited energy and peace guaranteed by deterrence: it delivered a share of that energy, and it left humanity with the permanent shadow of total destruction. The internet was supposed to democratize knowledge, disintermediate power structures, and give everyone a voice: it has fulfilled some of these promises, but it has also generated unprecedented economic concentrations, large-scale cognitive manipulation, and a fragmentation of public spaces whose consequences we are only beginning to grasp.

None of these technologies has been in vain. Each has genuinely transformed the world, sometimes for the better. But none has abolished the structures of domination, inequality, or predation that pre-existed its emergence. These structures have adapted, reconfigured, and redeployed themselves on the new terrain. The economic power of the 19th century found its home in steam power; that of the 20th in electricity, oil, and nuclear power; that of the 21st in digital infrastructures and artificial intelligence models. Technology changes. Power dynamics, however, are reconstituted on the new platforms.

Each technological revolution has delivered on some of its promises. None of them kept the promise that was projected onto them: to deliver humanity from itself.

There is no reason to believe that artificial general intelligence will be an exception. It will bring new capabilities, some precious, others formidable. It will not exempt humanity from answering the questions it has been asking itself since its inception: What is a good life? What do we owe to future generations? How should we share the Earth? No algorithm decides these questions. And as long as these questions remain without a shared answer, the most precise knowledge will not translate into proportionate action.

No tool can save a humanity that has not decided what it wants to be.

We often speak of our era as "technological adolescence." The metaphor is apt. We have acquired immense power over matter, over living beings, over information, without yet having developed the political and moral frameworks that would allow us to use it wisely. We know how to modify the planetary climate, and we do not know how to agree not to do so. We know how to edit the human genome, yet we don't yet know what limits to place on this power. We know how to build intelligences that may soon surpass our own, yet we don't know who will control them, nor for whose benefit they will be deployed.

This asymmetry between the power we have acquired and the wisdom with which we use it gives full meaning to Hans Jonas's Principle of Responsibility: we have entered an age where might precedes responsibility.

But adolescence is not marked by an increase in power. It is marked by learning moderation. The Greeks had a word for this: sōphrosynē. Wisdom as temperance, discernment as awareness of one's limitations, self-control as the condition for all just action. It was not a minor virtue in their thought. Plato considered it one of the cardinal virtues; Aristotle saw it as an essential form of temperance. The Ancients understood that the greatness of a civilization was not measured by the volume of what it could achieve, but by the righteousness of its will.

The Kardashev scale classifies civilizations according to the energy they control: their planet, their star, their galaxy. It is a measure of power. One could imagine another scale, which would measure not the power captured, but the power a civilization is willing to relinquish in order to preserve what is worth preserving.

Perhaps a mature civilization is not primarily defined by the amount of energy it harnesses, but by the degree of power it is willing to relinquish.

The political question remains, one that intersects with the anthropological without being subsumed by it. Every technological promise today comes with a discreet, almost invisible, ethical clause: provided it is used properly, subject to regulation, and if humanity acts responsibly. This clause, in reality, encapsulates the entire debate. Who defines proper use? Who determines the regulation? Who decides what acting responsibly means? Who controls artificial intelligence models, who profits from them, and who bears the costs?

These questions are not preliminaries that can be dealt with later. They are the subject at hand. To evade them is to condemn ourselves to discovering too late that the tool from which we expected salvation was shaped by those who have a vested interest in maintaining the status quo. The history of previous technological revolutions bears sufficient witness to this, so we must not repeat the mistake.

The challenge of our time is not to produce an intelligence that surpasses us. It is to collectively rise to the level of the intelligence we already possess. To move from knowing to wanting. To transform what we understand into what we do. This is not an engineering problem. It is the very work of human societies since their inception: to decide together what they want to be.

And what if the maturity of a civilization consisted less in producing a superior intelligence than in finding the courage to act upon what it already knows?

Link to Didier Aubourg's article

This article is part of a series that Didier Aubourg is preparing for the magazine Diasporadz, at the invitation of the poet Brahim Saci.