Return to Lecture Notes →

Big Picture

The following notes are adapted from a lecture originally delivered at Nanjing University in July 2026, as part of a summer school with the theme ‘AI and the In/Computable’. The original lecture used the idea of the ‘incomputable’ to ask what can and cannot be reduced to calculation, moving from questions of AI, knowledge and human judgement towards a broader understanding of computation as something embedded in biological, technical and planetary systems. It also challenged the idea that AI is somehow external to us, instead locating it within longer histories of tools, archives, infrastructures and distributed forms of cognition.


Some of the language and framing of the in/computable remains in what follows. However, the material I want us to explore, beginning with the planet, moving through scales of the universe, and later turning to energy (and sustainability) speaks directly to the wider concerns. I.e. how we understand scale, levels of meaning, interdependence, technological systems, and humanity’s place within much larger (and also minuscule) material and planetary processes.

Let’s start by looking at a big picture, referred to as ‘Blue Marble’. You will likely have seen this many times. 

Blue Marble

In the same way the computer can be mistaken for a model of the mind, there is a way in which the Earth leads us to a mistaken model of the world. 

Specifically, I am referring to the iconic photograph of our planet Earth known as ‘Blue Marble’. The photograph was taken in 1972, the same year James Lovelock published his first paper on Gaia → the hypothesis that Earth’s living organisms, atmosphere, oceans and geological systems interact as a self-regulating planetary system. 

This photograph was taken by the crew of the Apollo 17 space mission; a picture of Earth never previously seen before, dominated by blue oceans. As Nicholas Mirzoeff remarks, other than the Apollo crew, ‘[n]o human has seen that perspective in person since the photograph was taken, yet most of us feel we know how the Earth looks because of Blue Marble’ (Mirzoeff, 2015: 3).

Mirzoeff was writing in 2015. In April 2026, however, NASA’s Artemis II mission carried humans beyond Earth orbit and around the Moon for the first time since Apollo 17 in 1972, once again giving astronauts a direct view of the Earth as a whole from deep space. The crew returned striking new images of the planet, including views of Earth setting beyond the lunar horizon.

Nonetheless, fifty years on from the Apollo mission, it is clear the world has changed significantly, and more to the point, from a Gaian perspective, the planet itself has changed: ‘In 2013, carbon dioxide passed the signature threshold of four hundred parts per million in the atmosphere for the first time since the Pliocene era about 3 to 5 million years ago’ (5). The world today, then, is physically different to one we see in Blue Marble. Actually, since Mirzoeff’s remarks, now 55 years on, a few more humans have seen this view again. The Artemis II mission returned human beings to lunar space in 2026; the crew photographed Earth while flying around the Moon’s far side (which no one else has ever seen!). Nonetheless, Blue Marble remains a cultural icon of both planetary unity and fragility. It encouraged a fantasy of a ‘global view’: the idea that Earth can be seen, grasped, managed or governed as a single whole. This represents a whole new ‘unit’ of meaning or computation. 

In this way, Blue Marble represents a technical, cultural and epistemological event. It is the visual analogue of what Bernard Stiegler criticises as the hegemony of calculation: not only the use of technical systems to know the world, but the fantasy that the world can be totalised as a knowable object. Bruno Latour, for example, who takes up a particular interest in Lovelock’s notion of Gaia, argues the image misleads us. Gaia (as a complex environmental system) is not a sphere held in the hand. It is a surface, a critical zone, a thin and volatile field of interactions, dependencies, feedback loops, agencies and overlapping elements. It cannot be reduced to the fantasy of the globe: 

…when you pretend to consider Gaia as a whole you immediately summon the image of the ‘blue planet’ viewed from outer space—in addition to an ample circular gesture of the two hands. But no matter how powerful the influence of such an iconic photograph, no matter how often you agitate your hands, the impulse to globalize should be resisted because nobody who claims to have ‘a global view’ actually resides in any real space. They imagine themselves as if they were looking from the outside at the Earth taken as one body among all celestial bodies, just as Galileo did. The global view is strictly speaking a view from nowhere—or from an office looking at a computer screen. (Latour and Lenton (2019: 675)

In 2012, NASA produced a new version of Blue Marble, but this is only a composite image assembled from a series of satellite image. From the satellite’s orbit, approximately 580 miles above the surface, the full view of the planet is not visible (it is necessary to go beyond 7,000 miles to see the entire globe). The image is made to seem as if it were taken from one vantage point, but it was not. It is accurate in each detail, but it is false in that it gives the illusion of having been taken from a specific place at one moment in time. Such ‘tiled rendering’ is a standard means of constructing digital imagery. It is a good metaphor for how the world is visualized today. We assemble a world from pieces, assuming that what we see is both coherent and equivalent to reality. (Mirzoeff, 2015: 7-8). While Mirzoeff suggests the composite view is ‘false’ (i.e. not equivalent to reality), he might agree with Latour that the ‘composite’ view offers a means to overcome the illusions of control of a global system, albeit leads to a data-centric worldview. 

Nonetheless, when we turn away from what the photographer saw (in the 1972 Blue Marble) and instead build up an ‘analytics’ from the data gathered by satellites, we arguably turn from a political (a mythological) view to a scientific one. Climatology is undoubtedly a complex, interdisciplinary science, which uses advanced modelling techniques to understand and predict climate patterns. This involves nonlinear dynamics (e.g. small changes in one part of the system can lead to disproportionate and unpredictable effects elsewhere); chaotic behaviour (e.g. the ‘butterfly effect’, whereby small uncertainties in initial conditions can grow exponentially over time); coupled and multiscale models (to integrate various subsystems and to operate across different scales, from local to global, short-term to long-term); complex adaptive systems (e.g. autotrophy); all of which is data intensive and computationally complex. 

This matters for what we take to be computable and incomputable because it complicates the very thing we are trying to protect. If the incomputable is too quickly identified with human freedom, human judgment or human creativity, then it remains anthropocentric. But Gaia forces us to widen the frame. Law-making, agency and world-forming do not belong only to humans. Life forms make worlds. They alter atmospheres, produce climates, transform environments, generate conditions of possibility for other forms of life. In Latour and Lenton’s terms, Gaia opens the possibility of extending the domain of freedom.

At the same time, this extension remains largely bound to organic life. The ‘politics of life agents’ is still a politics of life as we ordinarily understand it: biological, ecological, planetary. The question I want to carry forward is whether this is still too narrow. What if the incomputable is not best understood as a protected human remainder? What if it is not simply the name for freedom against calculation, life against machine, or Gaia against data? What if the deeper challenge is that the very distinction between calculation and life has been posed at the wrong scale?

This is where the trope of the incomputable begins to reach its limit. It has done important work. It has allowed us to say that truth is not reducible to prediction, that knowledge is not reducible to data, that judgment is not reducible to optimisation, that the future is not reducible to extrapolated becoming, and that Earth is not reducible to a globe seen from nowhere. But if biological life is already technical, if knowledge is already exosomatic, if Gaia is already a distributed system of agencies, and if the world we inhabit is increasingly assembled through sensors, satellites, models, networks and simulations, then perhaps the question is no longer simply what escapes computation.

Perhaps the more difficult question is what it means to be in computation.

In-computable

I’ll begin with a bone. Specifically, the carvings by human hands upon a piece of bone that dates back 20,000 years, from the Upper Palaeolithic era. (I take my reading of this bone from Snezana Lawrence’s book, A Little History of Mathematics – but the archaeological account is widely known). 

Referred to today as the Ishango bone, it is believed to be the earliest known mathematical object. Discovered in the 1930s by Belgian archaeologist Jean de Heinzelin de Braucourt in Ishango, in what is now the Democratic Republic of Congo, the small, sharpened piece of baboon fibula – about the size of a pencil – was found among other ancient remains: tools, spears, ropes, traces of prehistoric life. What was curious and enduring about this particular object was the etching upon the bone; a sequence of marks: 168 notches, grouped into three distinct columns.

To the casual observer, these might look decorative, or even accidental. But studies show the arrangement is too deliberate, too organised to be random scribbling. Some have argued the etchings reflect sophisticated numeric patterns: the doubling of numbers; a column of prime numbers between 10 and 20; and groupings that sum around the number 20. In short, the Ishango bone appears to be a counting or recording device; a kind of prehistoric calculator. It is worth noting, prime numbers are those you can only divide by one and themselves. They were first described around 500 BCE, so many centuries after the Ishango bone was discovered. 

Whatever the bone was actually for, the general view is that it is the earliest known mathematical object in human history. But what does it mean to call something a mathematical object? If nothing else, I want us to consider this object as evidence of being in the world mathematically – both being part of a overall calculation and also enabling the means to be conscious of such calculations. 

In one sense, we might say a mathematical object refers to anything that has been shaped or created with a mathematical purpose, such as an abacus, a calculator, even a spreadsheet. But more broadly, and more abstractly, mathematicians speak of intangible objects: a point A, a line a, a number n. These are conceptual entities, forms that exist within systems of logic and thought, not within the material world. A mathematical object, then, can be entirely immaterial; it need only be something we can act upon, reason with, or define through operation. Inevitably, it is our ability to discover and make mathematical objects and structures that has enabled us to generate contemporary AI. But, also, more fundamentally, mathematics is a deep expression of our imaginative power. And let’s keep in mind, it is not simply that mathematics allows us to occupy a virtual, abstract realm, but that that imaginative space comes to act directly upon material, lived environment. It is because of maths that we can do most of what we do in our daily lives – whether setting a morning alarm, boiling a kettle (governed by thermostatic circuits), navigating a route, or splitting a bill at a restaurant. As much as these are simple acts requiring mathematics, they also help define our sociality: i.e. regularities, connectivity, and the divisions of resource. 

Let’s hold a little longer on the Ishango bone. Though it is physical, it encodes abstraction. It bridges material and mental domains. Its marks are real, but they refer to ideas – patterns, regularities, structures – that exist beyond it. And this is what makes it significant: not just as an archaeological artefact, but as a symbolic threshold. The twentieth-century anthropologist, Claude Lévi-Strauss, in his essay ‘The Mathematics of Man’, argued that mathematics is not a late addition to human culture but a fundamental feature of human cognition. Long before formal proofs or theorems, our ancestors sought to make sense of the world through pattern, repetition, and symbolic order. The Ishango bone is our longest surviving trace of this drive—a material expression of an immaterial logic.

But Lévi-Strauss goes further. He challenges the assumption that mathematics enters culture as a tool for quantification or measurement. Rather, he sees mathematics as a way of modelling relations: a ‘human mathematics’; concerned with form, structure, and transformation, not with size or scale. Just as early linguistics identified underlying systems of difference in phonemes, Lévi-Strauss proposes that human societies intuitively work with abstract relations: kinship systems, myths, rituals, each governed by rules of combination and transformation akin to algebra or group theory. Read this way, the Ishango bone does not simply count; it relates. It is suggestive of how numbers were not just marked, but that thought was structured, i.e. that early humans were already engaging with abstractions, discontinuity, iteration, and systematicity.

For Lévi-Strauss, this is a vital correction to the idea that mathematics enters anthropology only through the blunt force of statistics or measurement. ‘The reign of necessity,’ he writes, ‘is not necessarily coterminous with that of quantity’ (p.93). What matters is not the ability to count things precisely, but to grasp the logic of their connections. In the small numbers, in the jump from two to three, in the shift from duality to triangulation, he sees the seeds of symbolic thought. The Ishango bone, then, becomes not just a relic, but an early witness to the mathematical imagination. Not merely calculation, but cognition.

In my own work, I have been looking back at the work of anthropologist Claude Lévi-Strauss; well-known for his structuralist reading of culture. In his early work, for example, he studied kinship systems and marriage rules, which appear to differ all over the world, and yet equally suggest of some constants (the fact we marry in most parts of the world is one such constant, or invariance). He was interested in finding order from the apparent disorder; to understand what was invariant despite all of the variance. He then applied a similar kind of thinking to the study of myths – which again, appear all over the world. All cultures tell stories and hold myths about the long distant past, including origin stories (where did we come from, where did all of the universe come from?). I don’t have time to expand further, but essentially, I have been looking back at his work on mythological thinking to consider possible advances for the architectures of contemporary Large Language Models (LLMs). I’ve been interested in how structures of thought allow us to make new thoughts, or new meaningful constructions. 

This word ‘meaning’ is of note. In a public broadcast lecture series from 1977, Lévi-Strauss points out that the word ‘meaning’ is probably one of the most difficult to define. He says:

What does ‘to mean’ mean? It seems to me that the only answer we can give is that ‘to mean’ means the ability of any kind of data to be translated in a different language. I do not mean a different language like French or German, but different words on a different level. After all, this translation is what a dictionary is expected to give you – the meaning of the word in different words, which on a slightly different level are isomorphic to the word or expression you are trying to understand. Now, what would a translation be without rules? It would be absolutely impossible to understand. […] To speak of rules and to speak of meaning is to speak of the same thing; and if we look at all the intellectual undertakings of mankind, as far as they have been recorded all over the world, the common denominator is always to introduce some kind of order. If this represents a basic need for order in the human mind and since, after all, the human mind is only part of the universe, the need probably exists because there is some order in the universe and the universe is not a chaos.

You can begin to hear all sorts of philosophical quandaries, regarding order/disorder, meaning and rules; and the idea of ‘translation’ is noteworthy. He does not mean language translation per se, but rather the ability of forms to transform → for change to be accounted for within a system of rules or properties. E.g. the four lines of a square could be taken apart and each arched and placed together to form a circle. One shape transformed into another, but crucially the new shape could only be a specific size, based on the originally lengths of the sides of the square. This is a rule or property that has to adhere, even though we have managed to shift from square to a circle. 

When we start to think of reality as a system of transformations, we might begin to understand a decentring of the human, or put another way, transformations are the manner in which we make ‘worlds’ – there is not a single world, but an ongoing set of transformations or ‘world-making’. We can glimpse this in the opening pages of Paul Klee’s well-known Pedagogical Sketchbook (c.1925), which perhaps we can describe as a live witnessing of the mathematical imagination: 

In the multiple explorations of form and composition by German-Swiss artist Paul Klee, a special place is occupied by works where the line is the main protagonist. In a famous quote, the artist describes the line as ‘a dot that went for a walk’ underlining the dynamic nature of the line as the means to conduct the human eye across the space of a painting or a drawing. (Link →)

Klee does not begin with the finished image, but with the most elemental mark: a point. A point is set in motion, he suggests, and so becomes a line. From that line, a plane. From plane, form. Klee’s drawing is not a picture, it is a system, a generative grammar of relations. In both cases, the, whether Klee or the Ishango bone, we see a logic of emergence. A simple gesture becomes an entire conceptual world.

We might say that mathematics is both ancient and ever-new. It begins with the hand – scratching, marking, measuring – but it leads us into the realm of the abstract: numbers, vectors, algorithms. In turn these abstracts allow us to be generative, imaginative, productive. And this is the world into which artificial intelligence now steps.

AI, like the Ishango bone or Klee’s line, is not just a tool for doing mathematics, it is itself shaped by mathematical objects. Neural networks, decision trees, tensor fields: these are not things we can pick up or hold. But they are structures we manipulate, systems we train, environments we inhabit. Even as we look at our present day circumstances and to what we believe will unfold in the future, we can keep in mind this lineage that runs through the earliest human inscriptions right up to today’s most advanced generative models.

In other words, in thinking about today’s technologies, let us not be too quick to say we are entering an entirely new world. Rather, we are extending one that began long ago, with a bone, with a point, with patterns carved into time. → we are just another element in a continually shifting field of transformations. We are in a universal computation! 

Seen on the scale of millennia, human passions fuse. Time neither adds to nor removes anything from the loves and hates felt by people, their involvements, their struggles and their hopes. They remain the same today as they were in the past. Randomly removing ten or twenty centuries of history would not affect, in a meaningful way, our knowledge of human nature. The only irreplaceable loss would be the works of art which these centuries gave rise to: because humans only differ through their works, and even exist only through them. Just as a wooden statue attests to the prior existence of a tree, so works of art provide the proof that, throughout history, amongst people, something actually happened. (Claude Lévi-Strauss, Regarder, ecouter, lire)

Scales, Units of Meaning

Lévi-Strauss is referring to time, on a scale of millennia (we might say an incomputable time from a human perspective). At this scale we see different patterns. And it is not just time, but also space that we can consider a whole range of scales → measures of computation, of ‘order’. Scales ask critical questions of us, e.g. why do I refer to myself as a single entity, a ‘person’? Looked at another way, I am a consortium of cells, bacteria, molecules; or another way still, I am one in a crowd, community, a species etc.

From a systems point of view, and which I believe philosophically impacts upon how we need to consider AI, relates to a question of scales. Put another way, we need to consider ‘units of meaning’ – at what level or perspective do you define the unit of meaning. This in turn impacts on how we make decisions for how we relate to the world around and how we seek to ‘design’ it.  

Bricolage – New Units of Meaning 

In anthropology, bricolage refers to a process of improvisation using available materials and ideas. Claude Lévi-Strauss introduced the concept in The Savage Mind (1962) (now translated as Wild Thought), contrasting the bricoleur, who creatively recombines existing elements, with the engineer, who works from theoretical plans. While he linked the bricoleur to mythical thought, Lévi-Strauss emphasised that this way of thinking is not inferior; in fact, scientific thought often involves bricolage.

We might consider the designer as inherently a bricoleur: someone who ‘makes “new from old” by playing with the formal harmonies anddisharmonies suggested by the sensory effects of the signs collected. Bricolagetherefore presupposes that we must pay attention to the sensory world, a worldalready given by history and culture’ (Floch, p.5).

Two surrealist examples can be illuminate this process. The surrealist painter Max Ernst had a notable influence on Levi-Strauss, writing:

“I learned from the surrealists not to fear abrupt and unexpected associations — such as those Max Ernst indulged in in his collages. […He] constructed personal myths out of images borrowed from other cultures and taken from old nineteenth-century books, and he made these images convey more than they could mean when seen by innocent eyes. In the Mythologiques, I also cut out mythic matter and recomposed the fragments in order to generate additional meaning of the same kind.”

If we think back to Latour’s remarks on the Blue Marble photograph, that we can never get to see the ‘whole’; rather as we see Ernst’s paintings we inhabit ‘situations’; fragments; composites.

Another example is Philippe Starck’s iconic ‘Juicy Salif’, a lemon squeezer for Alessi.

“Sometimes, some designers can even seem to be wilfully obscure about how they work, and where their ideas come from. The renowned (perhaps even notorious French designer Philippe Starck is known to suggest that design ideas seem to come to him quite magically, as it from nowhere. He has said that he has designed a chair while sitting in an aircraft during take-off, in the few minutes while the ‘fasten seat belts’ sign was still on. Perhaps the instruction to fasten seat belts’ was an inspirational challenge to his designing. Of the design process of his iconic lemon squeezer for the Italian kitchenware manufacturer Alessi, he has said that, in a restaurant, ‘this vision of a squid-like lemon squeezer came upon me ..? And so, Juicy Salif, the lemon squeezer, was conceived, went into production and on to become a phenomenally successful product in terms of sales (if not necessarily in terms of its apparent function).” (Nigel Cross, p.6)

The design process is not necessarily straightforward to understand, but at stake is a sense that there is always some form of ‘order’ or process, albeit a complex one. Design is a fashioning of orders of meaning. As Floch explains:

“Again and again, we will see that the product of the bricoleur’s work can be considered as a structure, as an object of meaning with its own closure and its own system – and this is due to the semi-symbolic coupling of certain of its sensory qualities with certain of its content categories. By organizing and reorganizing the materials and the images provided by the signs he or she collects, the bricoleur produces meaning by super-segmentation and by establishing paradigms found in a semi-symbolic semiosis. This means that the bricoleur makes ‘new from old by playing with the formal harmonies and disharmonies suggested by the sensory effects of the signs collected. Bricolage therefore presupposes that we must pay attention to the sensory world, a world already given by history and culture.” (Floch, p.5)

At what level do we determine meaning, units of meaning; the ‘holding forms’ of things. A useful reference point is the 1977 film Powers of Ten, made by Charles and Ray Eames. The film begins with an ordinary human scene, then moves outwards by powers of ten, from a picnic blanket to the city, the planet, the solar system, the galaxy and the observable universe, before reversing direction and travelling inward, through the body, the cell, the molecule, the atom and the subatomic world. Its lesson is simple but profound: the human is not the measure of all things, but one scale among many.

Our contemporary view of the universe extends this lesson still further. Through developments in quantum physics, astrophysics, molecular biology, materials science, particle physics, satellite imaging and computational modelling, we inhabit a universe that can be pictured, measured and inferred across an extraordinary range of scales. We move from galaxy clusters, nebulae and the Milky Way, to the solar system, Earth, landscapes, animals and human bodies; then downwards again into cells, bacteria, skin, DNA, atoms, nuclei, protons, neutrons, quarks and finally toward the Planck length, the smallest meaningful scale in current physical theory.

This sequence is humbling. The human figure appears briefly, somewhere in the middle. We are neither the largest nor the smallest thing. We are not the centre of the universe, nor even the centre of scale. We are a temporary and situated form of organisation within a much larger field of relations.

And yet, we are also the beings able to work this out. We cannot see most of these scales directly. No one sees a galaxy cluster as a whole with the naked eye. No one sees DNA, or an atom, or a quark, or the Planck scale. These are known through instruments, models, traces, mathematics and inference. We reason inductively from observation, pattern and experiment. We reason deductively from theory, geometry and calculation. We build devices that extend perception, then use concepts to interpret what those devices disclose.

DNA is especially important in this sequence. It sits between the biological and the informational. It is not simply a molecule among molecules. It is a structure that stores, transmits and transforms instructions for life. In this sense, DNA already complicates any simple opposition between life and computation. Life is not digital computation in the narrow sense, but biological life is deeply bound up with coding, sequence, replication, error, variation and transmission. Long before computers, life was already organised through processes that invite informational description.

The point, then, is not that everything is a computer in the ordinary sense. Rather, the point is that reality appears to us through relations of scale, pattern, information and transformation. The universe becomes thinkable because it can be measured, modelled, compared, inferred and translated between scales. We never stand outside this process. We are in it. This is the shift marked by the hyphen in the in-computable. We are not simply asking what computation cannot reach. We are asking what it means to live within a universe that becomes intelligible through computation, modelling, measurement and information. The human is one scale among many, but it is also a scale at which the universe begins to ask questions about itself.

At the smallest end of this sequence, we arrive at the Planck scale. This is not simply another smaller object, like a cell, molecule or atom. It marks something closer to a conceptual limit: the scale at which our familiar picture of continuous space and time begins to break down. If the classical imagination pictures the universe as a continuum, infinitely divisible all the way down, the Planck scale suggests a different possibility: that reality may have a lowest meaningful resolution, a smallest scale at which distance, time, energy and information can be thought.

This matters for the argument because it gives us a different analogy for meaning. In semiotics, everyday culture is organised at the level of the sign. We speak, write, interpret, desire and remember through signs. Culture does not float freely; it is built from units of difference that can be combined, repeated and transformed. By analogy, if physical reality is not finally continuous but structured at the Planck scale, then the universe itself may be understood as composed from fundamental units of difference. Not signs in the cultural sense, but something like the minimal conditions from which physical meaning, relation and structure become possible.

On this view, reality is not an infinite smoothness but a finite, though unimaginably vast, structure. The universe is not made meaningful only when humans interpret it. Rather, human meaning appears much later, at one particular scale, within a universe already articulated through difference, relation, quantity and form. This is the deepest sense of the in-computable: we are not outside computation, looking in. We are one emergent layer within a finite but immense field of structured differences, from Planck-scale limits to cosmic-scale forms.

Energy

I will turn now to the topic of energy, which we might argue is at the heart of everything we do, indeed at the very heart of the universe.

What I will present is largely taken from Barnabas Calder’s book Architecture: From Prehistory to Climate Emergency, which I thoroughly recommend as a general read. It offers an excellent history of architecture through the lens of energy: asking how the amount and kinds of energy available to human societies have determined what they could build, how they could build it, and ultimately how they could live.

Calder’s book opens with the following thought-provoking statement:

This book tells the story of how fossil fuels made the world a much better place for humans.

It is a disturbing story because an overwhelming scientific consensus makes it clear that unless we can get away from our all-embracing dependence on fossil fuels by 2050, we will make the planet apocalyptically horrible for humans and many other creatures. The construction and running of buildings are currently responsible for 39 per cent of all human greenhouse gas emissions — the equivalent of some 9 billion tonnes of CO₂. This figure needs to drop to net zero by 2050. It is the toughest challenge the world of architecture has ever faced.

There is an important tension here. Calder does not begin by simply condemning fossil fuels. He begins by acknowledging what they have made possible. Fossil fuels have enormously increased human capacities: transportation, heating, food production, manufacturing, medicine, sanitation, construction and standards of living.

The problem is that precisely the energy system which has enabled modern life is also driving climate change. So, the challenge is not simply to stop using fossil fuels. Rather we need to ask: How do we retain the extraordinary capabilities of an energy-rich civilisation while transforming the energetic basis on which that civilisation depends? And for architecture this question is particularly acute because buildings are not merely passive objects within an energy system. They are some of its largest material manifestations, and which contain energy systems and people within them.

Calder makes the extraordinary scale of contemporary energy consumption tangible by comparing it with one of the greatest construction projects in human history: the Great Pyramid of Khufu. Around 4,500 years ago, Khufu commanded the construction of a building weighing more than 5 million tonnes and rising approximately 147 metres. It is estimated its construction equates to around 78 million days of human labour, distributed across tens of thousands of people over more than a decade. This seems almost incomprehensibly large. Yet compare it with the energy available to an ordinary person living within a contemporary industrial economy.

Calder’s calculation is that: the lifetime energy consumption of just seven average Americans exceeds the energy required to build Khufu’s pyramid. That is a revealing inversion. Khufu was possibly the most powerful individual in the world. Yet in energetic terms, an ordinary modern household commands resources comparable with those available to ancient rulers. If the lifetime energy consumption of the entire US population were translated into pyramid-building energy, it would amount to nearly 50 million Great Pyramids.

The point is not about pyramids. It is about recognising how historically abnormal our present situation is. We live surrounded by quantities of energy that previous societies could barely imagine.

Energy is the capacity to do anything

Calder gives us a neat simple definition: ‘Energy is the capacity to do anything.’ Nothing moves without energy. Nothing is heated without energy. Nothing grows without energy. Nothing is manufactured, transported, built or destroyed without energy. This means that the scale of human activity has always been constrained by the amount of energy humans can access. But the source of that energy has changed radically.

For most of human history energy came largely from:

food → human muscle

food → animal muscle

wood and biomass → heat

wind → sails and mills

water → waterwheels

These were all ultimately constrained by ecological processes occurring across relatively limited areas of land. Farming is therefore not simply a system for producing food. It is an energy system. Land converts solar energy into plants; plants provide calories; calories power humans and animals; wood supplies heat. This creates competition between different uses of land. Land used to grow fuel cannot simultaneously grow food. Animals used for mechanical work must themselves be fed. Energy before fossil fuels was therefore scarce, geographically dispersed and comparatively expensive.

Common units 

One of the intellectual breakthroughs of nineteenth-century physics was recognising that apparently different phenomena (heat, motion, electricity, chemical reactions) could all be understood as transformations of energy. They can therefore be measured in common units, which we know as joules, calories or kilowatt-hours. This allows a particularly productive conceptual move. In principle, we can describe an entire building energetically.

We could calculate:

the calories consumed by workers;

the energy required to transport stone;

the heat needed to fire bricks;

the energy used to make glass;

the heat required to smelt metal;

the energy used by cranes and excavators;

the electricity required to heat, cool and illuminate the completed building.

Architecture can therefore be understood not simply as form and material, but as material plus energy. A building is, in one sense, a vast accumulation of past energy expenditure. This connects to what we now commonly call embodied energy: the energy consumed through extraction, manufacture, transportation and construction before a building is even occupied.

From muscle to fossil force

The difference between pre-industrial and industrial civilisation becomes stark when we compare human muscle with machines.

Calder estimates that a person performing sustained physical labour produces around: 0.075 kW. An elite athlete can briefly produce considerably more, but cannot sustain it. By contrast, a car engine can generate thousands of times the sustained power output of a human being. A bulldozer can effectively replace the physical effort of thousands of labourers.

So, when we see a single person operating an excavator on a construction site, we should not simply see one worker. Energetically, we are seeing one worker commanding a vast quantity of fossilised labour. This is one of Calder’s most useful ways of reframing fossil fuels. He suggests imagining them as energy servants. A tonne of oil contains approximately 11,630 kWh of energy, the energetic equivalent, in of more than 150,000 hours of human labour, or over 19,000 eight-hour working days Industrial civilisation has effectively acquired billions of invisible workers.

This distinction is particularly important for architecture. Consider steel. Modern steel production requires enormous temperatures and therefore enormous quantities of energy; around 5,000 kWh to produce a tonne of steel using efficient modern processes. To obtain an equivalent quantity of energy from pre-industrial wood supplies would require the output of thousands of square metres of sustainably managed woodland — and historical furnaces would have been far less efficient.

So, there is a reason Romans constructed extraordinary buildings from stone, concrete and brick but did not construct steel-framed skyscrapers. It was not simply that nobody had thought of them. Their energy economy made them implausible.

This gives us an important principle: Every architecture is also an architecture of its available energy system. Or, every design is also a design of its available energy system. Different societies push architectural limits in different ways because different things are energetically expensive or cheap. Romans could command huge amounts of labour and shipping. They could therefore transport extraordinary quantities of stone across their empire Industrial Britain could burn huge quantities of coal. It could therefore manufacture unprecedented quantities of brick, iron, steel and glass.

Twentieth-century societies acquired cheap oil, abundant electricity and powerful machinery. Architecture could consequently become taller, lighter, mechanically ventilated, air-conditioned and globally supplied. Materials could travel thousands of kilometres. Huge panes of glass could be produced. Concrete could be mixed and pumped. Steel could be manufactured in immense quantities. Cranes could lift loads no workforce of humans could practically lift. And once completed, buildings themselves became machines consuming continuous flows of energy through: lifts; lighting; heating; cooling; ventilation; automatic doors; digital infrastructure; servers and networks. Cheap energy becomes so deeply embedded in everyday life that its extraordinary character disappears. What would have seemed miraculous to Khufu appears mundane to us.

Calder offers a pity observation: ‘Energy is like gravity — you do not need to theorize it to be bound by its rules.’ Historical architects did not describe their buildings in terms of thermodynamics or kilowatt-hours. But, nonetheless, energy constrained their decisions. I.e. Energy is not simply another consideration that designers can choose to add or not. Energy  establishes the material conditions of possibility within which architecture operates.

Architecture does not simply record technological capability. Buildings also record what societies decide is worth spending energy on. The pyramid represents an extraordinary concentration of labour and resources around a particular religious and political cosmology. Palaces demonstrate political hierarchy. Temples, churches and mosques channel immense resources towards religious belief. Factories embody industrial production. Railway stations express the energy and mobility of the coal age. Glass-and-steel towers belong to an economy capable of producing tremendous quantities of steel, glass, electricity and artificial cooling. Architecture therefore makes social priorities material. Energy tells us something about what a society can do. Architecture tells us something about what it chooses to do with that capacity.

Three regimes

Calder reduces the very long history of human societies to three broad energy systems:

1. Foraging: Humans depend primarily upon naturally occurring flows of biological energy. Architectural remains are relatively limited.

2. Farming: Humans reorganise landscapes to capture solar energy systematically through crops, animals and woodland. Settlements become permanent and architecture proliferates.

3. Fossil fuels: Coal, oil and natural gas unlock immense stores of ancient solar energy accumulated over geological time.

The crucial point is that fossil fuels are fundamentally different from annual biological energy flows. We are consuming stores of energy accumulated over millions of years. This produces an extraordinary historical acceleration. Within only a few centuries, fossil energy has transformed transport, industry, agriculture, cities and architecture.

The skyline of virtually every contemporary metropolis is evidence of that energetic transformation. This brings us back to the opening paradox. The modern world is not simply accidentally dependent upon fossil fuels. Its material structures were built through fossil fuels and built around the expectation of abundant energy.

Our cities assume transport.

Our buildings assume manufactured materials.

Our glass towers often assume mechanical cooling.

Our supply chains assume global freight.

Our construction processes assume diesel-powered machinery.

Even apparently minor components (carpets, insulation, plasterboard, ceiling tiles) contain complex histories of extraction, manufacture and transportation. And more recently we are now debating the value of constructing large data centres.

Decarbonisation is not simply a matter of replacing a coal-fired power station with a wind turbine. It means reconsidering the whole energetic metabolism of the built environment. Yet things are not necessarily pessimistic. Human beings have repeatedly reorganised their societies and architectures in response to changing energetic circumstances. Architecture is evidence not only of dependence but also of extraordinary human adaptability. And that provides the more hopeful proposition: If architecture has always changed with its energy system, then changing the energy system will necessarily mean changing architecture again.

Hence, the question for designers is therefore not merely how we can use less energy. The larger question: What kinds of buildings, infrastructures and ways of living, i.e. what kinds of systems, become possible (and desirable) in a post-fossil-energy society?

References

Barnabas Calder, Architecture: From Prehistory to Climate Emergency (Pelican, 2021)

Nigel Cross, Design Thinking: Understanding How Designers Think and Work (Berg, 2011)

Jean-Marie Floch, Visual Identities (Continuum, 2000)

Bruno Latour and Timothy M. Lenton, ‘Extending the Domain of Freedom, or Why Gaia is so Hard to Understand’, Critical Inquiry, No. 45, 2019, pp.659-680.

Snezana Lawrence, A Little History of Mathematics (Yale University Press, 2025)

Claude Lévi-Strauss, ‘The Mathematics of Man’ [1956], Social Analysis, 67 (2), 2023, pp.80-98. 

Claude Lévi-Strauss, Myth and Meaning (Routledge, 2001)

Claude Lévi-Strauss, Look, Listen, Read (BasicBooks, 1997)

Nicholas Mirzoeff, How to See the World (Penguin, 2015)