The Age of Hypertrophy

We have mistaken our ability to produce more, for evidence that more is what we need. Agriculture solved calorie scarcity and then, with some considerable ingenuity, produced supermarket shelves that contain enough food to feed a small medieval kingdom—at least in some parts of the world. For a great many of us, industry solved manufacturing scarcity and we are now literally drowning in oceans of stuff.

The so-called Information Age solved knowledge scarcity with the internet and social media as the perfect expression of what followed – that information is no longer difficult to find. On the contrary, it’s impossible to escape, through the ubiquitous screens — first at our office desk, then at home, now on our bodies — fighting for our undivided attention.

Artificial Intelligence may be the first technology to reveal what happens when information itself becomes hypertrophic — when, like a muscle responding to exercise, growth makes us stronger but only up to a point, before more growth becomes counterproductive.

We are entering what we might call the Age of Hypertrophy: an age in which abundance keeps growing without ever asking whether scarcity was the problem to begin with.

More information, more things to consume — more everything — until abundance itself turns into its own undoing. Put simply, there is only so much chocolate we can eat before we feel sick.

We have built systems whose capacity to produce has grown beyond our capacity to meaningfully attend to or make use of what they produce. And as machines take on more of the cognitive work, we may find ourselves with fewer opportunities — and motivation — to exercise the human capacities that work once demanded.

The great promise of AI rests on a simple proposition – that intelligence can now be produced at scale. We can generate essays, computer code, music, and equally convincing nonsense at a speed that leaves us awestruck. Struck enough to believe that abundant intelligence is what we need.

For centuries, however, education was built around the opposite condition – that knowledge was scarce. Books and expertise were scarce, communication was slow, access to information was expensive. Schools, universities and libraries emerged partly in response – gathering knowledge, curating it, preserving it, passing it on. But we never stopped to ask, what exactly was scarce, and who had made it so?

Karl Polanyi described the Great Transformation as the process by which economic life was dislocated from the social relations in which it had been deeply embedded. Land and labor, once part of social life, were reorganized as “fictitious commodities”. Something similar can be seen in what happened to learning. A farmer learned farming by farming, and work, learning and rest were part of the same life. Then labor became a job, learning became schooling, and leisure a luxury, while the ordinary places in which people learned from one another and from the work they did became incidental to what counted as legitimate education. Ivan Illich described these older forms of learning as embedded in concentric circles of meaningful structures — family, work, craft, community, language, religion. Learning happened through participation because life itself was educational.

What looked like a scarcity of knowledge was, in part, what led to further concentration. Learning was taken out of the world and out of the person and reorganized into subjects and certified expertise, administered by specialists and measured against ever-changing standards. Education made knowledge available, but it also helped define where legitimate learning could take place and what it should look like. For a moment the internet disrupted all that and promised anyone with a bit of interest to find pretty much anything they want to learn about and do regardless of their abilities or context.

What some tech salespeople call ‘the inevitable future’

AI arrives at this point promising to extend this disruption and generate abundance but equally also accelerate further extraction, this time of the cognitive labor of learning itself. The machine can now perform the work underneath the educational product – the first draft, the translation, even the judgement about whether the answer is any good.

Even if we grant the strongest claims made for AI — and call what it produces ‘intelligence’ — abundance doesn’t automatically create meaning or have purpose. A million AI-generated novels don’t create a million readers; infinite information doesn’t produce infinite attention.

And yet the institutional response is more supply of everything. More platforms, more personalized instruction, more data collection, more AI. Governments and institutions are building frameworks and programs for AI in education, while schools and universities are being encouraged to incorporate these systems into teaching and learning.

The emerging evidence, meanwhile, offers concerning glimpses of what this new abundance might do to our cognitive capacities. An MIT study recently found that people writing essays with AI show weaker brain connectivity alongside lower reported ownership of their work and poorer recall, an effect researchers called ‘cognitive debt’.

A 2025 randomized controlled trial found lower delayed knowledge retention among students who used ChatGPT as a study aid. These findings are early, but they make it harder to treat the outsourcing of cognitive work as educationally neutral. What happens when the “cognitive crutch” becomes permanent, and the muscle we were supposed to be exercising never gets the chance to grow?

The temptation of AI is that it makes the typical struggle around learning look inefficient – as though it’s wrong to take years to learn a language or write a beautiful love letter. Why struggle to formulate an argument when the machine can formulate ten in ten seconds? Why read the book when AI can summarize it?

Then again, where exactly are we in a rush to? We still have to live the lives that all this efficiency was supposed to give us more time to live. We count completed essays and hours saved, while the things formed through the process are discarded as though they never mattered. You were supposed to be studying with your friends, but instead you stayed up all night laughing about silly stuff; a child spends an afternoon building a den in the garden for no particular reason.

These are the apparently wasted hours through which we exercise being human. A muscle grows because it’s repeatedly asked to do difficult things; when the demand disappears, it can atrophy. Our cognitive capacities have their own forms of exertion. We become capable of thinking by thinking, of remembering by remembering, of being resilient by having something to endure.

AI promises to remove much of that exertion. The hypertrophy of our technological environment may end up inducing the atrophy of our human capacities it was meant to augment. Those apparently wasted hours leave something behind, experiences and relationships that become part of the structure we carry into the future where, in hardships it is those apparently insignificant fragments of the past that give us somewhere to stand.

AI can generate all kinds of things. The harder thing may be finding the time and attention to live through the hours that produce nothing, yet nothing is just as meaningful. For adults, perhaps this is an irritating feature of modern life, but for our children it’s the fabric from which they are now being shaped.

They have no memory of the world before this flood. We are constructing the environment in which the next generation will learn what thinking feels like, what being human and being humane means. Perhaps we should grab hold of the places where learning still happens as part of life — the kitchen, the street, the broken bicycle, the blank page — places where education happens without announcing itself as such.

The danger of this new Age is that its definition of progress assumes that enough efficiency will finally bring us to equilibrium, and that equilibrium will bring abundance. We spent centuries trying to make knowledge abundant and seemingly succeeded. We now need to ask what all this abundance is for — and give a plausible answer.