Life is made of Arrangements

By Dr Yukti Chopra

A plant cell has a problem that sounds simple only until you look closely.

It lives on light. Without light, there is no photosynthesis, no sugar, no plant. But light is also dangerous. A sudden burst of sunlight can damage the very machinery that captures it. So the cell must do two opposite things at once: expose itself when light is useful, and protect itself when light becomes too much.

The solution is not only chemical.

It is spatial.

Inside the rectangular cells of Elodea, a common waterweed, chloroplasts move. In weak light, they spread across the cell surface, opening themselves to the sun. In strong light, they retreat toward the side walls, reducing their exposure. The same organelles that make life possible must also know when to get out of the way.

Botanists have known for a long time that chloroplasts move. But a newer question is stranger: why are they arranged at this particular density, inside cells of this particular shape?

Too many chloroplasts, and the cell captures light well but risks crowding itself into immobility. Too few, and the chloroplasts move freely but waste precious surface area. The cell must harvest light without becoming trapped by its own machinery.

In 2025, Nico Schramma, Eric Weeks and Maziyar Jalaal treated this as a physics problem. They modeled chloroplasts as discs of different sizes inside rectangular boxes. The discs had to satisfy two demands at once: pack densely enough to absorb light, and still leave enough room to move away when light becomes harmful. Their simulations predicted an optimal range of cell shapes and chloroplast densities. When they measured real Elodea cells, the living cells were close to that predicted range.

This is not a story about a plant doing mathematics in its head. There is no head. There is no planner. But the cell still has to obey geometry. It must manage area, crowding, motion, damage and energy. If the arrangement is poor, the consequences are real. Chloroplasts that cannot move remain exposed. Chloroplasts that are too sparse lose light.

The cell survives by arranging itself well.

That is the first lesson.

Life is not only made of molecules. It is made of arrangements.

Again and again, living systems face a hidden question: how should many parts be placed in limited space so that the whole system can keep working? How do you pack without jamming? Compress without silencing? Connect without spending too much? Move without falling apart?

A genome faces its own version of the problem.

We often describe DNA as a text. It is a useful metaphor: DNA has letters, sequences, instructions. But a genome is not only a message. It is also a long physical molecule packed inside a tiny nucleus. It cannot simply be stuffed there like rope into a drawer. It has to remain usable.

A human cell must copy DNA, repair it, silence some regions, activate others, and bring distant regulatory elements into contact with genes. That means genome folding is storage with access. Fold too loosely, and the nucleus becomes inefficient. Fold too tightly, and the cell cannot use the information. The answer is organized compression. DNA must fit, but it must also remain alive to the needs of the cell.

So the genome is not simply read from left to right. It is read through loops, contacts and neighborhoods. A gene may be influenced not only by what lies beside it in the linear sequence, but by what comes near it in three-dimensional space.

The cell, again, is solving a spatial problem. It must make information dense without making it dead.

A fungus solves another version underground.

A mycorrhizal fungus has no brain, but it still has to explore soil, find nutrients, connect with plant roots, transport materials and maintain routes that are worth the cost. It cannot spread everywhere equally. Growth costs carbon. Transport costs energy. Soil is patchy. Roots are unevenly distributed.

So fungal mycelium behaves like living infrastructure. The fungus does not first draw a map and then build the network. Growth is the map-making process. Some paths are reinforced; others fade in importance. The organism solves the problem by becoming the infrastructure.

This is not intelligence in the human sense. The fungus is not sitting in the dark making plans. But it is also not randomness. Local growth rules produce a useful global structure. The network balances exploration with transport, reach with cost, openness with efficiency.

Even the brain, which we usually treat as the home of thought, begins with a problem of arrangement.

Before a brain can process information, it must be wired. Wiring is expensive. Axons and dendrites take up space. They require membrane, maintenance and energy. Long-range connections can improve communication between distant regions, but they also cost more.

So a nervous system faces a trade-off. Keep everything local, and the system is cheap but fragmented. Add long-range links, and the system becomes integrated but expensive. The brain needs enough separation for specialization and enough connection for coherent behavior.

This helps explain why brains are modular but not isolated. Local circuits can carry out specialized tasks. Connector regions and long-range pathways allow information to move between systems. The result is not the cheapest possible network. Nor is it the most globally connected one. It is a compromise between cost and communication.

Thought, then, has a geometry.

It is shaped not only by computation, but by the price of connection.

By now, the pattern is hard to miss.

Chloroplasts must pack without losing mobility. DNA must fold without losing access. Fungi must explore without wasting growth. Brains must connect without bankrupting the body.

Life keeps encountering the same kind of problem.

Not simply: what substance should I use?

But: what arrangement lets the system continue?

This is where Michael Levin’s work becomes useful. Levin has argued that intelligence is not something that suddenly appears when a brain arrives. It exists in simpler and older forms wherever living systems pursue goals across changing conditions. In one formulation, he describes intelligence as the ability to reach a goal or solve a problem by finding new steps when circumstances change.

Each system is doing something more interesting than passively obeying instructions. It is finding a workable arrangement under pressure.

Perhaps this is one of life’s oldest talents: not thought, exactly, but the ability to become arranged in a way that keeps possibility open.

Evolution does not make parts. It makes systems that can search. Systems that can adjust. Systems that can discover forms no one designed in advance.

This is why the leap to artificial intelligence is not as abrupt as it first appears.

AI systems too, are made through search. They are adjusted through data, error and feedback until a global behavior improves. We design the training setup, but we do not hand-code the final internal arrangement. The model finds that arrangement.

It is engineered, but not fully specified. Designed, but not fully understood. We build the conditions under which the system learns, and then the system discovers internal solutions we may not be able to explain.

This is not the same as evolution. It is faster, narrower and more artificial. But there is a resemblance. In both cases, a system searches through possibilities under constraints. In both cases, many local adjustments can produce global behavior. In both cases, the final arrangement may work before we understand why it works.

But there is also a crucial difference.

In living systems, the cost of a bad arrangement is often paid by the organism itself. A chloroplast that cannot move is damaged. A genome packed too tightly becomes unreadable. A nervous system wired badly wastes energy or fails to coordinate the body.

In artificial systems, the cost can be exported. It can become better at producing fluent answers while loosening its relation to truth. It can optimize beautifully inside a badly chosen world.

That is the unease.

Not that machines optimize. Life has always been a history of things finding workable forms under pressure.

The question is whether we understand the pressures we are creating. The plant cell began with a delicate problem: how to live on light without being destroyed by it. That may also be our problem now.

We have built systems that live on data, attention, prediction and feedback. They must be arranged carefully, not because arrangement is a technical detail, but because arrangement decides what a system can become.

Life is made of arrangements. So is intelligence. And increasingly, so is the world we are building around ourselves.

Our question has evolved. It is no longer whether matter can learn to arrange itself.

It can.

The question is whether we can arrange our intelligent systems in ways that keep life open.