Nations Must Die

Nations must die.

By that, I mean the political division of lands and seas should disappear. But I also mean something more specific: the nation-state as the highest container of power and control should go away completely.

We don’t need the idea of nations anymore, just as we no longer need kingdoms. The place where you are born should not determine your rights, your freedom of movement, your duties, or the limits of anyone else’s concern for you. That idea should die too.

Cultures, languages, food, music, memory, and the human need to travel or stay and belong should live and multiply. The idea of nations and a Government representing it has run its course.

The machine that nations were

A nation is an old answer to a hard problem: how do millions of people share land, food, work, risk, and responsibility? For most of history, geography was the database. Borders were the interface. Citizenship was the access key. Information moved slowly, so power needed layers. A village reported to a province. A province reported to a capital. The capital sent orders back down.

Companies grew the same way. In From Hierarchy to Intelligence, Jack Dorsey and Roelof Botha argue that corporate hierarchy is an information routing system. Managers became necessary because no leader could know what thousands of people were doing. Information went up. Orders came down. Every layer helped the company grow, and every layer slowed it down.

Nations are larger versions of the same machine. Ministries, departments, provinces, embassies, and border offices route facts, needs, money, permission, and blame. AI can now do much of that routing. It can translate languages, read public records, trace decisions, model consequences, match resources to needs, and keep a live picture of a large system. Geography no longer needs to decide who can understand whom or work with whom. Once the routing problem changes, the structure built around it can change too.

Assume abundance is possible

Let us make two large assumptions. First, a future of abundance is possible. Second, we arrive at it together.

By abundance, I do not mean infinite consumption or a machine that grants every wish. I mean something less magical and more useful. Every person can eat good food, live in a safe home, receive care, learn, move, and build a family without taking those things from someone else.

Sam Altman describes a similar future in Moore’s Law for Everything:

“As AI produces most of the world’s basic goods and services, people will be freed up to spend more time with people they care about.”

The hard part of that sentence is not production. It is making sure the freedom reaches everyone.

Energy is clean and plentiful. Food grows with little waste. Machines build homes and maintain water systems. Disease is found early. Knowledge is available to anyone who wants it. No child begins life already losing a contest for basic survival. Under those conditions, land stops being the prize beneath every conflict. A fertile valley does not need an army if food can be grown almost anywhere. A deposit of fuel does not need a flag if energy comes from the sun, wind, atoms, and better storage. Migration stops looking like a raid on a limited stockpile.

This does not make every disagreement disappear. Humans can invent a status contest even in paradise. It removes the oldest pressure under many of our fights: the fear that there will not be enough. Nations were built to manage that fear. Abundance weakens their reason to exist.

What work should machines inherit?

Machines should inherit the work of keeping civilization alive. Grow and move food. Balance energy grids. Build shelter. Maintain roads, sewers, and water systems. Watch for fire, flood, disease, and crop failure. Repair what breaks. Clean what we polluted. Do the dangerous work in mines, reactors, storms, oceans, and space.

They should also carry the burden of coordination. A machine can notice that one region has spare grain while another has a failed harvest. It can route medicine without asking which passport deserves it. It can model the effect of a decision before millions of people have to live with the mistake. This is where AI earns its place. It handles scale, memory, repetition, prediction, and maintenance. It makes the basic systems of life boring and reliable.

In Machines of Loving Grace, Dario Amodei imagines powerful AI as a “country of geniuses in a datacenter.” He also gives the idea a useful limit: “Intelligence may be very powerful, but it isn’t magic fairy dust.” Fields still need water. Houses still need materials. Medicine still needs experiments. Intelligence can improve these systems, but the physical world keeps its own clock.

The machines should manage abundance as infrastructure. They should not decide what abundance is for.

What work remains human?

Humans must choose the direction. We decide what is fair, what is beautiful, what deserves protection, and what risks are worth taking. We raise children. We care for people who cannot repay us. We make promises. We forgive. We make music, food, jokes, rituals, and strange new forms of life together.

We also decide when the machine is wrong.

Geoffrey Hinton puts the control problem plainly in his Nobel Prize conversation:

“Will we have a way of making sure they don’t want to take over?”

An abundant future fails if people hand the power to choose that future to the systems producing it. That work cannot be reduced to finding the most efficient answer. An efficient system can still be cruel. A perfect allocation can still deny someone a voice. Democracy may remain one useful method for keeping power answerable, but no method saves us from moral choice.

I made a similar argument about integrity in the post-AI era. Machines can make competence abundant. They cannot make character abundant. They can show us the consequences of a choice, but they cannot carry the responsibility for making it.

The right division of labor is clear enough. Machines should run the systems. Humans should remain responsible for their purpose and limits.

Abundance must include every living being

A human paradise on a dying planet would be another failure. Earth must heal as part of the same project. Forests should return where forests belong. Wetlands should fill again. Rivers should run clean. Coral reefs, grasslands, mangroves, tundra, and deserts should hold more life, not less. Species pushed toward extinction should have room to recover.

Positive terraforming begins here, on the planet we damaged. That might mean turning degraded, human-made wastelands green. It might mean helping rain return to places stripped of trees. It might mean rebuilding soil, restoring watersheds, removing invasive species, reintroducing native ones, and giving large animals enough connected land to move. It does not mean covering every desert with rainforest. Deserts are alive too. Healing Earth means helping each place become more fully itself.

Machines can measure soil, water, heat, migration, and loss at a scale no person can hold in mind. They can plant, monitor, and repair for centuries. They can show us where an intervention helps and where our good intentions cause harm. Humans still have to choose restraint. The goal is a thriving planet, not a perfectly managed garden.

What happens when survival is no longer the project?

For most of our history, survival took almost all our time. Then civilization turned survival into systems. Agriculture, cities, medicine, industry, and software each moved part of the burden away from the individual. AI and machines may finish much of that transfer.

Then what?

Perhaps we begin the next construction project.

The Dyson sphere is a useful symbol for it. Freeman Dyson imagined an advanced civilization capturing far more of its star’s energy. We would probably build a vast swarm of solar collectors rather than a solid shell. The exact shape is not the point. The scale is.

A species that no longer spends its intelligence fighting over patches of Earth could use that intelligence to build beyond Earth. Orbital habitats. Solar collectors. Scientific instruments spread across the solar system. Ships that take generations. Experiments we cannot yet name. The appetite for discovery will not end when hunger does. Google DeepMind’s essay A New Golden Age of Discovery makes the quieter claim that “future demand for science appears practically limitless.” That seems right. Solving survival gives curiosity more room, not less.

Machines would do much of the building. Space is cold, radioactive, distant, and hostile to the human body. Robots do not need air, sleep, or a return ticket. They can work for decades before people arrive. Humans would supply the reason to go. Curiosity. Wonder. The wish to know whether life exists elsewhere. The desire to make something that takes longer than one lifetime.

Why nations fade

Abundance will not dissolve nations by itself. Power has a habit of surviving the problem that created it. But the case for nations becomes much weaker when food, shelter, energy, knowledge, and care no longer depend on controlling scarce land. The border loses its practical purpose. Citizenship loses its power to assign human worth. The machinery of the state starts to look like a relic built for a shortage we no longer have.

People will still belong to places and cultures. They will still form communities, make rules, settle disputes, and stop anyone who tries to dominate others. Those functions do not require humanity to remain divided into armed territorial camps. The transition cannot come from conquest. One nation cannot defeat the others and call the result unity. It has to come from making the old structure less useful, year after year, until a border carries no more moral weight than an old county line.

Machines can take over the work of survival, maintenance, coordination, and planetary repair. Humans can care for Earth, raise families, make culture, govern shared power, and begin building beyond the planet. Once no child goes hungry and no forest must be destroyed for survival, that is enough work for us.