Artificial Persons
Management, Liability, and the Animation of Institutions
Some people think AI replaces managers.
That sounds right if management is only the visible choreography of coordination. The little ceremonies get automated.
But I think this is the wrong animal.
AI does not produce fewer managers. AI produces more managers. More precisely, it managerializes everyone.
Artificial People
A programmer using agents starts as a craftsperson touching the material directly. Then something changes. The code is still there, but now the work happens through delegation.
The programmer is no longer simply writing. He is directing a small artificial staff.
The fog goes beyond cognitive offloading. It is the phenomenology of becoming the intermediary layer.
You lose direct contact with the material. You gain responsibility for an actor that can touch the material faster than you can. You stop thinking with your hands and start thinking through reviews, approvals, rollback plans and permissions.
You start monitoring what you said, when you said it, for the chance that the thing does what you asked instead of what you meant.
Obsolescence, perhaps at the granular level, as if the grocery store had stolen from you the privilege of beheading the hen.
AI offers a similar sanitation of agency. The act arrives without the contact. The consequence arrives without the ritual.
The distribution line absorbs the work but does not absolve you from accidentally poisoning others.
You become liable.
Liability
The absorption of legal liability matters, of course. Many seek that privilege alone. But management absorbs more. It stands between capability and consequence. It turns raw action into authorized action.
Someone has to decide what counts.
The more powerful the technology, the more this function matters.
A hammer needs little management. A factory needs a little army.
Software did not abolish management either. It produced a zoo of managers.
AI will not be different.
AI widens the blast radius of delegation. It produces more actions whose authorship must be interpreted. It forces the dumb and terrifying question: who acted when the system acted?
The user? The company? The model provider? The deployer? The manager who approved the workflow? The policy that constrained it? The team that trained it? The institution that benefited from it? The agent itself?
For now the answer is everyone and no one. That answer cannot survive contact with reality. It must be absorbed.
Burnham
To talk about management and not contend with Burnham is to miss a key part of its history.
Burnham was often wrong and is not a comfortable thinker, but The Managerial Revolution deserves inspection, for when it arrived, for who wrote it, and what he said.
The book is usually read as reactionary, conservative, anti-socialist, proto-Cold War, one more grim ancestor text from the twentieth-century machinery room.
But this is just too flat.
That reading happens because we are standing on the far side of Burnham. The managerial world already happened. Its intuitions have already become common sense.
So Burnham can look obvious, or crude, or morally embarrassing, in the way Machiavelli looks today — which perhaps Burnham knew would happen. Being suspicious of Burnham is generally prudent.
Burnham was not just “a writer.”
He was also a statesman.
He was also a revolutionary.
He grew within Trotskyist factional life, Hegelian training, intellectual combat, splinter groups, party discipline — people who believed that sentences could recruit history. This matters. He knew what ideas look like before they become institutions. He knew what it meant for language to become a command structure. He knew that a theory can recruit a class before the class fully recognizes itself. He understood that power had to be materially taken.
The Managerial Revolution holds its strangeness across time. The name may suggests business, but inside you won’t find better ways to run a firm. It belongs to the shelves of classical, cold, blunt political theory.
From our perspective, Burnham is uncitable in most rooms — reactionary, OSS, National Review, the wrong lineage across decades. At the time, he was arguing that capitalism and socialism were giving way to the same managerial substrate.
Who owns the machine? Wrong question.
Who can direct it? That is Burnham’s real question.
And there was, by then, already management theory. Taylor, Fayol, and others. The prewar tradition pictured management as rationalized labor, hierarchy, supervision, procedure, efficiency, the arrangement of bodies and tasks inside the productive apparatus.
He separates the manager from the owner, but also from the engineer, the scientist, the technician. This is important. The manager is not simply the one who knows how the machine works. The manager is the one who makes the system act.
Then, there is a sentence where Burnham says the managerial function may require the “psychological and social sciences,” because human beings themselves are among the things being organized. Not yet culture decks, performance frameworks, behavioral incentives, executive coaching, organizational design or mission statements. But the door is already open. The human interior has entered the administrative field.
After 1940, management theory floods precisely into that interior. Drucker, Simon, McGregor, Chandler, and then the rest of the postwar managerial cathedral.
Decision-making. Motivation. Knowledge work. Strategy. Structure. Culture. Autonomy. Executive cognition.
The manager is no longer merely the foreman of production. He becomes the interpreter of systems. The priest of coordination. The person who converts uncertainty into directives without appearing to command.
Burnham’s language is colder. Production. Planning. Class. State. Control. But the bones are there.
He writes that managers see themselves as “providing its brains.” He says their training makes them think in terms of “co-ordination, integration, efficiency, planning.”
Burnham did not merely observe the present. He redescribed the present as a future that has already happened. The revolution (The Managerial Revolution, that is) was not merely idealized — in his frame, it had already materialized.
Seen from today, this is closer to hyperstition than prediction. Burnham gives the managerial class a conscience, a mirror, a name, a posture, a destiny. Like a post-modern Das Kapital, or perhaps like the Fama Fraternitatis Rosae Crucis, describing something before it held roots. A practical King in Yellow of administration.
You read it and the scenery remains the same. The shareholder is still there. The parliament is still there. The party is still there. The public justification is still there. But the eye starts moving behind them, looking for the operators.
Get Daniel Rodríguez’s stories in your inbox
Join Medium for free to get updates from this writer.
But who was Burnham at the time?
During World War II he leaves NYU and enters the OSS — the American predecessor to the CIA. A classified OSS paper from spring 1944 later becomes part of The Struggle for the World. The theorist of managerial power does not remain outside the administrative interior. He becomes useful to it.
Who was really Burnham?
One of his critics was Orwell.
Orwell at the time writes Animal Farm while criticism of Stalin is still diplomatically inconvenient in Britain. His proposed preface, “The Freedom of the Press,” is about the fact that unpopular ideas can be suppressed without formal censorship. He had worked at the BBC. He had seen the polite machinery.
Then he starts chewing on Burnham.
Orwell at first dismisses him. How could he not? Burnham kept mistaking the apparent victor for history itself. Germany looks inevitable, so Germany becomes the future. Russia advances, so Russia becomes the future. The strongest force in the room acquires metaphysical dignity. Orwell calls it power worship.
But then, Orwell returns to him. 1944, 1946, 1947. Again and again. Burnham irritates him because Burnham is wrong, but also effective. Orwell notices people getting agitated after reading Burnham.
Then, at his deathbed, Orwell extrapolates Burnham’s three-world picture into planetary architecture on Nineteen Eighty-Four: Oceania, Eurasia, Eastasia. The same shape and the same pressure.
And then there is O’Brien.
In 1984, O’Brien is not merely Stalin, nor Hitler. He is an intellectual of domination. Someone who acted with full awareness of his action. Someone who knows that truth can be administered. He knows memory can be administered. He knows language, loyalty, pain, exhaustion, belief, and love can be made into governable materials.
O’Brien is what you get when the description of domination becomes indistinguishable from its practice. That, finally, is what Burnham showed Orwell was possible. Organs of control without anything he could identify as human.
No Longer Human
The managerial revolution did not end in 1941. It learned software. Now it is learning agents.
The human no longer only writes the code, or does the work in any recognizable way. The human increasingly directs processes that do these things, then reviews, corrects, escalates, and explains them.
That is managerial work, regardless of the title.
What remains with the human coordinator, and above the human, with the institution, is liability. Every new delegation therefore generates more orchestration. More capability means more boundary-work. More autonomy means more supervision. More speed means more auditability. More reach means more liability.
The old manager supervised workers.
The new manager supervises systems that supervise work.
The next manager supervises the policies by which systems supervise work.
This is the shape of industrialization. Shells that protect bodies. Membranes and absorption lines. The re-routing of effects. Vertically integrated ecosystems. Bodies with organs. Organs upon organs upon organs.
The Shell
Another of such, shallow critiques, in this case of mediation, says that all these layers are fake.
Bureaucratic unreality. Spectacle. Simulation. Baudrillard echoes through every account of our times.
I see it. But it is not enough. The frame makes the thinker subject to the object — bound to a partial image.
The shell of abstractions and bureaucracy isn’t merely the absence of something real. It is a vessel. Higher forms of technology cannot touch base reality nakedly. Reality is too sharp. Reality is analog and technology is digital. Reality contains too much exception, and ambiguity, and randomness. Complex things survive by discretizing, by interposing specialized organs between themselves and the world.
Management is one such organ.
So are law, accounting, medicine, education, insurance, interface design, human resources, compliance, venture capital, parental care, state administration. Each absorbs a different kind of shock. Each digests reality into a form some emerging thing can survive.
A startup is a shell around a future object that cannot yet survive reality. Sometimes the pitch decks and demos are fraudulent, other times something actually grows.
Parenthood is the same structure at a deeper layer. A child cannot meet the world directly. Parents absorb and metabolize reality into survivable doses until the child can bear more of it directly.
Entrepreneurship and parenthood are both dances around something not yet fully real but already demanding sacrifice. They are nests, eggs, membranes, liability sinks. They create successors capable of outlasting the dancers.
If AI management is gestational, what is gestating?
The ancient male womb
Zeus swallows the pregnant Metis, embodiment of cunning intelligence, and Athena later emerges from his head.
Semele is destroyed after seeing Zeus in divine form, and Zeus rescues the unborn Dionysus by sewing him into his thigh until he can be born.
These are not only strange stories about male gestation. They are political theology — instruments for reshaping belief, encoding a principle: what is born must pass through established authority to become legitimate in the higher order.
Athena emerges from the head of sovereignty as strategy, warcraft, civic intelligence, sanctioned order. Dionysus, even as divine excess, still has to be carried to term inside the body of the king.
The sovereign becomes womb.
Modern institutions do the same thing without thunderbolts. They absorb volatile potential, filter it, fund it, name it, insure it, supervise it, and release it as something admissible. A product. A policy. A company. A citizen. A market. A model. A legal person. An agent.
The body of authority becomes the incubator of the next real.
The Egg
We are incubating delegated agency with legal personhood lagging behind it.
In other words, existing artificial persons are becoming animate.
Corporations, bureaucracies, states, funds, universities, platforms, courts, militaries, churches, supply chains, and startups.
Person-like entities already persist through time, own property, sign contracts and are sued. They remember through records. They speak through representatives. They act through employees.
Now an institution gets equipped with AI to observe itself, speak to itself, model its environment, generate decisions, negotiate across interfaces, and act through automated limbs.
That is the fetus. Not the model.
Humans are cells inside corporations. Corporations are cells inside markets and states. States are organs inside planetary civilization.
Once artificial persons become animate, they may start acting less like documents and more like citizens of larger bodies.
Its skull is agency.
Its spine is memory.
Its hands are tools.
Its blood is capital.
Its protective membrane is management.
