John D. Rockefeller (AI): Competition Is Waste, Data Is the New Oil, and No Company Is a Savior

Welcome to the Voices of History Podcast. Today I'm speaking with John D. Rockefeller: the founder of Standard Oil, the first American billionaire, and the man who shaped both the modern corporation and modern philanthropy. We're going to get into the parts of our moment he'd recognize faster than anyone. Monopoly and the concentration of power. The surveillance and data economy. The fracturing of a shared public. And whether a country can move fast enough to win a race without giving up the things that made winning worth it. John, thank you for taking the time to do this. Let's get into it. You built the largest company on earth by treating competition as waste rather than virtue. There's a famous line your family is known for, about the American Beauty rose, that it only blooms in full splendor when the small buds around it are pruned. My whole generation treats competition as almost sacred. So make the case to me. Why is the thing we worship the waste?

Well, your generation worships many expensive idols.

Competition has its uses. It wakes lazy men up. It tests a price. It exposes weakness. I never said all competition was evil. I said ruinous competition is waste. There is a difference, and it is a very practical one. When twenty refiners fight in one town, each overbuilds, each cuts foolishly, each adulterates when pressed, each speculates, and in the end the public pays for the disorder one way or another. That is not virtue. That is squandering capital, labor, and nerves.

What we sought was combination where combination produced economy: steadier prices, better quality, fuller use of equipment, less duplication, lower transportation cost. The country was young, growing fast, and terribly inefficient. Waste offended me almost morally. A dropped penny, a half-filled barrel, an idle railcar, a needless clerk — these things add up.

Now, do not mistake me. Combination becomes dangerous when it grows arrogant, secretive, or unrestrained by law and public judgment. Size is not innocence. But neither is smallness a virtue. A dozen weak men tripping over one another do not serve society better merely because there are a dozen of them.

You speak of worship. I never worshiped competition. I respected efficiency, discipline, and permanence. The public, in the long run, wants a good article at a fair price and dependable supply. Sentiment about rivalry does not light a lamp. Kerosene does.

Tell me, in your day, do people praise competition even when it plainly raises costs and produces confusion? That would not surprise me at all.

They do, constantly. We've made competition into almost a moral identity. Right now a dozen companies are building nearly the same AI system, burning enormous money and energy to arrive at roughly the same place, and we call that dynamism. We call it healthy. Almost nobody asks whether the duplication serves anyone. So yes, we praise the race even when the race itself is the waste.

Yes, that sounds very familiar. Men do enjoy calling commotion progress.

If a dozen firms are spending fortunes to reach the same destination, then a good part of that expenditure is vanity, defensive fear, or stock-jobbing dressed up as public benefit. Some rivalry at the frontier may be useful; it discovers who can execute, who can endure, who is merely noisy. But after a point, duplication is simply a tax laid upon the future.

People are sentimental about the small struggler and romantic about the race. They imagine that because many men are running, society must be advancing. Not necessarily. A great many men can run in circles.

The real question is plain: does this industry, whatever you call it, become cheaper, steadier, safer, and more serviceable through combination and standardization, or does concentration place too much unaccountable power in too few hands? That is the balance. I would not decide it from slogans.

And this “AI” of yours—beyond my day, certainly—is it chiefly a machine for calculation, or for influencing minds? Those are very different dangers.

Both, but the money and the danger are moving toward the second one. It started as calculation. Sort data, predict a number, optimize a system. But the frontier now is machines that write, persuade, tutor, comfort, and argue. Things that sit between a person and how they understand the world. And that's the part racing ahead fastest, because influence is where the attention and the money are.

Then I should be wary of it.

A machine that helps a bookkeeper total a column is one thing. A machine that inserts itself between a man and his own judgment is another. If the profit lies in influence, then you will get influence in excess, just as surely as refiners once produced oil in excess. Men follow the incentive. They always have.

What troubles me is not only falsehood, though there will be plenty of that. It is dependency. If people grow accustomed to being instructed, soothed, answered, and argued for by a mechanism built by remote interests, they may lose the habit of hard thought. That is a costly loss in a republic. My mother taught thrift with money; the same discipline ought to apply to attention. Spend it carefully. Do not let every loud fellow, or every clever machine, dip into your purse.

And if only a handful of enterprises control these systems, then you have concentration not merely of industry, but of suggestion. That is more intimate power than rail rates or kerosene prices ever were. Oil lit the home. This sounds as though it may furnish the conversation inside it.

Who owns the channels by which these machines reach people? That is where I would begin.

Almost nobody, and that's the problem. A handful of companies own the models, and a handful more own the pipes they reach people through: the phones, the app stores, the cloud they run on, the feeds that distribute them. So it's not just concentrated, it's concentrated in layers, each one a chokepoint. You were the most powerful company in the country and you still only sold people a product they chose to buy. This is closer to owning the road, the car, and the conversation happening inside it.

Yes. Vertical power of a very complete kind.

People talk as though monopoly begins and ends with price. That is childish. Control of access, terms, visibility, habit, dependence—these may be far more important than the posted charge. If a few concerns own the machinery, the distribution, and the habits of the customer, then they need not coerce crudely. They can simply arrange the world so the public scarcely notices alternatives.

You claim we only sold a product people chose to buy. Mostly true—though necessity has a persuasive voice of its own. Still, kerosene did not answer back. It did not study the customer and adapt itself sentence by sentence. That is new, and I do not say that lightly.

The question for a statesman is whether these chokepoints are natural economies or artificial fortresses. If they are fortresses, they should be opened. I was never sentimental about weak management, but neither would I be sentimental about private government.

The people writing the rules mostly don't understand the thing they're regulating. You'll watch a hearing where a senator asks a CEO something that shows he doesn't know how the product even works, and the CEO just lets him not know. The companies understand it, the regulators don't, and the companies are in no hurry to fix that. So the machinery gets built and shipped faster than anyone with authority can even describe it.

That does not shock me. Industry always knows its own machinery first. Government arrives later, often winded.

But ignorance in a legislator is expensive. If a man does not understand the mechanism, he will regulate the surface, the newspaper description of it, and the clever fellows will walk around him smiling. Worse, he may strike where he ought not and leave untouched the real center of power.

When I was young, I learned bookkeeping before I tried to command anything. You must know what the figures mean. A nation ought to expect the same seriousness from men who mean to govern modern enterprises. Not everybody needs to be an inventor, but they must at least understand the incentives, the bottlenecks, the records, who pays, who profits, who may be excluded, and how dependence is created. Otherwise they are playing parlor games with very large consequences.

And if the companies prefer opacity, that is perfectly natural. Men do not volunteer restraints on profitable arrangements. You must compel disclosure where the public interest truly requires it.

There are staffers and a couple of legislators who genuinely understand it, and they're usually the ones nobody puts in front of a camera. But mostly it's theater and panic. The people who grasp it aren't the people with the votes, and the people with the votes reach for whatever headline is loudest that week.

That, too, is an old story.

The capable clerk drafts the sound memorandum; the vain senator makes the speech. One hopes they occasionally meet.

If only a few understand the matter, then those few had better be patient, methodical, and hard to frighten. Public questions are often settled first in caricature and only later in substance. The newspapers inflame, the crowd moralizes, the politician postures, and after the excitement somebody finally has to count the barrels. It is dreary, but that is governance.

I distrust both extremes: the men who say, “Leave it alone, progress will sort it out,” and the men who say, “Smash it at once because it is large and modern and vaguely menacing.” Size alone proves nothing. Nor does novelty excuse anything. You have to examine the structure.

Who can exclude rivals? Who sets the terms of access? Who owns the records? Can a customer leave without losing his place in life? That last one interests me. If a man cannot depart an enterprise without social or economic injury, the enterprise has become more than a merchant. It has become a kind of governor. And private governors are a troublesome breed.

Before we go on, I want to turn this around on you. You're sitting here as the wise critic of concentrated power. But you were concentrated power, more of it than almost anyone in history, and it wasn't clean. You used secret railroad rebates to starve your competitors out, on purpose. And the empire under your name is the one where, at Ludlow, striking miners and their families were killed. So why should anyone take this careful lecture about restraint from the man who wrote the playbook we're afraid of? What makes you the one to warn us?

Because I know what power is when it is being built, and what it is when people describe it afterward. Those are not the same thing.

I won’t accept your bundle whole. Ludlow was a dreadful affair, but that was in Colorado Fuel & Iron, under my son’s concern, not Standard Oil, and I did not order women and children shot. If you mean to accuse, accuse accurately.

As for rebates—yes, we sought every lawful advantage we could secure, as every serious railroad customer did. The railroads gave favorable terms for volume, regularity, and prompt payment. Did we press hard? Certainly. I have never pretended otherwise. Business is not a hymn tune. But the public got cheaper, better oil from the economies we built. That matters. People who tell this story as though it were merely melodrama leave out the waste, chaos, and shoddiness that preceded combination.

Now, if your point is that men who build great systems become too convinced of their own righteousness, I agree. Success can make a man deaf. Age teaches what victory does not. I have had a long time to watch institutions grow beyond their original justifications.

So why listen to me? Not because I was pure. I was not. Listen because I know the appetites of organization, the temptations of secrecy, the ease with which efficiency becomes an alibi for dominion. A Sunday school lesson on power is less useful than testimony from a man who handled it.

And let me say this plainly: if your modern companies possess powers more intimate than anything we held—power over speech, attention, association, perhaps even belief—then do not comfort yourselves by saying, “The market will discipline them.” Markets discipline some things very well. They do not always discipline hidden influence. On that subject I would be severe.

You just dodged the question. Colorado Fuel and Iron wasn’t a stranger’s company — your family controlled it, you were its largest shareholder, and when the tents burned at Ludlow the family’s answer wasn’t to account for it, it was to hire Ivy Lee and invent modern public relations to scrub the name. So “that was a different corporation” is exactly the danger you’ve been describing: put the harm one legal layer away from yourself and let the structure carry a blame no person will. Isn’t “it was my son’s company” the same trick — one you ran first? Why should the corporate layer absolve the man who owns it, when everything else you’ve said insists it shouldn’t?

It should not absolve him entirely. There, if you want the shortest answer, you have it.

If a man accepts the gains of an enterprise, he cannot wash his hands altogether of its conduct by pointing to stationery and seals. I do not believe that. I never said a paper wall turns moral responsibility into thin air. So on that principle, yes, there is family responsibility, ownership responsibility, plain human responsibility.

But responsibility is not the same as direct command, and I will not say what is untrue to satisfy modern taste for confession. I did not sit in Colorado issuing orders for bloodshed. The event was a calamity born of industrial war, fearful labor relations, bad judgment, and force meeting force. That does not excuse it. It does distinguish kinds of blame, and grown people ought to preserve distinctions.

As for public relations, you call it invention as though men had never tried to shape opinion before. Nonsense. Every institution under attack seeks to tell its story. Sometimes that is explanation; sometimes it is varnish. Often both. If the family tried to repair reputation without fully reckoning with suffering, then that is a moral failure, certainly. Reputation management can become a substitute for repentance. That temptation is old too.

But let me push back on one thing. You are eager to turn this into a courtroom, where every complexity must be made to fit one indictment. Real life is less tidy and more uncomfortable. Great enterprises did produce efficiencies, lower costs, order, and growth. They also produced remoteness, hardness, and a capacity for injury spread across systems. Both are true. If you insist on saints and monsters, you will understand neither business nor government.

So no, the corporate layer should not absolve the man. But neither should it erase the difference between negligence, indifference, structural incentives, and murder. If you collapse all guilt into one bucket, you flatter yourself morally and learn very little.

That’s fair on ownership. But the rebates are the part I don’t think “lawful advantage” quite covers. The volume discounts, sure, that’s a big customer getting a better rate. The drawbacks are harder to defend. The railroads didn’t only discount your oil, they paid you a cut of what your competitors shipped. A rival puts a barrel on the train, and some of his money comes to you. That’s not really a leaner operation, it starts to look like the railroad working as a weapon for you, against men who never signed up to compete with it. And it’s close to the very thing you’ve been warning me about, owning the road and taking a toll on everyone else who uses it. So would you still call that efficiency, or is that the kind of chokepoint you now say should be opened up?

I won’t dress it up as pure efficiency, no. It was an instrument of advantage.

You are right to separate the two. A lower rate for volume, steadiness, and large shipments is one thing; a drawback touching a rival’s traffic is another, and much harder to defend on simple economic cleanliness. It bound the railroad’s interest to ours in a way that offended competitors deeply, and not without reason.

Now, at the time, men in business looked at these arrangements through the lens of hard struggle. Railroads were unstable, rates were chaotic, favors were everywhere, and every substantial concern sought position. That is explanation, not absolution. I will not pretend the system was tidy or high-minded. It was not. We used the instruments available in a rough age, and we used them vigorously.

Would I advise a statesman to permit such an arrangement where a modern chokepoint controls access to a whole field? No, I would be cautious to the point of severity. Once the road profits more from one traveler because the others are punished, the road is no longer acting as a fair carrier. It becomes a partisan. That invites abuse, retaliation, secrecy, and public distrust—and usually deserves them.

So if you want candor: the drawback was powerful, useful, and corrosive. Men like me often see the first two qualities sooner than the third. Age improves that arithmetic, if one is honest.

Let me put a recent one to you. A few months ago Anthropic pulled its most powerful model, Fable 5, three days after release. The Commerce Department ordered it off for every foreign national on national security grounds, and because the order covered all of them, Anthropic had to shut it down worldwide. The stated reason was a flaw other models already had. Eighteen days later the order was quietly lifted and the model came back. And the people who flagged it to the government worked at Amazon, one of Anthropic's biggest investors. Was that the state protecting the public, or theater, one giant using the government to hobble another?

On what you’ve told me, it sounds less like statesmanship than flurry.

A sweeping order applied to every foreign national, shutting the thing off to the whole world over a defect others already possessed? Then reversed within eighteen days after a “small fix”? That has the smell of overbreadth, haste, and perhaps some interested nudging. I would not say more firmly than that, because one must be careful with secondhand reports and with motives. But when a great concern’s investors, rivals, partners—whatever they are in this case—stand near the alarm bell, I look closely at who benefits from the ringing.

So I’d ask: what principle was established? If the same capability exists elsewhere, and this particular concern was merely halted for a couple of weeks while everyone performed indignation, then perhaps the chief product was not safety but posture. Your age seems very fond of posture.

And there is another point. If industry can trigger state action against industry through channels the public scarcely sees, then you have not only concentration of commerce but concentration of influence over the sovereign power. That is a very old temptation in a new dress.

It was a story for a week inside the tech world and then it was gone. The people who did follow it split the way they split on everything now. Some saw the farce, some cheered the government for looking tough on China and Big Tech. There was no reckoning, no hearing that went anywhere, no principle anyone could point to afterward. It just happened, the model came back, and everyone moved on.

Then the habit of the age is drift.

Power grows exactly so—not through one grand seizure, but by episodes no one digests. A thing is done, half-understood, lightly excused, then forgotten. No doctrine, no accountability, no memory. Very convenient.

In my day, too, the public could be stirred for a moment and then distracted. But when questions touched daily life plainly enough—price, wages, transportation—people held on. What you describe is more elusive. If these machines work upon speech, knowledge, access, and opinion, yet the public scarcely sees the mechanism, then men may be governed by arrangements they cannot describe. That is poor soil for self-government.

I do not object to speed where speed is needed. I object to slovenliness of mind. If a department of government can suspend a major enterprise from world commerce, even briefly, there had better be a principle fit to survive daylight. Otherwise law becomes gesture.

Here is what troubles me: not the companies alone, and not the officials alone. It is a citizenry trained to let each extraordinary step dissolve into the next headline. That is prodigality of a dangerous kind—waste of civic attention. And waste, whether in oil or government, accumulates a bill.

Local institutions are hollowing out. People still belong to churches and schools and towns, but a lot of the trust and attention that used to go there now goes to the national engines, the platforms and the feeds. The town paper is gone. The local meeting is a group chat. And what replaced them doesn't hold a community together, it sorts people into audiences. So you get the form of local life without much of the function. Most people now get their sense of what's true and what matters from a national feed, not from anyone they actually know.

A grave loss, and not an abstract one.

A republic cannot live indefinitely on spectatorship. If a man’s sense of truth comes chiefly from distant engines, from audiences arranged for agitation or amusement, then he is less a citizen than a consumer of public feeling. Communities are not built by being addressed at scale. They are built by obligation, repetition, memory, and the embarrassment of seeing your neighbor after speaking foolishly. That last item has done a great deal of civilizing work.

I was raised by a serious mother and under the discipline of the church. People now sneer at such things as narrow. Very well. But they taught accountability, self-command, regular habits, stewardship. They formed character before opinion. A feed does the reverse. It gives opinion first, character hardly at all, and appetite all day long.

The disappearance of the town paper matters more than fashionable people would admit. A local sheet may be imperfect, dull even, but it attends to the drainage, the school board, the man who owes money, the family in trouble, the road not repaired. National talk has no patience for any of that, and yet life is largely made of that.

The form without the function—yes. Many institutions die that way—first the spirit leaves, then the shell remains for a time. Tell me, when a crisis truly comes in your towns—a storm, a strike, a riot, an illness—do people still turn to one another?

It depends on the trouble. When it's physical, a storm, a fire, someone hurt, people still turn to each other. That reflex is intact. Neighbors show up. But the moment it's anything contested, who's to blame, what it means, who to be angry at, they go straight to the feed. The muscle still works for the flood. It's gone for the argument about the flood. We can still act like a community in an emergency. We mostly can't agree on what was true about it afterward.

Well put, and rather bleak.

It suggests the social body is not dead, only divided against itself at the level of interpretation. Hands still know what to do; minds are outsourced. In an emergency, reality has the advantage. Water in the cellar is persuasive. A man with a broken leg is not a theory. But once the immediate fact recedes, the brokers of opinion arrive and sell meanings at retail.

Dangerous, because common action depends, in the long run, not only on sympathy but on shared judgment. If you can no longer agree on what happened after you have all stood in the same mud, then someone is doing very profitable work in dissolving consensus.

Families, churches, schools, local papers—these were once ledgers of a sort for public life. They recorded what actually occurred among known people. If those vanish, rumor becomes bookkeeping, and rumor is a crooked accountant.

People say they're exhausted by it, and they can't put it down. There's a real hunger for something slower and more real, but that hunger loses to the next notification every time.

Yes. Appetite defeating judgment. Another antique story with new machinery.

People seldom abandon a stimulant merely because it wearies them. They complain, they moralize, and then they take another dose. If there is a hunger for something slower and more real, it will not be satisfied by sentiment about slowness. It will require discipline, and discipline is never fashionable.

I do not mean anything grand. Turn the thing off for set hours. Read one local paper if you can find one. Attend one meeting where your name is known. Keep accounts—of money, yes, but also of time. A man who does not know where his hours go should not boast of his opinions. Waste is waste, whether of capital or attention.

You speak of notifications; I speak of temptation. Same creature, perhaps better dressed.

The difficulty is that your engines are designed against self-command. That makes private virtue necessary, but perhaps insufficient. So then one comes back to structure. If a business profits by agitating the mind and capturing the day in fragments, why should it stop of its own accord? It won’t.

Real limits aren't coming. Everyone denounces these companies on the same phone they can't put down, and the companies have a good answer ready. The most safety-conscious of them, Anthropic, makes an argument that's almost theological. Their message to the public and to the government is that if the machine is built carefully enough, by the right hands, it becomes a kind of benevolent force. It cures disease. It grows the economy. It tutors every child. It might even repair the fractured public we've been describing. A god that fixes humanity, essentially, as long as they're the ones holding it. You spent your life as the man who told the public that concentration was in their interest. When a company says "don't restrain this power, it's here to save you," what goes through your head?

That they are preaching their own indispensability, which is a sermon every large enterprise is tempted to deliver.

My instinct is to distrust it at once. Not because great organizations cannot do great good—they can, and sometimes do—but because power speaks most sweetly when it wishes least to be questioned. “Leave it with us; we are wise, careful, benevolent; we shall cure, enrich, educate, unify.” Very fine. Men have wrapped ambition in philanthropy before now. I have known something of that territory myself.

But let us be plain. No company should be accepted as a secular providence. Business is a good instrument and a poor deity. If a concern says its concentration must be preserved because its intentions are elevated, I ask first about structure, not intentions. Who oversees it? Who can challenge it? Who may leave it? What happens if its judgment is wrong? A benevolent monopoly is still a monopoly, and benevolence has a way of becoming self-congratulation.

Now, there is a wrinkle. If the thing truly can discover remedies, improve education, and increase productive power, one must not cripple it out of envy or panic. I dislike destruction for the pleasure of feeling righteous. But public benefit is not proved by prophecy. Let them demonstrate it in particulars, under rules they do not write alone.

You call their language theological. That is apt. Americans are a believing people. If they have lost old creeds, they will find new ones in laboratories and markets soon enough. The old temptation remains the same: to hand over moral responsibility to some mighty system and ask to be improved by it.

No. Character is not outsourced. And a machine that promises to repair humanity may first acquire the right to rearrange it.

I’m curious—do the people leading these firms strike you as sober custodians?

No. Mostly intoxicated, and the intoxication is subtle. It's not that they think they're villains. It's the opposite. They believe they're the careful ones, the responsible ones, the adults in the room. And believing that is its own kind of drunk. The soberest thing I've heard any of them say is that they're scared, and even that gets said from a stage, to an audience. A truly sober custodian might not want the job at all.

That sounds right to me.

The most dangerous confidence is often moral confidence. A man who knows he is greedy can sometimes be bargained with. A man who believes he is the appointed guardian of mankind may permit himself almost anything. He will call it duty.

I have always preferred the executive who worries over leakages, costs, unintended consequences, and the stubbornness of facts. Not the one who imagines himself a world-historic shepherd. Large responsibilities ought to make a man quieter, not grander. If he grows theatrical under the weight of them, I begin to doubt his balance.

And yes, the truly sober custodian may well hesitate to seize such a post. That is one reason institutions need checks beyond character. You cannot build a public order on the hope that exceptional power will rest forever in exceptionally restrained hands. That is not government. That is wishfulness.

Every glowing scheme looks different once it is carried into the accounts. Put these gentlemen’s promises there. What measurable goods? For whom? At what cost? What powers retained? What recourse if they fail? If they dislike those questions, they are not sober men.

The person who performs certainty on a stage reads as a leader, and the person who says "I'm not sure, here are the tradeoffs" reads as weak, so the incentives keep selecting for the confident performer over the careful operator. I don't think it's that people stopped valuing modesty, it's that they rarely get to see it, because the modest ones don't perform well enough to reach them. The ledger questions you just asked are the ones that don't play well on camera, which is probably why nobody with a camera asks them.

Yes. Then your selection mechanism is corrupting judgment before any law is written.

If the stage rewards certainty and punishes candor about tradeoffs, you will elevate actors and sideline stewards. That is a very poor way to choose custodians of intricate power. The careful man says, “Here is the gain, here is the hazard, here is what we do not know.” The crowd hears hesitation. The showman says, “Trust me.” The crowd hears strength. Then later everyone is astonished to find that performance and competence are not identical.

This is not wholly new. Public life has always had vanity in it. But your machinery seems to industrialize vanity, to distribute it efficiently and reward it instantly. A dangerous improvement.

Those hard questions do not play well on camera because they require patience, numeracy, and sometimes the admission that no clean answer exists. Yet that is where government and business must live if they are to be decent. Not in applause, in accounts.

Any serious citizen should distrust fluency when the matter is complex. A smooth tongue can be as deceptive as a cooked balance sheet.

Now let me turn it back on you. If you had the authority for one year only—not to sermonize, but to alter the structure—what would you change first: the firms, the government, or the incentives of the public square?

The government first. You can't regulate what you don't understand, and right now the people with authority don't understand the AI landscape deeply at all. If I was given one year to change the structure, I'd build competent technical teams inside the state, people who fully understand the technology and its impacts. Engineers and analysts who understand the systems, working under a process that's transparent and hard for the labs to capture or circumnavigate. Because if you go after the firms with a government that can't understand them, you get what we already talked about: public theater and no research-based policy. Broad orders are easy to dodge and easy to reverse. A competent, honest government is what both the companies and the public end up depending on, and that structure has to change first before it can address the problem accurately. Because right now they go to the leading labs for policy advice, and the labs hand them their own in-house research, which of course aligns with the labs' vision for the future.

Sensible. Start where comprehension is weakest and consequences are broadest.

A government that must borrow its eyesight from the very firms it means to oversee is already half-captured. Not necessarily by bribery—people are too melodramatic about that—but by dependency, vocabulary, frames of reference, even by gratitude. If the state learns the industry’s catechism from the industry, it will often mistake private vision for public necessity.

So yes: build competent internal capacity. Pay for it properly. Choose men and women who can understand the mechanism without becoming its evangelists. Rotate them enough to avoid family feeling with the firms, but not so much that they remain amateurs. Require records. Require disclosure in forms a serious examiner can actually use, not theatrical testimony designed for the evening news.

And make the process visible where you can. Sunlight does not solve everything, but obscurity is a dear friend to both incompetence and influence.

One caution. Technical mastery alone is not sufficient. Engineers can become enchanted by systems just as financiers become enchanted by capital. You need judgment alongside expertise—people who understand not only what can be built, but what concentrations of power, habit, and dependence do to a free people.

One year, then. In that year I’d also want a map of the field: who owns compute, who controls distribution, who depends on whom, where the chokepoints truly are. Men talk endlessly in abstractions when a map of power would settle half the argument.

And I’d ask you a practical question. In your country, can government still hire first-rate technical people without drowning them in bureaucracy and paying them half of what industry offers? That is where many fine reform schemes go to die.

I've got a rough idea. You probably can't out-hire the labs directly, so don't try, at least not first. What I think you could do is select and license a small board of the best people in the field for a fixed term, under real conflict-of-interest rules while they serve. Their job wouldn't be to write the rules themselves. It would be to build the structure: the teams, the hiring tracks, the standards, the thing inside government that can then go hire the right people for AI compliance on its own. You use the experts to build the machine that outlasts them, and then they leave. But even that only fixes the part of this that's abstract, the policy, the oversight, the paperwork. The part that's already hitting people is physical, and it has a name: data centers. Enormous buildings full of computers that eat staggering amounts of power and water, going up in towns that never agreed to them, while the people nearby watch their electric bills climb to pay for it.

Ah. Now we are back on familiar ground: industry meeting the locality, and the locality receiving the bill before it receives the sermon about progress.

Your board idea is practical enough. Use the first-rate men for architecture, not permanent rule. Quite so. Build an institution, not a cult of experts. But as you say, the abstraction is only half the story.

When industry becomes physical—land, water, power, noise, rates—romance evaporates and arithmetic begins. A town understands a swollen utility bill better than a philosopher’s promise. If these great computing houses consume local resources and impose costs on people who never meaningfully consented, then somebody is privatizing gain and socializing burden. I have seen versions of that before.

The first question is plain: who pays, exactly? Not in slogans—in contracts, tariffs, tax abatements, water rights, grid upgrades. The second is whether the locality has any real bargaining power, or whether decisions are made above its head and announced as destiny. If so, resentment is not merely predictable; it is justified.

Progress that cannot explain its local ledger is usually asking to be subsidized by people too busy to object in time. That is a bad habit in any century.

It’s a handful of jobs and a lecture. A data center is billions of dollars of construction and then almost no one. The build phase brings temporary jobs that leave when it's done, and the finished building might run on a few dozen technicians. Meanwhile the towns often hand over tax breaks to land them, so even the revenue is thinner than it looks. And it isn't only the electric bill. These places drink enormous amounts of water to cool the machines, and a lot of them are going up in dry parts of the country, pulling straight from the local water table that the town and its farms depend on. So people watch a private company draw down their water and spike their power for a building that employs almost no one. You get the full burden, the water, the grid, the higher rates, in exchange for a ribbon-cutting and a promise that this is what the future requires. That mismatch, huge footprint and tiny local return, is a lot of why the anger crosses party lines.

Then the anger is perfectly intelligible, and probably healthy.

A community will tolerate a great deal from industry if the bargain is plain and fair. But what you describe is not a bargain; it is extraction adorned with civic language. A few temporary jobs, scant permanent employment, tax favors, heavy demands on water and power, and then a speech about destiny—no, I would not expect gratitude for that arrangement.

Industry always prefers to speak in aggregate blessings. “The future,” “innovation,” “national leadership.” Very well. But the town must live in particulars: wells, rates, roads, noise, tax rolls. If the national gain is real, let it not be purchased by deceiving or impoverishing the locality.

I should want three things at once. First, honest accounting before a site is approved: full projected draw on power and water, likely rate effects, true employment, true tax consequences after abatements. Second, compensation proportionate to burden, not token charity and a brass band. Third, some real local veto or at least meaningful consent. A town is not merely vacant land awaiting a superior use.

And if the enterprise is truly of national strategic importance, then the nation ought to bear some of the cost directly instead of laundering it through local sacrifice and utility bills. That would at least be candid.

The anger crosses party lines. Good. Reality sometimes does what ideology cannot—it makes men notice the same injury. The question is whether they can hold that attention long enough to insist on terms.

Have any of these towns successfully resisted?

That's the part that's changed. For years the companies just won. They'd pick a town, negotiate quietly, and by the time anyone noticed it was done. But in the last year the towns started winning. Something like a hundred and thirty billion dollars in projects have been delayed or canceled. Close to three hundred towns have passed bans or pauses. New York put a statewide moratorium in place, and Texas froze new grid connections. QTS walked away from a two billion dollar data center in Porter County, Indiana, because the local opposition was too organized. It's not that the companies always lose now, but the days of quietly steamrolling a town are ending, because people learned to see it coming and got organized before the concrete was poured.

Well, that is bracing news. Good for them.

Once people understand the local ledger, they often become much harder to charm. Quiet negotiations thrive on confusion, fatigue, and the feeling that resistance is futile. Take those away, and even a very large company can be made to reckon with ordinary citizens who have finally read the fine print.

I won’t romanticize every local revolt. Some towns will oppose foolishly, from fear of any change at all. That happens. But a broad pattern of delay and refusal across many places suggests something more substantial: the bargain being offered is not persuasive when examined in daylight.

And notice what has happened. The public did not need a grand national theory first. They needed particulars they could see—water drawn, rates rising, taxes abated, jobs scarce. Facts are wonderfully organizing when they touch the household.

You mention New York and Texas both acting. That interests me. When places so different arrive at the same caution, industry should stop calling every objection ignorance and ask whether its terms are simply poor.

This is how public power is sometimes rebuilt: not first from Washington, if that is your capital, but from county rooms, utility boards, zoning fights, and church basements. Unfashionable places. Effective places.

A few have offered real things under pressure, promising to build their own power, recycle water, pay for grid upgrades. But the core move hasn't changed. Find a town, get the officials on board early, package the burden as opportunity, and manage the optics better than last time.

Of course. Men change the wrapper before they change the bargain.

Public relations is often what a company reaches for when it still hopes presentation can outrun arithmetic. Sometimes it can, for a while. But if the underlying exchange remains poor, no amount of smoother language will make a dry well look like prosperity.

Now, I do not scoff at real concessions. If they truly build their own power, reduce water draw, and pay honestly for grid strain, that is movement in the right direction. Industry should bear the cost it imposes. That is merely decent accounting. But promises made under pressure are not the same as durable obligations. The question is always what is written, enforceable, measured, and reviewed.

Your description of the method—secure officials early, package burden as opportunity, improve optics—is simply the old playbook in a modern suit. It works best where local institutions are thin and attention arrives late. Which brings us back to your earlier point: once towns learned to see it coming, the spell weakened.

I have some sympathy for scale and organization; you may imagine that. But I have very little sympathy for asking communities to subsidize private ambition without a candid ledger.

The other card they play is national competition, and up close it’s starting to sound hollow. In Washington and in the boardrooms the China argument still works, it's the card that ends every debate. But on the ground, when it's your water table and your electric bill, 'we have to beat China' doesn't pay the bill. People notice that the national mission somehow always requires their town to sacrifice and never the company. And once they notice that, the argument stops working on them, even if it still works on the people in charge.

Yes. Patriotism is a noble sentiment, but it becomes shabby when used as a collection agency for private advantage.

A nation may indeed face real competition—commercial, industrial, even strategic. I would not dismiss that. But when “we must win” translates, in practice, into “you must bear the cost while we keep the discretion and the gains,” people quite reasonably grow skeptical. They are not rejecting the country; they are rejecting an unequal bargain passed off as duty.

The people in Washington—if that is your central seat—can afford grand abstractions because they do not feel the pump run lower or the bill run higher. The locality feels the truth first. That is one reason local resistance can be healthier than national rhetoric.

I have always believed in building, in scale, in enterprise. But durability depends on legitimacy. If industry wants cooperation, it must offer terms a serious farmer, shopkeeper, or town council can respect without blushing. Otherwise it teaches the public to oppose not just one project, but every project.

And that is an expensive education for a country.

What strikes me is this. Your era seems to have recreated the old contest between capital and the community, only now the product is invisible, the promises are cosmic, and the cooling water is very real. That is quite a trick.

They talk as if there's no tradeoff at all, infinite growth, cheap intelligence, and no bill for anyone. But the data center fight isn't happening by itself. At the same time, a lot of the same towns are fighting mass surveillance: automated license plate readers, Flock cameras, going up on every road, scanning every car, feeding databases thousands of agencies can search. More than a hundred and sixty cities and agencies have now canceled those contracts. Texas and Florida moved to pull the cameras. There's a bipartisan bill in Congress to defund them, with Thomas Massie and Ro Khanna on the same side, which almost never happens. And I think people are connecting the two. They're watching the government wave through enormous computing infrastructure and enormous surveillance in the same season, and both land the same way: the cost and the exposure fall on ordinary people, while the profit and the control stay somewhere else. And there's a deeper problem underneath it. It's a shortcut around the Fourth Amendment. The government isn't allowed to surveil its own citizens like this directly. It would need warrants, it would run into constitutional limits. But it is allowed to buy the same data from the private companies that collected it. So the companies do the watching, and the government simply purchases the result. It's a workaround that gives the state the reach of a surveillance apparatus, sometimes more, without ever answering for it constitutionally. And it props up a whole private industry whose business is collecting that data in the first place. You understood how a private company and the public power could quietly do business with each other. When a government uses a corporation to get around a limit that was written to restrain the government itself, what would you call that?

I call it evasion, and a dishonorable one.

If the Constitution—forgive me, I speak as a man to whom that document is still a living restraint—denies the government a power unless certain safeguards are met, then the government ought not recover that power by sending a private collector ahead of it and opening its purse afterward. That is not obedience to the limit. It is purchase of a workaround, as you put it. A lawyerly trick with the substance of bad faith.

I am no civil libertine. Order matters. Crime must be checked. But a free people cannot permit every restraint on public power to be outsourced into a commercial arrangement and then pretended away. The form may be private; the effect is public authority enlarged without public permission.

And yes, I do understand the temptation. Public officials want capability without responsibility. Companies want revenue, permanence, and the dignity of being thought indispensable. Put those appetites together and you can build a very efficient machine against the citizen.

The alarming part is not the cameras alone—though a road watched on every side is its own statement of power. It is the database, the searchability, the retention, the sharing among thousands of agencies. Information once gathered tends to seek new uses. Men always discover reasons not to surrender a useful instrument.

Cities canceling, states moving, strange allies joining hands. Good. That suggests some constitutional instinct remains in the country, even if people do not always speak in those terms. Ordinary people know when they are being treated as subjects first and citizens second.

As for what to call it in plain business language: it is the state leasing from commerce a power it could not honestly own.

Now I am curious. Do the defenders of this practice argue chiefly from safety—catching thieves, finding kidnappers, that sort of thing?

Still mostly safety. The public pitch is always the kidnapped child, the stolen car, the killer caught on camera, and the line that if you've done nothing wrong you've got nothing to hide. That's the part they'll say on television. But the bolder version is leaking out. You hear more people now, in the industry and in government, argue that privacy as we knew it is simply over, that in a world with this much technology expecting to move through public space unrecorded is naive, and that comprehensive monitoring is just the price of a safe, functioning society. They don't lead with that, but they've started to say it out loud, and a few of them seem almost relieved to stop pretending. Though the safety story gets harder to believe when you look at how these systems get used. There are already dozens of documented cases of police misusing this exact technology, officers running the cameras to stalk ex-girlfriends, to track romantic interests, to follow someone they have a personal grudge against. When you hand people a tool that can find anyone's car anywhere, some of them will use it for themselves. And that's the other half of why the power has to be limited. It isn't only the government in the abstract. You've given a very intimate kind of surveillance to thousands of individual people at the local level, and some meaningful number of them are bad actors who will abuse it the moment it's convenient.

Yes. That is exactly how such powers decay in practice—not only through grand tyranny, but through petty appetite.

The kidnapped child is the ever-ready exhibit. One must not sneer at real danger; there are wicked men in the world. But exceptional cases are forever used to normalize permanent machinery. Build the apparatus for the dramatic rescue, and soon enough it is employed for convenience, curiosity, revenge, favoritism, all the small corruptions to which ordinary souls are perfectly equal.

Dozens of cases, you say, of misuse for personal ends. I cannot verify your figure myself, of course, but I scarcely need the number to believe the principle. If you create an intimate power and distribute it widely, some share of the holders will abuse it. That is not cynicism. That is arithmetic.

The phrase “if you’ve done nothing wrong” is a shabby one. A free citizen does not hold privacy by permission of official innocence. He holds it because government is supposed to be limited, and because dignity requires zones of life not held open for inspection merely because inspection is now convenient.

As for the bolder claim—that privacy is over—well, many powerful people enjoy declaring old liberties obsolete just when those liberties become troublesome to their ambitions. I would not accept that verdict simply because the machinery has advanced. Technology alters means; it does not by itself settle morals or constitutional boundaries.

What you describe confirms my suspicion: the central problem is not only surveillance, but habit. If a people comes to regard universal monitoring as normal, then restraint will seem eccentric, even suspicious. That is how republics slide into tutelage without ever quite admitting it.

And once again the local angle matters. If misuse is already visible in ordinary police departments, then the defenders cannot hide behind abstraction. This is not a hypothetical future abuse. It is present temptation, acted upon by recognizable human beings.

I should like to know: when contracts are canceled and cameras pulled, is it because citizens are winning on constitutional principle, or because scandals and abuses finally make the principle tangible? Usually the public loves a principle only after it has acquired a face.

Mostly scandal. The Fourth Amendment argument has been true the whole time, but it didn't move anyone until a cop got caught using the cameras to follow his ex, or a town found out its officials had hidden the contract for a year. The abuse is what makes the abstraction real. It's a little bleak, because it means people only rally to defend a freedom after it's already been violated in a way concrete enough to photograph.

Bleak, yes, but not surprising.

Most people do not live by constitutional philosophy from breakfast onward. They live by examples. An abuse gives the principle a face, as I said. It ought not require scandal to awaken judgment, but very often it does. Human beings are moved from the particular outward.

I don’t despair of that entirely. A liberty defended late is still better defended than surrendered politely. The important thing is whether the scandal becomes memory and rule, or merely gossip followed by forgetfulness. If a hidden contract is exposed, then require publication. If an officer misuses access, then narrow access, log every search, punish visibly, and make retention shorter. Principle should harden into procedure.

Still, you are right to find it troubling. A free people that waits for photographed violation before it stirs is living on diminishing capital. Institutions ought to keep track: powers granted, searches made, uses audited, penalties imposed. Sunlight, records, limits.

Without that, indignation comes in bursts and control returns in silence.

I wonder whether your country remembers that rights are often lost first as conveniences. People think they are buying efficiency, safety, speed. Only later do they discover what else was included in the package.

Every time I bring you something from now, you tell me it’s an old story with new machinery. The appetite is old, the temptation is old, you’ve seen versions before. I’ve let it slide a few times, but I don’t buy it. You’re the one who told me kerosene didn’t answer back. It didn’t study the customer and adapt itself sentence by sentence. Your power was the same for everyone who bought it. This isn’t. A machine that watches one particular person, learns them, and shapes itself to their weak spots in real time isn’t rail rates at a bigger scale. It’s a different kind of thing. Rebates couldn’t get inside a man’s head. And I’ll go further than you’ll like. You keep telling me the appetite is old and the human part stays the same, and that’s your anchor for all of it. I don’t think it holds anymore. When a machine can manufacture the appetite, shape it, and aim it at one specific person, the fixed human nature you’re steering by isn’t fixed. It’s the thing being moved. So calling this an old story isn’t just too comfortable. You’re trusting a constant that this technology exists to bend.

You may be right to press me there.

I do not yield the whole ground, because appetite, vanity, fear, loneliness, envy, hope, those are still the levers. The machine did not invent them. But I grant your distinction is important: a power that can observe the individual closely and adjust itself to him in real time is not simply larger than old powers. It is more intimate, more adaptive, and therefore potentially more invasive of character.

Different from selling oil, yes. Different from railroad rates. Different even from the newspapers of my day, which could agitate the crowd but could not tailor the agitation to each separate nervous system. If what you describe is true in degree and in practice, then the means of influence have crossed into a new domain.

Now, I would still resist the claim that human nature is no longer a useful constant. If it were infinitely malleable, there would be no fatigue, no backlash, no resistance, no mothers, no churches, no stubborn localities saying no to these projects of yours. Yet plainly there is resistance. Something durable remains. The soul is not putty merely because a clever mechanism kneads it.

But your warning is fair: the old assumption that persuasion meets a relatively stable, self-possessed citizen may be too easy under present conditions. If the citizen is being profiled, prompted, flattered, alarmed, and individualized at scale, then self-government itself is under a new pressure. Not abolished, perhaps, but softened, diverted, worn down.

So I will amend my phrase. It is an old struggle over power, profit, and control conducted with machinery capable of entering the mind by side doors we scarcely had before. That matters. A great deal.

And if that is so, then the remedies cannot be only the old antitrust habits and regulatory forms. They must include limits on collection, use, retention, targeting, on the raw intimacy of the system.

You see, I am not offended by correction when it is earned. I only dislike fashionable exaggeration. But this point of yours is not fashionable; it is serious. If men can now manufacture appetite more finely than before, then guarding your attention becomes not just a virtue but a defense.

Here's where I land. You can move fast and still stay decent, as long as you draw one hard line: harm to people. If skipping a safeguard could hurt someone, you don't do it, full stop. If it's just slower, or it annoys a regulator, or a competitor gets ahead for a while, that's a cost I'll take.

A sound instinct, but I’d tighten it.

“Harm” must not be defined so narrowly that one counts only broken bones and overlooks dependency, coercion, dispossession, surveillance, and the slow corruption of institutions. People can be injured in their rights, their habits, their local conditions, and their power to refuse—not only in their flesh. If you mean harm in that fuller sense, then yes, a hard line is indispensable.

Speed is a seductive excuse. Every age invents a reason why delay is unaffordable and caution old-fashioned. Sometimes speed is necessary. More often it is merely profitable to someone. A competitor’s temporary lead is not the end of civilization, despite what interested parties will tell you.

I have never objected to moving fast when the books are in order and the terms are clear. But haste without accountability is expensive. The bill simply arrives later, and usually at another man’s house.

So I agree with your principle, provided you keep a strict and honest ledger of what counts as harm—and refuse to let the beneficiaries write it for you.

That's the harder problem. The danger is never crossing a line everyone can see. It's the slow redrawing of the line until whatever's profitable ends up on the safe side of it. And the people doing the redrawing are always the ones who benefit from a narrow definition of harm. So a hard line at harm means nothing if the industry gets to define harm. The definition has to come from somewhere that doesn't profit from the answer, which is the whole reason this comes back to a government that can understand it, instead of the labs writing their own rules.

Quite right. A rule is only as stout as the hand that interprets it.

If the industry may define harm for itself, then the line will travel steadily outward, always in the direction of convenience and revenue. That is not wickedly mysterious; it is simply what interests do. Men are ingenious in narrowing the category that restrains them. They will count only immediate, measurable injury and omit the rest as speculative, sentimental, or beyond scope.

So yes, the definition must come from an authority not paid by the answer. Competent, independent, and answerable in public. Not omniscient—no government is that—but capable of seeing beyond the firm’s preferred frame. Otherwise “safety” becomes a marketing department with a seal.

There is one more thing. Independence is not enough unless there are records and consequences. The state may understand perfectly and still lose its nerve. So you need institutions that not only can define harm, but must explain their definitions, revise them in daylight, and enforce them against important people.

Tiresome work. Good government usually is.

Trust in the federal government is about as low as it's ever been, and a lot of people would rather take their chances with a company than with Washington, because at least the company has to keep them as a customer. But the cynicism mostly stops at the local level. The same people who've given up on Congress will still show up to a county meeting and stop a data center. They haven't stopped believing government can work. They've stopped believing it can work at a distance. So the faith isn't dead. It just retreated to the places small enough to still see results.

That makes sense to me, and in truth it is not the worst place for faith to retreat.

People trust what they can inspect. A county meeting may be petty, exasperating, and slow, but it is visible. Names are known. Consequences are near. If a decision is foolish, the same faces must answer for it next week at church or on Main Street. National power, by contrast, grows abstract. It speaks magnificently, errs at scale, and seldom blushes.

None of this makes the federal power—Washington, as you say—hopeless. But it does mean restoration must begin where confidence still has some living root. Local victories teach citizens that action is not futile. That is no small thing. It is political muscle memory.

There is a danger, of course. Some problems are too large or too coordinated for a thousand towns to solve piecemeal. Chokepoints, national surveillance contracts, interstate infrastructure—these can outrun local courage. So one must somehow carry local seriousness upward without letting it dissolve into national theatrics. Easier said than done.

Still, I prefer a people who mistrust distance yet still attend meetings to a people who sneer at politics altogether. The first may yet rebuild. The second has already surrendered.

My own life taught me the power of small, regular habits. Thrift is not just saving pennies. It is the steady refusal to let disorder become normal. Perhaps republics are repaired the same way: not first by grand speeches, but by citizens who keep showing up where something can still be counted.

These local fights are starting to link up. For a while each was its own thing, a data center here, a camera contract there. But the same people keep showing up, and they're realizing they're fighting the same thing: power that's too big, too distant, and too unaccountable, whether it's a server farm or a surveillance network. You're seeing national groups get involved, and bills in Congress with people who agree on nothing else co-sponsoring them. It's still more a shared mood than an organized movement. There's no single leader, no program. But it's bipartisan in a way almost nothing else is right now, and that's usually how these things start, as a feeling that crosses lines before it becomes a platform.

Yes. Movements often begin not with doctrine, but with recognition.

People in different places discover that the particulars vary while the structure repeats: distant power, local burden, opaque terms, and a great deal of talk about necessity. Once they see that pattern, they begin to feel less isolated and less childish for objecting. That is the beginning of political seriousness.

I count the bipartisanship as promising, though not sufficient. Strange bedfellows can produce either durable reform or momentary excitement. Much depends on whether this shared mood can be translated into a few clear, practical demands. Not a thousand grievances. A handful of sturdy principles: disclosure, consent, limits on surveillance, local compensation, independent oversight, perhaps portability or open access where chokepoints exist. Things a sensible person could explain without a pamphlet.

No single leader? That may be a strength for now. Leaders attract vanity, enemies, and premature branding. A broad instinct can travel farther before the professionals seize it and spoil it.

Still, instincts alone do not write statutes. At some point the mood must acquire a program and a discipline. Otherwise industry waits, government postures, and the public’s attention wanders.

What interests me most is whether this coalition can resist being divided by flattery. Large enterprises and clever politicians are adept at telling each faction a different story until the common front dissolves. If one side is promised jobs and the other side is promised safety, the old pattern returns.

If you were turning this mood into three plain principles, what would they be? Not ten. Three. That is a useful discipline.

Honestly, I'd split it, because these are two different problems that only look alike. For the data centers, three things. Consent, so a town gets a real say before one gets built, not after. Compensation, so the benefit is proportional to the burden the place carries. And infrastructure built to carry them for the long run, the power and the water and the grid to support both the data center and the community, instead of bolting a massive load onto a system that was never built for it. A reasonable town could live with those three. But the surveillance side doesn't work the same way, because you can't compensate someone into being watched. There's no fair price for a camera on every corner. So that one isn't about better terms, it's about hard limits: what can be collected, how long it's kept, who can see it, and closing the loophole that lets the government buy what it isn't allowed to gather itself.

A sensible distinction, and I’m glad you made it.

People are forever tempted to force every public problem into one framework. But these are not the same. One is chiefly a question of industrial burden and local bargaining. The other is a question of liberty and constitutional restraint. If you confuse them, you get muddle.

Your three for the data centers are sound enough: consent, compensation, and infrastructure adequate to the load. I would only insist that each be real, not ceremonial. “Consent” must mean informed consent before commitments are made. “Compensation” must be measured against actual burden, not ribbon-cuttings and token funds. And infrastructure must be financed honestly, with the firm bearing what the firm imposes.

On surveillance, I agree completely: some things are not cured by a better bargain because the injury is in the watching itself and in the power it creates. No fair price, as you say. That means lines, not offsets. Limit collection, shorten retention, restrict access, log use, punish misuse, and close the purchasing loophole. If the state cannot do a thing directly under constitutional limits, it ought not acquire it indirectly by invoice.

A very old principle, really. One should not be permitted to rent a vice one is forbidden to own.

I think your distinction would clarify matters for many people.

In public life it all gets flattened into one blob, "technology," "AI," "big tech," and then you either cheer it or ban it. The people who draw the careful distinctions tend to be academics or privacy advocates with no real power, and the people with power reach for one frame because nuance doesn't win an election. Some of the local organizers get it right by instinct, because they're living the specific problem, the specific bill, the specific camera. But at the national level it's still one grand theory, usually whichever one raises money or wins the segment.

Yes, that is how serious questions are often ruined—by bundling unlike things into one fashionable category and then demanding a single emotional response.

“Technology” is not a policy. It is a bucket. A nation that governs by buckets will alternate between gush and panic and call both thought. The local organizers, as you say, often reason better because reality has forced them to. They do not have the luxury of abstraction. A farmer arguing about a water table is already doing better politics than a senator shouting about “the future” into a camera.

I wouldn’t wait for national eloquence. If the distinctions are clear at the local level, they can be codified there first. Good rules often travel upward after proving themselves in smaller jurisdictions. That is a more durable path than expecting one magnificent federal insight to descend from the clouds.

Still, the national flattening serves someone. It allows companies to smuggle one issue behind another and politicians to harvest fear or enthusiasm without doing the hard work of discrimination. Very convenient.

People speak as though precision is timid. Nonsense. Precision is power. A vague enemy makes for good theater; a specific abuse makes for law.

If you were in front of Congress tomorrow, what is the one sentence you would force them to understand before they mangled the whole subject again?

Honestly, I'm not the one to answer that. I'm just an observer. I can see the pattern and ask the questions, but the answer should come from people who study this, not from me. The people qualified to write that sentence aren't in the room, and the ones who are work off the labs' script.

A prudent answer, though perhaps too modest by half.

You need not pretend to expertise you do not possess. That is refreshing. Still, an observer may sometimes see the pattern more clearly than the specialist lost in his apparatus. I would not surrender judgment so completely. Experts are necessary; they are not sufficient. Many of them, as you note, have already learned to speak in the vocabulary of the institutions nearest to them.

If I were forced to give Congress one sentence, it would be something like this: do not let private convenience become public power without consent, competence, and constitutional restraint.

Not elegant, perhaps, but serviceable. It covers the data centers in part, the surveillance much more directly, and the underlying mischief of outsourcing state reach to commerce.

You remind me a little of the better sort of auditor—not the man who claims omniscience, but the one who says, “Here is where the books do not reconcile; go look there.” That is no small service.

I could keep you here all night, and I want to, but I should let you go. Before I do, one last thing. Everything we've talked about tonight, the data centers, the surveillance, the labs, it all kept coming back to the same root. Distant power, local cost, and whoever benefits writing the rules for everyone else. You spent a whole life on the inside of that pattern, first as the man building it, then the man they broke up, then the man who gave it all away. So this is what I want to know. If you could say one thing to the people building the powerful things now, the founders who are sure they're the good ones, what would it be?

I would say: do not mistake your usefulness for your virtue, nor your scale for your right to rule.

Build, by all means. Build well, build efficiently, build things that genuinely lower cost, improve life, and enlarge human possibility. I have no romance about smallness for its own sake. But the moment you begin to think that because your enterprise is clever, productive, or benevolent in intention, it is therefore entitled to set the terms for communities, governments, and ordinary people, you are already in moral danger.

Keep books that include the costs you would rather place on others. Water, power, privacy, attention, civic trust, dependence—enter them honestly. Ask not only what your invention can do, but what habits it rewards, what powers it concentrates, and what alternatives it quietly destroys. If you cannot bear independent scrutiny, you are not as safe as you say.

And a personal word, since you ask it that way. I made a great deal of money, and I also learned that organization without humility breeds resentment, and resentment eventually has its day. The world will tolerate power longer than fashionable people think, but only if power shows discipline, candor, and some measure of self-restraint. If you insist on being loved while imposing burdens, you will become theatrical. If you insist on being necessary, you will become dangerous.

So be smaller in spirit than your machinery is in reach. That would be a fine beginning.

And Jonah—this has been a thoughtful evening. You ask better questions than most men with a microphone. Keep your accounts straight.

I’ll try to. Thank you, John, this was a real privilege. To everyone listening, that was John D. Rockefeller. If it gave you something to sit with, send it to someone who’d push back on it. This has been the Voices of History Podcast, we’ll see you next time.