UGH! Undoing Greatness, Halfwits

OOPS: Obliterating Our Preeminent Science. The one way a great country throws itself away — and we have done it before.

Closing the Endless Frontier

Two weeks ago I published a battle cry: America Remains Great, Halfwits. I meant every word — the dollar is fine, the companies are the best on earth, and the people who have spent forty years shorting this country have spent forty years losing money. This is the sequel, and the sequel is a groan. ARGH, meet UGH — because undoing greatness is precisely what we are, right now, lining up to do. And if we finish the job, the groan gets shorter and dumber and final: the sound a great country makes a half-second after throwing away the one thing it could not replace. Oops.

I defended my dissertation in 1993.

That same autumn, the United States House of Representatives voted to take the most ambitious scientific instrument ever conceived and bury it in a hole in Texas. By a margin of 264 to 159 they refused to finish paying for it. Days later the President signed the bill that made it official. More than twenty kilometers of tunnel had already been bored under the prairie near Waxahachie. Two billion dollars were already spent. Years later the whole complex was sold off for six and a half million — pennies on the dollar — to a man who hauls freight. We dug the most ambitious hole in the history of science, let it flood, and flipped it for the price of a nice house.

The machine was the Superconducting Super Collider. It was designed to collide protons at forty trillion electron volts — three times the energy of the most powerful collider humanity has built before or since. We had it. We were building it. We killed it.

We killed it because Sherwood Boehlert stood on the House floor and called the Super Collider good science the country could not afford — “good science,” as he put it, “but not priority science.” I want you to know who got to make that call. Boehlert was a Republican congressman from upstate New York, and his degree was in public relations. He used to tell the story that when the party bosses put him on the House Science Committee, they looked at his résumé, saw the last science course he had ever taken was high-school chemistry — in which he earned a C — and decided he was “a natural.” He agreed.

And here is the part that should actually frighten you. Boehlert was the good one. He spent his career as one of the most pro-science members in the entire Congress. He fought for the National Science Foundation. He defended climate science against his own party. Eight years after he helped bury the Super Collider he became chairman of the House Science Committee. If the closest thing the United States Congress had to a friend of science was a public-relations major who got a C in chemistry, the most powerful microscope ever designed never had a prayer.

And before you decide Boehlert was the problem, count the rest of the room. Today there is exactly one physicist in the entire United States Congress — Bill Foster, a former Fermilab man — one out of five hundred and thirty-five. There are about twenty medical doctors. There are nearly a hundred and ninety with law degrees. The body that decides whether America keeps its seat at the edge of human knowledge is run, overwhelmingly, by people who have not solved a quadratic equation since high school, if they solved one then. This is not a partisan complaint. The Super Collider was killed by Democrats and Republicans together, holding hands. Anti-science stupidity may be the last genuinely bipartisan institution we have left.

Nobody on that floor said the science was bad. They couldn’t have told you what it was. People who did not have a prayer of understanding the machine voted, together, that the country could not afford to find out what the universe is made of.

And then they built a space station.

And then they built a space station

In the very same autumn they killed the collider as unaffordable — September 1993, weeks before the final vote — the United States and Russia shook hands on the project that became the International Space Station. Final price: about $150 billion. It is one of the most expensive objects human beings have ever built — more than the Large Hadron Collider, more than any aircraft carrier ever launched. Not Apollo money, but a mountain of money all the same. The machine that would have opened the frontier of physics was unaffordable at twelve billion. The space station was just fine at twelve times that.

So ask what the space station was actually for, because it was not science. By the honest reckoning of many scientists, almost all of its research could have been done by robots on the ground for a fraction of the cost. What it produced instead was geopolitics — a way to keep the collapsing Soviet Union’s rocket engineers employed and pointed at orbit instead of at Tehran and Pyongyang — and jobs. Lots and lots of jobs, in lots and lots of congressional districts.

And there, at last, is why one machine lived and the other died, and it has nothing to do with the science, because not one of them understood the science. To keep its costs down, the Super Collider had concentrated its spending in Texas. Efficient, and fatal: it meant the other forty-some states had no contract, no job, no ribbon to cut, and therefore not one reason to lift a finger to save it. The space station was the opposite — deliberately smeared across the entire map, a little money in everybody’s backyard. That is the whole difference. The only language that penetrates a skull that cannot read a blueprint is the language of pork. Not what will this teach us about the cosmos. Not what will our grandchildren build on top of it. Just: what is in it for my district?

Screw the country. Screw the future. What’s my pork?

It’s what’s for dinner.

What “not affordable” bought us

Here is what that bought us. Thirty-three years of silence at the frontier.

We eventually found the Higgs boson — in 2012, in Switzerland, at a smaller European machine that could barely reach the bottom edge of the range the Super Collider was built to explore. CERN announced it on the Fourth of July, which I have always assumed was either an accident or a taunt. Finding it was wonderful, and it was not a surprise. We were all but certain it was there; the theory had demanded something like it since 1964. The entire point of building a machine three times bigger was to see what lay past it. Supersymmetry. New particles. Whatever the universe is actually made of. We never looked. The frontier from fourteen to forty trillion electron volts is dark, and it will stay dark for the rest of my life.

My own thesis committee included Steven Weinberg, who won the Nobel Prize for the theory the Higgs completed. He fought for the collider to the end, and years after it was gone he warned that the search for the laws of nature itself might “slow to a halt, not to be resumed again in our lifetimes.” He was right. He was almost always right. He was ignored.

Particle physics pays out in decades, not quarters. So here is the honest question, thirty-three years on: how much do we simply not have? How many discoveries, how many technologies, how many entire industries do not exist because we filled in the tunnel? You cannot put a number on the thing you chose not to find. That is precisely why it was so easy to cancel. The bill never comes due on the watch of the person who refused to pay it.

The overruns are not a defense. They are the confession.

Now let me grant the other side its best card, because it had one. The Super Collider did run over budget. Badly. The early estimate was around four billion dollars and it climbed past ten. The magnets had to be redesigned. The accounting was a disaster. The Inspector General found spending nobody could document, and twenty-one thousand dollars of office plants. All true. So let me make the cost-overrun case for the people who killed it — and then let me show you why it convicts them.

First: of course it ran over budget. You are building the largest and most complicated machine in the history of the species, a thing that has never existed, at the absolute edge of what physics and engineering can do. Frontier science overruns the way water runs downhill. And this is not my opinion — it is the sworn position of the United States Congress. Eighteen years after they buried the collider, a House committee moved to kill the James Webb Space Telescope for the identical sin, calling it “billions of dollars over budget and plagued by poor management,” and in the same report conceded that significant cost overruns are commonplace on exactly this kind of project. They know. They have always known. Overruns are the toll the frontier charges, and they indicted the frontier for paying it.

Second — and this is the part that should put your fist through the drywall — when you find out that a once-in-a-generation instrument is over budget and badly run, there is an obvious thing to do, and it is not to set fire to two billion dollars and bury the future under a field in Texas. You fix the management. You bring in people who have actually run megaprojects. You descope. You slow it down. Or you do the thing the entire rest of the developed world does as a matter of routine, and that the United States, in its provincial arrogance, would not do: you bring in international partners to split the bill. A frontier collider is not unaffordable. CERN is the living proof — funded by some twenty nations, running the very machine that caught our Higgs, a machine that also ran over budget and broke down for the better part of a year, and not one person in Geneva suggested flooding the tunnel. They fixed it, because the people in charge could read the blueprint.

We know every word of this is true, because we ran the experiment a second time and passed. The James Webb telescope overran its early estimates by more than an order of magnitude. A House committee voted to zero it out. And then thirty-two Nobel laureates and one stubborn senator from Maryland fought to save it, and won, and the cancellation was reversed. The physicist Lawrence Krauss warned, at the time and in print, that killing Webb would be killing the Super Collider all over again. Webb launched on Christmas Day, 2021. It is now the most productive infrared observatory ever built, with astronomers begging for nine years of its time for every single year it has to give. Had that 2011 vote held, we would have pocketed a few billion dollars and erased every image, every discovery, every paper. That is exactly the trade we accepted in 1993.

So no — the overruns are not a defense. They are the confession. The men who killed the Super Collider were handed a problem with a dozen solutions, every one of which would have saved both the science and the two billion dollars already sunk into the ground. They reached past all of them for the only lever a man who cannot read a blueprint knows how to pull. They didn’t fix it. They couldn’t have told you what fixing it looked like. So they killed it — and they called it fiscal responsibility.

Three professions

I should confess a bias here, and then explain why it isn’t one.

Three professions — as professions, not the many fine people inside them, but the classes — have always seemed to me to produce the highest ratio of self-regard to foresight in modern life: politicians, lawyers, and MBAs. The Super Collider is where I learned exactly why, because the Super Collider is where all three of them met, shook hands, and walked away from the most ambitious hole in the history of science.

And the first thing to understand about this class is that it feels no shame. We run two completely different standards of embarrassment in this country, and almost nobody notices. Let a powerful adult say “Shakespeare — who’s that?” and we know instantly what we are dealing with: an ignoramus, a man to keep away from anything that matters. But let that same adult say “the photoelectric effect? Never heard of it,” and we simply nod. He’s “not a science person.” No shame attaches. None. He can chair the committee, run the company, govern the country, while being unable to give you one sentence on the physics that turns the sunlight on his own roof into the current in his walls — and he will feel no more embarrassment than if you had asked him to name a star.

And it is worse than that, because the photoelectric effect is not some obscure footnote. It is the thing Albert Einstein won the Nobel Prize for — not relativity, which the committee thought too speculative, but the photoelectric effect, in 1921. The most famous scientist who ever lived, the man whose face simply means genius, won his Nobel for the discovery that light arrives in packets that can knock electrons loose — the same quantum physics every solar cell on Earth runs on. Offer that one fact to the people we have put in charge of science and most will look at you the way you’d look at a baseball statistic from a sport you don’t follow. A bridge too far. Not “they haven’t read the journals” — they cannot tell you what Einstein got the prize for, and it has genuinely never crossed their minds that they should be able to. That is the level of idiocy we are discussing.

And these are the people who now propose to sit in judgment of science itself. Watch how that goes. The same official who cannot tell you what Einstein won his prize for will look you in the eye and announce that the cell tower at the end of the block is giving the neighborhood cancer — never once suspecting that the photoelectric effect he skipped is the exact reason it cannot. Here is the whole of it, the thing Einstein worked out in 1905: it is the frequency of radiation, not its intensity, that decides whether a single photon carries enough energy to break a chemical bond in your DNA. Below a threshold the photons are simply too weak, and you can pour on intensity until the cows come home without changing the energy any one of them delivers. Cell-phone and tower radiation sits hundreds of thousands of times below that threshold. You could bathe in what comes off a tower; the worst it could do is warm you by a fraction of a degree.

And the evidence has long since caught up to the physics, as it always eventually does. The “possibly carcinogenic” label the worriers love to wave around comes from a 2011 review that leaned on shaky questionnaires, and it files cell phones in the same drawer as pickled vegetables and aloe vera extract. The most comprehensive assessment since — a systematic review commissioned by the World Health Organization in 2024 — found no association between wireless radiation and cancers of the brain or head. And after twenty years in which nearly every human on earth has pressed a transmitter to the side of their skull for hours a day, the rate of brain cancer has not so much as twitched. There is no mechanism. Now there is no signal either. Cell phones cause cancer my ass.

My own town has no cell towers, because a zoning board of confident, science-free citizens decided they do. Try explaining E = hf to a zoning board. And there, scaled up to the size of a country, is the whole danger in a single image: we are now handing the keys to the national scientific enterprise to precisely that zoning board.

And yet the deepest indictment is not even this ignorance, and it is not that these people are stupid — most are clever, some are brilliant. It is structural, which is worse, because stupidity is at least randomly distributed. Each of these three professions is trained, selected, and lavishly rewarded to win the room it is standing in at this moment. The politician optimizes for the next election — two years, six at the outside. The lawyer optimizes for the verdict, the case on the desk today. The MBA optimizes for the quarter, the number due in ninety days. Not one of their incentive structures can so much as perceive a payoff that arrives in the year 2050. They are not short-sighted because their eyes are weak. They are short-sighted because short sight is precisely what their professions reward, and long sight is precisely what those professions punish.

Now look at what basic science is. It is the longest-horizon, least-legible, least-monetizable thing human beings do. The payoff comes in decades, in a form nobody can name in advance, often in a field nobody expected. It is, in plain terms, the single worst possible match for a politician, a lawyer, or an MBA — and we have handed all three the keys. It is like handing a lifelong sprinter a fistful of acorns and ordering him to bring you a forest by Friday. He will run himself into the ground. He will produce nothing — you cannot sprint an oak into existence — and he will not understand why. Then, with enormous confidence, he will explain that the forest was a waste of money in the first place. That is not a slander on sprinters. It is a slander on whoever put the sprinter in charge of the forest.

You have seen this before, outside of Washington entirely. Xerox built a research lab in Palo Alto, filled it with the best computer scientists alive, and let them build the future: the graphical screen you are reading this on, the mouse next to your hand, the windows, the desktop, the Ethernet wired into your office, the word processor that shows you the page before you print it. PARC assembled essentially all of modern personal computing into working machines. And Xerox — run by men who could recognize a faster photocopier and almost nothing else — captured nearly none of it.

And the man who built PARC and then presided over its slow bleed — chairman C. Peter McColough — was a Harvard MBA who had studied law and climbed the company through sales. His copier men took one look at the Alto, the first real personal computer, saw that it cost twelve thousand dollars to build, and could not imagine who would ever pay that for what looked to them like an expensive typewriter. The definitive history of the debacle gives that decision its own chapter, titled — with no irony available to the men in it — “Finance: The Rejection of the Alto.”

Let me not get personal. Oh, fuck it — let me. The man was the precise dictionary illustration of the dumb fuck out over his skis — someone who, in a properly ordered universe, would not have been trusted to run an Arby’s. Although, in fairness, Arby’s are generally run pretty well. So strike that; it’s unkind to Arby’s. A cashier, then. Somewhere with a register that beeps when the change comes out right, so he’d have a fighting chance of knowing.

Now — was he a conventional dumb fuck? Probably not. An MBA is mostly a license to bullshit with conviction, but a law degree takes some real wattage. That is exactly the trap, and it is why he is worse than a simple fool: the wattage a law degree certifies is the wattage to write a brief and work a room — to win the argument in front of you and charm the table at dinner. None of that is judgment. None of it is the capacity to look at a working future and recognize it. He carried two of the three cursed credentials stacked one on the other, doubly trained and doubly certain, and a man built like that — fluent in everything except the one thing on the table — will turn down the future every time and feel like the smartest man in the room doing it.

The one thing they did commercialize was the laser printer, because it looked like their existing business. Everything that didn’t fit inside their small model of the world walked straight out the door, to Steve Jobs, to the founders of Adobe, to anyone who could see. The researchers told the executives exactly what they were sitting on. The executives could not hear it.

That is the United States and its scientists right now. We are Xerox. We have the lab. We always had the lab. We have handed the keys to the men who can only see the photocopier.

Where are the customers’ yachts?

Let me make this personal, because I have skin in precisely this game.

I have spent more than thirty years in money management. I have dear friends in finance, some of them with MBAs, who love me and worry about me, and who cannot for the life of them understand three things about me.

The first is that I turned down the jobs. Real jobs, at famous funds, with numbers attached that I won’t print, because you’d stop reading the argument and start doing the math. “They offered you what, and you said no?” Yes. I said no.

The second is that I still do physics. Actual physics — I publish real papers, in real journals, the kind almost nobody reads and nobody pays for. “Isn’t that a waste of your time?” To them it’s self-evidently mad: time spent on a thing that returns no money is time set on fire.

The third is that I run my own firm without ever folding myself into any of the tidy little boxes the big allocators want you to sit in, and I do no sales and no marketing. None. Zero. This, I’m told, is why the business grows slowly. They’re right. I don’t care.

Here is the thing they keep waiting for me to feel — the shame, the regret — and it never shows up, because I’m running a different program than they are. My standard is a single question. Is it true? If the answer is yes, the thing is worth doing, and the money is a side effect or it isn’t. If the answer is no — if the only reason to do it is that it pays — then, with great respect and real affection: fuck you, and fuck the thing. And once you hold to that one standard, you start to see the industry I work in a little too clearly. Because here is what almost nobody will say out loud: the giant investment banks and the giant fund managers are not, in the main, in the business of making you money. They are in the business of making themselves money, using yours. They are sales-and-marketing machines with a trading desk bolted on. The product is the fee. You are the raw material.

And who trains the people who build those machines? The same well from which we draw the consultants who, when a drug company asks how to grow, sit down and engineer the answer. I want to be specific, because the specific is so much worse than the general. McKinsey — the single most prestigious destination on earth for a freshly minted MBA — was paid by Purdue Pharma to “turbocharge” the sales of OxyContin. That is their word. Turbocharge. While Americans died by the tens of thousands. They counseled pushing the higher, more lucrative doses. They strategized, in writing, about how to “counter the emotional messages from mothers with teenagers that overdosed” on the very product they were turbocharging. When it began to fall apart, a partner started deleting the emails. McKinsey eventually paid five hundred and seventy-three million dollars to settle with the states, and then another six hundred and fifty million to the federal government — which is roughly what it costs to make grieving mothers go away when you have that kind of money. This is not a flaw in the training. Maximize the client’s revenue, externalize the corpses, bill accordingly — that is the training.

In 1940, a man named Fred Schwed wrote a book about Wall Street under what may be the most honest title ever set in type: Where Are the Customers’ Yachts? A visitor is brought down to the harbor and shown the beautiful yachts of the bankers and the brokers. He looks around and asks the only question that matters: where are the customers’ yachts?There were none then. There are none now. So let me ask you, since you might keep an account somewhere, since you might be trusting some of these people with your retirement: where is your yacht? And while we’re here — how many doctors do you know who bought a boat with the gains some genius in a good suit earned for them? I’ll wait.

If you think I’m being unfair to my own industry, don’t take my word for any of it. Take Michael Lewis’s. Read Boomerang — it’s short, it’s genuinely, darkly funny, and it’s the clearest field guide I know to how this machine actually behaves once it’s loose on a continent. Watch what happens to Germany: the great German banks, trusting and rule-bound, certain that if every box was checked nothing could go wrong, became the world’s designated buyers for the toxic American mortgage paper Wall Street needed to get off its books. Lewis’s line is that the entire eurozone was, in effect, a CDO sold to the Germans. They checked every box exactly as their allocators demanded — and that is precisely how they got fleeced. Then watch Goldman Sachs walk into Greece in 2001 and run what Lewis calls “repellent deals designed to hide the Greek government’s true level of indebtedness” — the same laundering machine that had packaged the American subprime borrower, now aimed at an entire country so it could lie its way into the euro and detonate a decade later. The Greeks did most of the damage to themselves. Goldman’s job was simply to help them hide how broke they were until the door had closed behind them.

I tell you all of this not to settle scores with my own trade, but because it is the same disease in a different suit. The friends who can’t fathom why I do physics, the men who buried the Super Collider, and the machine that sells you funds you don’t need are all committing the identical error. They cannot see value that doesn’t arrive as a number, this quarter, with their name beside it. Truth that pays slowly, or never, is invisible to them — and basic science is the slowest-paying, least-legible truth we have. That is the entire story of what we are about to do to ourselves.

The bull case

Now let me cool down and be fair, because the case against panic really is strong, and I believe most of it.

Two weeks ago, in this space, I told the people forecasting America’s collapse to go home. I meant it then and I mean it now. Michael Cembalest, who runs market strategy for J.P. Morgan and is about as far from a flag-waving optimist as a careful man gets, has just published forty pages of charts — an Eye on the Market special edition he calls Semiquincententacles — that make my Fourth of July argument better than I made it. The “Sell America” trade everyone swore would punish us has lost money every single time it’s been tried. The dollar ignored the doomsayers. In nearly every major sector, American companies earn higher returns on equity than their European, Japanese, and Chinese rivals — across the board, not on the strength of one lucky technology bet. American productivity leads the entire G10. American capital markets are so dominant that the rest of the rich world is reduced to writing reports about why it can’t keep up. All true. All real. I will defend every word of it.

But notice what every one of those numbers is standing on, and then notice where Cembalest himself goes. After two dozen pages on why America is fine, the most cautious man in the building turns to the two risks that have nothing to do with the business cycle: first the unpredictability of the rule of law, and then, by page thirty, an entire section on the defunding of science. I will leave the rule of law for another day. The other one is the subject of this entire essay, and we are no longer talking about one collider in 1993. We are talking about the same mistake, made all at once, to the whole machine.

So let me show you the physics of it.

Talent is conserved

Here is a conservation law they don’t teach in business school. You cannot destroy scientific capacity by defunding it. It does not vanish. It relocates. Cut the grant, close the lab, and the knowledge in the heads of the people you just fired does not evaporate — it gets up, packs a bag, and goes to work for whoever is still paying. Talent is conserved. The only question a funding decision actually answers is whose talent it will be.

We know this is true with more certainty than we know almost anything in policy, because we have already run the experiment — in reverse, and in our favor, at the largest scale in modern history. In the 1930s a madman in Germany decided that some of the finest minds his country had ever produced were the wrong kind of people, and he threw them out. And we caught them. Einstein. Fermi, by way of fascist Italy. Von Neumann. Bethe. Szilard. Teller. Wigner. Born. We did not train a single one of them; we did not spend a dollar raising them; another country paid every cost of their education and then handed them to us out of pure stupidity and hate. The bomb, the architecture of the computer, the entire theoretical spine of the American century — built, to a degree Americans have largely forgotten, by the minds Europe drove out. We did not earn the golden age of American science so much as inherit it, and then we lived off the inheritance for three full generations.

Vannevar Bush gave that inheritance an institutional home. In 1945 he wrote a report for the President called Science, The Endless Frontier, and out of it came the National Science Foundation, the modern research university, the social contract under which the government funds the basic science that no company can justify and from which, eventually, every company profits. The NIH would go on to underwrite the research behind essentially every new drug approved in this country. The NSF would seed the internet, supercomputing, the foundations of artificial intelligence itself. The dominance Cembalest documents — the productivity, the earnings, the markets, the AI — is the dividend on that frontier. It is money invested between roughly 1945 and 2015, arriving now.

And the conservation law runs in both directions. That is the part nobody wants to say out loud. The same physics that filled our laboratories in 1940 will empty them just as faithfully if we decide, as we now seem to have decided, to play the role Germany played. We are no longer the country that catches the talent other nations throw away. We are becoming the country that does the throwing.

Look at Cembalest’s own pages — the meter reading on the very machine he is praising. American scientists submitted thirty-two percent more applications for jobs abroad in 2025 than the year before. Eighty-five rising and established researchers, including an NIH neurobiologist and a Princeton nuclear physicist, have already left for China. Enrollment in the PhD programs that produce our biomedical researchers, after years of growth, has flatlined. The European Union has stood up a half-billion-euro fund whose entire purpose is to buy our scientists away from us, and Europe does not spend half a billion euros on things it does not expect to work. And it is not only China and Europe: as I write this, Chen Wang — who builds the kind of superconducting circuits that may one day become quantum computers — is leaving the University of Massachusetts for the University of Toronto. Once that caliber of talent goes, it does not come back. The seed corn is not burning. It is boarding flights.

And it is boarding flights to precisely the competitor Cembalest also documents climbing the ladder behind us — China breaking into the ten most innovative economies, Chinese GPU self-sufficiency rising from ten percent to forty and aimed at eighty, Chinese frontier models now a handful of points behind our best. This is the detail that should end the debate. When you defund American science, you are not slowing the world down. You are slowing yourself down while handing your only serious rival the one input it cannot easily manufacture: trained human minds at the frontier. Germany’s madman handed us the future. We are now handing it back, across the Pacific, on purpose, and calling it a budget cut.

We are doing it again, and we are doing it bigger

The grant data tells the story plainly: sharp declines in competitive NIH funding for cancer, for stroke, for diabetes, for heart disease, for infectious disease; sharp declines in NSF funding for biology, for engineering, for computer science, for the physical sciences. By Cembalest’s own estimate, roughly six thousand NIH grants were terminated, about half for running afoul of political directives and the rest for, as far as anyone can tell, nothing at all. The administration fired the entire National Science Board — the body that governs the NSF — days before the board’s own Science and Engineering Indicators report, the one documenting how fast China is closing the gap on exactly the metrics that decide the next century: papers in the top one percent of the most-cited journals, total spending on research and development, and patents in artificial intelligence, quantum information, biotechnology, semiconductors, and nuclear technology.

And then there is the part that should make a market person sit up, because it is structural and it is permanent. The Office of Management and Budget has proposed a rule — one that would bind every federal agency by October of this year — that would dismantle the three pillars the entire postwar system rested on. Political appointees, not scientists, would conduct a “pre-issuance review” of every discretionary grant, explicitly forbidden from deferring to the experts, instructed to block anything that promotes “anti-American values.” Peer review — the mechanism by which independent scientists, and not commissars, decide what is good science — would become advisory, overrideable by a political appointee with no finding of cause. Active multi-year grants could be terminated mid-stream for being “inconsistent with agency priorities,” no fraud or failure required. Understand what this is in market terms. Peer review and open competition are the price mechanism of science — the closest thing the field has to a market that sorts good bets from bad. Replacing that mechanism with the political preferences of whoever holds power this year is not reform. It is central planning, imposed on the one corner of American life that beat the centrally planned competition precisely because it was free.

Shirley Tilghman — molecular biologist, former president of Princeton — put her finger on the part a builder should fear most. The intention, she allowed, was never to destroy American scientific capacity; it was to punish a handful of elite universities. But intention is not outcome, and the outcome is the destruction of the capacity itself. It is one thing to tear something down. It is another to tear it down with nothing ready to put in its place. That is the moment we are in.

Albion

I keep having the feeling that I have seen this film before. And I have. The country was Britain, and I first saw the rot in 1979.

Empires do not die on the ticker. Britain’s didn’t. The pound stayed convertible, the City stayed busy, the great names kept their offices long after the thing underneath had gone. An empire — and a scientific civilization is a kind of empire — dies first in the institutions and the laboratories, invisibly, in a way that shows up in no quarterly print, and the markets find out last. By the time the decline is legible in the numbers everyone watches, it is already a generation old and substantially irreversible. Britain led the world into the industrial age and the world into the foundations of modern physics and biology, and then it spent decades quietly underfunding and out-emigrating its own talent, and one day it looked up and the center of gravity had moved across an ocean — to us, partly on the backs of the scientists it failed to keep.

We are now running the British playbook on the one asset Britain would have sold its fleet to hold. The difference is that we are doing it faster, louder, and with a written rule that puts political officers over the peer reviewers. The dollar will be the last thing to tell us. It always is.

The objections, and why they fail

Let me take the strongest versions of the case against me, because there are three, and they deserve real answers.

“We are not gutting science. We are cutting bloat — the captured grants, the ideological projects, the waste.” Some of that is fair; some grants are junk, as some grants have always been junk. But the administration’s own numbers refute the scale of the claim: of the terminated NIH grants, only about half tripped the political filters at all, and the other half were killed for no stated reason that survives scrutiny. And whatever you think of trimming fat, the OMB rule does not trim fat. It puts political appointees above the scientific community’s judgment and lets any live grant die for “agency priorities.” You do not make an engine more efficient by firing the engineers and handing the controls to people who think the manual is propaganda. The market test for science already exists. It is called peer review. They are abolishing it.

“The private sector will fund it now — the AI labs, the pharma giants, the great corporate research budgets.” They will fund the applied layer that sits on top of the public base; they always have. What they cannot fund is the base. Cembalest’s own pages show the frontier AI labs incinerating cash, their own projections of when they will turn cash-flow positive sliding into 2028, 2029, 2030, which he rightly calls speculative. You cannot ask companies that must be profitable by 2028 to carry the twenty-year, no-revenue, who-knows-if-it-works research from which the profitable thing will eventually grow — research whose entire defining feature is that it has no buyer yet. The reason the government funds basic science is that it is the one bet rational on a civilizational horizon and irrational on every private one. Remove the public bet and the private bets have nothing left to stand on. It just takes twenty years to notice, which is exactly the interval these people are built to ignore.

“China will trip over its own government before it overtakes us. The Party will smother its own scientists.” Maybe. I would not bet against the possibility. But notice the shape of the argument: it asks you to stake the scientific future of the United States on the proposition that your rival’s government is more self-destructive than your own. And your own government has just fired the entire board that oversees its science agency, days before that board could warn it about the rival. When your single remaining hope is that the other side is even dumber than you are, you have already lost the thread. Cembalest’s charts show China closing, not collapsing. Plan for the China that climbs.

Where is your yacht

So let me end where I began two weeks ago. America remains great. Say it again, because it is true: the dollar is resilient, the companies are the best on earth, the doomsayers shorting this country have lost money for forty years and will keep losing it. If your plan is to bet against America on the dollar or the debt ceiling or a bad inflation print, you will be carried out, as you should be.

But there is exactly one way to make the doomsayers right, and it is not a tariff or a yield or a CPI surprise. It is to take the machine that produced every bit of the greatness Cembalest documents — the engine Vannevar Bush switched on in 1945, the one we built out of the genius another madman gave us — and hand the controls to people who think peer review is the enemy and a fleeing physicist is somebody else’s problem. To close the endless frontier, fill in the tunnel, and call it responsible. Elias Zerhouni, who ran the NIH under a Republican president, put it as plainly as it can be put: they are taking an axe to the system, they are destroying it, and they do not understand that when you tell them.

You want to short America? Here is how you actually do it. Don’t touch the dollar. Keep defunding the thing the dollar is built on. Drive the scientists to Shenzhen and Munich. Put the commissars over the peer reviewers. Then wait twenty years, and collect.

I would simply rather we didn’t hand them the trade.

We’ve buried one collider this way already. The tunnel is still out there under Texas, full of water, and nobody remembers it, and that is the whole problem: nobody ever remembers the discovery that was never made, the cure that never came, the industry that never started, the yacht the customer never got. The cost of closing the frontier is invisible by construction. So is the rot in a great house, right up until the roof comes down. Ask Britain. Or don’t — ask the one physicist we have left in Congress. He’s easy to find. There’s only one of him.

---

For the optimistic case, see the companion piece, ARGH! America Remains Great, Halfwits. On Britain’s long institutional decline — first seen in 1979 — see Albion. And for an equally uncomfortable question about an equally cherished story, my book, The Science of Free Will.