Keen On America

Ah, the American South. So romantic. A few days ago we were in Clayton, Georgia, hearing about the dissident Lillian Smith, whose house was burned down for telling the truth about Jim Crow. Today we go south-west into Alabama, to Ozark, in the Wiregrass, the boondocks of the boondocks — a region named for a scrub that grows where nothing else will. Except for murder.
The Wiregrass, as the Birmingham, Alabama–based journalist Matt Kessler explains, has poor soil, no cotton, almost no plantations, and therefore, none of the southern decorum either. He has spent nine years there researching The Wiregrass, a new true-crime investigation into a case nobody has satisfactorily explained.
On 31 July 1999, JB Beasley turned seventeen. She and her friend Tracie Hawlett set out from the Alabama town of Dothan (peanut capital of the world) to go party-hopping in the Wiregrass. Searching for a field party with a keg, the two young women got lost. They stopped at a closed filling station in Ozark to use the payphone, and someone gave them directions that didn’t sound suspicious.
“Right out of the lot, left at the McDonald’s.”
The surveillance tape shows them leaving in the right direction. Their car was found by the roadside the next morning. At two in the afternoon somebody finally opened the trunk. Both girls were inside, shot in the head. Southern gothic murders most foul.
So what happened? The bodies sat in that trunk for five hours in a steaming Alabama summer, before an officer from out of town had the smart idea to open it. Trained officers then searched the car without gloves. A lead investigator tried to tow it back to Dothan, which would have destroyed the crime scene. Years later the car was quietly released and resold, with no paperwork stating who authorised it. A man named Johnny Barrentine confessed and was let go when the DNA did not match. And when Kessler rang the Ozark police for the file, he was told that everything before 2000 had been “accidentally” destroyed. Oops.
Twenty years on, a genetic genealogy search matched semen from the scene to Coley McCraney, a married preacher with no criminal record. He was convicted on four counts of capital murder in 2023 and is now serving life without parole. McCraney admitted to the sex, but denied the killing. The DNA is the only material evidence since nobody found a weapon. Kessler believes in his innocence, and suspects the police are implicated, though the officer he calls a person of interest will stay unnamed until you buy the book. Ah, the South.

FURTHER READING

•   Matt Kessler, The Wiregrass: A Tale of Murder and Retribution in the Deep South (Grand Central, 6 October 2026). Kirkus calls it “a meticulous, urgent account of a case officially closed, but far from settled.”
•   The official record, against which his account should be read: the Alabama Attorney General’s statement on McCraney’s sentencing in June 2023, following conviction on four counts of capital murder.
•   In September 2024 the Alabama Court of Criminal Appeals upheld the conviction. The defence had argued that Rena Crumb should have been allowed to testify that an Ozark officer, not McCraney, killed the girls — the same account Kessler describes here. Crumb subsequently recanted, and the judge ruled her evidence hearsay. The case is now before the Alabama Supreme Court.
•   Keri Leigh Merritt on Lillian Smith and the Georgia side of the same story, the episode AK opens on: listen or watch.

FIVE TAKEAWAYS

1. The reporting is far stronger than the theory it is marshalled behind. The most valuable hour of this conversation is the texture, and it is first-hand: a Klansman soliciting donations on the town square in the eighties like the Salvation Army; black families teaching their children to carry what they called driving around money for the inevitable bogus stop; locals pressing handguns and window-shattering hammers on a reporter they thought was in danger for asking questions. None of this requires anyone to accept his conclusion. It establishes something narrower and more damning, which is that this was not a department capable of investigating anything.
2. The investigation failed at almost every point where it could fail. The bodies lay in a closed trunk for five hours in high summer before an officer from Dothan, there on his own initiative, opened it — several whistleblowers told Kessler that any experienced investigator would have known what he was smelling. Local officers then searched the vehicle without gloves. One tried to tow it to another jurisdiction. The heat destroyed the evidence so thoroughly that no time of death could be established. Some years later the car itself was released and sold, and no document records who authorised that. Kessler says he cannot tell whether this is incompetence or concealment, which is the honest answer.
3. A pattern of police sexual coercion runs underneath the case. A Police Explorers programme at the local high school, nominally training juniors and seniors for the job, was shut down by an incoming deputy chief because officers had been sleeping with the students. A school administrator told Kessler of pupils arriving in his office in tears after officers traded bogus speeding tickets for sexual favours. During the reporting, the police chief and a supervisor were forced out over coercing the wives of officers on parole. AK’s question is the sharp one: whether two girls from the bigger town simply refused to play along.
4. The conviction rests on DNA and almost nothing else. A genetic genealogy search of the kind that caught the Golden State Killer led police to Coley McCraney, who gave a sample voluntarily and was arrested for double murder and rape. There is no weapon and no other material evidence. Put on the stand by his own lawyer against his wishes, he placed himself in the girls’ car and could not satisfactorily account for the evening seventeen years earlier. Kessler accepts the DNA is his and argues the encounter was consensual, which is where his real intellectual claim sits: that DNA at a scene is accepted prima facie as guilt without anyone asking what it actually shows.
5. He accuses the police, and is candid that he cannot finish the sentence. Asked directly what happened, he says he believes a local officer was involved, names one as a person of interest, and then declines to identify him on air — you will have to read the book. He is scrupulous about the rival theory, having found nothing in a quarter of a million files to support the rumour that the girls stumbled onto a drug flight. Listeners should know what the hour does not mention: Rena Crumb, whose secondhand account of a drunken confession is central here, later recanted, and the appeals court that upheld the conviction in 2024 ruled her evidence hearsay....

On 16 October 1962, exactly sixty-four years ago, photos of Soviet missiles on Cuba were shown to President Kennedy and the world faced nuclear Armageddon. The definitive account of that thirteen-day crisis, One Minute to Midnight, was written by today’s guest, the presidential crisis historian Michael Dobbs. He has now written about a crisis that few remember, but which he calls the climactic crisis of the Cold War.
It happened in 1983. In September of that year, the Soviets shot down a South Korean airliner with 269 people aboard, among them the most anti-communist member of the United States Congress. A few weeks later a Soviet early-warning computer reported that America had launched its missiles, and a colonel named Petrov, alone in a bunker at one in the morning, decided on his own authority that the machine was lying. In November NATO ran a nuclear release exercise called Able Archer, which Moscow took for the real thing.
The difference between 1962 and 1983, Dobbs says, was twenty years of technology. Kennedy had thirteen days. Reagan, by his own estimate in his memoirs, would have had six minutes. Thus the title of his new book, The Six Minute War, subtitled Reagan, Russia and the Race to Automate Armageddon.
As the incoming Moscow correspondent for the Washington Post, Dobbs had a front seat on the crisis. He arrived in Moscow just after the airliner came down, and his best material is anecdotal. He saw Russians bussed to a state-organised protest outside the American embassy, chanting on cue while jealously eyeing the Buicks and Cadillacs parked on the street. In a Moscow cinema showing Three Days of the Condor as anti-American propaganda, he watched the audience gasp — not at the dastardly CIA, but at the lavish contents of Robert Redford’s fridge.
Fast forward to today, and we get to the most worrying part of Dobbs’ analysis. The presidential crisis historian reckons we are now in a more dangerous period than at any point he can remember. We have more weapons, less warning, with our split-second decisions dependent on machines whose reasoning is increasingly hard for us humans to understand. Call the compounding annual risk one per cent. Multiply it across thirty years and the arithmetic isn’t reassuring. We might now be just a few seconds to midnight.

FURTHER READING

•   Michael Dobbs, The Six Minute War: Reagan, Russia, and the Race to Automate Armageddon (Simon & Schuster)
•   His Cuban Missile Crisis book, One Minute to Midnight (2008) — the thirteen days against which he measures the six minutes. Down with Big Brother and Six Months in 1945 complete his Cold War trilogy.
•   The National Security Archive’s declassified sourcebook on Able Archer 83 — the NATO exercise the Kremlin read as cover for a first strike.
•   Eliot Cohen made the companion argument on this show: that the nuclear danger now sits with Pakistan and North Korea rather than Moscow and Washington. Listen or watch.
•   Harold James’s essay “Late Soviet America”, the gerontocracy parallel AK raises at 25:26 — among his Project Syndicate columns.

FIVE TAKEAWAYS

1. The headline is about now, not 1983. The historian who wrote the definitive account of the Cuban Missile Crisis says we are in a more dangerous period today than at any point in his memory of the Cold War, as a child or later as a reporter. His reasoning is that the warheads have proliferated, the warning times have been compressed, and the decisions are moving to computers whose workings we cannot inspect. The Cold War at least had a rough nuclear balance between two systems that understood each other. He is more frightened for his grandchildren than he was at the tensest moments he lived through.
2. What changed between 1962 and 1983 was twenty years of technology. Kennedy and Khrushchev had thirteen days to argue their way out of Cuba, which now looks almost leisurely. By 1983 bombers had given way to missiles and computers had entered the early-warning chain, and Reagan wrote in his memoirs that he might have six minutes to decide whether a blip on a screen was the end of the world. That line is where the book’s title comes from. Dobbs is candid that there was no six-minute war: the compression of time is the subject, not an event.
3. A Soviet colonel is the reason this is a book about automation. At one in the morning in the autumn of 1983, Soviet satellites and computers reported an American missile launch, and Colonel Petrov had to decide whether to pass it up the chain as real. He had written some of the algorithms himself and knew they could be wrong, and he reasoned that nobody opens a surprise attack with five missiles rather than two hundred. So he called it a false alarm on his own authority. Dobbs’s point is not that Petrov was a hero but that somebody else in that chair might have answered differently.
4. Reagan was the most anti-nuclear president since the bomb was invented. This is the claim least appreciated at the time and the one Dobbs presses hardest. The man denouncing the evil empire was also reading the Bible for prophecies of Armageddon and was genuinely tormented by the possibility of nuclear war, which is why he later stunned his own advisers at Reykjavik by proposing to abolish nuclear weapons outright. Asked which Reagan to believe, Dobbs refuses to choose: both were real, the militant anti-communist and the pragmatist who found someone he could deal with in Gorbachev.
5. Nobody starts a nuclear war on purpose, which is the frightening part. In both 1962 and 1983 the danger was never a deliberate first strike but an accident that escalates past the point where anyone can stop it — a stray airliner, a faulty algorithm, an exercise misread as cover. Andropov told the former ambassador Averell Harriman exactly that in the Kremlin, and Dobbs thinks it remains true with Pakistan, North Korea and now AI added to the mix. His prescription sounds modest and is not: elect sensible leaders, keep talking to the other side, and practise the empathy that let Kennedy ask what the world looked like from Khrushchev’s chair.

ABOUT THE GUEST

Michael Dobbs is a presidential crisis historian and a former foreign correspondent and Moscow bureau chief for the Washington Post, where he launched the paper’s Fact Checker column in 2007. He reported from Moscow through the 1983 war scare and was there in August 1991 when Boris Yeltsin climbed onto a tank. His Cold War trilogy comprises Down with Big Brother, One Minute to Midnight and Six Months in 1945; he is also the author of King Richard, The Unwanted, Saboteurs and a ...

Last month Adam Kirsch was on the show arguing that our current obsession with UFOs represents a new kind of religion. Today’s guest, the international relations expert Alexander Wendt, argues that today’s UFO obsession is the least of our problems. In his new book, The Last Humans (out today), Wendt’s worst-case scenario is that the UFOs are already here — and have been for some time.
While the Pentagon has conceded only that the objects are real and that it cannot identify them, Wendt extrapolates further. A recognition of the presence of UFOs would bring, he warns, both a crisis of global security and the demise of the traditional state system. Worst of all, he fears, would be a descent into a Hobbesian-style war-against-all.
Wendt is less worried about the invasion of the earth by little green men. If they had wanted to attack, he says, they would have done it when we fought with sticks rather than nukes. Instead, he worries about the social contract we have with the state. We obey the laws of the state because they keep us safe. Spacecraft that the Pentagon cannot identify, in airspace the Pentagon can’t defend, would leave that contract in tatters. What follows is not war with aliens but Hobbes’s great fear of a return to his dreaded state of nature.
So, as Wendt notes, “the threat of UFOs is not them. It’s actually us.” Rather than a new religion, they represent an old nightmare. Beam me up, Thomas Hobbes.

FURTHER READING

•   Alexander Wendt, The Last Humans: UFOs, Global Security, and the Threat to State Sovereignty (Oxford University Press, out 15 October 2026)
•   Adam Kirsch on We Want to Believe — the case that this is a religion, not a security problem: listen or watch

•   Wendt’s 1992 article “Anarchy Is What States Make of It”, the piece that launched constructivism, and his Social Theory of International Politics — both listed at his Mershon Center page

•   The June 2021 Preliminary Assessment: Unidentified Aerial Phenomena from the Office of the Director of National Intelligence — the report that, on Wendt’s account, changed the game

FIVE TAKEAWAYS

1. His thesis is that the danger is on our side of the encounter. He separates the craft from the question of who is flying them, and says the Defense Department has settled the first — after seventy-five years of denial it changed its mind, which he finds significant. On the second he is agnostic but unworried about attack: anything capable of getting here could have destroyed us long ago, when we were far weaker. The threat he describes is political. Our civilisation is organised around the assumption that we are alone, and the discovery that we are not would expose a state that cannot identify these objects, cannot stop them entering its airspace, and cannot protect anyone from them. That is the emperor with no clothes, and what it threatens is allegiance.
2. What he actually fears is Hobbes, not invasion. Once the news is out, people will disagree about what to do — shoot them down, worship them, ignore them — and there is no way to settle the argument, because nobody can get inside the craft to find out who is right. Disagreement with no arbiter is the oldest problem in political philosophy and the one Hobbes built the state to solve. His worst case is not conquest from outside but an implosion from within, which is why he thinks disclosure handled too quickly and too casually would itself be dangerous.
3. The title has three meanings, and the third is the interesting one. We could be the last humans physically, if they exterminate us. We could be the last biologically, if the hybridisation programme described by alien abduction experiencers is real. Or we could be the last in what he calls an experiential sense — the last people who will ever know what it was like to believe we were alone in the universe. He thinks the third is much the most realistic, calls it sad and potentially uplifting at once, and notes that in all three cases the anthropocentric world we have built does not survive.
4. It is bad news for America and possibly less bad for China. A country accustomed to sitting on top of the international system has the furthest to fall, and every scenario he sketches makes the United States smaller: against a superior intelligence the gap between the strongest state and the rest stops mattering. On China he is openly speculating, but the thought is that a more collectivist culture has more cohesion and loyalty to draw on in a shock like this, where an individualistic one has people going every which way. His guess is that Washington and Beijing would try to cooperate, and that the real question is whether either can keep a grip on its own population.
5. The UFOs are the occasion; the argument is about constructivism. The most fluent stretch of the hour has nothing to do with aliens. Wendt explains why idealism became a slur in international relations after Munich, how realism and then liberalism replaced it, and where his own school came from — sociologists rather than economists, George Herbert Mead and Erving Goffman rather than rational choice. His 1992 article “Anarchy Is What States Make of It” argued that the international system has no single logic and that everything depends on what actors do with it. That is the claim this book applies to the strangest case available.

ABOUT THE GUEST

Alexander Wendt is Mershon Professor of International Security and Professor of Political Science at The Ohio State University, and a founder of the constructivist school of international relations. He is the author of Social Theory of International Politics and of the 1992 article "Anarchy Is What States Make of It". In the 2017 TRIP survey his peers named him the most influential scholar in international relations of the previous twenty years. He shared the 2023 Skytte Prize in Political Science and was elected to the American Academy of Arts and Sciences in 2025. The Last Humans is published by Oxford University Press.

ABOUT KEEN ON AMERICA

Keen On America is a daily conversation about the ideas, books and arguments shaping the United States, hosted by Andrew Keen, author of The Internet Is Not the Answer and How to Fix the Future. New episodes every day.

“We don’t have to choose between AI and no AI. The bigger choice is what type of AI we want, and who we want in control of it.” — Ben Green

INTRO

In our age of AI, do we want to be ruled by all-knowing engineer kings? Not according to Ben Green, a professor of information and public policy at Michigan, whose new book, Algorithmic Realism, presents “data science practices to promote social justice”. Unsexy? Yes. And that, according to Green, is the whole point of algorithmic realism. Such a down-to-earth use of AI is super unsexy by design, he explains.
Green’s critique is of Silicon Valley utopians who promote an algorithmic formalism. So he’s dismissive of the idea of “superintelligence” with its promise of curing cancer at the snap of a finger. He borrows from the sociologist Erik Olin Wright the idea of real utopias reached concretely, step-by-step. So he’s not exactly against algorithmic idealism. Just the lazy type that sees the world as an infinitely scalable laboratory.
Green describes working as a data scientist for the city of Boston and being asked to cut EMS response times. Instead of building some sort of super sexy algorithm, he went and talked to people, and found what was actually driving the call volume: mental illness, addiction and homelessness. Shaving two minutes off an ambulance would have done nothing for any of those people. What they built instead — and he is careful to stress the they — was a team trained to handle someone in a crisis and connect them to counselling and shelter. Super unsexy; and super useful.

FURTHER READING

•   Ben Green, Algorithmic Realism: Data Science Practices to Promote Social Justice (MIT Press, 2026)
•   His earlier book, The Smart Enough City — the same argument, made about city government in 2019
•   Green on DOGE: “Using AI to Reform Government Is Much Harder Than It Looks”, at Tech Policy Press

•   My own case for the engineer kings, three days ago: Crowning Our Engineer Kings

•   Matthew Botvinick of Anthropic on AI and Political Freedom — the inside-the-lab answer that Green doubts: listen or watch

•   Sherry Turkle on Artificial Intimacy — the humanist case on AI: listen or watch

•   Stephen Witt, Jensen Huang’s biographer, on the engineer king par excellence: listen or watch

FIVE TAKEAWAYS

1. The engineer’s mindset is the thing he is actually arguing with. Not engineering itself, which he wants more of, but a particular habit of mind: that problems have neat optimal solutions, that facts and reason settle things, and that building more powerful tools will of itself improve society. What that leaves out is power — who gets to decide, what their incentives are, and whether the decision might be worth democratising. He puts the underlying assumption plainly: our goal is to build more and more powerful tools, and as our tools get more powerful they will inevitably work for the betterment of society. He thinks that is a fundamental misunderstanding of how technology and social impact are related.
2. He does not think you can fix it from inside the labs. Pressed twice on whether well-intentioned people at Anthropic and OpenAI can do good with their engineering brains, he gave the same answer both times. The companies are pushing as hard as they can towards superintelligence and, in OpenAI’s case, an enormous IPO, and in that setting safety is treated as a public-relations and liability problem rather than a deeply held ambition. He points to the recent run of departures and whistleblowing, and to reporting that engineers raising safety concerns were ignored. There are many well-intentioned people there; the business decisions lead.
3. An ambulance in Boston is his best argument. Working as a data scientist for the city, he was asked to improve emergency medical response times, and the obvious move was an algorithm that would optimise ambulance deployment and save a couple of minutes per incident. Instead he spent the time building relationships with the people doing the work, and found that the rising call volume was being driven by mental illness, drug addiction and homelessness. Optimising deployment would have left every one of those people exactly where they were. What came out of it instead was a new team, trained to handle someone in a moment of crisis and connect them to counselling, social services and shelter. Still data science — he did a great deal of analysis — but aimed at the problem rather than the metric.
4. He wants narrow AI, and he is rude about scale. There is, he argues, a very tenuous relationship between a system that tops a benchmark and a system that is useful to anyone in practice. The labs are chasing the first; what helps people is the second, tailored to a particular use. His unglamorous example from the city-government years is cleaning up your data, because most organisations have messy, disorganised datasets and nothing works on top of those. He is similarly unsparing about the money burned on what he calls token maxing over the past year. And on the American strategy of simply building bigger: our sausage factory is more powerful than your sausage factory, which he regards as shortsighted industrial policy.
5. Utopia is fine; the missing part is the route. He is not a fan of effective altruism but says he is not against utopian thinking at all — the book that shaped this one most was Erik Olin Wright’s Envisioning Real Utopias. What he objects to is the shape of the labs’ promise: that AI will cause enormous disruption and then, somehow, universal basic income and the end of disease and climate change will follow. No pathway, just the promise dangled in front of us. Wright’s term for what is missing is accessible way stations. So his own ambition is radical and his steps are deliberately small: what one person or one organisation can actually do, this month, in the job they already have.

ABOUT THE GUEST

Ben Green is an assistant professor at the University of Michigan School of Information and the Gerald R. Ford School of Public Policy, and a faculty associate at the Berkman Klein Center for Internet and Society at Harvard. He has worked as a data scientist for the City of Boston and writes on algorithms in government for Tech Policy Press. His previous book, The Smart Enough City, was published in 2019. Algorithmic Realism: Data Science Practices to Promote Social Justice is published by MIT Press.

ABOUT KEEN ON AMERICA

Keen On America is a daily conversation about the ideas, books and arguments shaping the United States, hosted by And...

Yesterday Janet Reitman described post-9/11 America, with its forever wars in Iraq & Afghanistan, as a tragedy in five acts. Today we meet a man who helped strategize one of those acts. Eliot A. Cohen was Counselor of the State Department under Condoleezza Rice from 2007 — in part responsible (“if that’s the right word” as he puts it) for America’s post-9/11 forever wars. Not that he feels deep remorse. Afghanistan was unavoidable, he says — no president would have refrained, and the Democrats would have done the same. While he was in favour of the Iraq adventure, he does admit that he never asked hard enough questions about whether America’s institutions were fit for that kind of war.
His new book, The Strategist, is a short introduction to thinking about war and politics. So is Cohen a realist or idealist when it comes to thinking about international relations? There are two kinds of people in this world, he responded, half jokingly, people who divide things into two and people who don’t. And given Cohen’s distrust of what he calls “purely linear descriptions of history where it’s all one thing or another”, he’s unambiguously in the latter category.
So who are the great strategists in history from whom we should learn about war and politics? For Cohen, Churchill ranks high, as does Lincoln. And how do contemporary American “statesmen” like Hegseth and Trump compare? “It’s like you’re dealing with two different species,” Cohen dryly notes.
There’s one question I half regret not asking. Janet Reitman’s new book, The Inheritance, is full of stories about people that the forever wars broke. Like the angry ex-Marine who was transformed into a leading neo-Nazi. So perhaps I should have asked Cohen about the human costs of Iraq and Afghanistan. If 21st-century America is to avoid more overseas tragedies, must the strategist also be a moralist? History isn’t so linear, he would probably have replied. As he says at the beginning of The Strategist: “you may wish to be done with war, but war is not done with you.”

FURTHER READING

•   Eliot A. Cohen, The Strategist: How to Think About War and Politics (Princeton University Press)
•   His weekly podcast with Eric Edelman, Shield of the Republic

•   Janet Reitman on The Inheritance — the five-act American tragedy Cohen declines to sign up to: listen or watch

•   Elliot Ackerman on The Fifth Act, another five-act tragedy of how the Afghan war ended: listen or watch

•   Jacob Weisberg on Profiles in Cowardice, which takes a much harder line on George W. Bush than Cohen: listen or watch

FIVE TAKEAWAYS

1. He will not call the whole thing a blunder, and he explains why. Asked straight out whether he shares Janet Reitman’s view that American foreign policy after 9/11 was one long miscalculation, he declined to comment on a book he hasn’t read and then gave the historian’s answer. A complicated mix. A great deal of miscalculation, which nobody can deny, and also some things that were successes. His objection is methodological as much as political: judgments made at the time always get amended by longer perspectives, and he dislikes purely linear accounts of history in which everything is one thing or another. He supported the Iraq war, says so in the new book and at length in an earlier one, and the place he faults himself is narrow and specific — he never asked rigorously enough whether American organisations were fit for that kind of war.
2. Realists and idealists is a distinction he thinks is useless. He took a doctorate in political science and then went off it, because the categories appeal to what he calls minds that swing on hinges. There is no pure realist and no pure idealist; what actually happens is that people’s values colour how they see the world, so the two are never separable. Churchill is his example of the mix — the man who identified Hitler years before his colleagues did, and the man who cut a deal with Stalin over Poland that he hated making. That is what conducting a country’s affairs in wartime is like: caught between what you would like to do and what the world will let you do.
3. Lincoln is his greatest American strategist, and had no military experience at all. Thirty days in the Black Hawk War, no fighting, and a famous moment when he marched his platoon up to a rail fence, could not remember the order for turning a column, and simply told them to fall out and fall in on the other side. What Lincoln had instead was a grip on human nature and what Cohen calls profound common sense. He does not think common sense can be taught — pressed on whether it is innate, he said that was a philosophical question and he didn’t know enough to answer it. What he did try to teach at Johns Hopkins was judgment, which he defines as learning to ask questions, listen carefully to the answers, and then ask some more. Along with how to write a clear memorandum, which he insists is not a mundane skill.
4. Rice hired him to tell her things she didn’t want to hear, and he did. Reading the North Korean estimates he decided something was squirrelly, laid out the evidence that Pyongyang was cheating on its plutonium declaration, and was given licence to go hunting. He came back with the circumstantial case for a covert uranium enrichment programme. The assistant secretary negotiating the agreement asked sarcastically whether he imagined somebody would walk into a building in Pyongyang and find a blast door opening onto spinning centrifuges — which is more or less what happened when Siegfried Hecker visited Yongbyon some time later. He also told her Afghanistan was in far worse shape than the briefings claimed, having unpacked a reassuring statistic and found it bogus. Earlier still, he told President Bush in the Oval Office that the wrong general was commanding in Iraq, and named David Petraeus as the replacement. He says he only found the nerve because his son was deploying as an infantry officer.
5. The useful question about a war is not how it ends but when it began. This is his wife’s idea, borrowed from her work running a photographic archive: every photograph is a double crop, in space and in time, and so is every war. Where you start the clock changes what you think the war is. The Iran war, on his reading, began in 1979 with the founding of a republic organised around hostility to the United States, and everything since is another campaign in it. Ukraine began in 2022 if you are reading the headlines, 2014 if you are Ukrainian, around 1990 if you are Vladimir Putin. He applies the same test to the Second World War — a continuation of 1914, or a new war against the Nazis beginning in 1933 — and says the answer changes how you fight and, more importantly, how you handle the end.

ABOUT THE GUEST

Eliot A. Cohen holds the Arleigh A. Burke Chair in Strategy at the Cente...