This is Atlantic Intelligence, an eight-week series in which The Atlantic’s leading thinkers on AI will help you understand the complexity and opportunities of this groundbreaking technology. Sign up here.

Earlier this year, The Atlantic published a story by Gary Marcus, a well-known AI expert who has agitated for the technology to be regulated, both in his Substack newsletter and before the Senate. (Marcus, a cognitive scientist and an entrepreneur, has founded AI companies himself and has explored launching another.) Marcus argued that “this is a moment of immense peril,” and that we are teetering toward an “information-sphere disaster, in which bad actors weaponize large language models, distributing their ill-gotten gains through armies of ever more sophisticated bots.”

I was interested in following up with Marcus given recent events. In the past six weeks, we’ve seen an executive order from the Biden administration focused on AI oversight; chaos at the influential company OpenAI; and this Wednesday, the release of Gemini, a GPT competitor from Google. What we have not seen, yet, is total catastrophe of the sort Marcus and others have warned about. Perhaps it looms on the horizon—some experts have fretted over the destructive role AI might play in the 2024 election, while others believe we are close to developing advanced AI models that could acquire “unexpected and dangerous capabilities,” as my colleague Karen Hao has described. But perhaps fears of existential risk have become their own kind of AI hype, understandable yet unlikely to materialize. My own opinions seem to shift by the day.

Marcus and I talked earlier this week about all of the above. Read our conversation, edited for length and clarity, below.

— Damon Beres, senior editor

“No Idea What’s Going On”

Damon Beres: Your story for The Atlantic was published in March, which feels like an extremely long time ago. How has it aged? How has your thinking changed?

Gary Marcus: The core issues that I was concerned about when I wrote that article are still very much  serious problems. Large language models have this “hallucination” problem. Even today, I get emails from people describing the hallucinations they observe in the latest models. If you produce something from these systems, you just never know what you're going to get. That’s one issue that really hasn’t changed.

I was very worried then that bad actors would get a hold of these systems and deliberately create misinformation, because these systems aren’t smart enough to know when they’re being abused. And one of the largest concerns of the article is that 2024 elections might be impacted. That’s still a very reasonable expectation.

Beres: How do you feel about the executive order on AI?

Marcus: They did the best they could within some constraints. The executive branch doesn’t make law. The order doesn’t really have teeth.

There have been some good proposals: calling for a kind of “preflight” check or something like an FDA approval process to make sure AI is safe before it’s deployed at a very large scale, and then auditing it afterwards. These are critical things that are not yet required. Another thing that I would really like to see is independent scientists as part of the loop here, in a kind of peer-review way, to make sure things are done on the up-and-up.

You can think of the metaphor of Pandora’s box. There are Pandora’s boxes, plural. One of those boxes is already open. There are other boxes that people are messing around with and might accidentally open. Part of this is about how to contain the stuff that’s already out there, and part of this is about what’s to come. GPT-4 is a dress rehearsal of future forms of AI that might be much more sophisticated. GPT-4 is actually not that reliable; we’re going to get to other forms of AI that are going to be able to reason and understand the world. We need to have our act together before those things come out, not after. Patience is not a great strategy here.

Beres: At the same time, you wrote on the occasion of Gemini’s release that there’s a possibility the model is plateauing—that despite an obvious, strong desire for there to be a GPT-5, it hasn’t emerged yet.  What change do you realistically think is coming?

Marcus: Generative AI is not all of AI. It’s the stuff that’s popular right now. It could be that generative AI has plateaued, or is close to plateauing. Google had arbitrary amounts of money to spend, and Gemini is not arbitrarily better than GPT-4. That’s interesting. Why didn’t they crush it? It’s probably because they can’t. Google could have spent $40 billion to blow OpenAI away, but I think they didn’t know what they could do with $40 billion that would be so much better.

However, that doesn’t mean there won’t be other advances. It means we don’t know how to do it right now. Science can go in what Stephen Jay Gould called “punctuated equilibria,” fits and starts. AI is not close to its logical limits. Fifteen years from now, we’ll look at 2023 technology the way I look at Motorola flip phones.

Beres: How do you create a law to protect people when we don’t even know what the technology looks like from here?

Marcus: One thing that I favor is having both national and global AI agencies that can move faster than legislators can. The Senate was not structured to distinguish between GPT-4 and GPT-5 when it comes out. You don’t want to go through a whole process of having the House and Senate agree on something to cope with that. We need a national agency with some power to adjust things over time.

Is there some criterion by which you can distinguish the most dangerous models, regulate them the most, and not do that on less dangerous models? Whatever that criterion is, it’s probably going to change over time. You really want a group of scientists to work that out and update it periodically; you don’t want a group of senators to work that out—no offense. They just don’t have the training or the process to do that.

AI is going to become as important as any other Cabinet-level office, because it is so pervasive. There should be a Cabinet-level AI office. It was hard to stand up other agencies, like Homeland Security. I don’t think Washington, from the many meetings I’ve had there, has the appetite for it. But they really need to do that.

At the global level, whether it’s part of the UN or independent, we need something that looks at issues ranging from equity to security. We need to build procedures for countries to share information, incident databases, things like that.

Beres: There have been harmful AI products for years and years now, before the generative-AI boom. Social-media algorithms promote bad content; there are facial-recognition products that feel unethical or are misused by law enforcement. Is there a major difference between the potential dangers of generative AI and of the AI that already exists?

Marcus: The intellectual community has a real problem right now. You have people arguing about short-term versus long-term risks as if one is more important than the other. Actually, they’re all important. Imagine if people who worked on car accidents got into a fight with people trying to cure cancer.

Generative AI actually makes a lot of the short-term problems worse, and makes some of the long-term problems that might not otherwise exist possible. The biggest problem with generative AI is that it’s a black box. Some older techniques were black boxes, but a lot of them weren’t, so you could actually figure out what the technology was doing, or make some kind of educated guess about whether it was biased, for example. With generative AI, nobody really knows what’s going to come out at any point, or why it’s going to come out. So from an engineering perspective, it’s very unstable. And from a perspective of trying to mitigate risks, it’s hard.

That exacerbates a lot of the problems that already exist, like bias. It’s a mess. The companies that make these things are not rushing to share that data. And so it becomes this fog of war. We really have no idea what’s going on. And that just can’t be good.

Related:

Why are we letting the AI crisis just happen? The new AI panic

P.S.

This week, The Atlantic’s David Sims named Oppenheimer the best film of the year. That film’s director, Christopher Nolan, recently sat down with another one of our writers, Ross Andersen, to discuss his views on technology—and why he hasn’t made a film about AI … yet.  

— Damon

Editor’s Note: This article is part of “If Trump Wins,” a project considering what Donald Trump might do if reelected in 2024.

If Donald Trump wins the next election, he will attempt to turn the men and women of the United States armed forces into praetorians loyal not to the Constitution, but only to him. This project will likely be among his administration’s highest priorities. It will not be easy: The overwhelming majority of America’s service people are professionals and patriots. I know this from teaching senior officers for 25 years at the Naval War College. As president, Trump came to understand it too, when he found that “his generals” were not, in fact, mere employees of a Trump property.

But the former president and the people around him have learned from that experience. The last time around, Trump’s efforts to pack the Defense Department with cranks and flunkies came too late to bring the military under his full political control. The president and his advisers were slow-footed and disorganized, and lacked familiarity with Washington politics. They were hindered as well by the courage and professionalism of the military officers and civilian appointees who, side by side, serve in the Defense Department.

Trump now nurses deep grudges against these officers and civilians, who slow-rolled and smothered his various illegal and autocratic impulses, including his enraged demand to kill the Syrian leader Bashar al-Assad in 2017, and his desire to deploy America’s military against its own citizens during the Black Lives Matter protests in the summer of 2020.

The 2020 election, of course, is the source of Trump’s chief grudge against senior military leaders. General Mark Milley, then the chairman of the Joint Chiefs of Staff, was especially determined to keep the armed forces out of the various schemes to stay in office devised by the Trump team and its allies, including a delusional plan, proposed by retired Lieutenant General Michael Flynn, to have the military go into swing states and seize voting machines. Trump has since implied (in response to a profile of Milley by The Atlantic’s Jeffrey Goldberg) that Milley should get the death penalty. Milley reportedly believes that Trump, if reelected, will try to jail him and other senior national-security figures, a concern shared by former Secretary of Defense Mark Esper and former Director of National Intelligence James Clapper.

[From the November 2023 issue: Jeffrey Goldberg on how Mark Milley held the line]

In a second term, Trump would combine his instincts for revenge and self-protection. He would seek not only to get even with an officer corps that he thinks betrayed him, but also to break the military as one of the few institutions able to constrain his attempts to act against the Constitution and the rule of law.

Publicly, Trump presents himself as an unflinching advocate for the military, but this is a charade. He has no respect for military people or their devotion to duty. He loves the pomp and the parades and the salutes and the continual use of “sir,” but as retired Marine General John Kelly, Trump’s former chief of staff, said in 2023, Trump “couldn’t fathom people who served their nation honorably” when he was in office. Privately, as Goldberg has reported, Trump has called American war dead “losers” and “suckers,” and has said that wounded warriors are disgusting and should be kept out of sight.

Trump instead prizes military people who serve his ego and support his antidemocratic instincts. He thinks highly of Flynn, for example, who had to resign after 22 days as national security adviser and is now the marquee attraction at various gatherings of Christian nationalists and conspiracy theorists around the country. In late 2020, angered by his election loss and what he saw as the disloyalty within the national-security community, Trump fired or forced out top Defense Department leaders and tried to replace them with people more like Flynn. The brazen actions that the 45th president took in his final, desperate weeks in office—however haphazard—illustrate the magnitude of the threat he may pose to the military if he is reelected.

On November 9, 2020, Trump dumped Esper and named Christopher Miller, a retired colonel and Pentagon bureaucrat, as acting secretary of defense. Miller took along Kash Patel, a Trump sycophant, as his chief of staff. Trump sent Douglas Macgregor, another retired colonel and a pro-Russia Fox News regular, to Miller as a senior adviser. (Earlier, Trump had attempted and failed to make Macgregor the ambassador to Germany.) Trump installed Anthony Tata—a retired one-star Army general who has claimed that Barack Obama is a Muslim and that a former CIA director was trying to have Trump assassinated—in the third-most-senior job at the Pentagon. A few months earlier, the Senate had wisely declined to confirm Tata’s appointment to that position, but in November, Trump gave him the job in an acting capacity anyway.

These moves, among others, led all 10 living former secretaries of defense to issue a startling and unprecedented joint statement. On January 3, 2021, they directly enjoined Miller and his subordinates to uphold their constitutional duty and “refrain from any political actions that undermine the results of the election or hinder the success of the new team.” The letter pointedly reminded Miller and his team that they were “bound by oath, law and precedent,” and called upon them, “in the strongest terms,” to honor “the history of democratic transition in our great country.”

If reelected, Trump would attempt to gain authoritarian control of the Defense Department’s uppermost levels from the very beginning. There are more Anthony Tatas and Douglas Macgregors out there, and Trump’s allies are likely already seeking to identify them. If the Senate refused to confirm Trump’s appointees, it wouldn’t matter much: Trump has learned that he can keep rotating people through acting positions, daring the Senate to stop him.

The career civil servants underneath these appointees—who work on everything from recruiting to nuclear planning—would disobey Trump if he attacked the constitutional order. These civilians, by law, cannot be fired at will, a problem Trump tried to remedy in the last months of his administration by proposing a new category of government appointments (Schedule F) that would have converted some of the most important civil-service positions into political appointments directly controlled by the White House. President Joe Biden immediately repealed this move after taking office, but Trump has vowed to reinstate it.

[Read: Trump’s open plot to dismantle the federal government]

In his two-pronged offensive to capture the military establishment while eviscerating the civil service, Trump would likely rely on former officers such as Miller and fringe-dwelling civilians such as Patel, but he would also almost certainly find at least a few serving senior officers—he would not need many—who would accept his offer to abandon their oath. Together, they would make a run at changing the nature of the armed forces.

This is not abstract theorizing. The Heritage Foundation recently released “Project 2025,” a right-wing blueprint for the next Republican president’s administration. The Defense Department chapter was written by none other than former Acting Secretary Christopher Miller. It is mostly a rationalization for more spending, but it includes a clear call for a purge of the military’s senior ranks to clean out “Marxist indoctrination”—an accusation he does not define—along with demands for expelling trans service members and reinstating those service members who were dismissed for refusing COVID vaccinations.

The problems of ideological polarization and extremism in the armed forces are not as extensive as some critics of the military imagine, but they are more worrisome than the military leadership would like to admit. Military officers tend to be more conservative than the public, and as far back as the Clinton and Obama administrations, I occasionally heard senior officers speak of these liberal presidents in deeply contemptuous terms (potentially a crime under military regulations). Today, military bases are subjected to a constant barrage of Fox News in almost every area with a television, and toward the end of my teaching career (I retired in 2022), I often heard senior officers repeating almost verbatim some of the most overheated and paranoid talking points about politics and national affairs from the network’s prime-time hosts. Some of these officers would be tempted to answer Trump’s call.

The rest of the members of the professional military, despite their concerns, would likely follow their instincts and default to the orders of their chain of command. The American political system was never intended to cope with someone like Trump; the military is trained and organized to obey, not resist, the orders of the civilian commander in chief.

Trump’s plans would likely use this obedience to the chain of command to exploit an unfortunate vulnerability in the modern American armed forces: The military, in my experience, has a political-literacy problem. Too many people in uniform no longer have a basic grounding in the constitutional foundation of American government and the civil-military relationship. (Some of my colleagues who teach in senior-military educational institutions share this concern, and over the years, some of us have tried, often in vain, to push more study of the Constitution into the curricula.) These men and women are neither unintelligent nor disloyal. Rather, like many Americans, they are no longer taught basic civics, and they may struggle with the line between executing the orders of the president as the commander in chief and obeying the Constitution.

Trump’s appointees also would be able to influence the future of the armed forces through assignments and promotions (and non-promotions) within each branch—and through their behavior as examples to the rest of the military. With top cover from the White House, Trump’s functionaries in the Pentagon, working with his supporters in the ranks, could poison the military for years to come by ignoring laws, regulations, and traditions as they see fit. (Recall, for example, that Trump is an admirer of the disgraced Navy SEAL Eddie Gallagher, and intervened to make sure Gallagher kept his SEAL Trident after he was charged with war crimes and found guilty of posing for photos with a captive’s dead body.) America’s military is built on virtues such as honor and duty, but abusing and discarding the norms that support those virtues would change the military’s culture—and faster than we may realize.

Even if only some of the actions I’ve described here succeed, any number of disasters might follow. Trump could jeopardize national security by surrounding himself with military and defense officials who would help him dissolve our alliances (especially NATO), weaken our military readiness, undermine our intelligence services, and abandon our friends around the world, all while he seeks closer relations with authoritarian regimes—especially Vladimir Putin’s Russia. He could issue illegal orders to engage in torture or to commit other war crimes overseas. And he could bring the entire planet to disaster should senior military leaders obey his unhinged orders to kill foreign leaders, start a war, or even use nuclear weapons.

[From the January/February 2024 issue: Anne Applebaum on how Trump will abandon NATO]

At home, Trump could order unconstitutional shows of military support for his administration to intimidate his opponents. He could order American soldiers into the streets against protesters. (Trump’s allies are reportedly drawing up plans to invoke the Insurrection Act on Inauguration Day to quell any demonstrations against his return to office.) Officers refusing such orders could be dismissed or reassigned, which in turn could provoke a political confrontation between the Trump loyalists in the high command and the rest of the armed forces, itself a frightening and previously unthinkable prospect.

And if Trump succeeds in simultaneously capturing the U.S. military while gutting the other key institutions that protect democracy—especially the courts and the Justice Department—nothing will stop him from using force to put down opposition and stay in power.

Some Americans fear that the United States is already in a struggle with fascism. The firm constitutional loyalty of the armed forces during Trump’s presidency was a reminder that such fears are overblown, at least for the moment. But Trump and his allies understand that by leaving the military outside their political control the last time around, they also left intact a crucial bulwark against their plans. They will not make the same mistake twice.

This article appears in the January/February 2024 print edition with the headline “A Military Loyal to Trump.”

Brilliant, witty, and ambitious, Henry Kissinger made diplomacy the stuff of unrivaled celebrity. He thrived on attention, and would have been thrilled by  the flood of coverage that marked his death last week. Whether the obituaries and commentaries put his record in a positive or negative light, almost all of them treated Kissinger as the master of events.

This may be how he wanted to be remembered, but it’s not what really happened. No matter how often Kissinger is described as the Cold War’s most powerful secretary of state and a peerless elder statesman, the truth is that his tenure was often rocky, as full of setbacks as acclaim. By the time he left government, he was viewed by many of his colleagues as a burden, not an asset. Once out of office, the advice he gave his successors was sometimes spectacularly wrong, and frequently ignored.

In President Richard Nixon’s first term, Kissinger presided over three big diplomatic transformations—withdrawal from Vietnam, the opening to China, and détente with the Soviet Union. When he became secretary of state, his policy dominance was virtually unchallenged.  He was the first (and, to this day, only) person ever to run the State Department while serving simultaneously as the president’s national security adviser. Outside of government, he enjoyed unprecedented global renown. Less than a month after his Senate confirmation, he won the Nobel Peace Prize.

[Shan Wang: Henry Kissinger’s real legacy]

Yet when Kissinger left office barely three years later, most of his ambitious schemes were unrealized. Others had simply been rejected. On the left, many revile Kissinger for the human costs of the policies he pursued; on the right, some still admire his unsentimental use of military force. In fact, the real story of Kissinger’s tenure as secretary of state is a tale in which, again and again, he encountered the limits of his power, and found himself unable to impose his will.

The policies Kissinger developed largely in secret to help wind down the Vietnam War enjoyed far less support once the war was over and they were subjected to more normal, open debate. His influence ebbed steadily. In 1975, Gerald Ford, who had succeeded Nixon a year earlier, forced Kissinger to give up the national-security job. Ford created further checks on Kissinger’s power by picking two former congressional colleagues, Donald Rumsfeld and George H. W. Bush, as secretary of defense and CIA director, respectively. Congress itself voted into law a series of challenges to Kissinger’s policies, something it had consistently failed to do under Nixon. Perhaps worst of all, the secretary of state bore some of the blame for Ford’s defeat in the 1976 election. The president’s campaign managers told reporters they saw him as a vulnerability. So did Ronald Reagan, whose bid for the Republican nomination centered in part on a promise to fire Kissinger.

Kissinger’s lost dominance was especially pronounced in what was arguably the central arena of his policy: the stable relationship—known as “détente”—that he sought to establish with the Soviet Union. His problems began with arms control. In November 1974, soon after Ford became president, Kissinger arranged a quick summit with the Soviet leader Leonid Brezhnev, hoping for a breakthrough in negotiating a long-term treaty to limit each side’s strategic nuclear forces. But he was never able to turn the framework they agreed on into a real treaty. One obstacle was a congressional requirement that U.S. and Soviet forces be equal—at a time when Soviet missiles were getting steadily bigger and more numerous. Outside experts claimed that Kissinger’s framework couldn’t meet that test. Writing in Foreign Affairs, Paul Nitze—a senior national-security official under Harry Truman, John F. Kennedy, and Lyndon B. Johnson—insisted it would give Soviet forces a three-to-one advantage. (Privately, Nitze was far angrier, calling the secretary of state a “traitor to his country.”)

Even harder for Kissinger to handle was opposition within Ford’s inner circle. Rumsfeld, once he became defense secretary, was ready to take disagreements with Kissinger right into the Oval Office, telling the president that the United States had been losing its nuclear edge for a decade. At the CIA, Bush approved an assessment largely endorsing Nitze’s critique. Outside the administration, Reagan echoed the same charges. No surprise, then, that Ford eventually put the talks aside.

Kissinger found the ideological dimension of Soviet-American relations still more vexing. He had promised Soviet leaders to expand trade ties by granting Moscow “Most Favored Nation” tariff status, but he could not manage congressional demands for freer emigration from the Soviet Union. The initiative collapsed, but not before senior figures in both Congress and the Kremlin concluded that Kissinger had been deceiving them. On human rights more generally, the secretary of state was isolated within his own administration. He did persuade the president not to meet with Aleksandr Solzhenitsyn, the most famous and outspoken Soviet dissident, but three other members of the Ford Cabinet defied him and conspicuously attended an AFL-CIO dinner in Solzhenitsyn’s honor. Even the young Dick Cheney, then the deputy White House chief of staff, dissented: Détente, he argued, didn’t have to be all “sweetness and light.”

[Gary J. Bass: The people who didn’t matter to Henry Kissinger]

Learning little from this opposition, Kissinger continued to hurt himself with scarcely concealed disdain for opponents of the Soviet regime. (“You know,” he once joked, “what would have happened to them under Stalin.”) The impact reached well beyond Washington. When Reagan delegates to the 1976 Republican convention wanted to repudiate Kissinger, they drafted a platform plank titled “Morality in Foreign Policy.” Ford and his advisers—who had already banned official use of the word détente—felt they had to allow it to pass.

Apart from arms control and human rights, Kissinger also had trouble imposing his views on Soviet-American competition in the Third World. When he wanted to launch a covert program to arm rebels against Moscow’s client regime in Angola, news quickly leaked to The New York Times. Congressional Democrats, predictably, voted to block the weapons transfer altogether. Less predictably, many Republican senators—liberals, moderates, and conservatives alike—also joined in, giving the measure a two-to-one majority. The president’s own party was deserting its celebrity diplomat.

Kissinger was furious, just as he had been earlier in 1975 when, with the fall of Saigon approaching, he proposed a big increase in arms supplies for South Vietnam. To make it happen, however, congressional approval was necessary—and again wanting. Ford ultimately chose not to fight the issue. Instead, in a speech at Tulane University, he declared the war “finished as far as America is concerned.” The White House did not even let Kissinger know that the game-over announcement was coming.

Much of the commentary on Kissinger’s career has presented him as the embodiment of unchecked presidential power over foreign policy. But the pushback against his policies grew steadily stronger as their downsides became better known. In the 1970s, Congress became far more assertive on foreign policy, legislating issues including arms control, human rights, foreign military sales, and covert action. Kissinger frequently railed against the decade-long decline in national-security budgets, but this too was part of his legacy. So were other institutional reforms, such as the Carter administration’s creation of a human-rights bureau in the State Department and the annual publication of global-human-rights reports. Other forms of pushback were less foreseeable: The “most powerful secretary of state in the post-World War II era” surely never imagined what Jimmy Carter’s high-profile envoy to China—Leonard Woodcock, the former head of the United Auto Workers—would tell his Beijing staff at their first meeting: “Never again shall we embarrass ourselves before a foreign nation the way Henry Kissinger did with the Chinese.”

After he left office, Kissinger kept much of the advice he gave his successors confidential, probably thinking that a little mystery about the extent of his influence would only help his new consulting business. But enough is known about some of his Oval Office meetings to challenge the common picture of presidents and advisers listening reverently while Henry Kissinger shared his wisdom. Kissinger’s sustained effort to reorient Reagan’s policies toward the Soviet Union provides a striking example. Together with Nixon, he argued that Mikhail Gorbachev was cynically exploiting the president’s naive antinuclear sentiments so as to tear apart the Western alliance. Under perestroika, they argued, the Soviet threat was actually increasing, not diminishing. Reagan ignored them—and over time harvested a global Soviet foreign-policy retreat.

Kissinger’s shortfalls in office and after are not the whole story, of course. In his first weeks as secretary of state, he was plunged into a crisis—Egypt’s surprise Yom Kippur attack on Israel, followed by the OPEC oil embargo. The cease-fire and disengagement agreements he negotiated bolstered American influence in the Middle East, a region to which he had paid little previous attention. He seemed, to quote the title of my colleague Martin Indyk’s recent book, the “master of the game.”

Yet here, too, the master’s record seems ripe for reassessment—and not just for his early, forgivable missteps. At the start of the Yom Kippur war, Kissinger thought it might be best to keep a low profile and meet Israel’s needs indirectly, by contracting with private companies to deliver arms. Nixon ordered his celebrity policy maker to stop dithering and organize a U.S. airlift. “Do it now!” he barked. More serious is the charge that, even at the height of his power, Kissinger had, of all things, a too-limited conception of what diplomacy could achieve. The most it should try to accomplish, he felt, was to stabilize the world, not to alter—much less transform—it. Hence, the secretary of state was reluctant to take on the hardest parts of the Middle East puzzle—above all, the clash between Israelis and Palestinians, still atop the headlines half a century later.

[From the December 2016 issue: The lessons of Henry Kissinger]

Indyk traces Kissinger’s hesitation to the same sources others have cited: his conservative view of history, his immersion as a scholar in the diplomacy of 19th-century Europe, and his personal experience of 20th-century totalitarianism. All of these drove home the value of stability. But, in looking to explain this conception of diplomacy, we should not leave out what Kissinger surely learned from his own bumpy record as secretary of state. No matter what the tributes and obituaries say, every day on the job confirmed the limits of his power, the difficulty of overcoming them, and his ability to make mistakes when he tried to do so.

US troops and installations have come under regular attacks due to Washington's support of Israel's bombardment of Gaza.