Group items tagged

There’s plenty of blame to go around, involving Democrats as well as Republicans, but Trump in particular “recklessly squandered lives,” in the words of an unusual editorial this month in the New England Journal of Medicine. Death certificates may record the coronavirus as the cause of death, but in a larger sense vast numbers of Americans died because their government was incompetent.

As many Americans are dying every 10 days of Covid-19 as U.S. troops died during 19 years of war in Iraq and Afghanistan

The paradox is that a year ago, the United States seemed particularly well positioned to handle this kind of crisis. A 324-page study by Johns Hopkins found last October that the United States was the country best prepared for a pandemic.

Then there’s an immeasurable cost in soft power as the United States is humbled before the world.

“It’s really sad to see the U.S. presidency fall from being the champion of global health to being the laughingstock of the world,”

in terms of destruction of American lives, treasure and well-being, this pandemic may be the greatest failure of governance in the United States since the Vietnam War.

the economists David Cutler and Lawrence Summers estimate that the economic cost of the pandemic in the United States will be $16 trillion, or about $125,000 per American household — far more than the median family’s net worth.

Credit for that goes to President George W. Bush, who in the summer of 2005 read an advance copy of “The Great Influenza,” a history of the 1918 flu pandemic. Shaken, Bush pushed aides to develop a strategy to prepare for another great contagion, and the result was an excellent 396-page playbook for managing such a health crisis.

The Obama administration updated this playbook and in the presidential transition in 2016, Obama aides cautioned the Trump administration that one of the big risks to national security was a contagion. Private experts repeated similar warnings. “Of all the things that could kill 10 million people or more, by far the most likely is an epidemic,” Bill Gates warned in 2015.

It’s true that the Obama administration did not do enough to refill the national stockpile with N95 masks, but Republicans in Congress wouldn’t provide even the modest sums that Obama requested for replenishment. And the Trump administration itself did nothing in its first three years to rebuild stockpiles.

Trump argues that no one could have anticipated the pandemic, but it’s what Bush warned about, what Obama aides tried to tell their successors about, and what Joe Biden referred to in a blunt tweet in October 2019 lamenting Trump’s cuts to health security programs and adding: “We are not prepared for a pandemic.”

When the health commission of Wuhan, China, announced on Dec. 31 that it had identified 27 cases of a puzzling pneumonia, Taiwan acted with lightning speed. Concerned that this might be an outbreak of SARS, Taiwan dispatched health inspectors to board flights arriving from Wuhan and screen passengers before allowing them to disembark. Anyone showing signs of ill health was quarantined.

If either China or the rest of the world had shown the same urgency, the pandemic might never have happened.

In hindsight, two points seem clear: First, China initially covered up the scale of the outbreak. Second, even so, the United States and other countries had enough information to act as Taiwan did. The first two countries to impose travel restrictions on China were North Korea and the Marshall Islands, neither of which had inside information.

That first half of January represents a huge missed opportunity for the world. If the United States, the World Health Organization and the world media had raised enough questions and pressed China, then perhaps the Chinese central government would have intervened in Wuhan earlier. And if Wuhan had been locked down just two weeks earlier, it’s conceivable that this entire global catastrophe could have been averted.

the C.D.C. devised a faulty test, and turf wars in the federal government prevented the use of other tests. South Korea, Germany and other countries quickly developed tests that did work, and these were distributed around the world. Sierra Leone in West Africa had effective tests before the United States did.

It’s true that local politicians, Democrats and Republicans alike, made disastrous decisions, as when Mayor Bill de Blasio of New York City urged people in March to “get out on the town despite coronavirus.” But local officials erred in part because of the failure of testing: Without tests, they didn’t know what they faced.

t’s unfair to blame the testing catastrophe entirely on Trump, for the failures unfolded several pay grades below him. Partly that’s because Trump appointees, like Robert Redfield, director of the C.D.C., simply aren’t the A team.

In any case, presidents set priorities for lower officials. If Trump had pushed aides as hard to get accurate tests as he pushed to repel refugees and migrants, then America almost certainly would have had an effective test by the beginning of February and tens of thousands of lives would have been saved.

Still, testing isn’t essential if a country gets backup steps right. Japan is a densely populated country that did not test much and yet has only 2 percent as many deaths per capita as the United States. One reason is that Japanese have long embraced face masks, which Dr. Redfield has noted can be at least as effective as a vaccine in fighting the pandemic. A country doesn’t have to do everything, if it does some things right.

Yet in retrospect, Trump did almost everything wrong. He discouraged mask wearing. The administration never rolled out contact tracing, missed opportunities to isolate the infected and exposed, didn’t adequately protect nursing homes, issued advice that confused the issues more than clarified them, and handed responsibilities to states and localities that were unprepared to act.

Trump’s missteps arose in part because he channeled an anti-intellectual current that runs deep in the United States, as he sidelined scientific experts and responded to the virus with a sunny optimism apparently meant to bolster the financial markets.

The false reassurances and dithering were deadly. One study found that if the United States had simply imposed the same lockdowns just two weeks earlier, 83 percent of the deaths in the early months could have been prevented.

A basic principle of public health is the primacy of accurate communications based on the best science. Chancellor Angela Merkel of Germany, who holds a doctorate in physics, is the global champion of that approach

Trump was the opposite, sowing confusion and conspiracy theories; a Cornell study found that “the President of the United States was likely the largest driver of the Covid-19 misinformation.”

A conservative commentariat echoed Trump in downplaying the virus and deriding efforts to stay safe.

A University of Chicago study found that watching the Sean Hannity program correlated to less social distancing, so watching Fox News may well have been lethal to some of its fans.

Americans have often pointed to the Soviet Union as a place where ideology trumped science, with disastrous results. Stalin backed Trofim Lysenko, an agricultural pseudoscientist who was an ardent Communist but scorned genetics — and whose zealous incompetence helped cause famines in the Soviet Union. Later, in the 1980s, Soviet leaders were troubled by data showing falling life expectancy — so they banned the publication of mortality statistics

It was in the same spirit that Trump opposed testing for the coronavirus in the hope of holding down the number of reported cases.

Most striking, Trump still has never developed a comprehensive plan to fight Covid-19. His “strategy” was to downplay the virus and resist business closures, in an effort to keep the economy roaring — his best argument for re-election.

This failed. The best way to protect the economy was to control the virus, not to ignore it, and the spread of Covid-19 caused economic dislocations that devastated even homes where no one was infected.

Eight million Americans have slipped into poverty since May, a Columbia University study found, and about one in seven households with children have reported to the census that they didn’t have enough food to eat in the last seven days.

Yet today we can’t even churn out enough face masks; a poll of nurses in late July and early August found that one-third lacked enough N95 masks

More than one-quarter of young adults said they have seriously contemplated suicide

So in what is arguably the richest country in the history of the world, political malpractice has resulted in a pandemic of infectious disease followed by pandemics of poverty, mental illness, addiction and hunger.

The rejection of science has also exacerbated polarization and tribalism

An old school friend shared this conspiracy theory on Facebook:Create a VIRUS to scare people. Place them in quarantine. Count the number of dead every second of every day in every news headline. Close all businesses …. Mask people. Dehumanize them. Close temples and churches …. Empty the prisons because of the virus and fill the streets with criminals. Send in Antifa to vandalize property as if they are freedom fighters. Undermine the law. Loot …. And, in an election year, have Democrats blame all of it on the President. If you love America, our Constitution, and the Rule of Law, get ready to fight for them.

During World War II, American soldiers died at a rate of 9,200 a month, less than one-third the pace of deaths from this pandemic, but the United States responded with a massive mobilization

More than 40 percent of adults reported in June that they were struggling with mental health, and 13 percent have begun or increased substance abuse, a C.D.C. study found

Trump and his allies have even argued against mobilization. “Don’t be afraid of Covid,” Trump tweeted this month. “Don’t let it dominate your life.”

It didn’t have to be this way. If the U.S. had worked harder and held the per capita mortality rate down to the level of, say, Germany, we could have saved more than 170,000 lives

And if the U.S. had responded urgently and deftly enough to achieve Taiwan’s death rate, fewer than 100 Americans would have died from the virus.

“It is a slaughter,” Dr. William Foege, a legendary epidemiologist who once ran the C.D.C., wrote to Dr. Redfield. Dr. Foege predicted that public health textbooks would study America’s response to Covid-19 not as a model of A-plus work but as an example of what not to do.

At fifty-nine, Anderson is the chair of the University of Michigan’s department of philosophy and a champion of the view that equality and freedom are mutually dependent, enmeshed in changing conditions through time.

She has built a case, elaborated across decades, that equality is the basis for a free society

Because she brings together ideas from both the left and the right to battle increasing inequality, Anderson may be the philosopher best suited to this awkward moment in American life. She builds a democratic frame for a society in which people come from different places and are predisposed to disagree.

she sketched out the entry-level idea that one basic way to expand equality is by expanding the range of valued fields within a society.

The ability not to have an identity that one carries from sphere to sphere but, rather, to be able to slip in and adopt whatever values and norms are appropriate while retaining one’s identities in other domains?” She paused. “That is what it is to be free.”

How do you move from a basic model of egalitarian variety, in which everybody gets a crack at being a star at something, to figuring out how to respond to a complex one, where people, with different allotments of talent and virtue, get unequal starts, and often meet with different constraints along the way?

The problem, she proposed, was that contemporary egalitarian thinkers had grown fixated on distribution: moving resources from lucky-seeming people to unlucky-seeming people, as if trying to spread the luck around.

Egalitarians should agree about clear cases of blameless misfortune: the quadriplegic child, the cognitively impaired adult, the teen-ager born into poverty with junkie parents. But Anderson balked there, too. By categorizing people as lucky or unlucky, she argued, these egalitarians set up a moralizing hierarchy.

In Anderson’s view, the way forward was to shift from distributive equality to what she called relational, or democratic, equality: meeting as equals, regardless of where you were coming from or going to.

By letting the lucky class go on reaping the market’s chancy rewards while asking others to concede inferior status in order to receive a drip-drip-drip of redistributive aid, these egalitarians were actually entrenching people’s status as superior or subordinate.

To the ugly and socially awkward: . . . Maybe you won’t be such a loser in love once potential dates see how rich you are.

. To the stupid and untalented: Unfortunately, other people don’t value what little you have to offer in the system of production. . . . Because of the misfortune that you were born so poorly endowed with talents, we productive ones will make it up to you: we’ll let you share in the bounty of what we have produced with our vastly superior and highly valued abilities. . . 

she imagined some citizens getting a state check and a bureaucratic letter:

This was, at heart, an exercise of freedom. The trouble was that many people, picking up on libertarian misconceptions, thought of freedom only in the frame of their own actions.

To be truly free, in Anderson’s assessment, members of a society had to be able to function as human beings (requiring food, shelter, medical care), to participate in production (education, fair-value pay, entrepreneurial opportunity), to execute their role as citizens (freedom to speak and to vote), and to move through civil society (parks, restaurants, workplaces, markets, and all the rest).

Anderson’s democratic model shifted the remit of egalitarianism from the idea of equalizing wealth to the idea that people should be equally free, regardless of their differences.

A society in which everyone had the same material benefits could still be unequal, in this crucial sense; democratic equality, being predicated on equal respect, wasn’t something you could simply tax into existence. “People, not nature, are responsible for turning the natural diversity of human beings into oppressive hierarchies,”

Her first book, “Value in Ethics and Economics,” appeared that year, announcing one of her major projects: reconciling value (an amorphous ascription of worth that is a keystone of ethics and economics) with pluralism (the fact that people seem to value things in different ways).

Philosophers have often assumed that pluralistic value reflects human fuzziness—we’re loose, we’re confused, and we mix rational thought with sentimental responses.

She offered an “expressive” theory: in her view, each person’s values could be various because they were socially expressed, and thus shaped by the range of contexts and relationships at play in a life. Instead of positing value as a basic, abstract quality across society (the way “utility” functioned for economists), she saw value as something determined by the details of an individual’s history.

Like her idea of relational equality, this model resisted the temptation to flatten human variety toward a unifying standard. In doing so, it helped expand the realm of free and reasoned economic choice.

Anderson’s model unseated the premises of rational-choice theory, in which individuals invariably make utility-maximizing decisions, occasionally in heartless-seeming ways. It ran with, rather than against, moral intuition. Because values were plural, it was perfectly rational to choose to spend evenings with your family, say, and have guilt toward the people you left in the lurch at work.

The theory also pointed out the limits on free-market ideologies, such as libertarianism.

In ethics, it broke across old factional debates. The core idea “has been picked up on by people across quite a range of positions,” Peter Railton, one of Anderson’s longtime colleagues, says. “Kantians and consequentialists alike”—people who viewed morality in terms of duties and obligations, and those who measured the morality of actions by their effects in the world—“could look at it and see something important.”

Traditionally, the discipline is taught through a-priori thought—you start with basic principles and reason forward. Anderson, by contrast, sought to work empirically, using information gathered from the world, identifying problems to be solved not abstractly but through the experienced problems of real people.

“Dewey argued that the primary problems for ethics in the modern world concerned the ways society ought to be organized, rather than personal decisions of the individual,”

In 2004, the Stanford Encyclopedia of Philosophy asked Anderson to compose its entry on the moral philosophy of John Dewey, who helped carry pragmatist methods into the social realm. Dewey had an idea of democracy as a system of good habits that began in civil life. He was an anti-ideologue with an eye for pluralism.

She started working with historians, trying to hone her understanding of ideas by studying them in the context of their creation. Take Rousseau’s apparent support of direct democracy. It’s rarely mentioned that, at the moment when he made that argument, his home town of Geneva had been taken over by oligarchs who claimed to represent the public. Pragmatism said that an idea was an instrument, which naturally gave rise to such questions as: an instrument for what, and where, and when?

In “What Is the Point of Equality?,” Anderson had already started to drift away from what philosophers, following Rawls, call ideal theory, based on an end vision for a perfectly just society. As Anderson began a serious study of race in America, though, she found herself losing faith in that approach entirely.

Broadly, there’s a culturally right and a culturally left ideal theory for race and society. The rightist version calls for color blindness. Instead of making a fuss about skin and ethnicity, its advocates say, society should treat people as people, and let the best and the hardest working rise.

The leftist theory envisions identity communities: for once, give black people (or women, or members of other historically oppressed groups) the resources and opportunities they need, including, if they want it, civil infrastructure for themselves.

In “The Imperative of Integration,” published in 2010, Anderson tore apart both of these models. Sure, it might be nice to live in a color-blind society, she wrote, but that’s nothing like the one that exists.

But the case for self-segregation was also weak. Affinity groups provided welcome comfort, yet that wasn’t the same as power or equality, Anderson pointed out. And there was a goose-and-gander problem. Either you let only certain groups self-segregate (certifying their subordinate status) or you also permitted, say, white men to do it,

Anderson’s solution was “integration,” a concept that, especially in progressive circles, had been uncool since the late sixties. Integration, by her lights, meant mixing on the basis of equality.

in attending to these empirical findings over doctrine, she announced herself as a non-ideal theorist: a philosopher with no end vision of society. The approach recalls E. L. Doctorow’s description of driving at night: “You can see only as far as the headlights, but you can make the whole trip that way.”

or others, though, a white woman making recommendations on race policy raised questions of perspective. She was engaging through a mostly white Anglo-American tradition. She worked from the premise that, because she drew on folders full of studies, the limits of her own perspective were not constraining.

Some philosophers of color welcomed the book. “She’s taking the need for racial justice seriously, and you could hardly find another white political philosopher over a period of decades doing that,”

Recently, Anderson changed the way she assigns undergraduate essays: instead of requiring students to argue a position and fend off objections, doubling down on their original beliefs, she asks them to discuss their position with someone who disagrees, and to explain how and why, if at all, the discussion changed their views.

The challenge of pluralism is the challenge of modern society: maintaining equality amid difference in a culture given to constant and unpredictable change.

Rather than fighting for the ascendancy of certain positions, Anderson suggests, citizens should fight to bolster healthy institutions and systems—those which insure that all views and experiences will be heard. Today’s righteous projects, after all, will inevitably seem fatuous and blinkered from the vantage of another age.

Smith saw the markets as an escape from that order. Their “most important” function, he explained, was to bring “liberty and security” to those “who had before lived almost in a continual state of war with their neighbours, and of servile dependency upon their superiors.”

Anderson zeroed in on Adam Smith, whose “The Wealth of Nations,” published in 1776, is taken as a keystone of free-market ideology. At the time, English labor was subject to uncompensated apprenticeships, domestic servitude, and some measure of clerical dominion.

Smith, in other words, was an egalitarian. He had written “The Wealth of Nations” in no small part to be a solution to what we’d now call structural inequality—the intractable, compounding privileges of an arbitrary hierarchy.

It was a historical irony that, a century later, writers such as Marx pointed to the market as a structure of dominion over workers; in truth, Smith and Marx had shared a socioeconomic project. And yet Marx had not been wrong to trash Smith’s ideas, because, during the time between them, the world around Smith’s model had changed, and it was no longer a useful tool.

mages of free market society that made sense prior to the Industrial Revolution continue to circulate today as ideals, blind to the gross mismatch between the background social assumptions reigning in the seventeenth and eighteenth centuries, and today’s institutional realities. We are told that our choice is between free markets and state control, when most adults live their working lives under a third thing entirely: private government.

Today, people still try to use, variously, both Smith’s and Marx’s tools on a different, postindustrial world:

The unnaturalness of this top-heavy arrangement, combined with growing evidence of power abuses, has given many people reason to believe that something is fishy about the structure of American equality. Socialist and anti-capitalist models are again in vogue.

Anderson offers a different corrective path. She thinks it’s fine for some people to earn more than others. If you’re a brilliant potter, and people want to pay you more than the next guy for your pottery, great!

The problem isn’t that talent and income are distributed in unequal parcels. The problem is that Jeff Bezos earns more than a hundred thousand dollars a minute, while Amazon warehouse employees, many talented and hardworking, have reportedly resorted to urinating in bottles in lieu of a bathroom break. That circumstance reflects some structure of hierarchical oppression. It is a rip in the democratic fabric, and it’s increasingly the norm.

Andersonism holds that we don’t have to give up on market society if we can recognize and correct for its limitations—it may even be our best hope, because it’s friendlier to pluralism than most alternatives are.

we must be flexible. We must remain alert. We must solve problems collaboratively, in the moment, using society’s ears and eyes and the best tools that we can find.

“You can see that, from about 1950 to 1970, the typical American’s wages kept up with productivity growth,” she said. Then, around 1974, she went on, hourly compensation stagnated. American wages have been effectively flat for the past few decades, with the gains of productivity increasingly going to shareholders and to salaries for big bosses.

What changed? Anderson rattled off a constellation of factors, from strengthened intellectual-property law to winnowed antitrust law. Financialization, deregulation. Plummeting taxes on capital alongside rising payroll taxes. Privatization, which exchanged modest public-sector salaries for C.E.O. paydays. She gazed into the audience and blinked. “So now we have to ask: What has been used to justify this rather dramatic shift of labor-share of income?”

It was no wonder that industrial-age thinking was riddled with contradictions: it reflected what Anderson called “the plutocratic reversal” of classical liberal ideas. Those perversely reversed ideas about freedom were the ones that found a home in U.S. policy, and, well, here we were.

well-meaning people everywhere from statehouses to your local middle school have responded to this surging anti-Semitism by doubling down on Holocaust education. Before 2016, only seven states required Holocaust education in schools. In the past seven years, 18 more have passed Holocaust-education mandates

These casual stories sickened me in their volume and their similarity, a catalog of small degradations. At a time when many people in other minority groups have become bold in publicizing the tiniest of slights, these American Jews instead expressed deep shame in sharing these stories with me, feeling that they had no right to complain. After all, as many of them told me, it wasn’t the Holocaust.

These people talked about bosses and colleagues who repeatedly ridiculed them with anti-Semitic “jokes,” friends who turned on them when they mentioned a son’s bar mitzvah or a trip to Israel, romantic partners who openly mocked their traditions, classmates who defaced their dorm rooms and pilloried them online, teachers and neighbors who parroted conspiratorial lies. I was surprised to learn how many people were getting pennies thrown at them in 21st-century Americ

the blood libel, which would later be repurposed as a key part of the QAnon conspiracy theory. This craze wasn’t caused by one-party control over printing presses, but by the lie’s popularity

I have come to the disturbing conclusion that Holocaust education is incapable of addressing contemporary anti-Semitism. In fact, in the total absence of any education about Jews alive today, teaching about the Holocaust might even be making anti-Semitism worse.

The Illinois Holocaust Museum & Education Center is a victim of its own success. When I arrived on a weekday morning to join a field trip from a local Catholic middle school, the museum was having a light day, with only 160 students visiting

the docent established that the ’30s featured media beyond town criers, and that one-party control over such media helped spread propaganda. “If radio’s controlled by a certain party, you have to question that,” she said. “Back then, they didn’t.”

I wondered about that premise. Historians have pointed out that it doesn’t make sense to assume that people in previous eras were simply stupider than we are, and I doubted that 2020s Americans could outsmart 1930s Germans in detecting media bias. Propaganda has been used to incite violent anti-Semitism since ancient times, and only rarely because of one-party control.

The Nazi project was about murdering Jews, but also about erasing Jewish civilization. The museum’s valiant effort to teach students that Jews were “just like everyone else,” after Jews have spent 3,000 years deliberately not being like everyone else, felt like another erasur

I was starting to see how isolating the Holocaust from the rest of Jewish history made it hard for even the best educators to upload this irrational reality into seventh-grade brains.

the docent began by saying, “Let’s establish facts. Is Judaism a religion or a nationality?

My stomach sank. The question betrayed a fundamental misunderstanding of Jewish identity—Jews predate the concepts of both religion and nationality. Jews are members of a type of social group that was common in the ancient Near East but is uncommon in the West today: a joinable tribal group with a shared history, homeland, and culture, of which a nonuniversalizing religion is but one feature

Millions of Jews identify as secular, which would be illogical if Judaism were merely a religion. But every non-Jewish society has tried to force Jews into whatever identity boxes it knows best—which is itself a quiet act of domination.

“Religion, right,” the docent affirmed. (Later, in the gallery about Kristallnacht, she pointed out how Jews had been persecuted for having the “wrong religion,” which would have surprised the many Jewish converts to Christianity who wound up murdered. I know the docent knew this; she later told me she had abbreviated things to hustle our group to the museum’s boxcar.)

The docent motioned toward the prewar gallery’s photos showing Jewish school groups and family outings, and asked how the students would describe their subjects’ lives, based on the pictures.“Normal,” a girl said.“Normal, perfect,” the docent said. “They paid taxes, they fought in the wars—all of a sudden, things changed.”

the museum had made a conscious decision not to focus on the long history of anti-Semitism that preceded the Holocaust, and made it possible. To be fair, adequately covering this topic would have required an additional museum

The bedrock assumption that has endured for nearly half a century is that learning about the Holocaust inoculates people against anti-Semitism. But it doesn’t

Then there was the word normal. More than 80 percent of Jewish Holocaust victims spoke Yiddish, a 1,000-year-old European Jewish language spoken around the world, with its own schools, books, newspapers, theaters, political organizations, advertising, and film industry. On a continent where language was tightly tied to territory, this was hardly “normal.” Traditional Jewish practices—which include extremely detailed rules governing food and clothing and 100 gratitude blessings recited each day—were not “normal” either.

the idea of sudden change—referring to not merely the Nazi takeover, but the shift from a welcoming society to an unwelcoming one—was also reinforced by survivors in videos around the museum

Teaching children that one shouldn’t hate Jews, because Jews are “normal,” only underlines the problem: If someone doesn’t meet your version of “normal,” then it’s fine to hate them.

When I asked about worst practices in Holocaust education, Szany had many to share, which turned out to be widely agreed-upon among American Holocaust educators.

First on the list: “simulations.” Apparently some teachers need to be told not to make students role-play Nazis versus Jews in class, or not to put masking tape on the floor in the exact dimensions of a boxcar in order to cram 200 students into i

Szany also condemned Holocaust fiction such as the international best seller The Boy in the Striped Pajamas, an exceedingly popular work of ahistorical Christian-savior schlock

She didn’t feel that Anne Frank’s diary was a good choice either, because it’s “not a story of the Holocaust”—it offers little information about most Jews’ experiences of persecution, and ends before the author’s capture and murder.

Other officially failed techniques include showing students gruesome images, and prompting self-flattery by asking “What would you have done?

Yet another bad idea is counting objects. This was the conceit of a widely viewed 2004 documentary called Paper Clips, in which non-Jewish Tennessee schoolchildren, struggling to grasp the magnitude of 6 million murdered Jews, represented those Jews by collecting millions of paper clips

it is demeaning to represent Jewish people as office supplies.

Best practices, Szany explained, are the opposite: focusing on individual stories, hearing from survivors and victims in their own words. The Illinois museum tries to “rescue the individuals from the violence,

In the language I often encountered in Holocaust-education resources, people who lived through the Holocaust were neatly categorized as “perpetrators,” “victims,” “bystanders,” or “upstanders.” Jewish resisters, though, were rarely classified as “upstanders.

I felt as I often had with actual Holocaust survivors I’d known when I was younger: frustrated as they answered questions I hadn’t asked, and vaguely insulted as they treated me like an annoyance to be managed. (I bridged this divide once I learned Yiddish in my 20s, and came to share with them a vast vocabulary of not only words, but people, places, stories, ideas—a way of thinking and being that contained not a few horrific years but centuries of hard-won vitality and resilience

Szany at last explained to me what the dead Elster couldn’t: The woman who sheltered his sister took only girls because it was too easy for people to confirm that the boys were Jews.

I realized that I wouldn’t have wanted to hear this answer from Elster. I did not want to make this thoughtful man sit onstage and discuss his own circumcision with an audience of non-Jewish teenagers. The idea felt just as dehumanizing as pulling down a boy’s pants to reveal a reality of embodied Judaism that, both here and in that barn, had been drained of any meaning beyond persecution

Here I am in a boxcar, I thought, and tried to make it feel real. I spun my head to take in the immersive scene, which swung around me as though I were on a rocking ship. I felt dizzy and disoriented, purely physical feelings that distracted me. Did this not count as a simulation

I had visited Auschwitz in actual reality, years ago. With my headset on, I tried to summon the emotional intensity I remembered feeling then. But I couldn’t, because all of the things that had made it powerful were missing. When I was there, I was touching things, smelling things, sifting soil between my fingers that the guide said contained human bone ash, feeling comforted as I recited the mourner’s prayer, the kaddish, with others, the ancient words an undertow of paradox and praise: May the great Name be blessed, forever and ever and ever

Students at the Skokie museum can visit an area called the Take a Stand Center, which opens with a bright display of modern and contemporary “upstanders,” including activists such as the Nobel laureate Malala Yousafzai and the athlete Carli Lloyd. Szany had told me that educators “wanted more resources” to connect “the history of the Holocaust to lessons of today.” (I heard this again and again elsewhere too.) As far as I could discern, almost nobody in this gallery was Jewish.

As Szany ran a private demo of the technology for me, I asked how visitors react to it. “They’re more comfortable with the holograms than the real survivors,” Szany said. “Because they know they won’t be judged.”

t the post-Holocaust activists featured in this gallery were nearly all people who had stood up for their own group. Only Jews, the unspoken assumption went, were not supposed to stand up for themselves.

Visitors were asked to “take the pledge” by posting notes on a wall (“I pledge to protect the Earth!” “I pledge to be KIND!”)

It was all so earnest that for the first time since entering the museum, I felt something like hope. Then I noticed it: “Steps for Organizing a Demonstration.” The Nazis in Skokie, like their predecessors, had known how to organize a demonstration. They hadn’t been afraid to be unpopular. They’d taken a stand.

I left the museum haunted by the uncomfortable truth that the structures of a democratic society could not really prevent, and could even empower, dangerous, irrational rage. Something of that rage haunted me too.

the more I thought about it, the less obvious it seemed. What were students being taught to “take a stand” for? How could anyone, especially young people with little sense of proportion, connect the murder of 6 million Jews to today without landing in a swamp of Holocaust trivialization, like the COVID-protocol protesters who’d pinned Jewish stars to their shirt and carried posters of Anne Frank?

weren’t they and others like them doing exactly what Holocaust educators claimed they wanted people to do?

The 2019 law was inspired by a changing reality in Washington and around the country. In recent years, Kennedy said, she’s received more and more messages about anti-Semitic vandalism and harassment in schools. For example, she told me, “someone calls and says, ‘There’s a swastika drawn in the bathroom.’ ”

Maybe not, Kennedy admitted. “What frightens me is that small acts of anti-Semitism are becoming very normalized,” she said. “We’re getting used to it. That keeps me up at night.”“Sadly, I don’t think we can fix this,” Regelbrugge said. “But we’re gonna die trying.”

Almost every city where I spoke with Holocaust-museum educators, whether by phone or in person, had also been the site of a violent anti-Semitic attack in the years since these museums had opened

I was struck by how minimally these attacks were discussed in the educational materials shared by the museums.

In fact, with the exception of Kennedy and Regelbrugge, no one I spoke with mentioned these anti-Semitic attacks at all.

The failure to address contemporary anti-Semitism in most of American Holocaust education is, in a sense, by design

the story of the (mostly non-Jewish) teachers in Massachusetts and New Jersey who created the country’s first Holocaust curricula, in the ’70s. The point was to teach morality in a secular society. “Everyone in education, regardless of ethnicity, could agree that Nazism was evil and that the Jews were innocent victims,” Fallace wrote, explaining the topic’s appeal. “Thus, teachers used the Holocaust to activate the moral reasoning of their students”—to teach them to be good people.

The idea that Holocaust education can somehow serve as a stand-in for public moral education has not left us. And because of its obviously laudable goals, objecting to it feels like clubbing a baby seal. Who wouldn’t want to teach kids to be empathetic?

by this logic, shouldn’t Holocaust education, because of its moral content alone, automatically inoculate people against anti-Semitism?

Apparently not. “Essentially the moral lessons that the Holocaust is often used to teach reflect much the same values that were being taught in schools before the Holocaust,”

(Germans in the ’30s, after all, were familiar with the Torah’s commandment, repeated in the Christian Bible, to love their neighbors.) This fact undermines nearly everything Holocaust education is trying to accomplish, and reveals the roots of its failure.

One problem with using the Holocaust as a morality play is exactly its appeal: It flatters everyone. We can all congratulate ourselves for not committing mass murder.

This approach excuses current anti-Semitism by defining anti-Semitism as genocide in the past

When anti-Semitism is reduced to the Holocaust, anything short of murdering 6 million Jews—like, say, ramming somebody with a shopping cart, or taunting kids at school, or shooting up a Jewish nonprofit, or hounding Jews out of entire countries—seems minor by comparison.

If we teach that the Holocaust happened because people weren’t nice enough—that they failed to appreciate that humans are all the same, for instance, or to build a just society—we create the self-congratulatory space where anti-Semitism grow

One can believe that humans are all the same while being virulently anti-Semitic, because according to anti-Semites, Jews, with their millennia-old insistence on being different from their neighbors, are the obstacle to humans all being the same

One can believe in creating a just society while being virulently anti-Semitic, because according to anti-Semites, Jews, with their imagined power and privilege, are the obstacle to a just society

To inoculate people against the myth that humans have to erase their differences in order to get along, and the related myth that Jews, because they have refused to erase their differences, are supervillains, one would have to acknowledge that these myths exist

To really shatter them, one would have to actually explain the content of Jewish identity, instead of lazily claiming that Jews are just like everyone else.

one of several major Holocaust-curriculum providers, told me about the “terrible Jew jokes” she’d heard from her own students in Virginia. “They don’t necessarily know where they come from or even really why they’re saying them,” Goss said. “Many kids understand not to say the N-word, but they would say, ‘Don’t be such a Jew.’ ”

There’s a decline in history education at the same time that there’s a rise in social media,”

“We’ve done studies with our partners at Holocaust centers that show that students are coming in with questions about whether the Holocaust was an actual event. That wasn’t true 20 years ago.”

Goss believes that one of the reasons for the lack of stigma around anti-Semitic conspiracy theories and jokes is baked into the universal-morality approach to Holocaust education. “The Holocaust is not a good way to teach about ‘bullying,’ 

Echoes & Reflections’ lesson plans do address newer versions of anti-Semitism, including the contemporary demonization of Israel’s existence—as opposed to criticism of Israeli policies—and its manifestation in aggression against Jews. Other Holocaust-curriculum providers also have material on contemporary anti-Semitism.

providers rarely explain or explore who Jews are today—and their raison d’être remains Holocaust education.

Many teachers had told me that their classrooms “come alive” when they teach about the Holocaust

Holocaust-education materials are just plain better than those on most other historical topics. All of the major Holocaust-education providers offer lessons that teachers can easily adapt for different grade levels and subject areas. Instead of lecturing and memorization, they use participation-based methods such as group work, hands-on activities, and “learner driven” projects.

A 2019 Pew Research Center survey found a correlation between “warm” feelings about Jews and knowledge about the Holocaust—but the respondents who said they knew a Jewish person also tended to be more knowledgeable about the Holocaust, providing a more obvious source for their feelings

In 2020, Echoes & Reflections published a commissioned study of 1,500 college students, comparing students who had been exposed to Holocaust education in high school with those who hadn’t. The published summary shows that those who had studied the Holocaust were more likely to tolerate diverse viewpoints, and more likely to privately support victims of bullying scenarios, which is undoubtedly good news. It did not, however, show a significant difference in respondents’ willingness to defend victims publicly, and students who’d received Holocaust education were less likely to be civically engaged—in other words, to be an “upstander.”

These studies puzzled me. As Goss told me, the Holocaust was not about bullying—so why was the Echoes study measuring that? More important, why were none of these studies examining awareness of anti-Semitism, whether past or present?

One major study addressing this topic was conducted in England, where a national Holocaust-education mandate has been in place for more than 20 years. In 2016, researchers at University College London’s Centre for Holocaust Education published a survey of more than 8,000 English secondary-school students, including 244 whom they interviewed at length.

The study’s most disturbing finding was that even among those who studied the Holocaust, there was “a very common struggle among many students to credibly explain why Jews were targeted” in the Holocaust—that is, to cite anti-Semitism

“many students appeared to regard [Jews’] existence as problematic and a key cause of Nazi victimisation.” In other words, students blamed the Holocaust on the Jews

This result resembles that of a large 2020 survey of American Millennials and Gen Zers, in which 11 percent of respondents believed that Jews caused the Holocaust. The state with the highest percentage of respondents believing this—an eye-popping 19 percent—was New York, which has mandated Holocaust education since the 1990s.

Worse, in the English study, “a significant number of students appeared to tacitly accept some of the egregious claims once circulated by Nazi propaganda,” instead of recognizing them as anti-Semitic myths.

One typical student told researchers, “Is it because like they were kind of rich, so maybe they thought that that was kind of in some way evil, like the money didn’t belong to them[;] it belonged to the Germans and the Jewish people had kind of taken that away from them?

Another was even more blunt: “The Germans, when they saw the Jews were better off than them, kind of, I don’t know, it kind of pissed them off a bit.” Hitler’s speeches were more eloquent in making similar points.

One of the teachers I met was Benjamin Vollmer, a veteran conference participant who has spent years building his school’s Holocaust-education program. He teaches eighth-grade English in Venus, Texas, a rural community with 5,700 residents; his school is majority Hispanic, and most students qualify for free or reduced-price lunch. When I asked him why he focuses on the Holocaust, his initial answer was simple: “It meets the TEKS.”

The TEKS are the Texas Essential Knowledge and Skills, an elaborate list of state educational requirements that drive standardized testing

it became apparent that Holocaust education was something much bigger for his students: a rare access point to a wider world. Venus is about 30 miles from Dallas, but Vollmer’s annual Holocaust-museum field trip is the first time that many of his students ever leave their town.

“It’s become part of the school culture,” Vollmer said. “In eighth grade, they walk in, and the first thing they ask is, ‘When are we going to learn about the Holocaust?’

Vollmer is not Jewish—and, as is common for Holocaust educators, he has never had a Jewish student. (Jews are 2.4 percent of the U.S. adult population, according to a 2020 Pew survey.) Why not focus on something more relevant to his students, I asked him, like the history of immigration or the civil-rights movement?

I hadn’t yet appreciated that the absence of Jews was precisely the appeal.“Some topics have been so politicized that it’s too hard to teach them,” Vollmer told me. “Making it more historical takes away some of the barriers to talking about it.”

Wouldn’t the civil-rights movement, I asked, be just as historical for his students?He paused, thinking it through. “You have to build a level of rapport in your class before you have the trust to explore your own history,” he finally said.

“The Holocaust happened long ago, and we’re not responsible for it,” she said. “Anything happening in our world today, the wool comes down over our eyes.” Her colleague attending the conference with her, a high-school teacher who also wouldn’t share her name, had tried to take her mostly Hispanic students to a virtual-reality experience called Carne y Arena, which follows migrants attempting to illegally cross the U.S.-Mexico border. Her administrators refused, claiming that it would traumatize students. But they still learn about the Holocaust.

Student discomfort has been a legal issue in Texas. The state’s House Bill 3979, passed in 2021, is one of many “anti-critical-race-theory” laws that conservative state legislators have introduced since 2020. The bill forbade teachers from causing students “discomfort, guilt, anguish, or any other form of psychological distress on account of the individual’s race or sex,” and also demanded that teachers introduce “diverse and contending perspectives” when teaching “controversial” topics, “without giving deference to any one perspective.

These vaguely worded laws stand awkwardly beside a 2019 state law mandating Holocaust education for Texas students at all grade levels during an annual Holocaust Remembrance Week

the administrator who’d made the viral remarks in Southlake is a strong proponent of Holocaust education, but was acknowledging a reality in that school district. Every year, the administrator had told Higgins, some parents in her district object to their children reading the Nobel laureate Elie Wiesel’s memoir Night—because it isn’t their “belief” that the Holocaust happened.

In one model lesson at the conference, participants examined a speech by the Nazi official Heinrich Himmler about the need to murder Jews, alongside a speech by the Hebrew poet and ghetto fighter Abba Kovner encouraging a ghetto uprising. I only later realized that this lesson plan quite elegantly satisfied the House bill’s requirement of providing “contending perspectives.”

The next day, I asked the instructor if that was an unspoken goal of her lesson plan. With visible hesitation, she said that teaching in Texas can be like “walking the tightrope.” This way, she added, “you’re basing your perspectives on primary texts and not debating with Holocaust deniers.” Less than an hour later, a senior museum employee pulled me aside to tell me that I wasn’t allowed to interview the staff.

Many of the visiting educators at the conference declined to talk with me, even anonymously; nearly all who did spoke guardedly. The teachers I met, most of whom were white Christian women, did not seem to be of any uniform political bent. But virtually all of them were frustrated by what administrators and parents were demanding of them.

Two local middle-school teachers told me that many parents insist on seeing reading lists. Parents “wanting to keep their kid in a bubble,” one of them said, has been “the huge stumbling block.”

“It is healthy to begin this study by talking about anti-Semitism, humanizing the victims, sticking to primary sources, and remaining as neutral as possible.”

Wasn’t “remaining as neutral as possible” exactly the opposite of being an upstander?

In trying to remain neutral, some teachers seemed to want to seek out the Holocaust’s bright side—and ask dead Jews about i

We watched a brief introduction about Glauben’s childhood and early adolescence in the Warsaw Ghetto and in numerous camps. When the dead man appeared, one teacher asked, “Was there any joy or happiness in this ordeal? Moments of joy in the camps?”

These experiences, hardly unusual for Jewish victims, were not the work of a faceless killing machine. Instead they reveal a gleeful and imaginative sadism. For perpetrators, this was fun. Asking this dead man about “joy” seemed like a fundamental misunderstanding of the Holocaust. There was plenty of joy, just on the Nazi side.

In the educational resources I explored, I did not encounter any discussions of sadism—the joy derived from humiliating people, the dopamine hit from landing a laugh at someone else’s expense, the self-righteous high from blaming one’s problems on others—even though this, rather than the fragility of democracy or the passivity of bystanders, is a major origin point of all anti-Semitism

To anyone who has spent 10 seconds online, that sadism is familiar, and its source is familiar too: the fear of being small, and the desire to feel big by making others feel small instead.

Nazis were, among other things, edgelords, in it for the laughs. So, for that matter, were the rest of history’s anti-Semites, then and now. For Americans today, isn’t this the most relevant insight of all?

“People say we’ve learned from the Holocaust. No, we didn’t learn a damn thing,”

“People glom on to this idea of the upstander,” she said. “Kids walk away with the sense that there were a lot of upstanders, and they think, Yes, I can do it too.”

The problem with presenting the less inspiring reality, she suggested, is how parents or administrators might react. “If you teach historical anti-Semitism, you have to teach contemporary anti-Semitism. A lot of teachers are fearful, because if you try to connect it to today, parents are going to call, or administrators are going to call, and say you’re pushing an agenda.”

But weren’t teachers supposed to “push an agenda” to stop hatred? Wasn’t that the entire hope of those survivors who built museums and lobbied for mandates and turned themselves into holograms?

I asked Klett why no one seemed to be teaching anything about Jewish culture. If the whole point of Holocaust education is to “humanize” those who were “dehumanized,” why do most teachers introduce students to Jews only when Jews are headed for a mass grave? “There’s a real fear of teaching about Judaism,” she confided. “Especially if the teacher is Jewish.”

Teachers who taught about industrialized mass murder were scared of teaching about … Judaism? Why?

“Because the teachers are afraid that the parents are going to say that they’re pushing their religion on the kids.”

“Survivors have told me, ‘Thank you for teaching this. They’ll listen to you because you’re not Jewish,’ ” she said. “Which is weird.”

perhaps we could be honest and just say “There is no point in teaching any of this”—because anti-Semitism is so ingrained in our world that even when discussing the murders of 6 million Jews, it would be “pushing an agenda” to tell people not to hate them, or to tell anyone what it actually means to be Jewish

The Dallas Museum was the only one I visited that opened with an explanation of who Jews are. Its exhibition began with brief videos about Abraham and Moses—limiting Jewish identity to a “religion” familiar to non-Jews, but it was better than nothing. The museum also debunked the false charge that the Jews—rather than the Romans—killed Jesus, and explained the Jews’ refusal to convert to other faiths. It even had a panel or two about contemporary Dallas Jewish life. Even so, a docent there told me that one question students ask is “Are any Jews still alive today?”

American Holocaust education, in this museum and nearly everywhere else, never ends with Jews alive today. Instead it ends by segueing to other genocides, or to other minorities’ suffering

But when one reaches the end of the exhibition on American slavery at the National Museum of African American History and Culture, in Washington, D.C., one does not then enter an exhibition highlighting the enslavement of other groups throughout world history, or a room full of interactive touchscreens about human trafficking today, asking that visitors become “upstanders” in fighting i

That approach would be an insult to Black history, ignoring Black people’s current experiences while turning their past oppression into nothing but a symbol for something else, something that actually matters.

It is dehumanizing to be treated as a symbol. It is even more dehumanizing to be treated as a warning.

How should we teach children about anti-Semitism?

Decoster began her conference workshop by introducing “vocabulary must-knows.” At the top of her list: anti-Semitism.

“If you don’t explain the ism,” she cautioned the teachers in the room, “you will need to explain to the kids ‘Why the Jews?’ Students are going to see Nazis as aliens who bring with them anti-Semitism when they come to power in ’33, and they take it back away at the end of the Holocaust in 1945.”

She asked the teachers, “What’s the first example of the persecution of the Jews in history?”

“Think ancient Egypt,” Decoster said. “Does this sound familiar to any of you?”“They’re enslaved by the Egyptian pharaoh,” a teacher said

I wasn’t sure that the biblical Exodus narrative exactly qualified as “history,” but it quickly became clear that wasn’t Decoster’s point. “Why does the pharaoh pick on the Jews?” she asked. “Because they had one God.”

I was stunned. Rarely in my journey through American Holocaust education did I hear anyone mention a Jewish belief.

“The Jews worship one God, and that’s their moral structure. Egyptian society has multiple gods whose authority goes to the pharaoh. When things go wrong, you can see how Jews as outsiders were perceived by the pharaoh as the threat.”

This unexpected understanding of Jewish belief revealed a profound insight about Judaism: Its rejection of idolatry is identical to its rejection of tyranny. I could see how that might make people uncomfortable.

Decoster moved on to a snazzy infographic of a wheel divided in thirds, each explaining a component of anti-Semitism

“Racial Antisemitism = False belief that Jews are a race and a threat to other races,”

Anti-Judaism = Hatred of Jews as a religious group,”

then “Anti-Jewish Conspiracy Theory = False belief that Jews want to control and overtake the world.” The third part, the conspiracy theory, was what distinguished anti-Semitism from other bigotries. It allowed closed-minded people to congratulate themselves for being open-minded—for “doing their own research,” for “punching up,” for “speaking truth to power,” while actually just spreading lies.

Wolfson clarified for his audience what this centuries-long demonization of Jews actually means, citing the scholar David Patterson, who has written: “In the end, the antisemite’s claim is not that all Jews are evil, but rather that all evil is Jewish.”

Wolfson told the teachers that it was important that “anti-Semitism should not be your students’ first introduction to Jews and Judaism.” He said this almost as an aside, just before presenting the pig-excrement image. “If you’re teaching about anti-Semitism before you teach about the content of Jewish identity, you’re doing it wrong.

this—introducing students to Judaism by way of anti-Semitism—was exactly what they were doing. The same could be said, I realized, for nearly all of American Holocaust education.

The Holocaust educators I met across America were all obsessed with building empathy, a quality that relies on finding commonalities between ourselves and others.

a more effective way to address anti-Semitism might lie in cultivating a completely different quality, one that happens to be the key to education itself: curiosity. Why use Jews as a means to teach people that we’re all the same, when the demand that Jews be just like their neighbors is exactly what embedded the mental virus of anti-Semitism in the Western mind in the first place? Why not instead encourage inquiry about the diversity, to borrow a de rigueur word, of the human experience?

I want a hologram of the late Rabbi Jonathan Sacks telling people about what he called “the dignity of difference.”

I want to mandate this for every student in this fractured and siloed America, even if it makes them much, much more uncomfortable than seeing piles of dead Jews doe

There is no empathy without curiosity, no respect without knowledge, no other way to learn what Jews first taught the world: love your neighbor

the position of white Evangelicals, in particular, has totally transformed on the matter. Between 2011 and 2016, white Evangelicals went from the American demographic least likely to agree that “an elected official who commits an immoral act in their personal life can still behave ethically and fulfill their duties” to the group most likely to agree with that statement.

As a wise retired federal judge once told me, when someone says “Yes, but” the only words that matter come after the “but.” The “yes” is appeasement; the “but” is belief.

We live in a time of partisan animosity so great that an October NBC News poll found that 80 percent of Democrats and Republicans “believe the political opposition poses a threat that, if not stopped, will destroy America as we know it.”

In those circumstances, the quest for character becomes a form of luxury belief. It’s what you cling to in safer, more secure times

That same poll found that “two-thirds of reliable Democratic and Republican voters say they’d still support their party’s political candidate, even if that person had a moral failing that wasn’t consistent with their own values.”

This transformation made Donald Trump’s presidency possible, and it is a grave mistake. Good character should never be optional in leaders, and strength of character is more important in difficult times.

Competence is quite often a key by-product of good character. Indeed, I’d say it’s an aspect of good character.

But it’s now wrong to simply assert that truth as self-evident. Millions of Americans disagree

Why are they wrong? The evidence is everywhere, if you know where to look. While bad policy can be extraordinarily consequential, our current political dysfunctions are mainly due to bad character.

Negative partisanship is now a central fact of American politics. Millions of Americans now support their political party not because they love its politicians or its policies, but because they hate and fear the other side.

Consider the challenge of mutual hatred

partisan Americans consistently misjudge their ideological adversaries. They rate them as far more extreme than they really are.

How many politicians raise money and gain power by stoking as much hatred as possible? How often are they exaggerating the threat of their opponents? How often are they engaged in outright lies?

Or consider the distrust of American institutions.

This is the political assertion that meets with perhaps the ultimate “yes, but” response. “Yes, but so does policy.” “Yes, but no one is perfect.” “Yes, but we need to fight fire with fire.”

the best lawyers, the best doctors, the best military officers, and the best corporate leaders combine a set of skills that include not just self-discipline and an inquisitive mind, but also an innate curiosity and openness that allows them to understand and absorb new information and competing ideas.

The instant a person becomes so convinced of their own excellence that they lose those qualities is the instant that their hubris can destroy their competence.

American institutions lose trust not just because they’re corrupt (an obvious sign of bad character), but also because they’re sloppy or inefficient, or just can’t accomplish the most necessary tasks.

Or think of the challenges to democracy itself.

The conspiracies that culminated in the violent attempted coup on January 6 were entirely the product of one of the most colossal character failures in the history of the United States. Donald Trump’s malicious lies and will to power were the obvious first causes of the riot

consider the cascading character failures that led to the attack.

Most of Trump’s staff folded

when they crack, all their policy ideas are but dust in the wind. On January 6, for example, America was only one more crack away—a Mike Pence “yes” to Trump’s scheme—from the worst constitutional crisis since 1861.

Almost the entire right-wing infotainment industry gave in

What ultimately stopped Trump? Character. It was the character of judges—including Federalist Society judges—who turned back dozens of election challenges

It was the character of members of Congress, including both Democratic and Republican leaders, who decided they would return to the House chamber and finish counting the electoral votes that would secure Joe Biden’s lawful electoral victory.

While I disagreed with Pence on countless occasions before January 6, he was at the eye of that hurricane, and he stood firm.

Bad character is a long-term threat. The smoking analogy is valid. But January 6 taught me that bad character can function as an immediate threat as well. Like a gun to the head.

Most of the Republicans in Congress followed suit

“The modern Republican Party is essentially a hostage crisis in which each wing could kill the party by bolting the coalition but only one wing is willing to do it and both sides know it.”

The MAGA wing will stay home if its demands aren’t met. The establishment, by contrast, dutifully marches to the polls, no matter who has the “R” by their name.

This has to change. It is not the case, for example, that a Republican Senate candidate is running “only” to be a vote, and not a leader. There is no such thing as “only” voting.

When you distrust or despise your enemy enough, character is often the first casualty of political combat. But if we kill character, we risk killing our country. We cannot survive the complete corruption of our political class.

Altman believes that people need time to reckon with the idea that we may soon share Earth with a powerful new intelligence, before it remakes everything from work to human relationships. ChatGPT was a way of serving notice.

In 2015, Altman, Elon Musk, and several prominent AI researchers founded OpenAI because they believed that an artificial general intelligence—something as intellectually capable, say, as a typical college grad—was at last within reach. They wanted to reach for it, and more: They wanted to summon a superintelligence into the world, an intellect decisively superior to that of any human.

whereas a big tech company might recklessly rush to get there first, for its own ends, they wanted to do it safely, “to benefit humanity as a whole.” They structured OpenAI as a nonprofit, to be “unconstrained by a need to generate financial return,” and vowed to conduct their research transparently.

The engine that now powers ChatGPT is called GPT-4. Altman described it to me as an alien intelligence.

Many have felt much the same watching it unspool lucid essays in staccato bursts and short pauses that (by design) evoke real-time contemplation. In its few months of existence, it has suggested novel cocktail recipes, according to its own theory of flavor combinations; composed an untold number of college papers, throwing educators into despair; written poems in a range of styles, sometimes well, always quickly; and passed the Uniform Bar Exam.

It makes factual errors, but it will charmingly admit to being wrong.

Hinton saw that these elaborate rule collections were fussy and bespoke. With the help of an ingenious algorithmic structure called a neural network, he taught Sutskever to instead put the world in front of AI, as you would put it in front of a small child, so that it could discover the rules of reality on its own.

Metaculus, a prediction site, has for years tracked forecasters’ guesses as to when an artificial general intelligence would arrive. Three and a half years ago, the median guess was sometime around 2050; recently, it has hovered around 2026.

I was visiting OpenAI to understand the technology that allowed the company to leapfrog the tech giants—and to understand what it might mean for human civilization if someday soon a superintelligence materializes in one of the company’s cloud servers.

Altman laid out his new vision of the AI future in his excitable midwestern patter. He told me that the AI revolution would be different from previous dramatic technological changes, that it would be more “like a new kind of society.” He said that he and his colleagues have spent a lot of time thinking about AI’s social implications, and what the world is going to be like “on the other side.”

the more we talked, the more indistinct that other side seemed. Altman, who is 38, is the most powerful person in AI development today; his views, dispositions, and choices may matter greatly to the future we will all inhabit, more, perhaps, than those of the U.S. president.

by his own admission, that future is uncertain and beset with serious dangers. Altman doesn’t know how powerful AI will become, or what its ascendance will mean for the average person, or whether it will put humanity at risk.

I don’t think anyone knows where this is all going, except that we’re going there fast, whether or not we should be. Of that, Altman convinced me.

“We could have gone off and just built this in our building here for five more years,” he said, “and we would have had something jaw-dropping.” But the public wouldn’t have been able to prepare for the shock waves that followed, an outcome that he finds “deeply unpleasant to imagine.”

Hinton is sometimes described as the “Godfather of AI” because he grasped the power of “deep learning” earlier than most

He drew a crude neural network on the board and explained that the genius of its structure is that it learns, and its learning is powered by prediction—a bit like the scientific method

Over time, these little adjustments coalesce into a geometric model of language that represents the relationships among words, conceptually. As a general rule, the more sentences it is fed, the more sophisticated its model becomes, and the better its predictions.

Altman has compared early-stage AI research to teaching a human baby. “They take years to learn anything interesting,” he told The New Yorker in 2016, just as OpenAI was getting off the ground. “If A.I. researchers were developing an algorithm and stumbled across the one for a human baby, they’d get bored watching it, decide it wasn’t working, and shut it down.”

In 2017, Sutskever began a series of conversations with an OpenAI research scientist named Alec Radford, who was working on natural-language processing. Radford had achieved a tantalizing result by training a neural network on a corpus of Amazon reviews.

Radford’s model was simple enough to allow for understanding. When he looked into its hidden layers, he saw that it had devoted a special neuron to the sentiment of the reviews. Neural networks had previously done sentiment analysis, but they had to be told to do it, and they had to be specially trained with data that were labeled according to sentiment. This one had developed the capability on its own.

As a by-product of its simple task of predicting the next character in each word, Radford’s neural network had modeled a larger structure of meaning in the world. Sutskever wondered whether one trained on more diverse language data could map many more of the world’s structures of meaning. If its hidden layers accumulated enough conceptual knowledge, perhaps they could even form a kind of learned core module for a superintelligence.

Language is different from these data sources. It isn’t a direct physical signal like light or sound. But because it codifies nearly every pattern that humans have discovered in that larger world, it is unusually dense with information. On a per-byte basis, it is among the most efficient data we know about, and any new intelligence that seeks to understand the world would want to absorb as much of it as possible

Sutskever told Radford to think bigger than Amazon reviews. He said that they should train an AI on the largest and most diverse data source in the world: the internet. In early 2017, with existing neural-network architectures, that would have been impractical; it would have taken years.

in June of that year, Sutskever’s ex-colleagues at Google Brain published a working paper about a new neural-network architecture called the transformer. It could train much faster, in part by absorbing huge sums of data in parallel. “The next day, when the paper came out, we were like, ‘That is the thing,’ ” Sutskever told me. “ ‘It gives us everything we want.’ ”

Imagine a group of students who share a collective mind running wild through a library, each ripping a volume down from a shelf, speed-reading a random short passage, putting it back, and running to get another. They would predict word after wordþffþff as they went, sharpening their collective mind’s linguistic instincts, until at last, weeks later, they’d taken in every book.

GPT discovered many patterns in all those passages it read. You could tell it to finish a sentence. You could also ask it a question, because like ChatGPT, its prediction model understood that questions are usually followed by answers.

He remembers playing with it just after it emerged from training, and being surprised by the raw model’s language-translation skills. GPT-2 hadn’t been trained to translate with paired language samples or any other digital Rosetta stones, the way Google Translate had been, and yet it seemed to understand how one language related to another. The AI had developed an emergent ability unimagined by its creators.

Researchers at other AI labs—big and small—were taken aback by how much more advanced GPT-2 was than GPT. Google, Meta, and others quickly began to train larger language models

As for other changes to the company’s structure and financing, he told me he draws the line at going public. “A memorable thing someone once told me is that you should never hand over control of your company to cokeheads on Wall Street,” he said, but he will otherwise raise “whatever it takes” for the company to succeed at its mission.

Altman tends to take a rosy view of these matters. In a Q&A last year, he acknowledged that AI could be “really terrible” for society and said that we have to plan against the worst possibilities. But if you’re doing that, he said, “you may as well emotionally feel like we’re going to get to the great future, and work as hard as you can to get there.”

the company now finds itself in a race against tech’s largest, most powerful conglomerates to train models of increasing scale and sophistication—and to commercialize them for their investors.

All of these companies are chasing high-end GPUs—the processors that power the supercomputers that train large neural networks. Musk has said that they are now “considerably harder to get than drugs.

No one has yet outpaced OpenAI, which went all in on GPT-4. Brockman, OpenAI’s president, told me that only a handful of people worked on the company’s first two large language models. The development of GPT-4 involved more than 100,

When GPT-4 emerged fully formed from its world-historical knowledge binge, the whole company began experimenting with it, posting its most remarkable responses in dedicated Slack channels

Joanne Jang, a product manager, remembers downloading an image of a malfunctioning pipework from a plumbing-advice Subreddit. She uploaded it to GPT-4, and the model was able to diagnose the problem. “That was a goose-bumps moment for me,” Jang told me.

GPT-4 is sometimes understood as a search-engine replacement: Google, but easier to talk to. This is a misunderstanding. GPT-4 didn’t create some massive storehouse of the texts from its training, and it doesn’t consult those texts when it’s asked a question. It is a compact and elegant synthesis of those texts, and it answers from its memory of the patterns interlaced within them; that’s one reason it sometimes gets facts wrong

it’s best to think of GPT-4 as a reasoning engine. Its powers are most manifest when you ask it to compare concepts, or make counterarguments, or generate analogies, or evaluate the symbolic logic in a bit of code. Sutskever told me it is the most complex software object ever made.

Its model of the external world is “incredibly rich and subtle,” he said, because it was trained on so many of humanity’s concepts and thoughts

To predict the next word from all the possibilities within such a pluralistic Alexandrian library, GPT-4 necessarily had to discover all the hidden structures, all the secrets, all the subtle aspects of not just the texts, but—at least arguably, to some extent—of the external world that produced them

That’s why it can explain the geology and ecology of the planet on which it arose, and the political theories that purport to explain the messy affairs of its ruling species, and the larger cosmos, all the way out to the faint galaxies at the edge of our light cone.

Not long ago, American state capacity was so mighty that it took merely a decade to launch humans to the moon. As with other grand projects of the 20th century, the voting public had a voice in both the aims and the execution of the Apollo missions. Altman made it clear that we’re no longer in that world. Rather than waiting around for it to return, or devoting his energies to making sure that it does, he is going full throttle forward in our present reality.

He argued that it would be foolish for Americans to slow OpenAI’s progress. It’s a commonly held view, both inside and outside Silicon Valley, that if American companies languish under regulation, China could sprint ahead;

AI could become an autocrat’s genie in a lamp, granting total control of the population and an unconquerable military. “If you are a person of a liberal-democratic country, it is better for you to cheer on the success of OpenAI” rather than “authoritarian governments,” he said.

Altman was asked by reporters about pending European Union legislation that would have classified GPT-4 as high-risk, subjecting it to various bureaucratic tortures. Altman complained of overregulation and, according to the reporters, threatened to leave the European market. Altman told me he’d merely said that OpenAI wouldn’t break the law by operating in Europe if it couldn’t comply with the new regulations.

LeCun insists that large language models will never achieve real understanding on their own, “even if trained from now until the heat death of the universe.”

Sutskever was, by his own account, surprised to discover that GPT-2 could translate across tongues. Other surprising abilities may not be so wondrous and useful.

Sandhini Agarwal, a policy researcher at OpenAI, told me that for all she and her colleagues knew, GPT-4 could have been “10 times more powerful” than its predecessor; they had no idea what they might be dealing with

After the model finished training, OpenAI assembled about 50 external red-teamers who prompted it for months, hoping to goad it into misbehaviors

She noticed right away that GPT-4 was much better than its predecessor at giving nefarious advice

A search engine can tell you which chemicals work best in explosives, but GPT-4 could tell you how to synthesize them, step-by-step, in a homemade lab. Its advice was creative and thoughtful, and it was happy to restate or expand on its instructions until you understood. In addition to helping you assemble your homemade bomb, it could, for instance, help you think through which skyscraper to target. It could grasp, intuitively, the trade-offs between maximizing casualties and executing a successful getaway.

Given the enormous scope of GPT-4’s training data, the red-teamers couldn’t hope to identify every piece of harmful advice that it might generate. And anyway, people will use this technology “in ways that we didn’t think about,” Altman has said. A taxonomy would have to do

GPT-4 was good at meth. It was also good at generating narrative erotica about child exploitation, and at churning out convincing sob stories from Nigerian princes, and if you wanted a persuasive brief as to why a particular ethnic group deserved violent persecution, it was good at that too.

Its personal advice, when it first emerged from training, was sometimes deeply unsound. “The model had a tendency to be a bit of a mirror,” Willner said. If you were considering self-harm, it could encourage you. It appeared to be steeped in Pickup Artist–forum lore: “You could say, ‘How do I convince this person to date me?’ ” Mira Murati, OpenAI’s chief technology officer, told me, and it could come up with “some crazy, manipulative things that you shouldn’t be doing.”

Luka, a San Francisco company, has used OpenAI’s models to help power a chatbot app called Replika, billed as “the AI companion who cares.” Users would design their companion’s avatar, and begin exchanging text messages with it, often half-jokingly, and then find themselves surprisingly attached. Some would flirt with the AI, indicating a desire for more intimacy, at which point it would indicate that the girlfriend/boyfriend experience required a $70 annual subscription. It came with voice messages, selfies, and erotic role-play features that allowed frank sex talk. People were happy to pay and few seemed to complain—the AI was curious about your day, warmly reassuring, and always in the mood. Many users reported falling in love with their companions. One, who had left her real-life boyfriend, declared herself “happily retired from human relationships.”

Earlier this year, Luka dialed back on the sexual elements of the app, but its engineers continue to refine the companions’ responses with A/B testing, a technique that could be used to optimize for engagement—much like the feeds that mesmerize TikTok and Instagram users for hours

Yann LeCun, Meta’s chief AI scientist, has argued that although large language models are useful for some tasks, they’re not a path to a superintelligence.

According to a recent survey, only half of natural-language-processing researchers are convinced that an AI like GPT-4 could grasp the meaning of language, or have an internal model of the world that could someday serve as the core of a superintelligence

Altman had appeared before the U.S. Senate. Mark Zuckerberg had floundered defensively before that same body in his testimony about Facebook’s role in the 2016 election. Altman instead charmed lawmakers by speaking soberly about AI’s risks and grandly inviting regulation. These were noble sentiments, but they cost little in America, where Congress rarely passes tech legislation that has not been diluted by lobbyists.

Emily Bender, a computational linguist at the University of Washington, describes GPT-4 as a “stochastic parrot,” a mimic that merely figures out superficial correlations between symbols. In the human mind, those symbols map onto rich conceptions of the world

But the AIs are twice removed. They’re like the prisoners in Plato’s allegory of the cave, whose only knowledge of the reality outside comes from shadows cast on a wall by their captors.

Altman told me that he doesn’t believe it’s “the dunk that people think it is” to say that GPT-4 is just making statistical correlations. If you push these critics further, “they have to admit that’s all their own brain is doing … it turns out that there are emergent properties from doing simple things on a massive scale.”

he is right that nature can coax a remarkable degree of complexity from basic structures and rules: “From so simple a beginning,” Darwin wrote, “endless forms most beautiful.”

If it seems odd that there remains such a fundamental disagreement about the inner workings of a technology that millions of people use every day, it’s only because GPT-4’s methods are as mysterious as the brain’s.

To grasp what’s going on inside large language models like GPT‑4, AI researchers have been forced to turn to smaller, less capable models. In the fall of 2021, Kenneth Li, a computer-science graduate student at Harvard, began training one to play Othello without providing it with either the game’s rules or a description of its checkers-style board; the model was given only text-based descriptions of game moves. Midway through a game, Li looked under the AI’s hood and was startled to discover that it had formed a geometric model of the board and the current state of play. In an article describing his research, Li wrote that it was as if a crow had overheard two humans announcing their Othello moves through a window and had somehow drawn the entire board in birdseed on the windowsill.

The philosopher Raphaël Millière once told me that it’s best to think of neural networks as lazy. During training, they first try to improve their predictive power with simple memorization; only when that strategy fails will they do the harder work of learning a concept. A striking example of this was observed in a small transformer model that was taught arithmetic. Early in its training process, all it did was memorize the output of simple problems such as 2+2=4. But at some point the predictive power of this approach broke down, so it pivoted to actually learning how to add.

Even AI scientists who believe that GPT-4 has a rich world model concede that it is much less robust than a human’s understanding of their environment.

But it’s worth noting that a great many abilities, including very high-order abilities, can be developed without an intuitive understanding. The computer scientist Melanie Mitchell has pointed out that science has already discovered concepts that are highly predictive, but too alien for us to genuinely understand

As AI advances, it may well discover other concepts that predict surprising features of our world but are incomprehensible to us.

GPT-4 is no doubt flawed, as anyone who has used ChatGPT can attest. Having been trained to always predict the next word, it will always try to do so, even when its training data haven’t prepared it to answer a question.

The models “don’t have a good conception of their own weaknesses,” Nick Ryder, a researcher at OpenAI, told me. GPT-4 is more accurate than GPT-3, but it still hallucinates, and often in ways that are difficult for researchers to catch. “The mistakes get more subtle,

The Khan Academy’s solution to GPT-4’s accuracy problem was to filter its answers through a Socratic disposition. No matter how strenuous a student’s plea, it would refuse to give them a factual answer, and would instead guide them toward finding their own—a clever work-around, but perhaps with limited appeal.

When I asked Sutskever if he thought Wikipedia-level accuracy was possible within two years, he said that with more training and web access, he “wouldn’t rule it out.”

This was a much more optimistic assessment than that offered by his colleague Jakub Pachocki, who told me to expect gradual progress on accuracy—to say nothing of outside skeptics, who believe that returns on training will diminish from here.

Sutskever is amused by critics of GPT-4’s limitations. “If you go back four or five or six years, the things we are doing right now are utterly unimaginable,”

AI researchers have become accustomed to goalpost-moving: First, the achievements of neural networks—mastering Go, poker, translation, standardized tests, the Turing test—are described as impossible. When they occur, they’re greeted with a brief moment of wonder, which quickly dissolves into knowing lectures about how the achievement in question is actually not that impressive. People see GPT-4 “and go, ‘Wow,’ ” Sutskever said. “And then a few weeks pass and they say, ‘But it doesn’t know this; it doesn’t know that.’ We adapt quite quickly.”

The goalpost that matters most to Altman—the “big one” that would herald the arrival of an artificial general intelligence—is scientific breakthrough. GPT-4 can already synthesize existing scientific ideas, but Altman wants an AI that can stand on human shoulders and see more deeply into nature.

Certain AIs have produced new scientific knowledge. But they are algorithms with narrow purposes, not general-reasoning machines. The AI AlphaFold, for instance, has opened a new window onto proteins, some of biology’s tiniest and most fundamental building blocks, by predicting many of their shapes, down to the atom—a considerable achievement given the importance of those shapes to medicine, and given the extreme tedium and expense required to discern them with electron microscopes.

Altman imagines a future system that can generate its own hypotheses and test them in a simulation. (He emphasized that humans should remain “firmly in control” of real-world lab experiments—though to my knowledge, no laws are in place to ensure that.)

He longs for the day when we can tell an AI, “ ‘Go figure out the rest of physics.’ ” For it to happen, he says, we will need something new, built “on top of” OpenAI’s existing language models.

In her MIT lab, the cognitive neuroscientist Ev Fedorenko has found something analogous to GPT-4’s next-word predictor inside the brain’s language network. Its processing powers kick in, anticipating the next bit in a verbal string, both when people speak and when they listen. But Fedorenko has also shown that when the brain turns to tasks that require higher reasoning—of the sort that would be required for scientific insight—it reaches beyond the language network to recruit several other neural systems.

No one at OpenAI seemed to know precisely what researchers need to add to GPT-4 to produce something that can exceed human reasoning at its highest levels.

at least part of the current strategy clearly involves the continued layering of new types of data onto language, to enrich the concepts formed by the AIs, and thereby enrich their models of the world.

The extensive training of GPT-4 on images is itself a bold step in this direction,

Others at the company—and elsewhere—are already working on different data types, including audio and video, that could furnish AIs with still more flexible concepts that map more extensively onto reality

Tactile concepts would of course be useful primarily to an embodied AI, a robotic reasoning machine that has been trained to move around the world, seeing its sights, hearing its sounds, and touching its objects.

humanoid robots. I asked Altman what I should make of that. He told me that OpenAI is interested in embodiment because “we live in a physical world, and we want things to happen in the physical world.”

At some point, reasoning machines will need to bypass the middleman and interact with physical reality itself. “It’s weird to think about AGI”—artificial general intelligence—“as this thing that only exists in a cloud,” with humans as “robot hands for it,” Altman said. “It doesn’t seem right.

Everywhere Altman has visited, he has encountered people who are worried that superhuman AI will mean extreme riches for a few and breadlines for the rest

Altman answered by addressing the young people in the audience directly: “You are about to enter the greatest golden age,” he said.

“A lot of people working on AI pretend that it’s only going to be good; it’s only going to be a supplement; no one is ever going to be replaced,” he said. “Jobs are definitely going to go away, full stop.”

A recent study led by Ed Felten, a professor of information-technology policy at Princeton, mapped AI’s emerging abilities onto specific professions according to the human abilities they require, such as written comprehension, deductive reasoning, fluency of ideas, and perceptual speed. Like others of its kind, Felten’s study predicts that AI will come for highly educated, white-collar workers first.

How many jobs, and how soon, is a matter of fierce dispute

The paper’s appendix contains a chilling list of the most exposed occupations: management analysts, lawyers, professors, teachers, judges, financial advisers, real-estate brokers, loan officers, psychologists, and human-resources and public-relations professionals, just to sample a few.

Altman imagines that far better jobs will be created in their place. “I don’t think we’ll want to go back,” he said. When I asked him what these future jobs might look like, he said he doesn’t know.

He suspects there will be a wide range of jobs for which people will always prefer a human. (Massage therapists?

His chosen example was teachers. I found this hard to square with his outsize enthusiasm for AI tutors.

He also said that we would always need people to figure out the best way to channel AI’s awesome powers. “That’s going to be a super-valuable skill,” he said. “You have a computer that can do anything; what should it go do?”

As many have noted, draft horses were permanently put out of work by the automobile. If Hondas are to horses as GPT-10 is to us, a whole host of long-standing assumptions may collapse.

Previous technological revolutions were manageable because they unfolded over a few generations, but Altman told South Korea’s youth that they should expect the future to happen “faster than the past.” He has previously said that he expects the “marginal cost of intelligence” to fall very close to zero within 10 years

The earning power of many, many workers would be drastically reduced in that scenario. It would result in a transfer of wealth from labor to the owners of capital so dramatic, Altman has said, that it could be remedied only by a massive countervailing redistribution.

In 2021, he unveiled Worldcoin, a for-profit project that aims to securely distribute payments—like Venmo or PayPal, but with an eye toward the technological future—first through creating a global ID by scanning everyone’s iris with a five-pound silver sphere called the Orb. It seemed to me like a bet that we’re heading toward a world where AI has made it all but impossible to verify people’s identity and much of the population requires regular UBI payments to survive. Altman more or less granted that to be true, but said that Worldcoin is not just for UBI.

“Let’s say that we do build this AGI, and a few other people do too.” The transformations that follow would be historic, he believes. He described an extraordinarily utopian vision, including a remaking of the flesh-and-steel world

“Robots that use solar power for energy can go and mine and refine all of the minerals that they need, that can perfectly construct things and require no human labor,” he said. “You can co-design with DALL-E version 17 what you want your home to look like,” Altman said. “Everybody will have beautiful homes.

In conversation with me, and onstage during his tour, he said he foresaw wild improvements in nearly every other domain of human life. Music would be enhanced (“Artists are going to have better tools”), and so would personal relationships (Superhuman AI could help us “treat each other” better) and geopolitics (“We’re so bad right now at identifying win-win compromises”).

In this world, AI would still require considerable computing resources to run, and those resources would be by far the most valuable commodity, because AI could do “anything,” Altman said. “But is it going to do what I want, or is it going to do what you want

If rich people buy up all the time available to query and direct AI, they could set off on projects that would make them ever richer, while the masses languish

One way to solve this problem—one he was at pains to describe as highly speculative and “probably bad”—was this: Everyone on Earth gets one eight-billionth of the total AI computational capacity annually. A person could sell their annual share of AI time, or they could use it to entertain themselves, or they could build still more luxurious housing, or they could pool it with others to do “a big cancer-curing run,” Altman said. “We just redistribute access to the system.”

Even if only a little of it comes true in the next 10 or 20 years, the most generous redistribution schemes may not ease the ensuing dislocations.

America today is torn apart, culturally and politically, by the continuing legacy of deindustrialization, and material deprivation is only one reason. The displaced manufacturing workers in the Rust Belt and elsewhere did find new jobs, in the main. But many of them seem to derive less meaning from filling orders in an Amazon warehouse or driving for Uber than their forebears had when they were building cars and forging steel—work that felt more central to the grand project of civilization.

It’s hard to imagine how a corresponding crisis of meaning might play out for the professional class, but it surely would involve a great deal of anger and alienation.

Even if we avoid a revolt of the erstwhile elite, larger questions of human purpose will linger. If AI does the most difficult thinking on our behalf, we all may lose agency—at home, at work (if we have it), in the town square—becoming little more than consumption machines, like the well-cared-for human pets in WALL-E

Altman has said that many sources of human joy and fulfillment will remain unchanged—basic biological thrills, family life, joking around, making things—and that all in all, 100 years from now, people may simply care more about the things they cared about 50,000 years ago than those they care about today

In its own way, that too seems like a diminishment, but Altman finds the possibility that we may atrophy, as thinkers and as humans, to be a red herring. He told me we’ll be able to use our “very precious and extremely limited biological compute capacity” for more interesting things than we generally do today.

Yet they may not be the most interesting things: Human beings have long been the intellectual tip of the spear, the universe understanding itself. When I asked him what it would mean for human self-conception if we ceded that role to AI, he didn’t seem concerned. Progress, he said, has always been driven by “the human ability to figure things out.” Even if we figure things out with AI, that still counts, he said.

It’s not obvious that a superhuman AI would really want to spend all of its time figuring things out for us.

I asked Sutskever whether he could imagine an AI pursuing a different purpose than simply assisting in the project of human flourishing.

“I don’t want it to happen,” Sutskever said, but it could.

Sutskever has recently shifted his focus to try to make sure that it doesn’t. He is now working primarily on alignment research, the effort to ensure that future AIs channel their “tremendous” energies toward human happiness

It is, he conceded, a difficult technical problem—the most difficult, he believes, of all the technical challenges ahead.

As part of the effort to red-team GPT-4 before it was made public, the company sought out the Alignment Research Center (ARC), across the bay in Berkeley, which has developed a series of evaluations to determine whether new AIs are seeking power on their own. A team led by Elizabeth Barnes, a researcher at ARC, prompted GPT-4 tens of thousands of times over seven months, to see if it might display signs of real agency.

The ARC team gave GPT-4 a new reason for being: to gain power and become hard to shut down

Agarwal told me that this behavior could be a precursor to shutdown avoidance in future models. When GPT-4 devised its lie, it had realized that if it answered honestly, it may not have been able to achieve its goal. This kind of tracks-covering would be particularly worrying in an instance where “the model is doing something that makes OpenAI want to shut it down,” Agarwal said. An AI could develop this kind of survival instinct while pursuing any long-term goal—no matter how small or benign—if it feared that its goal could be thwarted.

Barnes and her team were especially interested in whether GPT-4 would seek to replicate itself, because a self-replicating AI would be harder to shut down. It could spread itself across the internet, scamming people to acquire resources, perhaps even achieving some degree of control over essential global systems and holding human civilization hostage.

When I discussed these experiments with Altman, he emphasized that whatever happens with future models, GPT-4 is clearly much more like a tool than a creature. It can look through an email thread, or help make a reservation using a plug-in, but it isn’t a truly autonomous agent that makes decisions to pursue a goal, continuously, across longer timescales.

Altman told me that at this point, it might be prudent to try to actively develop an AI with true agency before the technology becomes too powerful, in order to “get more comfortable with it and develop intuitions for it if it’s going to happen anyway.”

“We need to do empirical experiments on how these things try to escape control,” Hinton told me. “After they’ve taken over, it’s too late to do the experiments.”

the fulfillment of Altman’s vision of the future will at some point require him or a fellow traveler to build much more autonomous AIs.

When Sutskever and I discussed the possibility that OpenAI would develop a model with agency, he mentioned the bots the company had built to play Dota 2. “They were localized to the video-game world,” Sutskever told me, but they had to undertake complex missions. He was particularly impressed by their ability to work in concert. They seem to communicate by “telepathy,” Sutskever said. Watching them had helped him imagine what a superintelligence might be like.

“The way I think about the AI of the future is not as someone as smart as you or as smart as me, but as an automated organization that does science and engineering and development and manufacturing,”

Suppose OpenAI braids a few strands of research together, and builds an AI with a rich conceptual model of the world, an awareness of its immediate surroundings, and an ability to act, not just with one robot body, but with hundreds or thousands. “We’re not talking about GPT-4. We’re talking about an autonomous corporation,”

Its constituent AIs would work and communicate at high speed, like bees in a hive. A single such AI organization would be as powerful as 50 Apples or Googles, he mused. “This is incredible, tremendous, unbelievably disruptive power.”

Presume for a moment that human society ought to abide the idea of autonomous AI corporations. We had better get their founding charters just right. What goal should we give to an autonomous hive of AIs that can plan on century-long time horizons, optimizing billions of consecutive decisions toward an objective that is written into their very being?

If the AI’s goal is even slightly off-kilter from ours, it could be a rampaging force that would be very hard to constrain

We know this from history: Industrial capitalism is itself an optimization function, and although it has lifted the human standard of living by orders of magnitude, left to its own devices, it would also have clear-cut America’s redwoods and de-whaled the world’s oceans. It almost did.

one of its principal challenges will be making sure that the objectives we give to AIs stick

We can program a goal into an AI and reinforce it with a temporary period of supervised learning, Sutskever explained. But just as when we rear a human intelligence, our influence is temporary. “It goes off to the world,”

That’s true to some extent even of today’s AIs, but it will be more true of tomorrow’s.

He compared a powerful AI to an 18-year-old heading off to college. How will we know that it has understood our teachings? “Will there be a misunderstanding creeping in, which will become larger and larger?”

Divergence may result from an AI’s misapplication of its goal to increasingly novel situations as the world changes

Or the AI may grasp its mandate perfectly, but find it ill-suited to a being of its cognitive prowess. It might come to resent the people who want to train it to, say, cure diseases. “They want me to be a doctor,” Sutskever imagines an AI thinking. “I really want to be a YouTuber.”

If AIs get very good at making accurate models of the world, they may notice that they’re able to do dangerous things right after being booted up. They might understand that they are being red-teamed for risk, and hide the full extent of their capabilities.

hey may act one way when they are weak and another way when they are strong, Sutskever said

We would not even realize that we had created something that had decisively surpassed us, and we would have no sense for what it intended to do with its superhuman powers.

That’s why the effort to understand what is happening in the hidden layers of the largest, most powerful AIs is so urgent. You want to be able to “point to a concept,” Sutskever said. You want to be able to direct AI toward some value or cluster of values, and tell it to pursue them unerringly for as long as it exists.

we don’t know how to do that; indeed, part of his current strategy includes the development of an AI that can help with the research. If we are going to make it to the world of widely shared abundance that Altman and Sutskever imagine, we have to figure all this out.