Group items tagged

During that same period, 29 states had, on average, at least one federally declared disaster a year somewhere within their borders. Five states have experienced at least 20 disasters since 2011.

The numbers exclude disaster declarations related to the coronavirus pandemic.

“Climate change is here,”

“Every single taxpayer is paying for climate change.”

By focusing on federally declared disasters, the report is able to equalize those differences, offering something close to a true accounting of which places are most exposed to climate shocks they cannot cope with on their own.

At the top of that list are five counties that have each experienced, on average, more than a disaster a year since 2011. Those counties are concentrated in two areas: Southern Louisiana (where counties are called parishes) and eastern Kentucky.

Louisiana outpaces the rest of the United States in another regard. Over the past decade, the state has received more federal disaster money per capita — $1,736 for each resident — than anywhere else in the nation, the report found. Only New York State comes close, at $1,348.

Since 2011, California has received 25 federal disaster declarations, including for wildfires in 2017 and 2018 that resulted in $2.5 billion in federal money to rebuild public infrastructure. Mississippi and Oklahoma have each suffered 22 disasters. Iowa has had 21, mostly for severe storms and flooding

Not every type of disaster is associated with climate change. For example, it’s unclear whether there is a link between rising temperatures and earthquakes. But scientists have become increasingly convinced that a warming world is contributing to worsening floods, hurricanes, wildfires and other extreme weather events.

The data also shows the areas least exposed to unmanageable climate shocks, at least so far. States in the Midwest, including Illinois, Indiana, Ohio and Michigan, are among those with the smallest number of federal disaster declarations, with an average of roughly one disaster every two years.

At the bottom of the list is Nevada, which has had just three federal disaster declarations since 2011. Next door, Arizona has had just six. Yet, Nevada and Arizona ranked highest for heat-related deaths from 2018 to 2021, according to the report

“Heat has the highest mortality of all climate impacts, but their disaster declarations were so low,” Ms. Chester said. The reason: Federal disaster declarations focus on property damage more than direct human consequences like illness, injury or death.

“By better understanding risk,” she said, “we can more effectively take action together to accelerate resilience and adaptation in our nation’s most at-risk and disadvantaged communities.”

To pay for that new spending, Rebuild by Design proposes, states should impose a 2 percent surcharge on insurance premiums.

Using an insurance surtax to pay for disasters is a strategy that is already in use, in a sense. As the report notes, Florida levies surcharges on private insurance policies to make up for shortfalls in its state-run insurance program — something that’s likely to happen in the aftermath of Hurricane Ian.

Rebuild by Design suggests reversing the chronology. Rather than taxing insurance payments to pay for disaster recovery, a state would come up with additional funds before a storm, then use that money to better prepare communities before a disaster strikes, perhaps making it unnecessary for the federal government to declare a disaster at all.

Lesson #5: Proportionality should be a guideline in war.

McNamara talks about the proportions of cities destroyed in Japan by the US before the dropping of the nuclear bomb, comparing the destroyed Japanese cities to similarly-sized cities in the US: Tokyo, roughly the size of New York City, was 51% destroyed; Toyama, the size of Chattanooga, was 99% destroyed; Nagoya, the size of Los Angeles, was 40% destroyed; Osaka, the size of Chicago, was 35% destroyed; Kobe, the size of Baltimore, was 55% destroyed; etc. He says LeMay once said that, had the United States lost the war, they would have been tried for war crimes, and agrees with this assessment.

Lesson #7: Belief and seeing are both often wrong. McNamara affirms Morris' framing of lesson 7 in relation to the Gulf of Tonkin incident: "We see what we want to believe."

Lesson #8: Be prepared to reexamine your reasoning. McNamara says that, even though the United States is the strongest nation in the world, it should never use that power unilaterally: "if we can't persuade nations with comparable values of the merit of our cause, we better reexamine our reasoning."

We, the richest nation in the world, have failed in our responsibility to our own poor and to the disadvantaged across the world to help them advance their welfare in the most fundamental terms of nutrition, literacy, health and employment.

we are not omniscient. If we cannot persuade other nations with similar interests and similar values of the merits of the proposed use of that power, we should not proceed unilaterally except in the unlikely requirement to defend directly the continental U.S., Alaska and Hawaii.

War is a blunt instrument by which to settle disputes between or within nations, and economic sanctions are rarely effective. Therefore, we should build a system of jurisprudence based on the International Court—that the U.S. has refused to support—which would hold individuals responsible for crimes against humanity.

If we are to deal effectively with terrorists across the globe, we must develop a sense of empathy—I don't mean "sympathy," but rather "understanding"—to counter their attacks on us and the Western World.

We underestimated the power of nationalism to motivate a people to fight and die for their beliefs and values.

Our misjudgments of friend and foe, alike, reflected our profound ignorance of the history, culture, and politics of the people in the area, and the personalities and habits of their leaders.

We failed then—and have since—to recognize the limitations of modern, high-technology military equipment, forces, and doctrine. We failed, as well, to adapt our military tactics to the task of winning the hearts and minds of people from a totally different culture.

We did not recognize that neither our people nor our leaders are omniscient. Our judgment of what is in another people's or country's best interest should be put to the test of open discussion in international forums. We do not have the God-given right to shape every nation in our image or as we choose.

We did not hold to the principle that U.S. military action … should be carried out only in conjunction with multinational forces supported fully (and not merely cosmetically) by the international community.

5. Proportionality should be a guideline in war

8. Be prepared to re-examine your reasoning

1. The human race will not eliminate war in this century, but we can reduce the brutality of war—the level of killing—by adhering to the principles of a "Just War," in particular to the principle of "proportionality."

5. We, the richest nation in the world, have failed in our responsibility to our own poor and to the disadvantaged across the world to help them advance their welfare in the most fundamental terms of nutrition, literacy, health and employment.

8. War is a blunt instrument by which to settle disputes between or within nations, and economic sanctions are rarely effective. Therefore, we should build a system of jurisprudence based on the International Court—that the U.S. has refused to support—which would hold individuals responsible for crimes against humanity.

9. If we are to deal effectively with terrorists across the globe, we must develop a sense of empathy—I don't mean "sympathy," but rather "understanding"—to counter their attacks on us and the Western World.

3. We underestimated the power of nationalism to motivate a people to fight and die for their beliefs and values.

5. We failed then — and have since — to recognize the limitations of modern, high-technology military equipment, forces, and doctrine.

. We failed, as well, to adapt our military tactics to the task of winning the hearts and minds of people from a totally different culture

9. We did not recognize that neither our people nor our leaders are omniscient. Our judgment of what is in another people's or country's best interest should be put to the test of open discussion in international forums. We do not have the God-given right to shape every nation in our image or as we choose.

10. We did not hold to the principle that U.S. military action … should be carried out only in conjunction with multinational forces supported fully (and not merely cosmetically) by the international community.

11. We failed to recognize that in international affairs, as in other aspects of life, there may be problems for which there are no immediate solutions … At times, we may have to live with an imperfect, untidy world.

we can’t simply replace one simplistic narrative, about the super power of mitigation policy, for another, focused only on the burdens it imposed and not at all on the costs of doing much less — or nothing at all.

Let’s start with the title. What is the big failure, as you see it?

McLean: I think it gets at things that had happened in America even before the pandemic hit. And among those things were, I think, a failure to recognize the limits of capitalism, a failure of government to set the right rules for it, particularly when it comes to our health care system; a focus on profits that may have led to an increase in the bottom line but created fragility in ways people didn’t understand; and then our growing polarization that made us incapable of talking to each other

How big is the failure? When I look at The Economist’s excess mortality data, I see the U.S. had the 53rd-worst outcome in the world — worse than all of Western Europe, but better than all of Eastern Europe.

McLean: I think one way to quantify it is to take all those numbers and then put them in the context of our spending on health care. Given the amount we spend on health care relative to other countries, the scale of the failure becomes more apparent.

o me, the most glaring example is the schools. They were closed without people thinking through the potential consequences of closing down public schools, especially for disadvantaged kids.

to compound it, in my view, public health never made the distinction that needed to be made between the vulnerabilities of somebody 70 years old and the vulnerabilities of somebody 10 years old.

In the beginning of the book you write, in what almost feels like a thesis statement for the book: “A central tenet of this book is that we could not have done better, and pretending differently is a dangerous fiction, one that prevents us from taking a much needed look in the mirror.”

This claim, that the U.S. could not have done any better, runs against your other claim, that what we observed was an American failure. It is also a pretty extreme claim, I think, and I wanted to press you on it in part because it is, in my view, undermined by quite a lot of the work you do in the book itself.

Would the U.S. not have done better if it had recognized earlier that the disease spread through the air rather than in droplets? Would it not have done better if it hadn’t bungled the rollout of a Covid test in the early months?

McLean: Everything that you mentioned — the point of the book is that those were set by the time the pandemic hit.

in retrospect, what we were doing was to try to delay as much spread as we could until people got vaccinated. All the things that we did in 2020 were functionally serving or trying to serve that purpose. Now, given that, how can you say that none of that work saved lives?

McLean: I think that the test failure was baked into the way that the C.D.C. had come to operate

But the big question I really want to ask is this one: According to the C.D.C., we’ve had almost 1.2 million deaths from Covid. Excess mortality is nearly 1.4 million. Is it really your contention that there was nothing we might’ve done that brought that total down to 1.1 million, for instance, or even 900,000?

McLean: It’s very — you’re right. If you went through each and every thing and had a crystal ball and you could say, this could have been done, this could have been moved up by a month, we could have gotten PPE …

When I came to that sentence, I thought of it in terms of human behavior: What will humans put up with? What will humans stand for? How do Americans act? And you’ve written about Sweden being sort of average, and you’ve written about China and the Chinese example. They lock people up for two years and suddenly the society just revolts. They will not take it anymore. They can’t stand it. And as a result, a million and a half people die in a month and a half.

Well, I would tell that story very differently. For me, the problem is that when China opened up, they had fully vaccinated just under two-thirds of their population over 80. So to me, it’s not a failure of lockdowns. It’s a failure of vaccinations. If the Chinese had only achieved the same elderly vaccination rate as we achieved — which by global standards was pretty poor — that death toll when they opened up would have been dramatically lower.

What do you mean by “lockdown,” though? You use the word throughout the book and suggest that China was the playbook for all countries. But you also acknowledge that what China did is not anything like what America did.

Disparities in health care access — is it a dangerous fiction to think we might address that? You guys are big champions of Operation Warp Speed — would it not have been better if those vaccines had been rolled out to the public in nine months, rather than 12

. But this isn’t “lockdown” like there were lockdowns in China or even Peru. It’s how we tried to make it safer to go out and interact during a pandemic that ultimately killed a million Americans.

McLean: I think that you’re absolutely right to focus on the definition of what a lockdown is and how we implemented them here in this country. And I think part of the problem is that we implemented them in a way that allowed people who were well off and could work from home via Zoom to be able to maintain very much of their lives while other people couldn’t

And I think it depends on who you were, whether you would define this as a lockdown or not. If you were a small business who saw your small business closed because of this, you’re going to define it as a lockdown.

n the book you’re pretty definitive. You write, “maybe the social and economic disasters that lockdowns created would have been worth it if they had saved lives, but they hadn’t.” How can you say that so flatly?

I think there are still open questions about what worked and how much. But the way that I think about all of this is that the most important intervention that anybody did anywhere in the world was vaccination. And the thing that determined outcomes most was whether your first exposure came before or after vaccination.

Here, the shelter-in-place guidelines lasted, on average, five to seven weeks. Thirty nine of the 40 states that had issued them lifted them by the end of June, three months in. By the summer, according to Google mobility data, retail and grocery activity was down about 10 percent. By the fall, grocery activity was only down about 5 percent across the country

Nocera: Well, on some level, I feel like you’re trying to have it both ways. On the one hand, you’re saying that lockdowns saved lives. On the other hand, you said they weren’t real lockdowns because everybody was out and about.

I don’t think that’s having it both ways. I’m trying to think about these issues on a spectrum rather than in binaries. I think we did interrupt our lives — everybody knows that. And I think they did have an effect on spread, and that limiting spread had an effect by delaying infections until after vaccination.

Nocera: Most of the studies that say lockdowns didn’t work are really less about Covid deaths than about excess mortality deaths. I wound up being persuaded that the people who could not get to the hospital, because they were all working, because all the doctors were working on Covid and the surgical rooms were shut down, the people who caught some disease that was not Covid and died as a result — I wound up being persuaded about that.

We’re in a pandemic. People are going to die. And then the question becomes, can we protect the most vulnerable? And the answer is, we didn’t protect the most vulnerable. Nursing homes were a complete disaster.

There was a lot of worry early on about delayed health care, and about cancer in particular — missed screenings, missed treatments. But in 2019, we had an estimated 599,600 Americans die of cancer. In 2020, it was 602,000. In 2021, it was 608,000. In 2022, it was 609,000.

Nocera: See, it went up!But by a couple of thousand people, in years in which hundreds of thousands of Americans were dying of Covid.

Nocera: I think you can’t dispute the excess mortality numbers.I’m not. But in nearly every country in the world the excess mortality curves track so precisely with Covid waves that it doesn’t make sense to talk about a massive public health problem beyond Covid. And when you add all of these numbers up, they are nowhere near the size of the footfall of Covid. How can you look back on this and say the costs were too high?

Nocera: I think the costs were too high because you had school costs, you had economic costs, you had social costs, and you had death.

McLean: I think you’re raising a really good point. We’re making an argument for a policy that might not have been doable given the preconditions that had been set. I’m arguing that there were these things that had been put in place in our country for decades leading up to the pandemic that made it really difficult for us to plant in an effective way, from the outsourcing of our PPE to the distrust in our health care system that had been created by people’s lack of access to health care with the disparities in our hospital system.

How would you have liked to see things handled differently?Nocera: Well, the great example of doing it right is San Fran

I find the San Francisco experience impressive, too. But it was also a city that engaged in quite protracted and aggressive pandemic restrictions, well beyond just protecting the elderly and vulnerable.

McLean: But are we going to go for stay-at-home orders plus protecting vulnerable communities like San Francisco did? Or simply letting everybody live their lives, but with a real focus on the communities and places like nursing homes that were going to be affected? My argument is that we probably would’ve been better off really focusing on protecting those communities which were likely to be the most severely affected.

I agree that the public certainly didn’t appreciate the age skew, and our policy didn’t reflect it either. But I also wonder what it would mean to better protect the vulnerable than we did. We had testing shortages at first. Then we had resistance to rapid testing. We had staff shortages in nursing homes.

Nocera: This gets exactly to one of our core points. We had spent 30 years allowing nursing homes to be owned by private equity firms that cut the staff, that sold the land underneath and added all this debt on

I hear you saying both that we could have done a much better job of protecting these people and that the systems we inherited at the outset of the pandemic would’ve made those measures very difficult, if not impossible, to implement.

But actually, I want to stop you there, because I actually think that that data tells the opposite story.

And then I’m trying to say at the same time, but couldn’t we have done something to have protected people despite all of that?

I want to talk about the number of lives at stake. In the book, you write about the work of British epidemiologist Neil Ferguson. In the winter of 2020, he says that in the absence of mitigation measures and vaccination, 80 percent of the country is going to get infected and 2.2 million Americans are going to die. He says that 80 percent of the U.K. would get infected, and 510,000 Brits would die — again, in the abs

In the end, by the time we got to 80 percent of the country infected, we had more than a million Americans die. We had more than 200,000 Brits die. And in each case most of the infections happened after vaccination, which suggests that if those infections had all happened in a world without vaccines, we almost certainly would have surpassed two million deaths in the U.S. and almost certainly would’ve hit 500,000 deaths in the U.K.

In the book, you write about this estimate, and you endorse Jay Bhattacharya’s criticism of Ferguson’s model. You write, “Bhattacharya got his first taste of the blowback reserved for scientists who strayed from the establishment position early. He co-wrote an article for The Wall Street Journal questioning the validity of the scary 2 to 4 percent fatality rate that the early models like Neil Ferguson’s were estimating and that were causing governments to panic. He believed, correctly as it turns out, that the true fatality rate was much lower.”

Nocera: I know where you’re going with this, because I read your story about the nine pandemic narratives we’re getting wrong. In there, you said that Bhattacharya estimated the fatality rate at 0.01 percent. But if you actually read The Wall Street Journal article, what he’s really saying is I think it’s much lower. I’ve looked at two or three different possibilities, and we really need some major testing to figure out what it actually is, because I think 2 percent to 4 percent is really high.

He says, “if our surmise of 6 million cases is accurate, that’s a mortality rate of 0.01%. That is ⅒th the flu mortality rate of 0.1%.” An I.F.R. of 0.01 percent, spread fully through the American population, yields a total American death toll of 33,000 people. We have had 1.2 million deaths. And you are adjudicating this dispute, in 2023, and saying that Neil was wrong and Jay was right.

hird, in the Imperial College report — the one projecting two million American deaths — Ferguson gives an I.F.R. estimate of 0.9 percent.

Bhattacharya’s? Yes, there is some uncertainty around the estimate he offers. But the estimate he does offer — 0.01 percent — is one hundred times lower than the I.F.R. you yourselves cite as the proper benchmark.

Nocera: In The Wall Street Journal he does not say it’s 0.01. He says, we need to test to find out what it is, but it is definitely lower than 2 to 4 percent.

Well, first of all, the 2 percent to 4 percent fatality rate is not from Neil Ferguson. It’s from the W.H.O.

But I think that fundamentally, at the outset of the pandemic, the most important question orienting all of our thinking was, how bad could this get? And it turns out that almost all of the people who were saying back then that we shouldn’t do much to intervene were extremely wrong about how bad it would be

The argument then was, more or less, “We don’t need to do anything too drastic, because it’s not going to be that big a deal.” Now, in 2023, it’s the opposite argument: “We shouldn’t have bothered with restrictions, because they didn’t have an impact; we would have had this same death toll anyway.” But the death toll turned out to be enormous.

Now, if we had supplied all these skeptics with the actual numbers at the outset of the pandemic, what kind of audience would they have had? If instead of making the argument against universal mitigation efforts on the basis of a death toll of 40,000 they had made the argument on the basis of a death toll of more than a million, do you think the country would’ve said, they’re right, we’re doing too much, let’s back off?

McLean: I think that if you had gone to the American people and said, this many people are going to die, that would’ve been one thing. But if you had gone to the American people and said, this many people are going to die and a large percentage of them are going to be over 80, you might’ve gotten a different answer.

I’m not arguing we shouldn’t have been trying to get a clearer sense of the true fatality rate, or that we shouldn’t have been clearer about the age skew. But Bhattacharya was also offering an estimate of fatality rate that turned out to be off by a factor of a hundred from the I.F.R. that you yourselves cite as correct. And then you say that Bhattacharya was right and Ferguson was wrong.

And you, too, Joe, you wrote an article in April expressing sympathy for Covid skeptics and you said ——Nocera: This April?No, 2020.Nocera: Oh, oh. That’s the one where I praised Alex Berenson.You also cited some Amherst modeling which said that we were going to have 67,000 to 120,000 American deaths. We already had, at that point, 60,000. So you were suggesting, in making an argument against pandemic restrictions, that the country as a whole was going to experience between 7,000 and 60,000 additional deaths from that point.

when I think about the combination of the economic effects of mitigation policies and just of the pandemic itself and the big fiscal response, I look back and I think the U.S. managed this storm relatively well. How about each of you?

in this case, Congress did get it together and did come to the rescue. And I agree that made a ton of difference in the short term, but the long-term effects of the fiscal rescue package were to help create inflation. And once again, inflation hits those at the bottom of the socioeconomic distribution much harder than it does those at the top. So I would argue that some of what we did in the pandemic is papering over these long-term issues.

I think as with a lot of the stuff we’ve talked about today, I agree with you about the underlying problems. But if we take for granted for a moment that the pandemic was going to hit us, when it did, under the economic conditions it did, and then think about the more narrow context of whether, given all that, we handled the pandemic well. We returned quickly to prepandemic G.D.P. trends, boosted the wealth of the bottom half of the country, cut child poverty in half, pushed unemployment to historical lows.

What sense do you make of the other countries of the world and their various mitigation policies? Putting aside China, there’s New Zealand, Australia, South Korea — these are all places that were much more aggressive than the U.S. and indeed more than Europe. And had much, much better outcomes.

Nocera: To be perfectly honest, we didn’t really look, we didn’t really spend a lot of time looking at that.

McLean: But one reason that we didn’t is I don’t think it tells us anything. When you look at who Covid killed, then you have to look at what the pre-existing conditions in a country were, what percentage of its people are elderly. How sick are people with pre-existing conditions?

I just don’t think there’s a comparison. There’s just too many factors that influence it to be able to say that, to be able to compare America to any other country, you’d have to adjust for all these factors.

But you do spend a bit of time in the book talking about Sweden. And though it isn’t precisely like-for-like, one way you can control for some of those factors is grouping countries with their neighbors and other countries with similar profiles. And Sweden’s fatality rate in 2020 was 10 times that of Norway, Finland and Iceland. Five times that of Denmark. In the vaccination era, those gaps have narrowed, but by most metrics Sweden has still done worse, overall, than all of those countries.

On the matter of omniscience. Let’s say that we can send you back in time. Let’s put you both in charge of American pandemic response, or at least American communication about the pandemic, in early 2020. What would you want to tell the country? How would you have advised us to respond?

McLean: What I would want is honesty and communication. I think we’re in a world that is awash in information and the previous methods of communication — giving a blanket statement to people that may or may not be true, when you know there’s nuance underneath it — simply doesn’t work anymore

o I would’ve been much more clear — we think masks might help, we don’t know, but it’s not that big of an ask, let’s do it. We think the early data coming out of Italy shows that these are the people who are really, really at risk from Covid, but it’s not entirely clear yet. Maybe there is spread in schools, but we don’t know. Let’s look at this and keep an open mind and look at the data as it comes in.

The eerily humanlike chatbot, which went viral last year by generating its own answers to prompts, blindsided E.U. policymakers. The type of A.I. that powered ChatGPT was not mentioned in the draft law and was not a major focus of discussions about the policy. Lawmakers and their aides peppered one another with calls and texts to address the gap, as tech executives warned that overly aggressive regulations could put Europe at an economic disadvantage.

Even now, E.U. lawmakers are arguing over what to do, putting the law at risk. “We will always be lagging behind the speed of technology,” said Svenja Hahn, a member of the European Parliament who was involved in writing the A.I. law.

Lawmakers and regulators in Brussels, in Washington and elsewhere are losing a battle to regulate A.I. and are racing to catch up, as concerns grow that the powerful technology will automate away jobs, turbocharge the spread of disinformation and eventually develop its own kind of intelligence.

Nations have moved swiftly to tackle A.I.’s potential perils, but European officials have been caught off guard by the technology’s evolution, while U.S. lawmakers openly concede that they barely understand how it works.

The absence of rules has left a vacuum. Google, Meta, Microsoft and OpenAI, which makes ChatGPT, have been left to police themselves as they race to create and profit from advanced A.I. systems

At the root of the fragmented actions is a fundamental mismatch. A.I. systems are advancing so rapidly and unpredictably that lawmakers and regulators can’t keep pace

That gap has been compounded by an A.I. knowledge deficit in governments, labyrinthine bureaucracies and fears that too many rules may inadvertently limit the technology’s benefits.

Even in Europe, perhaps the world’s most aggressive tech regulator, A.I. has befuddled policymakers.

The European Union has plowed ahead with its new law, the A.I. Act, despite disputes over how to handle the makers of the latest A.I. systems.

The result has been a sprawl of responses. President Biden issued an executive order in October about A.I.’s national security effects as lawmakers debate what, if any, measures to pass. Japan is drafting nonbinding guidelines for the technology, while China has imposed restrictions on certain types of A.I. Britain has said existing laws are adequate for regulating the technology. Saudi Arabia and the United Arab Emirates are pouring government money into A.I. research.

A final agreement, expected as soon as Wednesday, could restrict certain risky uses of the technology and create transparency requirements about how the underlying systems work. But even if it passes, it is not expected to take effect for at least 18 months — a lifetime in A.I. development — and how it will be enforced is unclear.

Many companies, preferring nonbinding codes of conduct that provide latitude to speed up development, are lobbying to soften proposed regulations and pitting governments against one another.

“No one, not even the creators of these systems, know what they will be able to do,” said Matt Clifford, an adviser to Prime Minister Rishi Sunak of Britain, who presided over an A.I. Safety Summit last month with 28 countries. “The urgency comes from there being a real question of whether governments are equipped to deal with and mitigate the risks.”

Europe takes the lead

In mid-2018, 52 academics, computer scientists and lawyers met at the Crowne Plaza hotel in Brussels to discuss artificial intelligence. E.U. officials had selected them to provide advice about the technology, which was drawing attention for powering driverless cars and facial recognition systems.

as they discussed A.I.’s possible effects — including the threat of facial recognition technology to people’s privacy — they recognized “there were all these legal gaps, and what happens if people don’t follow those guidelines?”

In 2019, the group published a 52-page report with 33 recommendations, including more oversight of A.I. tools that could harm individuals and society.

By October, the governments of France, Germany and Italy, the three largest E.U. economies, had come out against strict regulation of general purpose A.I. models for fear of hindering their domestic tech start-ups. Others in the European Parliament said the law would be toothless without addressing the technology. Divisions over the use of facial recognition technology also persisted.

So when the A.I. Act was unveiled in 2021, it concentrated on “high risk” uses of the technology, including in law enforcement, school admissions and hiring. It largely avoided regulating the A.I. models that powered them unless listed as dangerous

“They sent me a draft, and I sent them back 20 pages of comments,” said Stuart Russell, a computer science professor at the University of California, Berkeley, who advised the European Commission. “Anything not on their list of high-risk applications would not count, and the list excluded ChatGPT and most A.I. systems.”

E.U. leaders were undeterred.“Europe may not have been the leader in the last wave of digitalization, but it has it all to lead the next one,” Ms. Vestager said when she introduced the policy at a news conference in Brussels.

In 2020, European policymakers decided that the best approach was to focus on how A.I. was used and not the underlying technology. A.I. was not inherently good or bad, they said — it depended on how it was applied.

Nineteen months later, ChatGPT arrived.

The Washington game

Lacking tech expertise, lawmakers are increasingly relying on Anthropic, Microsoft, OpenAI, Google and other A.I. makers to explain how it works and to help create rules.

“We’re not experts,” said Representative Ted Lieu, Democrat of California, who hosted Sam Altman, OpenAI’s chief executive, and more than 50 lawmakers at a dinner in Washington in May. “It’s important to be humble.”

Tech companies have seized their advantage. In the first half of the year, many of Microsoft’s and Google’s combined 169 lobbyists met with lawmakers and the White House to discuss A.I. legislation, according to lobbying disclosures. OpenAI registered its first three lobbyists and a tech lobbying group unveiled a $25 million campaign to promote A.I.’s benefits this year.

In that same period, Mr. Altman met with more than 100 members of Congress, including former Speaker Kevin McCarthy, Republican of California, and the Senate leader, Chuck Schumer, Democrat of New York. After testifying in Congress in May, Mr. Altman embarked on a 17-city global tour, meeting world leaders including President Emmanuel Macron of France, Mr. Sunak and Prime Minister Narendra Modi of India.

, the White House announced that the four companies had agreed to voluntary commitments on A.I. safety, including testing their systems through third-party overseers — which most of the companies were already doing.

“It was brilliant,” Mr. Smith said. “Instead of people in government coming up with ideas that might have been impractical, they said, ‘Show us what you think you can do and we’ll push you to do more.’”

In a statement, Ms. Raimondo said the federal government would keep working with companies so “America continues to lead the world in responsible A.I. innovation.”

Over the summer, the Federal Trade Commission opened an investigation into OpenAI and how it handles user data. Lawmakers continued welcoming tech executives.

In September, Mr. Schumer was the host of Elon Musk, Mark Zuckerberg of Meta, Sundar Pichai of Google, Satya Nadella of Microsoft and Mr. Altman at a closed-door meeting with lawmakers in Washington to discuss A.I. rules. Mr. Musk warned of A.I.’s “civilizational” risks, while Mr. Altman proclaimed that A.I. could solve global problems such as poverty.

A.I. companies are playing governments off one another. In Europe, industry groups have warned that regulations could put the European Union behind the United States. In Washington, tech companies have cautioned that China might pull ahead.

In May, Ms. Vestager, Ms. Raimondo and Antony J. Blinken, the U.S. secretary of state, met in Lulea, Sweden, to discuss cooperating on digital policy.

“China is way better at this stuff than you imagine,” Mr. Clark of Anthropic told members of Congress in January.

After two days of talks, Ms. Vestager announced that Europe and the United States would release a shared code of conduct for safeguarding A.I. “within weeks.” She messaged colleagues in Brussels asking them to share her social media post about the pact, which she called a “huge step in a race we can’t afford to lose.”

Months later, no shared code of conduct had appeared. The United States instead announced A.I. guidelines of its own.

Little progress has been made internationally on A.I. With countries mired in economic competition and geopolitical distrust, many are setting their own rules for the borderless technology.

Yet “weak regulation in another country will affect you,” said Rajeev Chandrasekhar, India’s technology minister, noting that a lack of rules around American social media companies led to a wave of global disinformation.

“Most of the countries impacted by those technologies were never at the table when policies were set,” he said. “A.I will be several factors more difficult to manage.”

Even among allies, the issue has been divisive. At the meeting in Sweden between E.U. and U.S. officials, Mr. Blinken criticized Europe for moving forward with A.I. regulations that could harm American companies, one attendee said. Thierry Breton, a European commissioner, shot back that the United States could not dictate European policy, the person said.

Some policymakers said they hoped for progress at an A.I. safety summit that Britain held last month at Bletchley Park, where the mathematician Alan Turing helped crack the Enigma code used by the Nazis. The gathering featured Vice President Kamala Harris; Wu Zhaohui, China’s vice minister of science and technology; Mr. Musk; and others.

The upshot was a 12-paragraph statement describing A.I.’s “transformative” potential and “catastrophic” risk of misuse. Attendees agreed to meet again next year.

The talks, in the end, produced a deal to keep talking.

How could these two seemingly contradictory statements both be true?

respondents in 2016 and 2020 reported more moderate views, on average, than in previous elections. As a result, Trump improved the most over previous Republicans by capturing the votes of a larger number of people who report racially moderate views

the number of people scoring at moderate levels of racial resentment increased. Trump was not as popular among this voting bloc, compared to those with high racial resentment. But because this group is larger, whites with moderate racial resentment scores ended up contributing more net votes to Trump.

The point about Trump voters being less racially resentful on average than voters for previous Republican candidates, while likely true, should, I think, be interpreted as a statement about why it’s important to be mindful of over-time changes in groups’ sizes in the population,

Trump’s supporters were less xenophobic than prior Republican candidates’, less sexist, had lower animus to minority groups, and lower levels of racial resentment. Far from deplorables, Trump voters were, on average, more tolerant and understanding than voters for prior Republican candidates.

The data, Grimmer continued,point to two important and undeniable facts. First, analyses focused on vote choice alone cannot tell us where candidates receive support. We must know the size of groups and who turns out to vote. And we cannot confuse candidates’ rhetoric with the voters who support them, because voters might support the candidate despite the rhetoric, not because of it.

Several pieces of research into the 2016 election, including our book, “Identity Crisis,” and this interesting paper by Grimmer, Marble and Tanigawa-Lau, find that people’s vote choices in that election were more strongly related to their views on “identity-inflected issues” than they had been in prior elections. That is why our book argues that these issues are central to how we interpret the outcome in 2016.

the Grimmer paper in fact provides a key corrective to the debate over the 2016 election. In an email, Kane pointed to a key section that reads:

election-night pundits and even some academics have claimed that Trump’s victory was the result of appealing to white Americans’ racist and xenophobic attitudes. We show this conventional wisdom is (at best) incomplete

The Grimmer paper, Engelhardt continued, “encourages us to take a step back and focus on the big picture for understanding elections: where do most votes come from and are these patterns consistent across elections?”

understanding election outcomes requires not just understanding what contributes to vote choice (e.g., racial group attachments, racial prejudice), but also how many people with that particular attitude turned out to vote and what share of the electorate that group makes up.

Trump, Westwood concluded, “found support from both racists and moderates, but with the pool of racist voters shrinking, it is clear this is not a path to future victory.”

Discussion of racial resentment driving support for Trump could miss how folks low in racial resentment were actually critical to the election outcome. The paper makes just this clarifying point, noting, for instance, that White Democrats low in racial resentment were even more influential in contributing votes to Clinton in 2016 than to Obama in 2012. Change between 2012 and 2016 is not exclusively due to the behavior of the most prejudiced.

“It is a nice reminder for scholars and, especially, the media, that it is important to think carefully about base rates.”

Donald Trump’s candidacy in 2016 was a stress test for Republican partisanship, and Republican partisanship passed with flying colors

The election was close enough for Trump to win because the vast majority of G.O.P. voters found the idea of either sitting it out or voting for a Democrat they had spent 20+ years disliking so distasteful that Trump’s limitations, liabilities and overt racism and misogyny were not a deal-breaker.

Theodoridis noted that his oneminor methodological and measurement critique is that this sort of analysis has to take seriously what the racial resentment scale actually means

It may be that race is actually quite salient for those in the middle part of the scale, but they are just less overtly racist than those at the top of the scale.

and NOT as a statement about Trump being successful in attracting racially liberal voters (indeed, those lowest in racial resentment turned away from him, per Grimmer-Marble-Tanigawa-Lau’s own findings).

It is an interesting academic exercise to predict who will win the vote within a specific group, but it is more fundamental to elections to understand how many voters candidates will gain from each group

the important contribution from Grimmer et al is that there was a big change in the attitudes of the white electorate. A small number of whites with high levels of racial resentment did support Trump in 2016 at a higher rate than in prior elections, but the bulk of support for Trump came from more moderate whites. Trump managed to pull in support from racists, but he was able to pull in much more support from economically disadvantaged whites.

The Grimmer paper, according to Westwood, has significant implications for those making “general claims about the future Republican Party,” specifically challenging those who believe

that Republicans can continue to win by appealing to white Americans’ worst attitudes and instincts. While it is true Trump support is largest for the most racist voters, this group is a shrinking part of the electorate

Republicans, as Grimmer et al. show, must figure out how to appeal to moderate whites who hold more moderate attitudes in order to win

Also, the meaning of the racial resentment scale changes over time in ways that are not independent of politics, and especially of presidential politics. Position on the scale is not immutable in the way some descriptive characteristics may be.

it’s critical to avoid the idea that there is a single skeleton key that can explain all the varied undercurrents that led to Trump’s 2016 victory, or that any one paper will provide a definitive explanation

One clear benefit emerging from the continuing study of Trump’s 2016 victory is a better understanding of the complexity and nuance of what brought it about.

the presence of racial resentment among Republican voters emerged long before Trump ran for president, while such resentment among Democratic voters has been sharply declining

racial resentment didn’t do more for Trump than it did for Romney. The highly racially resentful have, with reason, been voting for Republicans for a long time

Trump’s more explicit use of race didn’t make supporters more racially resentful. Levels of racial resentment among Republicans are no higher now than they were before Trump. In fact, they are slightly lower

And the highly racially resentful already knew full well that their home was in the G.O.P.

While the focus of attention has been on those who fall at the high end of the distribution on racial resentment

Almost all the change has taken place among Democrats, as they moved to lower and lower levels of resentment

In a statistical sense, the fact that there are now so many more people at the low end of the distribution than before will produce a larger coefficient for the effect of racial resentment on voting behavior.

that does not mean that those high in racial resentment are now even more likely to vote for Republicans or that there are more people high in resentment

there are more people low in resentment than before and that they are even less likely to vote for Republicans than before. So the low end of the scale is doing the work.

my own view is that Grimmer, Marble and Tanigawa-Lau have made a significant contribution to understanding the Trump phenomenon.

Most important, they make the case that explanations of Trump’s victory pointing to the role of those at the extremes on measures of racial resentment and sexism, while informative, are in their own way too comforting, fostering the belief that Trump’s triumph was the product of voters who have drifted far from the American mainstream.

In fact, the new analysis suggests that Trumpism has found fertile ground across a broad swath of the electorate, including many firmly in the mainstream. That Trump could capture the hearts and minds of these voters suggests that whatever he represents beyond racial resentment — anger, chaos, nihilism, hostility — is more powerful than many recognize or acknowledge

“It is unfortunate to frame this as an arms race,” Mr. Tegmark said. “It is more of a suicide race. It doesn’t matter who is going to get there first. It just means that humanity as a whole could lose control of its own destiny.” 

Messrs. Musk and Wozniak have both voiced concerns about AI technology. Mr. Musk on Wednesday tweeted that developers of the advanced AI technology “will not heed this warning, but at least it was said.”

Yann LeCun, chief AI scientist at Meta Platforms Inc., on Tuesday tweeted that he didn’t sign the letter because he disagreed with its premise. 

Mr. Mostaque, Stability AI’s CEO, said in a tweet Wednesday that although he signed the letter, he didn’t agree with a six-month pause. “It has no force but will kick off an important discussion that will hopefully bring more transparency & governance to an opaque area.”

Mr. Tegmark said many companies feel “crazy commercial pressures” to add advanced AI technology into their products. A six-month pause would allow the industry “breathing room,” without disadvantaging ones that opt to move carefully

The letter said a pause should be declared publicly and be verifiable and all key actors in the space should participate. “If such a pause cannot be enacted quickly, governments should step in and institute a moratorium,” it said

AI labs and experts can use this time to develop a set of shared safety rules for advanced AI design that should be audited and overseen by outside experts, the authors wrote.

“I don’t think we can afford to just go forward and break things,” said Mr. Bengio, who shared a 2018 Turing award for inventing the systems that modern AI is built on. “We do need to take time to think through this collectively.”

But to understand why Republicans have died at higher rates, you can’t look at vaccine status alone. Congressional districts controlled by a trifecta of Republican leaders—state governor, Senate, and House—had an 11 percent higher death rate, according to the Lancet study. A likely explanation, the authors write, could be that in the post-vaccine era, those leaders chose policies and conveyed public-health messages that made their constituents more likely to die. Although we still can’t say these decisions led to higher death rates, the association alone is jarring.

One of the most compelling studies comes from researchers at Yale, who published their findings as a working paper in November. They link political party and excess-death rate—the percent increase in deaths above pre-COVID levels—among those registered as either Democrats or Republicans, providing a more granular view. They chose to analyze data from Florida and Ohio from before and after vaccines were available. Looking at the period before the vaccine,  researchers found a 1.6 percentage-point difference in excess death rate among Republicans and Democrats, with a higher rate among Republicans. But after vaccines became available, that gap widened dramatically to 10.4 percentage points, again with a higher Republican excess death rate. “When we compare individuals who are of the same age, who live in the same county in the same month of the pandemic, there are differences correlated with your political-party affiliation that emerge after vaccines are available,”

What’s most concerning about all of this is that partisan disparities in death rates were also apparent before COVID. People living in Republican jurisdictions have been at a health disadvantage for more than 20 years. From 2001 to 2019, the death rate in Democratic counties decreased by 22 percent, according to a recent study; in Republican counties, it declined by only 11 percent. In the same time period, the political gap in death rates increased sixfold.

over the decades, state policy decisions on health issues such as Medicaid, gun legislation, tobacco taxes, and, indeed, vaccines have likely had a stronger impact on state health trajectories than other factors

the long-term decline of health in red states indicates that there is an ongoing problem at a high level in Republican-led places, and that something has gone awry. “If you happen to live in certain states, your chances for living a long life are going to be much higher than if you’re an American living in a different state,”

o acknowledge how many Republicans didn’t have to die would mean giving credence to scientific and medical expertise. So long as America remains locked in a poisonous partisan battle in which science is wrongly dismissed as being associated with the left, the death toll will only rise.

At the same time, the share of high school students who say they have had sexual intercourse has fallen 29 percent since 1991,

The share of female teens who did not use birth control the last time they had sex dropped by more than a third over the last decade, according to an analysis of government surveys by the Guttmacher Institute. The share using the most effective form, long-acting reversible contraception (delivered through an intrauterine device or arm implant), rose fivefold to 15 percent. The use of emergency contraception also rose.

“They’re going to school and seeing new career paths open,” said Melissa S. Kearney, an economist at the University of Maryland. “Whether they are excited about their own opportunities or feel that unreliable male partners leave them no choice, it leads them in the same direction — not becoming a young mother.”

Abortion does not appear to have driven the decline in teen births. As a share of teenage pregnancy, it has remained steady over the past decade,

If adolescent girls are more cautious with sex and birth control, what explains the caution? A common answer is that more feel they have something to lose. “There is just a greater confidence among young women that they have educational and professional opportunities,” Mr. Wilcox said.

t women in their mid-30s were nearly 25 percent more likely than men to have a four-year college degree, and at every educational level earnings had grown faster for women than men.

Skeptics see limits in the data and note that the payoff to education is growing.“I strongly disagree with the argument that teen births have no effect on social mobility,” said Isabel V. Sawhill of the Brookings Institution. “It’s a lot easier to move out of poverty if you’re not responsible for a child in your teenage years.”

The debate is more than academic. Some progressives worry that a narrow focus on preventing teen births will undermine broader anti-poverty plans and risks blaming adolescents for their poverty. Other see reducing poverty and teen births as complementary causes meant not to blame young women but empower them.