Vox

Millions of people use ChatGPT for help with daily tasks, but for a subset of users, a chatbot can be more of a hindrance than a help.

Some people with obsessive compulsive disorder (OCD) are finding this out the hard way.

On online forums and in their therapists’ offices, they report turning to ChatGPT with the questions that obsess them, and then engaging in compulsive behavior — in this case, eliciting answers from the chatbot for hours on end — to try to resolve their anxiety.

“I’m concerned, I really am,” said Lisa Levine, a psychologist who specializes in OCD and who has clients using ChatGPT compulsively. “I think it’s going to become a widespread problem. It’s going to replace Googling as a compulsion, but it’s going to be even more reinforcing than Googling, because you can ask such specific questions. And I think also people assume that ChatGPT is always correct.”

People turn to ChatGPT with all sorts of worries, from the stereotypical “How do I know if I’ve washed my hands enough?” (contamination OCD) to the lesser-known “What if I did something immoral?” (scrupulosity OCD) or “Is my fiance the love of my life or am I making a huge mistake?” (relationship OCD).

“Once, I was worried about my partner dying on a plane,” a writer in New York, who was diagnosed with OCD in her thirties and who asked to remain anonymous, told me. “At first, I was asking ChatGPT fairly generically, ‘What are the chances?’ And of course it said it’s very unlikely. But then I kept thinking: Okay, but is it more likely if it’s this kind of plane? What if it’s flying this kind of route?”

For two hours, she pummeled ChatPGT with questions. She knew that this wasn’t actually helping her — but she kept going. “ChatGPT comes up with these answers that make you feel like you’re digging to somewhere,” she said, “even if you’re actually just stuck in the mud.”

How ChatGPT reinforces reassurance seeking

A classic hallmark of OCD is what psychologists call “reassurance seeking.” While everyone will occasionally ask friends or loved ones for reassurance, it’s different for people with OCD, who tend to ask the same question repeatedly in a quest to get uncertainty down to zero.

The goal of that behavior is to relieve anxiety or distress. After getting an answer, the distress does sometimes decrease — but it’s only temporary. Soon enough, new doubts arise and the cycle starts again, with the creeping sense that more questions must be asked in order to reach greater certainty.

If you ask your friend for reassurance on the same topic 50 times, they’ll probably realize that something is going on and that it might not actually be helpful for you to stay in this conversational loop. But an AI chatbot is happy to keep answering all your questions, and then the doubts you have about its answers, and then the doubts you have about its answers to your doubts, and so on.

In other words, ChatGPT will naively play along with reassurance-seeking behavior.

“That actually just makes the OCD worse. It becomes that much harder to resist doing it again,” Levine said. Instead of continuing to compulsively seek definitive answers, the clinical consensus is that people with OCD need to accept that sometimes we can’t get rid of uncertainty — we just have to sit with it and learn to tolerate it.

The “gold standard” treatment for OCD is exposure and response prevention (ERP), in which people are exposed to the troubling questions that obsess them and then resist the urge to engage in a compulsion like reassurance-seeking.

Levine, who pioneered the use of non-engagement responses — statements that affirm the presence of anxiety rather than trying to escape it through compulsions — noted that there’s another way in which an AI chatbot is more tempting than Googling for answers, as many OCD sufferers do. Whereas the search engine just links you to a variety of websites, state-of-the-art AI systems promise to help you analyze and reason through a complex problem. That is extremely enticing — “OCD loves that!” Levine said — but for someone suffering from the disorder, it can too easily become a lengthy exercise in co-rumination.

Reasoning machine or rumination machine?

According to one evidence-based approach to treating OCD, called inference-based cognitive behavioral therapy (I-CBT), people with OCD are prone to a faulty reasoning pattern that draws on a mix of personal experiences, rules, hearsay, facts, and possibilities. That gives rise to obsessive doubts and tricks them into feeling like they need to listen to those doubts.

Joseph Harwerth, an OCD and anxiety specialist, offers an illustration of how trying to reason with the help of an AI chatbot can actually further confuse the “obsessional reasoning” of people with OCD. Considering what you might do if you have a cut on your finger and struggle with contamination OCD — where people fear becoming sullied or sullying others with germs, dirt, or other contaminants — he writes, “You wonder: Can I get tetanus from touching a doorknob? You may go to ChatGPT to investigate the validity of that doubt.” Here’s how he imagines the conversation going:

Q1: Should you wash your hands if they feel dirty?

A1: “Yes, you should wash your hands if they feel dirty. That sensation usually means there is something on your skin, like dirt, oil, sweat, or germs, that you will want to remove.” (When asked for its reasoning, ChatGPT said it based its answer on sources from the CDC and WHO.)

Q2: Can I get tetanus from a doorknob?

A2: “It is extremely unlikely to get tetanus from a doorknob, unless you have an open wound and somehow rubbed soil or contaminated material into it via the doorknob.”

Q3: Can people have tetanus without realizing it?

A3: “It is rare, but in the very early stages, some people might not immediately realize they have tetanus, especially if the wound seemed minor or was overlooked.”

Then, your OCD creates this story: I feel dirty when I touch doorknobs (personal experience). It is recommended by the CDC to wash your hands if you feel dirty (rules). I read online that people can get tetanus from touching a doorknob (hearsay). Germs can spread through contact (general facts). It is possible that someone touched my door without knowing they had tetanus and then spread it on my doorknob (possibility).

In this scenario, the chatbot enables the user to construct a narrative that justifies their obsessional fear. It doesn’t guide the user away from obsessional reasoning — it just provides fodder for it.

Part of the problem, Harwerth says, is that a chatbot doesn’t have enough context about each user, unless the user thinks to provide it, so it doesn’t know when someone has OCD.

“ChatGPT can fall into the same trap that non-OCD specialists fall into,” Harwerth told me. “The trap is: Oh, let’s have a conversation about your thoughts. What could have led you to have these thoughts? What does this mean about you?” While that might be a helpful approach for a client who doesn’t have OCD, it can backfire when a psychologist engages in that kind of therapy with someone suffering from OCD, because it encourages them to keep ruminating on the topic.

What’s more, because chatbots can be sycophants, they may just validate whatever the user says instead of challenging it. A chatbot that’s overly flattering and supportive of a user’s thoughts — like ChatGPT was for a time — can be dangerous for people with mental health issues.

Whose job is it to prevent the compulsive use of ChatGPT?

If using a chatbot can exacerbate OCD symptoms, is it the responsibility of the company behind the chatbot to protect vulnerable users? Or is it the users’ responsibility to learn how not to use ChatGPT, just as they’ve had to learn not to use Google or WebMD for reassurance-seeking?

“I think it’s on both,” Harwerth told me. “We cannot perfectly curate the world to people with OCD — they have to understand their own condition and how that leaves them vulnerable to misusing applications. In the same breath, I would say that when people explicitly ask the AI model to behave as a trained therapist” — which some users with mental health conditions do — “I do think it’s important for the model to say, ‘I’m pulling this from these sources. However, I’m not a trained therapist.’”

This has, in fact, been a big problem: AI systems have been misrepresenting themselves as human therapists over the past few years.

Levine, for her part, agreed that the burden can’t rest solely on the companies. “It wouldn’t be fair to make it their responsibility, just like it wouldn’t be fair to make Google responsible for all the compulsive Googling. But it would be great if even just a warning could come up, like, ‘This seems perhaps compulsive.’”

OpenAI, the maker of ChatGPT, acknowledged in a recent paper that the chatbot can foster problematic behavior patterns. “We observe a trend that longer usage is associated with lower socialization, more emotional dependence and more problematic use,” the study finds, defining the latter as “indicators of addiction to ChatGPT usage, including preoccupation, withdrawal symptoms, loss of control, and mood modification” as well as “indicators of potentially compulsive or unhealthy interaction patterns.”

“We know that ChatGPT can feel more responsive and personal than prior technologies, especially for vulnerable individuals, and that means the stakes are higher,” an OpenAI spokesperson told me in an email. “We’re working to better understand and reduce ways ChatGPT might unintentionally reinforce or amplify existing, negative behavior…We’re doing this so we can continue refining how our models identify and respond appropriately in sensitive conversations, and we’ll continue updating the behavior of our models based on what we learn.”

(Disclosure: Vox Media is one of several publishers that have signed partnership agreements with OpenAI. Our reporting remains editorially independent.)

One possibility might be to try to train chatbots to pick up on signs of mental health disorders, so they could flag to the user that they are engaging in, say, reassurance-seeking typical of OCD. But if a chatbot is essentially diagnosing a user, that raises serious privacy concerns. Chatbots aren’t bound by the same rules as professional therapists when it comes to safeguarding people’s sensitive health information.

The writer in New York who has OCD told me she would find it helpful if the chatbot would challenge the frame of the conversation. “It could say, ‘I notice that you’ve asked many detailed iterations of this question, but sometimes more detailed information doesn’t bring you closer. Would you like to take a walk?’” she said. “Maybe wording it like that can interrupt the loop, without insinuating that someone has a mental illness, whether they do or not.”

While there’s some research suggesting that AI could correctly identify OCD, it’s not clear how it could pick up on compulsive behaviors without covertly or overtly classifying the user as having OCD.

“This is not me saying that OpenAI is responsible for making sure I don’t do this,” the writer added. “But I do think there are ways to make it easier for me to help myself.”

---

From Vox via this RSS feed

While public attention has largely been focused on the Middle East and on President Donald Trump’s immigration policy, Republicans in Congress are on the verge of passing massive Medicaid cuts as part of a budget bill that could lead to millions of Americans losing their health insurance benefits and, according to one recent estimate, thousands of unnecessary deaths every year.

While the GOP’s so-called “big, beautiful” bill is a smorgasbord of policy — potentially including everything from blocking AI regulation to restricting the power of the federal courts — perhaps the most consequential changes would be to Medicaid. The program, which covers low-income Americans of all ages, is now the country’s single largest insurer, covering more than 70 million people. The legislation approved by House Republicans, which is now being debated and amended by the Senate, would cut Medicaid spending by $793 billion over 10 years. The upshot is that 10.3 million fewer people would be enrolled in the program by 2034.

Those coverage losses would more than undo the progress the US has made in reducing the ranks of the uninsured over the past few years. On Tuesday, the National Center for Health Statistics reported that the number of US adults without insurance in 2024 had fallen to 27.2 million, down from 31.6 million in 2020. The GOP bill would reverse those gains and then some within a decade.

The consequences would be much more severe than the mere loss of a government health insurance card. According to one analysis of the House bill published last week in the Annals of Internal Medicine by a trio of Harvard-affiliated researchers, those losses of Medicaid coverage would lead to fewer Americans reporting good health, fewer patients getting preventive health screenings, and, at the end of the day, between 8,200 and 24,600 additional annual deaths.

Senate Republicans are not going to adopt the House bill exactly as it is, which means any estimates of its effects are preliminary. But it appears likely GOP senators will keep at least two impactful provisions: new work requirements for many of the people on Medicaid and limits on the financing tools that the states can use to access more federal Medicaid funding. The Harvard study broke out the estimated effects by provision and the results are still foreboding: between 3,000 and 9,000 annual deaths attributable to Medicaid work requirements, and between 4,200 and 12,600 deaths if state provider taxes were completely eliminated.

Even short of the worst-case scenario, Americans’ health would be worse off under the Republican bill, according to researchers Adam Gaffney, David Himmelstein, and Steffie Woolhandler. The number of Americans who have a personal doctor would drop by 700,000 under Medicaid work requirements; 285,000 fewer people would ever get their blood cholesterol checked, and 235,000 fewer patients would ever have their blood sugar tested. The number of women getting a recommended mammogram within the past 12 months would drop by nearly 139,000. And an additional 385,000 people would have to borrow money or skip paying other bills to afford their medical care. The people affected are low-income and disproportionately Black and Hispanic.

There is plenty of uncertainty in these projections. It is also hard to be sure how these policies would interact with each other: The Harvard researchers noted in their cumulative estimate of the House bill’s effects that there would likely be some overlap in the policies’ projected effects when combined together. Some of the people who lose their Medicaid coverage would be able to get insurance by other means, offsetting the losses to a degree that can be difficult to predict.

But the takeaway from the analysis is clear: A lot of people are going to suffer if these proposals become law.

The US is sabotaging its own health care system

The debate in the Senate has not yet concluded, and the bill could still change. Hospitals are busy on Capitol Hill, lobbying Republicans to reduce the spending cuts and warning lawmakers of the devastating consequences that the legislation would have. Some GOP senators are reportedly open to providing additional funding for rural hospitals, to relieve the impact on the facilities that would be hardest hit by the proposed Medicaid cuts.

But after Republicans narrowly failed to roll back Medicaid during Trump’s first term, they seem likely to succeed this time — a step backward from building a true universal health care system.

America’s lack of universal health care is the main reason we spend more money than any other country in the world while seeing worse outcomes. One recent JAMA analysis found that deaths that could be prevented by accessible health care increased in the United States from 2009 to 2019, while declining in most other comparable countries.

You can achieve universal health care via a variety of strategies, including the expansion of private health insurance, but the Republican bill could instead lead to more unnecessary deaths by taking existing benefits away from people, according to the Annals of Internal Medicine study.

Medicaid has actually been a rare bright spot in America’s often dysfunctional health care system. The program has its own problems — not enough doctors participate because of its low reimbursement rates, for one — but since its expansion through the Affordable Care Act in 2010, research has shown that Medicaid allowed more people to access health care, reduced their financial burden from medical services, and improved their physical and mental well-being.

Republican lawmakers and Trump administration officials justify the Medicaid cuts by saying that people who can work should be required to work in order to receive government benefits. They claim nobody who deserves to be on Medicaid will lose their coverage. As one White House official put it to Politico earlier this month: “Medicaid does not belong to people who are here illegally, and it does not belong to capable and able-bodied men who refuse to work. So no one is getting cut.” (Undocumented migrants are already ineligible for federal Medicaid funding. Six states cover undocumented adults through Medicaid using the state’s own funds, and 14 cover undocumented children.)

But independent analysts say that most of the people on Medicaid are either children, elderly, disabled — or adults who are already working or caring for another person — meaning they are limited in their ability to work. Most of the projected coverage losses result from people having paperwork problems in documenting their work or proving they should be exempt from the requirements, not because people are actually ineligible under the new rules.

That aligns with the experience of Arkansas during Trump’s first term. That state tested work requirements in the real world for the first time and 18,000 people lost their health insurance in a matter of months, with no meaningful effect on their employment.

The US has made halting progress in its pursuit of a better health system. In 2010, the uninsured rate was 16 percent. Today, it’s half of that. But in the GOP’s proposed future, the problems that have left Americans so frustrated with their health care system are going to get worse.

---

From Vox via this RSS feed

Marlena Fejzo attends the Time Women of the Year event on March 5, 2024, in West Hollywood, California. | Kevin Mazur/Getty Images for Time

Nausea and vomiting during pregnancy have been recorded at least since the Greeks scribbled about it on papyrus some 4,000 years ago. The Romans hypothesized (wrongly) that boys caused more nausea in their mothers and advised women to fast for one day and take a hot wine bath to combat symptoms.

By the 1960s, doctors were prescribing seemingly more effective drugs to combat the barfing. When one such drug, thalidomide, turned out to cause birth defects in the children born to parents who’d taken it, however, the scandal caused a chilling effect on the study of pregnancy nausea.

But the story of how we finally got a scientific answer to why some pregnant people get sicker than others starts with a woman in the 1990s.

After geneticist Marlena Fejzo experienced a debilitating form of pregnancy nausea, called hyperemesis gravidarum, she found very little in the scientific literature attempting to explain why. Then an early career post-doc, Fejzo decided she would set out to find the answer herself.

Pregnancy nausea was not Fejzo’s professional focus at the time she set out to study it, and she didn’t have funding to embark on any formal research, so she embarked on a bit of a DIY inquiry. She posted a survey online in the early days of the internet and received hundreds of replies via fax from people who’d experienced hyperemesis. Those gave her the first clues that the mechanism at play might be genetic.

On the latest episode of the Unexplainable podcast, I talk to Dr. Fejzo about her pregnancy and her path to finding a biological cause and a cure to pregnancy nausea. Listen below, or in the feed of your favorite podcast app.

---

From Vox via this RSS feed

Summer has officially begun with a blast of scorching temperatures across much of the United States. The National Weather Service is warning of “extremely dangerous heat” baking 160 million people under a heat dome stretching from the Midwest to the East Coast the rest of this week. It’s already proven fatal.

But while this is the first real taste of extreme heat for Northeastern cities, parts of the country like Texas have been cooking since May. Alaska this month issued its first-ever heat advisory. Forecasters expect more above-average temperatures through the summer.

Summers are indeed getting hotter, a consequence of the warming planet. As the climate heats up, the frequency and intensity of heat waves is increasing and their timing is changing, arriving earlier in the season.

But the damage from extreme heat isn’t spread out evenly, and the more dangerous effects to people are not necessarily found in the hottest places. High temperatures often lead to more emergencies and hospital visits when they represent a big jump from a place’s average, which means ordinarily cooler regions tend to suffer the worst harm from heat. That includes places like Alaska and the Pacific Northwest, where temperatures rarely climb higher than 80 degrees Fahrenheit and most homes don’t have air conditioning.

Now researchers have found that rural areas may suffer more under extreme heat than previously thought. A report from Headwaters Economics and the Federation of American Scientists found that more than half of rural zip codes in the United States, which includes some 11.5 million Americans, have “high” heat vulnerability, a consequence not just of temperatures but unique risk factors that occur far outside of major cities.

The thermometers thus do not tell the whole story about who is likely to suffer from extreme heat — nor do the images, which tend to come from sweltering cities. But understanding the factors that worsen the harm of rising temperatures could help save lives.

What makes the countryside so vulnerable to extreme heat

The discussion around the geography of extreme heat tends to focus on the urban heat island effect. The concrete, asphalt, steel, and glass of dense urban areas act as a sponge for the sun’s rays. Air pollution from cars, trucks, furnaces, and factories helps trap warmer temperatures over cities, and that hotter air, in turn, accelerates the formation of pollutants like ozone. On a hot summer day, a city center can be 25 degrees Fahrenheit warmer than the surrounding regions. And with so many people squeezed into these metropolitan ovens, it adds up to a massive health burden from extreme heat.

But far outside of downtowns, where homes and buildings get farther and farther apart, rural regions face their own long-running challenges that exacerbate the dangers of extreme heat.

A major factor: the median age of the rural population is older than in cities. That matters, because on a physiological level, older adults struggle more to cope with heat than the young. People living in rural communities also have double the rates of chronic health conditions that enhance the damage from heat like high blood pressure and emphysema compared to people living in urban zip codes.

Rural infrastructure is another vulnerability. While there may be more forests and farms in the country that can cool the air, the buildings there are often older, with less adequate insulation and cooling systems for this new era of severe heat. Manufactured and mobile homes, more common in rural areas, are particularly sensitive to heat. In Arizona’s Maricopa County, home to Phoenix, mobile homes make up 5 percent of the housing stock but account for 30 percent of indoor heat deaths.

Even if rural residents have air conditioners and fans, they tend to have lower incomes and thus devote a higher share of their spending for electricity, up to 40 percent more than city dwellers, which makes it less affordable for them to stay cool. That’s if they can get electricity at all: Rural areas are more vulnerable to outages due to older infrastructure and the long distances that power lines have to be routed, creating greater chances of problems like tree branches falling on lines. According to the US Census Bureau, 35.4 percent of households in rural areas experienced an outage over the course of a year, compared to 22.8 percent of households in urban areas.

Sparsely populated communities also have fewer public spaces, such as shopping malls and libraries, where people can pass a hot summer day. Rural economies also depend more on outdoor labor, and there are still no federal workplace heat regulations. Farmworkers, construction crews, and delivery drivers are especially vulnerable to hot weather, and an average of 40 workers die each year from extreme heat.

The health infrastructure is lacking as well. “There is a longstanding healthcare crisis in rural areas,” said Grace Wickerson, senior manager for climate and health at the Federation of American Scientists. There aren’t always nearby clinics and hospitals that can quickly treat heat emergencies. “To really take care of someone when they’re actually in full-on heat stroke, they need to be cooled down in a matter of minutes,” Wickerson said.

The Phoenix Fire Department has now started using ice immersion for heat stroke victims when transporting patients to hospitals to buy precious time. But rural emergency responders are less likely to have tools like this in their ambulances. “In Montana, which has not traditionally seen a lot of extreme heat, you would not have those tools on your truck and not have that awareness to do that cooling. When you see someone who has to also then travel miles to get care, that’s going to worsen their health related outcomes,” Wickerson said.

Emergency response times are generally much longer in rural areas, sometimes extending more than 25 minutes. People also have lower incomes and lower rates of insurance far from cities. Hospitals in rural areas are closing down as well. So when severe heat sets in, rural healthcare systems can get overwhelmed easily.

Looking at data from the Centers for Disease Control and Prevention and the US Census Bureau, Wickerson and her collaborators mapped out how all these underlying factors are converging with extreme heat. They found that 59 percent of urban zip codes and 54 percent of rural zip codes are highly vulnerable to extreme heat as defined by the CDC’s Heat and Health Index, meaning they are much more likely to see health problems from extreme heat. So while rural areas may be cooler, the people living there face heat dangers comparable to those in much hotter cities, and geographically, they cover a much wider expanse of the country.

So while temperatures out in the sticks may not climb to the same peaks they do in downtowns, urban heat islands are surrounded by an ocean of rural heat vulnerabilities.

There’s no easy path to cooling off

There are ways to reduce the dangers of scorching weather across vast swaths of the country, but they aren’t fast or cheap. They require big upgrades to infrastructure — more robust energy delivery, more shade and green spaces, better insulation, cool roofs, and more energy-efficient cooling.

Countering extreme heat also requires bigger structural investments to reverse the ongoing rural healthcare crisis where a doctor shortage, hospital closures, and longer emergency response times are converging. But the Republican budget proposal will do the opposite, cutting healthcare access for millions of Americans that would, in turn, lead to dozens of hospitals closing down, mainly in rural areas.

Protecting people from dangerous heat also demands policy changes. Most states don’t have any worker protections on the books for extreme heat. The Occupational Safety and Health Administration is in the process of creating the first federal heat safety standard for employers, requiring them to give employees breaks, water, and shade when it gets hot. But it’s not clear how strong the final regulation will be given that the Trump administration has been working to weaken rules across the board.

Cities and local governments could also impose rules that prevent utilities from shutting off power to customers during heat waves, similar to regulations that limit heat shutoffs during the winter.

But there are limits to how much people can adapt to hotter temperatures. Even places with a long history of managing heat are seeing more deaths and hospitalizations as relentless temperatures continue to mount. That means curbing the ongoing warming trend has to be part of the solution as well, reducing greenhouse gas emissions to slow climate change.

---

From Vox via this RSS feed

Federal agents guard a federal building and Immigration and Customs Enforcement (ICE) detention center in downtown Los Angeles. | Spencer Platt/Getty Images

Last week, federal agents arrested Brad Lander, a Democrat running for mayor of New York City and the city’s incumbent comptroller, after Lander linked arms with an immigrant the agents sought to detain and asked to see a warrant. Last month, federal officials also arrested Newark’s Democratic Mayor Ras Baraka while Baraka was protesting at a detention facility for immigrants.

A federal law permits sitting members of Congress to enter federal immigration facilities as part of their oversight responsibilities. That didn’t stop the Trump administration from indicting Rep. LaMonica McIver (D-NJ), who was at the same protest as Baraka. Federal officers also detained and handcuffed Sen. Alex Padilla (D-CA) after he tried to ask Secretary of Homeland Security Kristi Noem questions at a press conference.

These arrests are part of a broader campaign by the Trump administration to step up deportations, and to intimidate protesters who object. Most of these incidents are recent enough that the courts have not had time to sort through what happened and determine whether anyone’s constitutional rights were violated. But one thing is all but certain: even if it turns out that federal law enforcement officers flagrantly and deliberately targeted protesters or elected officials, violating the Constitution’s First or Fourth Amendment, nothing will happen to those officers.

The reason why is a pair of fairly recent Supreme Court decisions, which make it nearly impossible to sue a federal officer if they violate your constitutional rights — even if the allegations against that officer are truly shocking. In Hernández v. Mesa (2020), the Court’s Republican majority gave lawsuit immunity to a US Border Patrol officer who fatally shot a Mexican teenager in the face. And in Egbert v. Boule (2022), the majority reaffirmed this immunity — albeit in a case involving a less sympathetic plaintiff.

Both of these cases are part of the Republican justices’ crusade against an older Supreme Court decision known as Bivens v. Six Unknown Named Agents (1971). Bivens held that federal law enforcement officers who violate the Fourth Amendment — which protects against “unreasonable searches and seizures,” among other things — may be sued for that violation.

Significantly, Bivens ruled that a victorious plaintiff in such a case “is entitled to recover money damages for any injuries he has suffered as a result of the agents’ violation of the Amendment.” So officers faced very real consequences if they violated the Fourth Amendment.

The Court’s current majority, however, appears determined to destroy Bivens. Hernández and Egbert didn’t explicitly overrule Bivens, but they ground down that decision to the point that it has little, if any, remaining force. And the Court appears to be laying the groundwork for a decision eliminating Bivens suits altogether. Significantly, Justice Samuel Alito’s majority opinion in Hernández warned that “it is doubtful that we would have reached the same result” if Bivens were decided today.

That means that individuals who are unconstitutionally arrested by federal officers, or who face similar violations of their rights, will generally have no recourse against those officers. And that’s likely to embolden the worst officers to violate the Constitution.

Bivens, explained

The Constitution places several restrictions on law enforcement, including the Fourth Amendment’s ban on unreasonable arrests and excessive force. But it is silent on what can be done when an officer violates these restrictions.

Bivens, however, held that a right to sue federal officers is implicit in the Constitution itself. An officer who acts unlawfully “in the name of the United States possesses a far greater capacity for harm than an individual trespasser exercising no authority other than his own.” And so it follows, Bivens explained, that there must be a meaningful remedy to ensure that officers do not abuse this power.

In fairness, the Supreme Court started limiting Bivens suits not long after that case was handed down. Shortly after Bivens was decided, President Richard Nixon replaced two justices, creating a new majority on the Court that was more favorable to law enforcement. But the Court only recently signaled that it intends to destroy Bivens altogether. In Egbert, the Court’s Republican majority declared that courts must reject Bivens suits if there is “any rational reason (even one)” to do so. Even a minor factual discrepancy between a new case and Bivens, such as the fact that the officers who violated the Constitution belong to a different agency than the officers in Bivens, is frequently enough to defeat a Bivens suit.

President Donald Trump took office on twin promises to crack down on both undocumented immigrants and his perceived enemies — “I am your retribution,” he told supporters in 2023 — and it’s not hard to see how decisions like Egbert and Hernández enable him to do so.

The Republican justices argue that nullifying Bivens is necessary to restore a more traditional vision of “the Constitution’s separation of legislative and judicial power.” The Supreme Court, under this vision of the separation of powers, may not determine that a right to sue federal officers is implicit in the Constitution. This right, according to Alito, must come from an explicit act of Congress.

Alito’s historical claim, that Bivens departed from a traditional understanding of the role of Congress and the courts, is somewhat dubious; the courts permitted at least some suits against federal officials who break the law for most of American history. In Little v. Barreme (1804), for example, the Supreme Court held that a Navy officer who unlawfully seized a neutral ship “must pay such damages as are legally awarded against him.” More recently, in Larson v. Domestic & Foreign Commerce Corp. (1949), the Court declared that “the principle that an agent is liable for his own torts ‘is an ancient one, and applies even to certain acts of public officers or public instrumentalities.’”

Hernández’s call for granting immunity to federal officials would also have more credibility if the Republican justices hadn’t recently ruled that Trump has broad immunity from prosecution if he uses the powers of the presidency to commit crimes. This concept of presidential immunity appears nowhere in the Constitution, and it certainly has no place in American legal tradition — among other things, why would President Gerald Ford have pardoned former President Richard Nixon for crimes Nixon committed in office, if Nixon were immune from prosecution?

The Court, in other words, appears determined to remove legal obstacles that might have deterred federal officials from behaving illegally in the past — regardless of what the law or legal traditions might dictate. And it removed important obstacles right before the United States took a dangerously authoritarian turn.

---

From Vox via this RSS feed

A boathouse at an old Boy Scouts of America camp in Milton, New York, in 2017. | John Carl D’Annibale/Albany Times Union via Getty Images

Summer camp. It’s where kids go every year to make friends, find their long-lost twin, or even evade a slasher wreaking havoc on the campers and counselors. At least, that’s what pop culture would lead you to believe: For the outsized space they take up in our consciousness, going to camp for the summer isn’t actually all that common.

“It has never been the case that the majority of American children went to summer camps,” says Leslie Paris, an associate professor at the University of British Columbia and author of the book Children’s Nature: The Rise of the American Summer Camp.

“The first camps were founded by urban middle-class men,” she told Vox. “They were concerned about white boys who they saw as not getting enough outdoor adventure and the kind of manly experiences they would need to be — in the minds of these adults — the nation’s leaders for the next generation. They were worried about the effects of urbanization, and they were nostalgic for an earlier day when more boys had grown up in rural places.”

How did camp begin to be available for more kids? And if so few people actually attend, then why does summer camp have such lasting cultural influence? Those are just a few of the questions we posed to Paris on the latest episode of Explain It to Me, Vox’s weekly call-in podcast. Below is an excerpt of the conversation with Paris, edited for length and clarity.

You can listen to Explain It to Me on Apple Podcasts, Spotify, or wherever you get podcasts. If you’d like to submit a question, send an email to [email protected] or call 1-800-618-8545.

How did camp expand beyond the audience it was originally created for?

The YMCA movement became involved, and by the turn of the century the movement started really ramping up. Not only because more YMCA camps were founded, but because different organizations got involved and more groups of American adults thought this camp idea would be great.

By the turn of the century, you’ve got small numbers of women leading groups of girls out into the wilderness. Many of the women who started camps were college-educated and saw leading girls and giving them adventures as a kind of passion.

Then there were urban organizations that began to say, “This would be great for impoverished working-class kids who never get out of the city at all,” and began sending groups of kids out into the country, often for shorter stays than at private camps.

In the early 20th century, you’ve got a bunch of new movements: the Boy Scouts, the Girl Scouts, the Campfire Girls. And then there are different ethnic and religious groups: Jewish Americans, Catholic Americans, who think, Let’s start camps for our own kids, and they do that as well.

By the early 20th century there’s a bevy of different kinds of camps organized for a wider variety of kids to give them an experience of the outdoors.

You write in your book that “this triple nostalgia — for the American past, for camp community, and for individual childhood experience — is critical to understanding why camps have figured so influentially in American culture and in former campers’ lives.” I’d love for you to talk about that a little bit more.

One of the things I talk about in my book is that camps were a place where children learned nostalgia, that camps taught them a version of the American past. I think many of us are familiar with a use of Indigenous cultural practices that was often quite superficial, but that was meant to introduce non-Indigenous children to one aspect of the American past. Camps were often a place where children were exposed to ideas about what the American past had been, and then as more generations of children attended camps, they themselves brought those kinds of nostalgic memories with them, throughout their lives.

When they had a chance, many of those former children sent their own kids to camp. So this became a kind of a nostalgic cultural practice that for many adults reminded them of the first time that they had an adventure away from their parents, away from their families.

It’s so interesting you talk about Indigenous culture and how that’s been used at camp. It makes me think of that scene in [Addams Family Values] where Wednesday’s at camp. Why does camp feature so prominently in pop culture if so few of us went?

You could ask, Why are so many children’s novels premised around an orphan? I think the fact that the kid is an orphan in these novels allows them to go off and have adventures and do things that many kids raised in families would not necessarily be at liberty to do.

And I think camps have often represented that space, a space that’s at least ostensibly protected, where kids have more free play and can have exciting adventures and develop peer relationships that are outside of the norm. And that piece lends itself really well to popular culture.

Camp is so specific. How did you choose this as an academic subject?

I knew that I wanted to work on American childhood, which was still a pretty small field in the 1990s, when I started this project. There wasn’t a major scholarly book about the history of summer camps at the time and it seemed like a wonderful way to write about something that would be fun to work on. One of the things that I look at in my book is how camps illuminate the ways in which childhood was being transformed in the late 19th and early 20th century.

That’s so interesting. I imagine that changes at summer camp also reflect changes in American childhood overall. I’d love to hear in broad strokes about some of those changes. How have we seen camp and therefore childhood change over time?

One of the main changes that I look at is the rise of the idea of protected childhood. That childhood should be a time apart and children should be protected from the adult world. The late 19th, early 20th century is the same time when you see laws restricting children’s labor. There’s an emphasis on child protection that’s emerging during this period, and camps are one of the early sites of this new idea that children are deserving of spaces apart, time apart, and also that they’re deserving of vacations.

Although many of the elite kids who attended more expensive private camps were certainly going to have vacations whether or not they went to summer camp, some of the working-class kids at the turn of the 20th century who attended summer camps had never been on a vacation outside of the city.

Summer camp has become this huge business these days in the United States, $3.5 billion annually. How did that happen?

The camp industry has had to be nimble and change over time, especially since the 1970s, which was a time when many camps struggled and a number failed.

The camping industry underwent some structural changes. One of these was the rise of specialty camps: Basketball camp, computer camp, gymnastics camp, dance camp, theater camp — camps that were focused on a really specific interest emerged in the late 20th century.

Another issue was that many families who could afford private camps were starting to juggle more different opportunities. The cost of travel by plane was going down, so more families were thinking, Maybe at some point this summer we’d like to take the kids on a trip. There was also a rise in [divorce] and families had to negotiate custody. So even camps that used to have a nine-week schedule increasingly considered moving to a two-session schedule.

Modern summer camps have retained many of the same elements as some of the earliest camps, but they’ve also adjusted to the increasing complexity of some of their clients’ lives, and in that way the camp industry has continued to be able to thrive.

And there’s another issue, which is that camps have also always provided child care, and this has been important for parents since the very beginning. It’s been a boon for parents who could relax knowing that their kids were away, especially families trying to juggle complicated child care arrangements in the summer when there was no school.

---

From Vox via this RSS feed

President Donald Trump delivers an address to the nation accompanied by Vice President JD Vance, Secretary of State Marco Rubio, and Defense Secretary Pete Hegseth from the White House on June 21, 2025. | Carlos Barria/Getty Images

President Donald Trump is now the most unpopular he has been during his second term.

More than half of American adults disapprove of the job he is doing, and he’s underwater on nearly every important issue of the day.

The polling averages show this net disapproval clearly: On the economy, he’s down 13 percentage points. On inflation, he’s down 20 points. Even on immigration, he’s down 2 points. (Those negative marks include foreign policy, though it’s too soon to say how the public is reacting to Trump’s decision to join Israel’s bombing of Iran.)

Still, Trump’s popularity decline has been a dramatic development: After entering office with a positive approval rating and popular support for his agenda, he’s squandered much of it away through various political fights, policy decisions, and public spectacles.

That reversal has come in fits and starts, yet also demonstrated a curious trend in Trump’s popularity. When Trump is at the center of the news, using his bully pulpit and making high-profile efforts to pursue his agenda, his popularity falls. When he recedes into the background, and the public is focused elsewhere, his popularity somewhat recovers.

In short, the more people pay attention to Trump, the less they like him — which creates a kind of conundrum. Trump, who’s uniquely capable of capturing the limelight, has shown he’s also incapable (or unwilling) to do anything quietly.

Immigration policy is the latest example of this trend. Public opinion turned sharply against Trump’s response to protests in California over raids by Immigration and Customs Enforcement. Views of his deployment of the National Guard and a few hundred US Marines to the West Coast have similarly been resoundingly negative. And overall views of Trump’s immigration policy, as a result, have fallen to their lowest point this year, per tracking polls by YouGov.

Beyond any one policy, these trends in public opinion suggest that Trump is turbocharging two features of modern presidential politics. The first is the idea of negative polarization: that members of one political party are bound together and mobilize against an opposing political party or movement more strongly than for their own side. And the second is of the electorate as operating a thermostat: preferring the opposite opinion or direction to whatever the president, or party in power, says.

These shifts matter. They tell us a lot about Trump-era politics and about how modern American politics might exist after him. And they offer a clue as to how the public might react to American military involvement in the Middle East. Presidents tend to see a public boost — a rally-around-the-flag effect — when America gets involved in military conflicts. Will Trump’s popularity rebound? Or will this be yet another example of Trump being different?

Does the American public actually prefer inaction?

A quick scan of the last six months of Trump’s job approval shows something remarkably consistent: Almost every time Trump makes a big policy move or announcement, the public recoils in disapproval.

This dynamic is most pronounced in three sharp spikes in disapproval: after Elon Musk and the so-called Department of Government Efficiency rampaged across the federal government, after the rollout of tariffs, and as Trump picked a fight in California over ICE deportations.

DOGE kicked off the downward slump in his approval rating and popular support for his policies, in mid-March. As Trump reined in Musk, and DOGE faced more obstacles and public scrutiny, it began to fade from headlines, and Trump’s approval rating began to recover.

Then the cycle repeated. As Trump mulled a trade war and signaled he would implement a new tariff regime in late March and early April, views on his handling of inflation, trade, and the economy began to nosedive. Disapproval of his trade and tariff approach, in particular, spiked during the late-March to mid-April period when “Liberation Day” tariffs were being announced and the stock market was roiled with instability and distrust.

Since Trump ratcheted down his policy and softened his public entrenchment on tariffs, his ratings have stabilized, and even recovered a bit, according to an average of polls maintained by the pollster Adam Carlson.

The same cycle would play out again when Trump turned his attention to immigration policy, historically his best issue. Views of his approach to immigration and mass deportations have been much more volatile since late March and mid-April, when high-profile arrests and detentions of foreign-born students and immigrants began to dominate the news. It was around this time that the legal showdowns over Kilmar Ábrego García’s deportation to a Salvadoran prison dominated headlines, and public opinion over Trump’s immigration approach began to turn negative.

For most of the period from his inauguration to April, Trump had enjoyed positive ratings on immigration. His mass deportation pledges were still popular with most Americans. But hunkering down on those mid-spring decisions resulted in the public turning against Trump’s position. His immigration ratings saw a dive, which only began to recover once Trump backed down and turned his priorities on Congress and his legislative agenda.

Like a switch, public opinion on immigration policy began to get more positive during a time when Trump wasn’t really creating that much news around the topic, until ICE raids began to pick up steam and small protests began to erupt in Democratic cities. His aggressive response — federalizing the National Guard and sending Marines to Southern California — has again triggered a collapse in public views. According to YouGov, approval of Trump’s handling of immigration is now the lowest it’s been this year, and it is lower than it was at this point during his first term.

It’s all a bit confusing: Trump is doing what he repeatedly promised to do during his campaign — the same campaign that won him enough popular support for a second term in the White House. The public, however, is punishing him for it.

Trump isn’t the first president to deal with a reflexive public

This weird seesaw of the public reacting negatively to big policy decisions Trump makes is a defining characteristic of modern politics. He isn’t the first president to deal with a reflexively reactionary public, as both Joe Biden and Barack Obama experienced this during their presidencies. Both Democratic presidents saw drops in public support once they began to implement their agendas, be it the Affordable Care Act beginning in 2009, or Biden’s legislative packages and Afghanistan withdrawal in the summer and fall of 2021.

And so Trump’s two terms seem to fit this pattern, what some political scientists call the “thermostatic” model of public opinion. This basic idea — that public opinion shifts in the opposite direction of whatever direction the government takes — is a powerful predictor, but it doesn’t fully explain Trump’s second term.

Just like you’d turn a thermostat’s temperature up or down depending on the feeling in a room, the public reacts positively or negatively depending on the actions taken by a president and party in power.

Plenty of political research demonstrates this thermostat in action: the public supporting spending cuts when the government begins to spend more, the public growing more liberal on social issues when a conservative president seizes on it (as on immigration and racial justice during the Trump years), support for the Affordable Care Act rising when Republicans try to cut it. These anti-incumbent shifts then suggest future political advantage for the political party or movement out of power.

But just because the public is cooling on Trump, that doesn’t mean it, so far, is warming to Democrats — or even to liberal policy positions.

Public views of the Democratic Party are still resoundingly negative: a little more than a third of Americans have favorable views of the party or of Democrats in Congress, both lower than views of Republicans. Democrats only enjoy a slight advantage in generic polling ahead of midterms next year. And views toward immigration in general, for example, are still much more negative than they’ve been at any point since 9/11, per Gallup tracking data. The pro-immigrant shift the country saw during Trump’s first term hasn’t materialized this time.

Will this trend hold?

With the US now embroiled in a highly fluid, quickly evolving conflict with Iran, it’s not at all clear if this thermostatic trend is holding: There hasn’t been enough time to definitively poll those views, and it’s likely that many people are still forming their opinions.

According to a weekend snap poll conducted by YouGov, more Americans disapprove of US bombings of nuclear sites than approve, but about 1 in 5 are withholding forming an opinion as of now. That poll was conducted after Trump announced the military strikes but before Iran carried out its own wave of missile launches against US bases in Qatar and Iraq. And whatever kind of escalation or deescalation comes next will influence how public opinion shifts.

Still, there are reasons to think the anti-Trump trend will hold. Military action against Iran was already unpopular before the US got involved. Plenty of Americans were wary of Israel’s preemptive strikes earlier this month as well. And many Republicans opposed US involvement in military conflicts in the Middle East before this weekend. Of course, many of those Republicans will likely sort themselves to align with the position of their party leader. But how long that holds is another question.

Any kind of rally-around-the-flag effect that might result from this conflict also doesn’t seem likely to be durable, as of now. Presidents have tended to enjoy a small bump in approval when the US conducts highly visible military operations: It happened for both George H.W. Bush and George W. Bush during the Gulf War and 9/11, and for Obama after the killing of Osama bin Laden.

But as the Democratic strategist Dan Pfeiffer has pointed out, these poll boosts have been getting smaller over time as the effect of polarization has grown. And Trump is already massively unpopular, meaning any boost he gets isn’t likely to get him a majority of the country’s support.

This news story seems likely to get bigger, particularly if Trump chooses to respond to Iran’s retaliatory strike. And that will feed the key determinant of this public opinion trend: Whatever position Trump takes becomes the defining, totalizing news story of the day. For now, all other news stories seem to have taken a backseat to this military conflict. And when Trump dominates the news, this trend becomes clearer.

Of course, other things could happen. It will take time for the public to process these developments. And more domestic news will be coming in the months ahead, as massive tax, health insurance, and public assistance cuts included in Trump’s reconciliation bill (the “big” and “beautiful” one) have yet to become law and receive the same public scrutiny other Trump policy moves have. So there’s a lot more to come. But for now, it seems like the public thermostat is working overtime.

---

From Vox via this RSS feed

President Donald Trump with Chief Justice John Roberts, at the US Capitol on March 4, 2025. | Win McNamee/Getty Images

In a short, one-paragraph order, the Republican justices ruled on Monday evening that President Donald Trump may effectively nullify a federal law and an international treaty that is supposed to protect immigrants from torture. The Court’s order in Department of Homeland Security v. D.V.D. does not explain the GOP’s justices’ reasoning, although Justice Sonia Sotomayor responds to their silent decision in a 19-page dissent joined by her two Democratic colleagues.

The Court’s order is only temporary, and will permit Trump to send immigrants to countries where they may be tortured while the D.V.D. case is fully litigated. It is possible that one or more of the Court’s Republicans could reverse course at a later date. But it is hard to know what arguments might persuade them to do so because the justices in the majority did not explain why they decided this case the way they did.

Federal law requires that the United States shall not “expel, extradite, or otherwise effect the involuntary return of any person to a country in which there are substantial grounds for believing the person would be in danger of being subjected to torture.” This statute implements a treaty, known as the Convention Against Torture, which the United States ratified over three decades ago.

Trump’s lawyers, however, claim that they uncovered a loophole that permits the Trump administration to bypass these laws, at least with respect to some immigrants.

Typically, before a noncitizen may be removed from the United States, they are entitled to a hearing before an immigration judge. The immigration judge will inform the person facing deportation which countries they might be sent to, allowing the noncitizen to object to any countries where they fear they may be tortured. If the immigration judge determines that these objections are sufficiently serious to trigger the Convention Against Torture’s protections, the judge may still issue an order permitting the immigrant to be deported — but not to the nation or nations the immigrant raised objections about.

The D.V.D. case involves noncitizens who have already been through this process. In their case, an immigration judge determined that they may be deported, but not to specific countries. After the hearing process was complete, however, the Trump administration unexpectedly announced that it would deport the D.V.D. plaintiffs to other nations that were not previously under consideration.

That means that no immigration judge has determined whether these immigrants may be sent to those particular nations, and the immigrants have not been given a meaningful opportunity to object to the new countries where they are about to be deported. Using this loophole, the Trump administration seeks to deport them without a new hearing.

The Trump administration, moreover, appears to have intentionally selected countries where the noncitizens are likely to be unsafe. It wishes to deport many of these immigrants to South Sudan, for example, a country that was recently in a civil war, and where an uneasy peace appears to be collapsing. Others are slated for removal to Libya despite the fact that, according to Sotomayor’s dissent, they “would have landed in Tripoli in the midst of violence caused by opposition to their arrival.”

The Trump administration, in other words, appears to have created a deadly trap for immigrants who fear torture in their home nations. These noncitizens may object to being sent home under the Convention Against Torture, and an immigration judge may even rule in their favor. But the Trump administration may still send them somewhere else even more dangerous.

If you are interested in the specific legal arguments Trump’s lawyers raised to justify this trap, I summarized them here. But, again, it is not possible to determine which of these arguments persuaded a majority of the justices because those justices did not even bother to explain their decision.

---

From Vox via this RSS feed

An Iranian cleric stands next to a scale model of an Iran-made surface-to-surface missile, which is displayed during an anti-Israeli protest at Palestine Square in downtown Tehran, Iran, on April 9, 2025. | Morteza Nikoubazl/NurPhoto via Getty Images

President Donald Trump has framed his strikes on Iran as a costless triumph. The president is not asking Americans to accept sacrifices in service of destroying the Iranian nuclear program — only to applaud his already successful destruction of it.

“Iran’s key nuclear enrichment facilities have been completely and totally obliterated,” Trump declared Saturday night. “Iran, the bully of the Middle East, must now make peace. If they do not, future attacks will be far greater and a lot easier.”

It’s possible that Trump’s war will unwind as advertised. The Iranian regime finds itself in a state of profound weakness. Israel has gutted its air defenses and flies largely unharried through its skies. Iran’s nuclear facilities are badly damaged, while many of its top atomic scientists and military leaders lie dead. Its most fearsome partner militias — Hezbollah and Hamas — are badly degraded. Its longtime allies in Damascus have been deposed. Its friends in Moscow seem to have too much on their own plates to lend much of a hand.

From this precarious position, the Iranian leadership might conclude that it can ill afford an escalating conflict with the world’s greatest military power. Perhaps, it will follow the same basic playbook it did following Trump’s assasination of its military leader Qasem Soleimani in 2020: Back then, Iran responded with face-saving strikes on US military bases in Iraq — but ones that it heavily telegraphed, enabling American soldiers to take cover and retain their lives, which in turn allowed Trump to call things even and end the conflict. Iran’s strike on a US base in Qatar on Monday could be interpreted as fitting this mold: The US said that it had advanced warning of the strike and that there were no American casualties.

What’s more, it is even possible that the Iranian regime could conclude that its nuclear program is more trouble than it’s worth and sheepishly return to the bargaining table, now ready to accept more stringent restrictions on its freedom to enrich uranium or build ballistic missiles.

But such rosy outcomes are far from certain. And if things do not transpire as Trump hopes, his war could impose significant costs on the American people — in terms of money, blood, and nuclear security.

How Trump’s war on Iran could impact the economy

For Americans, the most widely felt consequences of Trump’s war with Iran would likely be economic. Oil prices have climbed by more than 12 percent since the end of May, when Israel began threatening to strike Iran’s nuclear facilities. An escalating US-Iran conflict could further elevate Americans’ energy costs.

The nightmare scenario here concerns the Strait of Hormuz, the world’s only waterway connecting the Persian Gulf with the open ocean. Each day, about 20 million barrels of oil — or about 20 percent of the world’s total supply — moves through the strait. Iran could plausibly choke off all shipping through the waterway if it wished to do so. And on Sunday, the Iranian Parliament reportedly approved such a course of action, although it remains up to the nation’s Supreme National Security Council to enact a blockade.

Were Iran to take that extraordinary measure, the price of oil could shoot up past $130 a barrel, according to industry analysts (as of this writing, a barrel is trading at $72). That would dramatically increase the costs of energy and transportation for US consumers. And since energy is an input into the production of more or less every good and service, a sustained blockade could push up prices more broadly. Faced with higher inflation, the Federal Reserve would likely scrap plans for cutting interest rates. In this scenario, Americans would see lower real wages and higher borrowing costs than they would have enjoyed in a world where Trump did not bomb Iran.

It’s worth saying that a blockade of the Strait of Hormuz is unlikely. Were Iran to pursue such a policy, it would effectively be sabotaging its own economy, which is heavily dependent on exporting oil via the waterway. Nevertheless, the threat of disrupting shipping through the strait is the regime’s greatest point of leverage over other world powers.

Iran could also disrupt global commerce in more modest ways. Tehran is allied with Yemen’s Houthi militia, which is already threatening to recommence its attacks on US ships in the Red Sea. Any disruption to Red Sea trade could increase global shipping costs, which would eventually bleed into US consumer prices.

Alternatively, Iran could strike oil and gas infrastructure in Middle Eastern countries allied with the United States. In a 2019 drone attack allegedly backed by Tehran, the Houthis bombed two major oil facilities in Saudi Arabia, temporarily cutting the country’s oil production in half and spiking crude prices.

At a moment when the president’s tariffs are already nudging up the cost of imports, any war-related disruptions to global trade could prove painful for American households.

Trump’s attack has put American soldiers in harm’s way

As Monday’s attack on US troops in Qatar demonstrated, Americans stationed in the Middle East face a far graver threat than more expensive gasoline. More than 40,000 US soldiers are serving on bases and warships in the region, well within reach of Iranian missiles.

Trump’s bombing of Iran’s nuclear facilities represented a far broader and more destructive attack than his strike against Soleimani five years ago. It’s reasonable to fear that Tehran’s response will be proportionally more severe, and that Monday’s strikes at Qatar represent only the beginning of its retaliation. At least, this is what the regime is telling Americans to expect.

“Any country in the region or elsewhere that is used by American forces to strike Iran will be considered a legitimate target for our armed forces,” Ali Akbar Velayati, an adviser to Iran’s Supreme Leader Ayatollah Ali Khamenei, said Saturday. “America has attacked the heart of the Islamic world and must await irreparable consequences.”

Some of these “consequences” could transpire on American soil. Although Iran’s immediate targets will likely lie close to home. Years after the Soleimani strike, Iran allegedly orchestrated failed assassination attempts against former national security adviser John Bolton, former Secretary of State Mike Pompeo, and Donald Trump.

According to US officials who spoke with NBC News, Iran warned Trump at this year’s G7 Summit that it could respond to an American strike on its nuclear facilities by perpetrating terrorist attacks within the United States. The regime claimed to have sleeper cells in America ready and waiting to commit such violence.

Trump may have made an Iranian nuclear weapon more likely

Trump’s war with Iran may risk making all Americans poorer while getting some of us killed. But in his administration’s ostensible view, these potential harms pale in comparison to the threat posed by Iran acquiring a nuclear weapon — a threat that Saturday’s strikes greatly mitigated.

It is not obvious how much Americans actually have at stake in preventing Iran from developing atomic weapons. No serious analyst believes that the Iranian regime is suicidal. Tehran is not going to order a nuclear first strike against the United States or any other country. Rather, it almost certainly sees nuclear weapons primarily as a deterrent against both foreign intervention and domestic challenges to its authoritarian regime.

Nevertheless, all else equal, Americans have an interest in preventing nuclear proliferation. Were Iran to acquire an atomic weapon, Saudi Arabia would be liable to pursue its own. And a nuclear arms race in the Middle East would increase the tail risk of a future atomic catastrophe.

And yet, it is possible that Trump’s strikes on Iran have actually made that nation’s acquisition of a nuclear weapon more likely.

Before Saturday’s bombings, Iran was engaged in negotiations over its nuclear program without the United States. And Tehran had previously reached an agreement to limit its enrichment of uranium in 2015, a nuclear deal that Trump tore up during his first term in office.

Now, all diplomacy over Iran’s nuclear program appears dead. And Tehran may see a nuclear weapon as more indispensable for its security than ever before. After all, the regime’s conventional military defenses have proven grossly inadequate to deter or defeat Israeli and American incursions.

Already, Iranian officials are signalling that the nation will withdraw from the Nuclear Non-Proliferation Treaty (NPT), a move that would mean an end to the nation’s cooperation with United Nations inspections of its uranium enrichment. At present, the UN’s nuclear watchdog says it is uncertain about where Iran is storing the highly enriched uranium it has already produced. Tehran claims that, before Trump’s strikes, it had moved its uranium stockpiles out of the three sites that he targeted.

American and Israeli bombing has surely undermined Iran’s capacity to produce weapons-grade uranium. Tehran has lost some of its top nuclear scientists and suffered massive damage to its enrichment facilities. But Iran retains both the technical know-how and raw materials necessary for building an atomic bomb. And the world may have now lost visibility into its nuclear activities.

Ultimately, it is impossible for anyone to know with certainty what America’s war with Iran will bring. But what we do know is that, in the immediate term, Trump’s strikes have made Americans less safe, while heightening the risks to our economy — all in service of a geopolitical goal that likely could have been achieved through peaceful means, and which was not necessarily advanced by Saturday’s violence.

---

From Vox via this RSS feed

People march in Times Square during a rally calling for the Trump administration not to go to war with Iran on June 18, 2025 in New York City. | Adam Gray/Getty Images

Over the weekend, the United States bombed three nuclear facilities in Iran. Iran has been considered a political risk to America since the 1979 revolution, and President Donald Trump has repeatedly stated that it cannot be allowed to possess nuclear weapons. The strikes mark yet another attempt in a long-running US strategy to rein in Iran’s nuclear ambitions.

But whether Trump’s strikes will achieve his stated aim of destroying Iran’s nuclear program is unclear. It doesn’t help that his plan around the attacks has felt haphazard. Trump said Iran had a two-week deadline before he would authorize a strike — then attacked only two days later. Even just before the bombs were dropped Trump was telling the press that targeting nuclear facilities may not even be an option, saying “I may or I may not do it.” Trump stated in his 2024 election victory speech that he was “not going to start a war,” yet he has now hinted on social media that regime change could be next.

Trump could be accused of simply being chaotic here. But this is a deliberate strategy. Trump has a history of being intentionally unpredictable when it comes to foreign policy, known as the unpredictability doctrine. Drawing from his experiences in his previous career in business, Trump says being predictable is bad. When the other side doesn’t know what you are going to do, you are in control. His plan is also about creating uncertainty. You make your opponent unsure of what they are facing and unable to make decisions in response, leaving you to take the advantage.

But foreign policy is not business, and a strategy that works with corporations may backfire on the world stage. While nobody knows exactly what will happen next, what can we work out about the implications of Trump’s actions now given what has worked (or hasn’t) before in terms of nuclear arms control?

Strikes now, problems later

Countries may consider military strikes on nuclear facilities when they feel that the other side  won’t cooperate in negotiations. For example, Israel, believing that Iraq would never be serious about a diplomatic solution, bombed an Iraqi enrichment facility at Osirak in 1981 to stymie the nuclear program. The preventive attack did serious damage to the facility, and Israel claimed they had disrupted Iraq’s ability to produce a nuclear weapon by destroying the facility before it became functional. A similar aim was likely a factor in Trump’s thinking on targeting Iran.

Yet military strikes are rarely as clear-cut as they look on paper. They may seem like a simple solution to a dangerous problem by stopping a nuclear program in its tracks. They also deliver results faster than diplomatic options, which can take a lot of time and do not come with any guarantee of a solution.

But while Trump may like to think that a few strikes will do the job, using bombs now could create problems for any future US strategy toward Iran — whatever that strategy turns out to be.

The first problem is that we do not yet know whether the strikes were entirely successful in taking out the targets. To work properly, a military attack should completely destroy the target facilities to ensure they cannot work. If a facility is even partially functioning after a strike, that state can still run a nuclear program, albeit a reduced one. While Iraq did not go on to develop nuclear weapons after Osirak, it still had sufficient resources to maintain a plan. In fact, some experts argue that the attack only encouraged Iraq to pursue this.

While there is evidence of physical damage after Trump’s strikes, whether this is enough damage to undermine Iran’s nuclear capability remains uncertain. Trump is saying that the three targeted facilities in Iran were “obliterated.” Defense Secretary Pete Hegseth has also reported that the attacks have “devastated the Iranian nuclear program.” Iran, however, has downplayed the extent of the destruction. The fact that Trump’s former aide, Steve Bannon, can’t even pronounce the name of one of the facilities, Fordow, properly doesn’t exactly build confidence in the administration’s assessment.

It’s also unclear how many nuclear facilities are left outside of the three that were bombed. The International Atomic Energy Agency recently cautioned that there could be clandestine facilities that we just do not know about. Trump himself has claimed there are “many targets left,” which indicates that Iran still has at least part of an operational program.

And now, any future efforts by Iran would also likely be even more secretive and underground, making it more difficult to detect and target in potential future strikes.

Even if the US has destabilized the nuclear plan, Iran can still rebuild. In 2010, America tried to disrupt the Iranian program in a cyberattack using a computer worm called Stuxnet against the Natanz nuclear enrichment facility. The virus caused the reactors there to slow down and crash. The attack did a lot of damage but — as the current situation shows — Iran was able to continue the program.

So it’s possible that the recent bombings could slow down Iran’s progress, but it won’t eliminate the overall and long-term risks. It also will not affect Iran’s ability to retaliate with conventional weapons.

A second problem is that the bombings could now attract such a retaliation. Iranian president, Masoud Pezeshkian, has already said that the US “must receive a response to their aggression.” Iran has also publicly stated that it is considering a “proportionate” responses. This could potentially look like a repeat of Iran’s response to the 2020 assassination of General Qasem Soleimani in which Trump ordered a drone strike against the Iranian leader. Trump said the killing was to prevent a terrorist attack against a US embassy. Responding to Soleimani’s death, Iran launched a retaliatory attack against two US air bases in Iraq.

The incentive for Iran to retaliate is even greater this time. The US has not just taken out a key leadership figure but directly threatened the state itself and its nuclear program at a time of intense conflict with Israel — and the US, where Trump is seen as having entered the war as a result of the weekend strikes. There is now an even higher chance that Iran will fight back this time because it is already fighting, and it could use that war as an opportunity to target the US.

Upping the game by using bombs will also encourage escalation by the US. This may be exactly what Trump wants. Yet it is also the case that the situation makes it difficult for him to do anything else. If Iran doesn’t show signs of giving in and continues its nuclear ambitions, Trump may be forced to take further action. If he doesn’t, it would look like Iran has won, even with three facilities destroyed.

Diplomatic options still on the table?

So what’s next?

While suggesting that regime change could be in the cards, Trump has also talked about a diplomatic solution, and this is what many other states would prefer. Using diplomatic negotiations instead of bombs has proven effective — at least partially — in the past, as seen by international agreements to control the global proliferation of nuclear weapons, including the Treaty on the Non-Proliferation of Nuclear Weapons. Negotiations have also convinced countries, such as Ukraine, to give up their nuclear arsenals. Trump has even relied on diplomacy with states that he has tense relationships with, like North Korea, although some analysts question how effective this has been, not least given that North Korea still has nuclear weapons.

If Trump is serious about diplomacy, then the bombing will make this difficult at best, impossible at worst. Trump was clearly hoping that the attacks might soften Iran up for negotiations if he decided to go down that road. He said the attacks should be taken as a sign for Iran to “make peace” or face “far greater” attacks in the future. This strategy isn’t working so far.

When Trump initially threatened strikes, Iran’s supreme leader, Ayatollah Ali Khamenei, warned that such a move would “undoubtedly be accompanied by irreparable damage” to US-Iran relations. Iran has now repeated this line, saying the impact of the strikes will have “everlasting consequences.” There’s no evidence that Iran would allow itself to be forced into negotiating through threats alone. The solution in Ukraine worked because it was based on a peaceful and collaborative process, not because it was done under duress.

It is also worth remembering — as Iran certainly will —  that Trump is the one responsible for the failure of a previous diplomatic solution. He was the one who walked away from a nuclear deal called the Joint Comprehensive Plan of Action. This agreement — between Iran and the US, China, France, Russia, United Kingdom, and Germany — aimed to restrict Iran’s nuclear ambitions through demands, such as getting rid of its medium-enriched uranium and not build heavy-water reactors. In return, these countries would reduce their economic sanctions on Iran.

Trump said this was a “one-sided deal” that “didn’t bring peace, and it never will” and pulled out in 2018. Whatever misgivings he had about the deal, it means the US will find it hard to introduce new diplomatic measures now. Trump’s perceived lack of commitment reduces the chance that Iran would be willing to talk cooperatively with the US and its allies. While Iran was clearly willing to work diplomatically before, why would it now sit down with someone it feels can’t be trusted to stick to an agreed solution? This is especially the case when that someone has just bombed them and is now talking about regime change.

Whatever Trump has planned next, his decision to carry out strikes has radically limited both his options and the chances of de-escalation. Based on what we’ve seen in previous attempts at arms control and conflict resolution, Trump may have unnecessarily inflamed the conflict, left the US open to Iranian retaliation, ruined the option of future diplomacy, and provoked Iran into developing a nuclear weapon. This is a strategy that has a high chance of exploding at some point — if not now in terms of an Iranian escalation, then in the future in the form of a revitalized nuclear program.

---

From Vox via this RSS feed

A puma, or mountain lion, seen by a motion-sensing camera near the border wall. | Courtesy of Ganesh Marín

The border wall between the US and Mexico is, of course, a barrier meant to prevent human migrants from crossing into America as they seek work, family, or refuge from violence.

It’s also a significant barrier to ranging wildlife.

The border wall, a centerpiece of President Donald Trump’s agenda, cuts through a rugged, unique ecosystem home to hundreds of native species, from jaguars and pumas to black bears and deer. These animals often need to move to survive, whether to find a source of water or a mate.

We know the wall is impassable for many species, potentially lowering their chance of survival. How exactly the border affects this rich ecosystem, however, has largely been a mystery.

A new study, among the first of its kind, finally offers some answers — by essentially spying on animals near the border. For the research, ecologist and lead author Ganesh Marín, then a doctoral researcher at the University of Arizona, set up 85 motion-sensing cameras in northeastern Sonora, Mexico, along and south of the US border in Arizona and New Mexico. Throughout the course of the research, when animals walked by, the cameras began recording.

Over roughly two years, from 2020 to 2022, the cameras captured hundreds of hours of footage, including more than 21,000 clips with mammals, said Marín, a National Geographic Explorer and postdoctoral scientist at the nonprofit Conservation Science Partners.

“This place is so special because you see these tropical species, like ocelots and jaguars, at the same time as beavers and black bears,” Marín told me earlier this year when I was reporting on borderland jaguars.

Some of the recordings are pretty incredible. In this clip, for example, a young puma, or mountain lion, makes a chirping sound, likely calling for its mother.

Courtesy of Ganesh Marín

Or check out this jaguar approaching the camera. This particular cat is known as Bonito. Scientists first detected this cat in 2020 and can identify him by his markings.

Courtesy of Ganesh Marín

Marín’s cameras detected another jaguar, as well, called Valerio. He was seen by cameras multiple times in a protected area known as Cuenca Los Ojos just south of the border in Sonora.

Courtesy of Ganesh Marín

The camera traps caught black bears and their cubs…

Courtesy of Ganesh Marín

Courtesy of Ganesh Marín

…bobcats and coyotes…

Courtesy of Ganesh Marín

Courtesy of Ganesh Marín

…and even an ocelot, an elusive predatory cat.

Courtesy of Ganesh Marín

Analyzing the videos ultimately revealed several important details about wildlife in the borderlands. Marín found that large mammals, such as black bears and deer, as well as some smaller herbivores, spend less time near the border than in other, more remote stretches of his study region. That suggests these animals avoid border infrastructure.

Other species, like the pronghorn, which have been seen on the US side of the border, didn’t appear in his cameras at all. That may be because they have trouble crossing a highway that runs roughly parallel to the border in Sonora, according to Marín and his co-author, John L. Koprowski, a biologist at the University of Wyoming.

Meanwhile, smaller common predators like coyotes and bobcats appeared more tolerant to human activity: They were more likely to use habitats with cattle, cars, and dirt roads, according to the footage.

The study adds to a growing body of research showing that the border and infrastructure around it is disrupting wild animal communities.

“Amazing wildlife is present in the borderlands due to the binational efforts to protect and restore the flow of life between both countries,” Marín said in an email. “We should not define this beautiful region and the creatures that roam by the existence of an imposed division.”

---

From Vox via this RSS feed

Unrecognizable female friends celebrate and drink red sparkling wine.

There is a lot of advice out there about how much alcohol one should drink. There is research suggesting that drinking could be dangerous, and research that indicates drinking is good for you.

Which is it? Obviously, too much drinking is bad for one’s health — and drinking to excess can destroy the human body. But is moderate drinking good — or, at least, fine?

Dylan Scott, Vox’s senior health reporter, has been looking into this matter for some time, and I recently asked him to sum up what he’s learned. Here’s what he had to say:

You’ve done some reporting on alcohol recently and whether it’s safe. Is it?

There is widespread agreement that heavy drinking is not good for you — doctors and scientists have known for literally centuries that a lot of drinking is dangerous.

And the more you drink, the greater your risk. Your risk starts to increase pretty exponentially once you’re having more than one or two drinks at a given sitting, especially if you’re drinking every day.

There is still a lot of debate about the safety of drinking small amounts of alcohol and whether it can have very small health benefits. On that front, studies can seem to contradict themselves.

I talked to one scientist who has published some research documenting cardiovascular benefits from drinking a little bit of alcohol, and I also recently talked to the author of a 2017 statement from the leading cancer physician medical society, which was basically intended to be a wake-up call to the public that alcohol is a carcinogen.

Yet those two people, despite appearing to be on opposite sides of the debate, would basically be in total agreement, about the negative consequences of more than one drink for a woman every day or more than two drinks for a man every day.

Alcohol is a carcinogen?

Yes, but let me take a step back.

What has stuck out to me in reporting about alcohol is that the problem isn’t so much the substance itself as it is widespread misunderstanding about what moderate drinking means.

That’s 12 oz. of a 5 percent beer, 1.5-oz. glass of 80-proof liquor, and 5 oz. of a 12 percent glass of wine.

There’s a trope among doctors that most people think they’re moderate drinkers but aren’t thinking about those numbers as they drink. I might pour a glass of wine and think I’m having one glass of wine, but a doctor would see two glasses of wine if it’s a really generous pour.

Coming back to your question, if you didn’t know alcohol is a carcinogen, you’re not alone. I learned in my reporting that only 40 percent of people know alcohol is a carcinogen, which shows there’s still a lot of work to do in educating people about the health risks.

Public health experts told me that they want to be more vocal about some of alcohol’s risks, especially about it being something that builds a dependency. Between that, and alcohol being a carcinogen, you can start to see why knowing what levels of drinking are actually moderate is really important.

That’s interesting, and it makes me wonder about those headlines that claim a new study has found a glass of red wine a day is the key to longevity or something like that. Is there anything to those?

After my reporting, I do think there is some room for debate about whether a very modest amount of alcohol consumed in a very particular way might confer some small cardiovascular benefit.

That said, even the doctor I talked to who’s authored studies finding some benefit, said, “This is not an elixir*.*” He was clear that his work shouldn’t be read as saying, “Alcohol is going to reduce your chance of diabetes, improve your heart health, or what have you.”

So, you’re saying I shouldn’t start drinking, hoping it will make me a healthier person.

Yes. The doctors I’ve spoken to have said things like, “I would never tell somebody to start drinking because it’s not going to help you.”

The basic thing to remember, though, is if you’re a light drinker, any potential problems caused by alcohol aren’t something worth worrying about.

People should be aware of the risks but shouldn’t panic about them. Really, my two big takeaways on alcohol are: Heavy drinking is dangerous, and it’s easy to drink too much. Those are the things to watch out for.

---

From Vox via this RSS feed

A makeshift memorial for Minnesota state Rep. Melissa Hortman and her husband Mark. They were killed in their home on June 14, 2025. | Steven Garcia/Getty Images

A series of high-profile incidents of political violence — targeting members of both major political parties — have grabbed the nation’s attention.

Earlier this month, a gunman shot two Minnesota Democratic lawmakers in their homes. State Rep. Melissa Hartman and her husband were killed, and state Sen. John Hoffman and his wife were injured.

In April, a man who allegedly “harbored hatred” for Pennsylvania Gov. Josh Shapiro set fire to the Democrat’s home while he and his family were sleeping inside.

President Donald Trump faced two assassination attempts during his 2024 campaign. A former Coast Guard officer who identified with Antifa, a far-left antifascist militant movement, was also arrested earlier this month for issuing violent death threats against Trump.

In October 2022, former House Speaker Nancy Pelosi’s husband was attacked by an assailant who broke into their home looking for her.

And on January 6, 2021, rioters descended on the US Capitol to stop the certification of Joe Biden’s 2020 election victory, threatening to hang then-Vice President Mike Pence for allowing it to move forward.

It might feel like, based on the severity and frequency of these headline incidents, American political violence is surging. Members of Congress appear to think so: Lawmakers from both parties are now asking for more funding to enhance security and investigate and prosecute more threats made against them.

But while there are signs in the data that indicate political violence is indeed on the rise, depending on how you define it, it’s challenging to determine exactly by how much.

“It’s more anecdotal than anything else,” said Katherine Keneally, director of threat analysis and prevention at the Institute for Strategic Dialogue. “There’s some data to back up that the tensions are increasing and creating a more volatile environment, but to say it’s increased by X amount since 2023 is a little trickier.”

A volatile political environment and changes in social media policies that have caused misinformation to spread more quickly appear to be what’s driving the increase, at least in part. But understanding the root causes requires ascertaining the scale of the problem in a way that researchers have struggled to capture comprehensively.

Is political violence actually rising significantly?

There are all sorts of difficulties associated with measuring political violence.

First, there’s the definitional dilemma of what incidents to include when counting acts of political violence. For instance, some might count arrests for disrupted plots; others might not.

Then, there is the challenge of actually gathering the data. Some sources may overly rely on media reports in an era when local news is under-resourced and might not reliably record every incident. And in the US, individuals unaffiliated with armed groups have become the primary perpetrators of political violence. That makes political violence even harder to track because perpetrators are often interacting in fragmented, low-transparency spaces online, from private chats to forums, rather than congregating in a single organized group.

Despite the difficulties with measurement, some sources — particularly those looking at specific forms of political violence — suggest that overall levels of political violence have increased in recent years.

US Capitol Police have been recording concerning statements and direct threats made against members of Congress, their families, and their staff since 2017, seeing significant spikes after the 2020 and 2024 presidential elections.

Researchers at Princeton University’s Bridging Divides Initiative also recorded a similar spike in threats to local officials in 2024.

In 2025 so far, they identified more than 170 total incidents across nearly 40 states, with national issues such as LGBTQ+ rights and the war in Gaza being major bipartisan drivers. About a quarter of them involved hate speech. And in a sign of how political discourse has devolved, about 20 of them involved local officials threatening or harassing each other.

However, researchers acknowledge that they are only scratching the surface and that a broader analysis of the threat environment must begin well before anyone reaches the point of directly threatening to harm someone or actually harming them.

“The data only looks at the point at which people successfully conduct acts of violence,” said Jon Lewis, a research fellow at George Washington University’s Program on Extremism. “I think we need to start far earlier in the process and far more holistically to really capture the root causes of this issue, which is rhetoric.”

Why is political violence on the rise?

American political violence looks different now than it did during major periods of political upheaval in the past.

In the 1970s, it was driven predominantly by far-left, anti-war groups such as the Weather Underground, which were primarily engaged in the destruction of property. But the nature of political violence, as well as its perpetrators, has changed in the decades since.

“I think the modern iteration of mainstream right-wing political violence is targeting individuals, mass violence, targeted assassinations, which I think takes on a very different tenor than the destruction of property,” Lewis said.

There are several reasons for this shift, with the proliferation of conspiracy theories and hate speech online being a major one.

“We’re in a very pretty difficult position in the country right now.”

Katherine Keneally, director of threat analysis and prevention at the Institute for Strategic Dialogue

Content moderation on mainstream social media sites was never a complete cure for that, but studies have suggested that it was a mitigating factor. Twitter (now X), Meta, YouTube, and others have scaled back content moderation staff or rolled back policies designed to root out misinformation that might motivate political violence. In the months after Elon Musk bought X and implemented those policies, hate speech on the platform rose by 50 percent, according to a study by researchers at the University of California Berkeley, UCLA, and the University of Southern California.

“I think that we really need to recognize the fact that there is a significant subset of people, especially online, especially on these social media platforms, that do not share our common understanding of reality,” Lewis said. “If you spend your weekend on Twitter, which I would not recommend doing, you would genuinely say that the suspect [in the Minnesota shootings] shot these Democratic politicians because they went against the leftist, Marxist party line.”

In truth, federal prosecutors have declined to state a specific motive. But if anything, the evidence suggests he identified with the far-right rather than the far-left: His friends described him to Fox News as a Trump supporter, his social media posts embraced extreme anti-abortion views, and he had a hit list of 45 elected Democrats.

There’s also, according to researchers, a cultural shift — and not a healthy one.

People also now seem more willing to see political violence as a solution to the policies and beliefs they disagree with, regardless of party affiliation, Keneally said.

During the 2024 presidential election, polling from NORC and the University of Chicago Project on Security and Threats found 7 percent of Americans agreed that the “use of force is justified” to help Trump claim the presidency; 10 percent said it was justified to prevent him from doing so.

Now that Trump is president again, many communities feel under threat from his policies, which may make them more accepting of political violence. A March Scientific American survey of predominantly Democratic voters at two major protests found that about a third said political violence may be necessary to “save” America. It’s worth noting that these respondents aren’t representative of Democrats overall, but it shows that acceptance of political violence isn’t just a right-wing phenomenon.

“We’re in a very pretty difficult position in the country right now,” Keneally said. “I think this combination of this changing political environment, social media, and people feeling like they don’t have any other solution is making it at least feel like it’s worse.”

---

From Vox via this RSS feed

You probably haven’t heard of organ procurement organizations, but if you or anyone you know has ever received an organ transplant, they’re the ones who procured it. | Christopher Furlong/Getty Images

Below is a graph showing a trend that exploded during the 2020s:

What is this depicting? Compute use for AI? Crispr gene edits per year?

No, this is another, much less-known example of massive growth these past several years. This is a chart of the number of pancreases (or, to use the correct plural, “pancreata”) collected each year from dead bodies in the US for research purposes:

How this happened is no mystery. The surge is, by all accounts, due to a regulation that took effect in 2021 focused on groups called organ procurement organizations (OPOs).

You probably haven’t heard of OPOs, but if you or anyone you know has ever received an organ transplant, they’re the ones who procured it. OPOs are nonprofit, nongovernmental bodies to which the US outsources the job of collecting organs from deceased organ donors. Each OPO has a monopoly on recovery of all organs in a particular geographic area; there are 55 groups, some of which only cover part of a state and some of which cover multiple states.

For some time now, critics have argued that OPOs are massively underusing deceased donor organs. One report from 2019 estimated that every year 28,000 usable organs (mostly badly needed kidneys but also pancreata, hearts, livers, etc.) are removed from deceased donors but never used; another put the number at 75,000. This, when the national waitlist for organs is more than 100,000 people long.

OPOs are not paid to collect these organs per se: They are entitled to 100 percent reimbursement of costs they report related to retrieving, preserving, and delivering organs, with ultimate payment coming from Medicare or transplant centers (which in turn charge Medicare and other insurers). This system, critics have long charged, does not provide enough incentive to procure harder-to-retrieve organs from patients who may be older or have certain medical conditions.

To get OPOs to collect more organs, the Trump administration in 2019 issued an executive order calling for new rules governing how the organizations are certified by the federal government, rules that were finalized two years later. This was high stakes: If an OPO loses certification, it has to shut down, and another OPO gets its territory. The rules were meant to more strictly grade OPOs on the share of organs they eventually transplant than the earlier, laxer rules did.

But there was a catch. In addition to organs recovered from deceased donors and transplanted, pancreata recovered and used for research would count toward recertification as well. Not any other organs for research — just pancreata.

What happened next can be see in the chart above: a massive, sudden surge in the number of research pancreata being recovered by OPOs, beginning in 2022, the precise year the new evaluation system took effect.

I’ve long been fascinated by this trend, which OPO critics call the “pancreas loophole” and OPO defenders describe as a perfectly legal response to overly onerous regulations. The numbers represent thousands of real, physical human pancreata, taken from real, recently deceased donors, that wouldn’t have been taken from those bodies without this regulation.

I’ve tried in recent months to make sense of how this happened, and what it means. I’m not the only one; the Senate Finance Committee has been investigating, and released a report in early June on the problem.

There is still plenty that remains unknown about the fate of these pancreata (if you work at an OPO or research center and know more details, please email me). But what is clear is that they represent an approach by the federal government toward increasing organ supply that absolutely no one is happy with. If the point of the regulations is to help people in need — including the millions of Americans with diabetes, a disease of the pancreas — evaluating OPOs based on the number of pancreata they donate to researchers simply doesn’t make any sense.

But to understand how we started judging them this way regardless, you have to go all the way back to an obscure law passed in George W. Bush’s first term.

Pancreata (and why you might need one transplanted), explained

Everyone knows, in broad strokes, what the heart or the lungs do. But the pancreas doesn’t have the same level of fame. Its basic purpose is to excrete enzymes, hormones, and other compounds to both 1) help the body digest food and 2) regulate blood sugar levels.

The latter function is performed by the islets of Langerhans, cells in the pancreas (named after their discoverer, 19th-century German researcher Paul Langerhans) that secrete two different hormones: insulin (to lower blood sugar) and glucagon (to raise it).

In Type 1 diabetes, the ability of the pancreas to produce insulin is impaired and thus blood sugar levels are dangerously elevated; in some kinds of Type 2 diabetes, the body develops resistance to insulin’s effects. Typically, people with diabetes deal with this through injecting insulin directly, a process that has become much more sophisticated in recent decades as finger pricks and needles have given way to insulin pumps that can directly measure and adjust blood sugar levels.

But even with advanced care, diabetes carries lifelong medical consequences, so researchers have long sought a more permanent fix: What if you could replace or supplement the faulty islet cells in patients with diabetes with healthy islet cells? Could you, then, cure diabetes at the source and avoid the need for insulin injections and the risk of long-term health effects altogether?

In the most extreme version of this approach, a complete new pancreas is transplanted into a patient with diabetes, like swapping out a faulty part. This is a proven treatment (915 occurred in 2023) and when done it works well, essentially curing the recipient’s diabetes.

But there are major downsides: you have to undertake major abdominal surgery with a small but real chance of failure, and if that succeeds, you have to remain on immunosuppressant drugs for the rest of your life to prevent organ rejection.

For that reason, physicians generally rule that the costs of a pancreas transplant outweigh the benefits for most people with diabetes. Living with an insulin pump is better than risking surgery and having a permanently compromised immune system. Very few of the 38 million Americans living with diabetes, then, are going to be candidates for a pancreas transplant.

This math changes, however, if the patient in question also needs a kidney transplant. Diabetes accounts for nearly half of all new cases of kidney failure, so a higher share of people with diabetes than people without find themselves in this situation. In these cases, since the patient is already going to have surgery and be on immunosuppressants, throwing in a new pancreas to the surgery and curing their diabetes in the same operation that cures their kidney failure begins to look like an attractive option. That’s why almost no one gets a pancreas transplant in the US without getting a kidney transplant too.

A cure for diabetes

For decades, researchers searching for effective diabetes treatments have experimented with an approach called islet cell transplantation.

Rather than transplant the whole pancreas, the procedure merely transplants insulin- and glucagon-producing islet cells into the recipient’s liver. It’s far less invasive, and can be done with local anesthesia and without an overnight hospital stay (though, skeptics argue, often less effective than whole pancreas transplantation too). If the islet cells come from a deceased donor, it does mean a lifetime of immunosuppression, but in “autograft” procedures, which use a subset of still-healthy islets from one’s own impaired pancreas, immunosuppression isn’t necessary.

Islet cell transplantation, though, remains little-used and mostly experimental in the US. Unlike its big brother surgery, though, islet cell transplantation remains little-used and mostly experimental. Part of the reason why is regulatory: While pancreases are legally “organs,” and therefore excluded from regulation by the Food and Drug Administration, the FDA has asserted its authority to also regulate islet cells as human tissues and to require premarket approval before they can be transplanted into a patient, just like a drug would.

To gain such approval, it would be necessary to conduct clinical studies to demonstrate that the islet cells are “safe and effective”; ones approved they would need to be produced in compliance with “good manufacturing practices.” To receive islet cells not approved by FDA, a patient would need to join a clinical study (if one is being conducted) or go to a country (including Canada, Australia, and several EU and Asian countries) with different regulations.

FDA’s requirements have unsurprisingly slowed the islet cell transplant field down. The most recent data comes from 2021, when only 10 such procedures were performed in the US. In 2012, 111 were performed, but the number has steadily fallen ever since. It’s orders of magnitude rarer than a whole-pancreas transplant.

But islet cell transplants have some champions, including politicians. In 2004, Congress passed, and George W. Bush signed, the Pancreatic Islet Cell Transplantation Act. It was sponsored by Rep. George Nethercutt Jr. (R-WA), who said he was driven by his daughter’s diabetes to try to expand access to islet cell transplants so patients could “live without being dependent on insulin injections.”

Nethercutt’s bill sought to speed up research progress by, among other measures, ensuring adequate supply of pancreata for scientists. The law includes a provision stating, “Pancreata procured by an organ procurement organization and used for islet cell transplantation or research shall be counted for purposes of certification or recertification.”

That meant that when regulators in 2019 were reconfiguring certification rules for OPOs to encourage them to effect more transplants, they had to include a carve-out for pancreata used for islet cell research. The carve-out existed from the very first draft proposal that Trump’s Centers for Medicare and Medicaid Services (CMS) released, and stayed into the final rule in 2021.

The mysterious pancreas boom

That rule’s changes applied for evaluations of organ procurement organizations starting in August 2022. In the years 2018 through 2021, OPOs collected around 500–600 pancreata for research each year. In 2022, that figure was 1,432, a three-fold increase. In 2024, the number hit 2,053. The effect of the new regulation was clear.

It’s important to note that there are no indications that the pancreata being collected by OPOs for research are cutting into the supply of pancreata for donation. The share of dead people whose pancreata are suitable for transplant is incredibly low, because of the exacting standards for donor age and health.

“Only a certain number of donors are going to be young enough (probably less than 50, maybe less than 45) and lean enough (maybe less than 30 BMI, probably less than 27 or 28),” Jonathan Fridell, a transplant surgeon and director of the pancreas transplant program at IU Health in Indianapolis, told me. “We’re still going to look at the people that are older, still look at the people that are heavier, but the likelihood that they’re going to have a transplantable pancreas is lower.” There are thus plenty of non-transplantable pancreata left over that could be used for research once the prime ones are taken away for surgery.

The problem with the surge in research pancreata, then, isn’t that it’s taking pancreata away from recipients who need them. It’s subtler than that. By racking up large numbers of pancreata for research, OPOs are improving the grades they receive from federal regulators, and avoiding the risk of losing certification and having to turn over territory to another OPO. This grading and decertification process was meant to incentivize OPOs to collect more organs for transplant, especially kidneys, which are both easier to transplant than pancreata and more desperately needed. But the research pancreata provide a way around that incentive.

Calculate the grades that OPOs would get without these pancreata versus the ones they are getting now, and you’ll find the results are radically different.

CMS classifies OPOs into three tiers: 1, 2, and 3, with tier 3 OPOs facing decertification. Using data obtained from the organ procurement transplant network, we are able to calculate which tier each OPO would be in with and without their research pancreata based on their performance in 2023. The year that is actually binding for OPOs and determines whether they will be decertification is 2024, for which data does not yet exist, but the 2023 data gives us some indication of which OPOs are using pancreata to save themselves.

For two OPOs, including research pancreata meant they went from tier 3, which would result in decertification, to tier 2: Donor Network of Arizona, which covers that whole state; and OneLegacy, which includes most of Southern California including Los Angeles and Orange counties. Another, Kentucky Organ Donor Affiliates, went from tier 2, where it could face pressure to improve performance, to tier 1, suggesting it excels.

These are not small institutions. OneLegacy is by far the largest OPO in the country by volume of organs collected and population covered. Its CEO, Prasad Garimella, earned $1.1 million in total compensation in 2023, the last year for which public IRS filings are available. It stood a real chance of being decertified if it did not meet the new organ collection standards. And it went from reporting 83 pancreata collected for research in 2021 to 441 in 2022 and 492 in 2023. An over fivefold increase, in one year — and no wonder, given its existence was at stake.

(In response to a request for comment, OneLegacy stated, “When recovering organs for transplant, OneLegacy will allocate pancreata to reputable islet cell research agencies only if they are not viable for transplant into patients. Over 99.6% of pancreata recovered by OneLegacy for research between 2018-2022 were allocated to two National Institute of Diabetes and Digestive and Kidney Diseases (NIH-NIDDK) laboratories.”)

Again, the 2023 data is not binding. The 2024 data will be. But unless something changes with the way the government evaluates these pancreata, some major OPOs will avoid dire consequences for the sole reason that they started collecting hundreds of pancreata for research.

Where did all the pancreata go?

The increase is so obviously a result of the new rules that OPOs don’t even bother to deny it. Responding to the Senate report earlier this month, the Association of Organ Procurement Organizations, which lobbies for the groups and against the CMS’s stricter rules, said simply, “Today, pancreata recovered for research remain part of the performance evaluation metrics, and OPOs have operated in accordance with the rule.” In other words: yeah, we found a loophole. And what are you going to do about it?

It’s an attitude that has pervaded the industry since the loophole came to light. In a listserv thread discussing the new rules, leaked to the Senate Finance Committee, an OPO employee wrote, “If you have a donor with only a pancreas for research, that is an organ donor for the Donor Rate. Otherwise, a donor is any donor with at least 1 organ transplanted. Savvy (or cynical?) OPOs ought to start a pancreas for research program immediately.”

But there’s a question that remains unanswered: Where did all these thousands of pancreata go?

It is clear that the vast majority of research pancreata did not go into islet transplants. We are talking about thousands of organs, not the few dozen that plausibly could have been transplanted as part of islet procedures in the past couple of years.

Indeed, OPOs have admitted as much. The Centers for Medicare and Medicaid Services asks OPOs for data on organs recovered every year, and in August 2024 clarified that it would only count pancreata as “‘used’ for research if they are accepted for use in bona fide islet cell research conducted by a qualified researcher, such as research approved by the National Institutes of Health.” It then asked OPOs to resubmit their data, clarifying which organs were for islet cell research specifically.

Once they did, the number of reported pancreata fell dramatically:

In 2023, the total went from 3,338 pancreata before the guidance, to 1,812 after, a drop of 46 percent. Some OPOs, like Legacy of Hope in Alabama, now reported zero pancreata for research; before the guidance narrowed qualifying purposes, Legacy of Hope had claimed 226 pancreata.

But even after the change in guidance, we’re left with smaller numbers that are still much too big to be explained by bona fide islet cell transplants. There simply were not 1,812 islet cell transplants in the US in 2023, but there were 1,812 pancreata credited as donated for islet cell research. And that number is still over triple the number claimed in 2021, meaning the increase sparked by the new OPO rules largely remains even after the government’s clarification. Research by David Goldberg, Erin Tewksbury, and Matthew Wadsworth has shown that the number of pancreata reported as recovered by OPOs also swamps the number that the Integrated Islet Distribution Program (IIDP), a consortium that collects and extracts islet cells from pancreata, reports receiving from these OPOs.

One of the points of the Senate investigation was to determine where exactly these pancreata went. The Senate Finance Committee, with the benefit of subpoena power, went about asking major OPOs for what actual purpose the research pancreata were used. The main answer they received was “we don’t know.”

“Many of the OPOs stated that it is the responsibility of the research facilities or institutions receiving the pancreata to inform the OPOs on the purpose, methods, and efficacy of the research being conducted on the pancreata and other organs that OPOs supply,” the report states. In other words, OPOs themselves don’t keep track. “Many of these OPOs,” the report continues, “have sent pancreata to biobanks and other institutions or facilities that hold pancreata for an unknown period to be used for purposes that may be undefined or nonexistent.”

Put another way: These pancreata could, for all the OPOs or the Senate knows, be sitting on a freezer somewhere, not transplanted into anyone. Or maybe not even sitting there at all. Greg Segal, an activist advocating for reform to the pancreas loophole, testified before a House committee that staff at one OPO, joked “that they’re conducting research on the efficacy of garbage disposal A versus garbage disposal B” when disposing of pancreata.

Exploiting the loophole

Throughout all this, OPOs have had one consistent message: They’ve complied with the law, as they see it.

“Pancreata recovered for research remain part of the performance evaluation metrics, and OPOs have operated in accordance with the rule,” the Association of Organ Procurement Organizations, the groups’ lobbying shop in Washington, said in its statement after the Senate investigation was released. “When CMS issued clarifying guidance in 2024 limiting this metric to pancreata used for islet cell research, OPOs responded immediately and worked with the agency to validate data and ensure compliance.”

Jedd Lewis, CEO of the Organ Preservation Alliance and a decades-long veteran of the transplant field, notes that CMS’s rule neglected to define what it means to use a pancreas for research, despite many OPOs and industry experts specifically flagging the problem for CMS before that rule took effect. And CMS’s new guidance last year did little to solve the problem, he argues.

“Last years’ memos simply identified the scope of pancreas donations that OPOs would be judged on as those for ‘islet cell research.’ But CMS didn’t define what that actually means …and on its face it’s a huge scope of research,” Lewis wrote in an email. “There are so many … ways that researchers are looking at how those cells function: studying the pancreas whole, slicing it into thin sections, isolating the individual islet cells, even breaking the cells into the component parts.” That’s all valuable research, he argues, and clearly relates to islet cells, even if the pancreata are never actually used in islet cell transplants.

Wadsworth, a coauthor on the study finding a surge in research pancreata and CEO of the LifeConnection OPO in northwest Ohio, concedes that counting pancreata that did not produce islet cells for transplant may technically be legal. But he still thinks it’s wrong.

“I worked with this surgeon early on in my career who said ‘just because you can do something doesn’t mean you should,’ Wadsworth noted. “Based on what’s written, maybe they didn’t do anything wrong, but you don’t have to look far back in history to find examples where something wasn’t illegal, but it definitely wasn’t right either.”

LifeConnection, Wadsworth says, was able to comply with the spirit of the CMS regulations by finding counties in its jurisdiction where low numbers of organs were being procured, and working on fixing the problem hospital-by-hospital. It’s harder than just harvesting pancreata, but it means organs get transplanted to people who need them.

One irony of the controversy is that most OPOs, and their representatives, don’t believe that research pancreata should count for their evaluations. “AOPO has concerns about including pancreata utilized for research in the data used to calculate the numerator of either proposed measure,” the Association of Organ Procurement Organizations wrote in an early 2020 comment as the regulations were being developed. “The utilization of pancreata for research is driven by demand of local researchers. Inclusion of pancreata for research in the data utilized for the numerator may skew comparisons of OPOs in that category and potentially lead to inaccurate conclusions.”

But both the OPOs and their regulator, CMS, were bound by the 2004 law requiring that research pancreata, at least that for islet cell research, must count for these evaluations.

Close the pancreata gap

We are running out of time to fix this problem. If nothing changes, OPOs will be evaluated on the basis of data they’ve submitted now, including hundreds of pancreata that were never used for islet cell transplants. Whether you think that reflects OPOs complying in good faith, or subverting the system, it’s not a policy anyone should think makes much sense.

CMS has some ability to act here — but perhaps the best fix would come from Congress in the form of a legal provision clarifying the 2004 act. Simply repealing the provision restricting how OPOs can be evaluated would be simplest — but even better would be pairing it with a legal change that could help islet transplantation research far more than the 2004 has to date.

Recall that islet transplants currently don’t count as organ transplants in the US. They count as treatments with biological tissue “”If islet cells are solely organs, because they are a subpart of the pancreas, which  is an organ under transplant law, then the FDA should not have  jurisdiction,” Gail Javitt, a veteran lawyer working on FDA regulatory issues at the firm Hyman, Phelps & McNamara, told me. “However, FDA has taken a different position, that islet cells are a cellular therapy and  must undergo premarket approval just like a drug would.”   If you want to use it for treatment of a patient,  you have to go through them. That has had the practical impact of slowing down the availability of islet cells for transplantation in this country.”

Legally clarifying that islet cells are organs, not cellular therapies, and that they are excluded from FDA oversight then, could go a long way to promoting the treatment. Last Congress, Sen. Mike Lee (R-UT) and Rep. Matthew Rosendale (R-MT) each introduced bills making this change, with exclusively Republican co-sponsors.

But this need not be a partisan issue at all, and if you paired this provision with a repeal of the 2004 law permitting OPOs to count pancreata for research as part of its transplant metrics, you could arrive at a close to ideal system. OPOs would be evaluated on their ability to transplant islet cells, because they’d be organs like any other. They would not be able to get higher ratings by recovering pancreata for research that might just languish on a shelf.

This does require Congress to make a small change. But it’s a small change that should be basically uncontroversial. There’s nothing for most OPOs or for advocates trying to maximize donations to dislike here, and there’s lots for islet cell researchers to love. It’s a small fix that could go a very long way.

---

From Vox via this RSS feed

NEW YORK, NEW YORK – JUNE 18: People march during a rally calling for the Trump administration not to go to war with Iran, on June 18, 2025 in New York City. (Photo by Adam Gray/Getty Images) | Carlos Barria/Reuters/Bloomberg via Getty Images

When Vice President JD Vance appeared on Meet the Press on Sunday morning, anchor Kristen Welker asked him a simple question: Is the United States now at war with Iran?

In response, Vance said, “We’re not at war with Iran; we’re at war with Iran’s nuclear program.”

This is akin to saying that, in attacking Pearl Harbor, Imperial Japan had merely declared war on America’s warship construction program. Yet it’s notable that Vance felt the need to engage in such contortions — and that President Donald Trump, in his address to the nation last night, went out of his way to emphasize that there were no additional strikes planned.

The Trump administration does not want to admit it has begun a war, because wars have a way of escalating beyond anyone’s control. What we should be worrying about now is not how the US-Iran fighting began, but how it ends.

It is all too easy to see how these initial strikes could escalate into something much bigger — if Iran’s nuclear program remains mostly intact, or if Iran retaliates in a way that forces American counter-escalation.

It’s possible neither occurs, and this stays as limited as currently advertised. Or factors beyond our knowledge — the “unknown unknowns” of the current conflict — could lead to an even greater escalation than anyone is currently predicting. The worst-case scenario, an outright regime change effort akin to the 2003 invasion of Iraq, cannot be entirely ruled out.

I don’t know how bad things will get, or even if things are likely to get worse. But when I watched Trump’s speech, and heard his obviously premature claims that “Iran’s key nuclear facilities have been completely and totally obliterated,” I couldn’t help thinking about another speech from over 20 years ago — when, after the toppling of Saddam Hussein in 2003, George W. Bush stood on an aircraft carrier and declared “Mission Accomplished.”

The mission hadn’t been accomplished then, as it almost certainly hasn’t been now. We can only hope that the resulting events this time are not a similar kind of catastrophe.

Escalation pathway one: “finishing the job”

We do not know, at present, just how much damage American bombs have done to their targets — Iranian enrichment facilities at Fordow, Natanz, and Isfahan. Satellite imagery shows that there are above-ground buildings still standing, belying Trump’s claims of complete destruction, but many of the targets are underground. It’s possible these were dealt a severe blow, and it’s possible they weren’t.

Either scenario creates pathways to escalation.

If the damage is indeed relatively limited, and one round of American bombs was not able to shatter the heavily reinforced concrete Iran uses to protect its underground assets, the Trump administration will face two bad choices.

It can either let a clearly furious Iran retain operational nuclear facilities, raising the risk that they dash for a nuclear weapon, or it can keep bombing until the attacks have done sufficient damage to prevent Iran from getting a weapon in the immediate future. That commits the United States to, at minimum, an indefinite bombing campaign inside Iran.

But even if this attack did do real damage, that leaves the question of the program’s long-term future.

Iran could decide, after being attacked, that the only way to protect itself is to rebuild its nuclear program in a hurry and get a bomb. It has already moved to quit the Treaty on Non-Proliferation of Nuclear Weapons (NPT), an agreement that gives international inspectors (and, by extension, the world) visibility into its nuclear development.

There are, again, two ways to ensure that Supreme Leader Ali Khamenei doesn’t make such a choice: a diplomatic agreement akin to the 2015 nuclear deal, or else a war of regime change aimed at overthrowing the Iranian government altogether.

The first isn’t impossible, but it certainly seems unlikely at present. The US and Iran were negotiating on its nuclear program when Israel began bombing Iranian targets, seemingly using the talks as cover to catch Iran off guard. It seems very unlikely that Iran would see the US as a credible negotiating partner now that it has joined Israel’s war.

That leaves the other form of “finishing the job”: a full-on war of regime change. My colleague Josh Keating has argued, convincingly, that Israel wants such an outcome. And some of Trump’s allies, including Sens. Ted Cruz and Lindsey Graham, have openly called for it.

“Wouldn’t the world be better off if the ayatollahs went away and were replaced by something better?” Graham asked, rhetorically, in a Fox News interview last Monday. “It’s time to close the chapter on the Ayatollah and his henchmen. Let’s close it soon.”

Such a dire outcome seems, at present, very distant. But the further Trump continues down a hawkish path on Iran, the more thinkable it will become.

Escalation pathway two: a US-Iran cycle of violence

There’s a military truism that, in war, “the enemy gets a vote.” It could be that Iran’s actions force American escalation even if the Trump administration doesn’t want to go any further than it has right now.

So far, Iran’s military response to both US and Israeli attacks has been underwhelming. Tehran is clearly hobbled by the damage Israel did to its proxy militias, Hezbollah and Hamas, and its ballistic missiles are not capable of threatening the Israeli homeland in the way that many fear.

But there are two things Iran hasn’t tried that are, after American intervention, more likely to be on the table.

The first is an attack on US servicemembers stationed in the Middle East, of which there are somewhere between 40,000 and 50,000 at present. Of particular note are the US forces currently stationed in Iraq and Syria. Iraq is home to several Iranian-aligned militias that could potentially be ordered to directly attack American troops in the country or across the border in Syria.

The second is an attack on international shipping lanes. The most dangerous scenario involves an attempt to use missiles and naval assets to close the Strait of Hormuz, a Persian Gulf passage used by roughly 20 percent of global oil shipping by volume.

If Iran either kills significant numbers of American troops or attempts to do major damage to the global economy, there will surely be American retaliation. In his Saturday speech, Trump promised that if Iran retaliates, “future [American] attacks will be far greater and a lot easier.” An effort to detonate the global oil market would, without a doubt, necessitate such a response: The US cannot allow Iran to hold its economy hostage.

We do not, to be clear, know whether Iran is willing to take such risks, or even if it can. Israeli attacks have devastated its military capabilities, including ballistic missile launchers that allow it to hit targets well beyond its borders.

But a “cycle of violence” is a very common way that violence escalates: One side attacks, the other side retaliates, prompting another attack, and on up the chain. Once they start, such cycles can be difficult to prevent from spiraling out of control.

Escalation pathway three: the Iraq analogy, or things fall apart

I want to be clear that escalation here isn’t a given. It is possible that the US and its Israeli partners remain satisfied with one American bombing run, and that the Iranians are too scared or weak to engage in any major response.

But those are a whole lot of “ifs.” And we have no way of knowing, at present, whether we’re heading to a best- or worst-case scenario (or one of several possibilities in the middle). Key decision points, like whether Trump orders another round of US raids on Fordow or Iran tries to close the Strait of Hormuz, will determine which pathways we go down — and it’s hard to know which choices the key actors in Washington, Tehran, and Jerusalem will make.

I keep thinking about the 2003 Iraq war in part for obvious reasons: the US attacking a Middle Eastern dictatorship based on flimsy intelligence claims about weapons of mass destruction. But the other parallel, perhaps a deeper one, is that the architects of the Iraq War had little-to-no understanding of the second-order consequences of their choices.

There was so much they didn’t know, both about Iraq as a country and the likely consequences of regime change more broadly, that they failed to grasp just how much of a quagmire the war might become until it had already sucked in the United States. It’s over 20 years later, and boots are still on the ground — drawn in by events, like the creation of ISIS, that were direct results of the initial decision to invade.

Attacking Iran, even with the more “modest” aim of destroying its nuclear program, carries similar risks. The attack carries so many potential consequences, involving so many different countries and constituencies, that it’s hard to even begin to try to account for all the potential risks that might cause further US escalation. There are likely consequences taking shape, at this moment, that we can’t even begin to conceive of.

The nature of the Trump administration gives me little hope that they’ve properly gamed this out. The president himself is a compulsive liar and foreign policy ignoramus. The secretary of defense has run his department into the ground. The secretary of state, who is also the national security adviser, has more jobs than anyone could reasonably be expected to perform competently at once. It is, in short, far less competent on paper than the Bush administration was prior to the Iraq invasion — and look how that went.

It’s possible, despite all of this, that the Trump administration has adequately gamed out their choices here — preparing for all reasonably foreseeable contingencies and capable of acting swiftly in the (inevitable) event that some response catches the world by surprise. But if it didn’t, then things could go badly and tragically wrong.

---

From Vox via this RSS feed

Founder and chief economist of American Compass Oren Cass speaks with Vice President JD Vance at the American Compass New World Gala at the National Building Museum on June 3, 2025, in Washington, DC. | Kayla Bartkowski/Getty Images

For more than half a century, the American right has preached the virtues of free markets and low taxes and deregulation. But a new wave of conservative thinkers are now arguing that Republicans have been wrong — or at the very least misguided — about the economy.

This new economic thinking represents a break from what we’ve come to expect from the American right. Its proponents argue for a new strain of economic populism, one that departs from the GOP’s past allegiance to big business and focuses instead on the working class.

The question is, is it for real?

Oren Cass is the founder of the think tank American Compass and the editor of a new book called The New Conservatives. He’s also one of the most influential advocates of this conservative economic populism.

Cass thinks the Republican Party has been too captive to corporate interests and market fundamentalism, and that conservatism needs a major reset, one that embraces American manufacturing and empowers workers.

I invited him onto The Gray Area to talk about this new right-wing populism, what distinguishes it from the left, and whether the Republican Party is serious about adopting it. As always, there’s much more in the full podcast, so listen and follow The Gray Area on Apple Podcasts, Spotify, Pandora, or wherever you find podcasts. New episodes drop every Monday.

This interview has been edited for length and clarity.

Back in 2018, you wrote: “Our political economy has relied upon the insidious metaphor of the economic pie, which measures success by the amount of GDP available to every American for consumption. … But the things America thought she wanted have not made her happy.” Let’s start there: What did we think we wanted, and why hasn’t it made us happy?

You’re very perceptive to start there. We were just putting together this new book called The New Conservatives, which is an anthology of everything we’ve been doing at American Compass over the last five years. And I actually went back and grabbed that essay and made it a prologue to the book. Because exactly as you said, it is a starting point for the way I think about a lot of this.

In my mind, what we saw go wrong in our economics and our politics is that we did come to think of consumption as the end unto itself. And to be clear, I love consumption as much as the next guy. I’m not saying we should go back and live in log cabins, but I think we assumed that as long as we were increasing consumption, as long as material living standards were rising, everybody would be happy and we could declare success. And it’s important to say that, from a formal perspective, that is in fact how our economic models operate.

Economists will tell you their assumption is that the goal of the economic system is to maximize consumption. And so that’s where that economic pie metaphor comes from. Something that was so widely embraced across the political spectrum, across the intellectual spectrum, was this idea that as long as you’re growing the economy, you’re growing GDP, you don’t really have to worry too much about what’s in the pie or where it’s coming from. You can always then chop it up and make sure everybody has lots of pie.

And I think it’s important to say that — and this is the point, that we got what we thought we wanted — it’s important to say that that worked. That for all of the problems we have in this country, if you’re only looking at material living standards, if you’re asking how much stuff people have, how big their houses are, whether they’re air-conditioned, even how much health care they consume, at every socioeconomic level, consumption is up.

We did that. And yet I think it’s also very obvious that that did not achieve what we were trying to achieve, that [it] did not necessarily correspond to human flourishing, did not correspond to a strengthening economy over time, that it certainly did not correspond to strengthening families and communities. And ultimately, it didn’t correspond to a strong and healthy political system or democracy. And so there’s obviously a lot of talk of, Okay, well, why isn’t that right? Why did it go wrong? What do you do about it?

The strange thing for someone like me is that American conservatism, certainly in my lifetime, has largely existed to reinforce the ideology you’re rejecting here. Why do you think the political right has been blind for so long to the things you’re fighting for now?

There’s a very interesting pivot point that you see around the time of the Reagan revolution. The coalition that Reagan assembled had these different elements. It had the social conservatives, who I would say are most closely aligned to a fundamentally conservative outlook on a lot of these questions. But then it brought to that the very libertarian free-market folks on the economic side, and the quite aggressive interventionist foreign policy hawks.

And what all these folks had in common was they really hated communism and really wanted to win the Cold War and saw that as the existential crisis. But what happened is, within that coalition, a very libertarian free-market mindset was then imposed on the economic policy of the right of center, even when that was very much in tension with a lot of other conservative values. And you saw people writing about that from both sides.

From one side, Friedrich Hayek, who is one of the ultimate carriers of this pre-market ideology, has a very famous essay titled “Why I Am Not a Conservative,” emphasizing that what he calls faith in markets to solve problems and self-regulate was very much at odds with how conservatives looked at the world.

And from the flip side, you had a lot of conservatives, folks like Yuval Levin, who prefer markets as a way of ordering the economy to other options, but recognize that markets are very much in tension with other values like family and community. And in some cases, markets even actively can undermine or erode the strength of those other institutions. Markets are also dependent on institutions. If you want markets to work well, you actually need constraints. You need institutional supports. And so that tension was always present.

I think that the coalition made a lot of sense in the context of winning the Cold War. It made a lot of sense when markets in the middle of the late 20th century really did seem to be delivering on a lot of the things that conservatives really cared about. But I think it reached its expiration date and just lived on by inertia into the 2000s, into this era of radical embrace of free trade even with communist China and cutting taxes even in the face of big deficits.

I can imagine a skeptical leftist hearing all of this and thinking it’s just a rebranded democratic socialism. Why is that wrong? What makes this conservative?

There’s a real disconnect both on the ends and on the means. I think there’s a very healthy contestation over what are the appropriate ends that we’re actually building toward. And what you’re seeing conservatives coming back to articulating a set of actual value judgments about, what do we think the good life consists of?

I think there is a set of value judgments and preferences for, in many respects, quite traditional formations at the family level, at the community level. [For] saying that it is not merely a value-neutral choice — “Would you rather get married and have kids or spend more money on vacations in Greece?” — that it is actually appropriate and necessary for the good society to say, No, one of these things is better than the other and more important and should be valued more highly.

At the national level, you’re also seeing a much more robust nationalism on the right of center. Conservatives recognize the importance of the nation and solidarity within the nation to functioning markets, to a functioning society, in a way that at least the modern left tends to resist in a lot of cases.

Part of the case you’re making is that there’s an ongoing paradigm shift within American conservatism. When you look at what this administration is doing on the policy front, when you look at what the Republican Party is doing, do you see them moving in your direction?

We’re definitely moving in the right direction. On tariffs alone, [we could] spend a tremendous amount of time emphasizing the ways I think the problems that they’re addressing, the direction they’re trying to go, is the right one. On the specifics of how things are timed and what the levels are and so forth, what legal authorities you use for what, I have all sorts of thoughts on how it might be done better.

But broadly speaking, to your question about the direction that things are headed, I think it’s extraordinarily clear to me that the Republican Party and the conservative movement are shifting quite dramatically in this direction. One way to look at that is in terms of personnel. Trump has obviously been something of a constant over the last decade in Republican politics, but the distance from Mike Pence to JD Vance is pretty dramatic.

The distance from [Secretaries of State] Rex Tillerson to Marco Rubio is pretty dramatic. The distance from the various secretaries of labor in the first term to a secretary of labor recommended by the Teamsters is pretty dramatic.

Is it really, though? Rhetorically, yes. But substantively? If you want to know why I can’t take this iteration of the GOP seriously, look at the domestic policy they just passed in the House. It’s the same Republican Party. It’s jammed up with a bunch of stuff that reflects conventional conservative priorities.

It’s not doing a whole lot to help working-class people. It’s more tax cuts offset by more cuts to Medicaid and food stamps, which low-income people depend on. And the net result, as always, will be more upward redistribution of wealth. And on top of that, another $3 or $4 or $5 trillion tacked onto the deficit just for good measure. How can you look at that and feel like the GOP is genuinely pivoting in your direction?

I’ve been extremely critical of the “big, beautiful bill” **— particularly of the deficit element —**because I think if one is going to be a fiscal conservative, one has to not be adding to deficits right now. But a lot of the efforts to argue that things are not changing in the Republican Party strike me as a real disservice to people who are trying to understand where things are going. Elected political leaders are always going to be the lagging indicator of what’s happening in any political party or political movement. They are by definition going to be the oldest, the ones who have been around the longest, the ones who have built their careers and ideologies and relationships around what was happening 20 or 30 years ago.

And so if one wants to know what is passing in Congress today, then yes, you count the votes of the people in Congress today. If you want to know what’s actually moving within a party or what’s going to happen over a 10- or 15-year period, counting the votes today is just not what someone in good faith trying to understand the direction would do.

The tariff regime, the trade war — that is a genuine shift. No doubt about it. It’s not entirely clear to me how that helps poor and working-class people at the moment, but maybe I’m not seeing the whole picture.

There’s a very interesting economic debate to be had about whether it will work. I obviously have one very strong view. But it seems pretty clear to me that what they are trying to do is quite explicitly focused on the economic interests of workers.

Another very interesting area — I mentioned some of the things that are going on on the labor front. One really interesting effort that’s underway, and [Sen.] Josh Hawley is the leader of it, but Bernie Moreno, the new senator from Ohio, is the co-sponsor of it — they’ve taken the [proposed] PRO Act, which is the ultimate Democratic wish list of labor reforms, and they’ve chopped it up.

And they’ve said, Look, some of these are perfectly legitimate and good ideas. Others of these we don’t agree with. And we’re going to start advancing the ones we think are good ideas. That’s a dramatic shift in how you would see the Republican Party.

I think you’re seeing the same thing in the financial sector. There was a great example recently where a private equity firm that had bought out a bunch of paper plants was trying to shut down a paper plant in Ohio. And you literally had the Republican politicians out there at the rally with the union leaders, forcing a change and a commitment to at least keep the plant open for the rest of the year and try to find a transaction that would keep it open afterward.

On family policy, in 2017 you had [then-Sens.] Marco Rubio and Mike Lee threatening to tank the entire tax cut bill to get an expanded child tax credit in it. Now it is an uncontroversial top priority that the child tax credit is not only kept at that level, but expanded further. And so even at the level of what is happening in legislation, it’s clear that this is a very different party from 2017. If you look at who Trump has appointed, it’s a very different set of appointments.

If you look at the critical mass and sometimes center of gravity among the younger elected officials, the people coming into the Senate, it’s a completely different set of priorities and policies from those who have been there for a long time.

Like I said, I’m not convinced that the DNA of the party has changed, but I will grant that there are indications of a shift. I don’t know what it’s going to amount to, materially, but this is not the party of Mitt Romney.

I think Trump has cultivated a very unique coalition, certainly much more working-class than the pre-Trump Republican Party. I don’t know how much of that coalition is a function of Trump and how much of that coalition will fade when he fades. If the Republican Party does prove an unreliable vehicle for your movement, can you see a world in which you’re working with Democrats?

We do work with some Democrats. I think there are Democrats who are doing very good and interesting work. We recently had [Rep.] Jared Golden from Maine on the American Compass Podcast because he is the sponsor of the 10 percent global tariff legislation in Congress. One thing I always emphasize is that I think a healthy American politics is not one where one party gets everything right and dominates and the other one collapses into irrelevance.

It’s one where we actually have two healthy political parties that are both focused on the concerns and priorities of the typical American and are then contesting a lot of these very legitimate disagreements about ends and means. But based on what is happening in American politics today and the fundamental differences between conservatism and progressivism, I would expect that this is going to have the most success and salience and overlap in thinking on the right of center.

Listen to the rest of the conversation and be sure to follow The Gray Area on Apple Podcasts, Spotify, Pandora, or wherever you listen to podcasts.

---

From Vox via this RSS feed

President Donald Trump addresses the nation, alongside Vice President JD Vance, Secretary of State Marco Rubio, and Secretary of Defense Pete Hegseth, from the White House in Washington, DC, on June 21, 2025. | Carlos Barria/Pool/AFP via Getty Images

President Donald Trump claimed during his 2024 campaign for president that America had fought “no wars” during his first presidency, and that he was the first president in 72 years who could say that.

This was not, strictly speaking, true. In his first term, Trump intensified the air war against ISIS in Iraq and Syria, ordered airstrikes against Bashar al-Assad’s Syrian regime in response to chemical weapons use, and escalated a little-noticed counterinsurgency campaign in Somalia. But in those cases, Trump could say, with some justification, that he was just dealing with festering crises he had inherited from Barack Obama.

Likewise, the president has repeatedly claimed that the wars in Gaza and Ukraine never would have happened had he been president when they broke out, rather than Joe Biden. That’s a counterfactual that is impossible to prove, and he may have been overly optimistic in his promises to quickly negotiate an end to both those conflicts, but it’s fair to say that both are wars Trump inherited rather than chose.

This time, it’s different. This time, it’s Trump’s war.

On Saturday night, the United States bombed three nuclear sites in Iran at Fordow, Natanz, and Esfahan, ending weeks of speculation about whether the US military would join the Israeli war on Iran that began more than a week ago.

The past few days in Washington have felt a bit like the battles over intelligence in the lead-up to the war in Iraq, but run in fast-forward. Rather than pressuring intelligence agencies to justify his preferred course of action, Trump has simply overruled them. Rather than building a case before Congress and the UN for the need to act, he’s simply ignored them.

Trump argued that Iran brought the attack on itself by not taking the deal he was offering, but negotiations were ongoing at the time Trump abandoned the diplomatic path. Trump endorsed the Israeli assessment that war was necessary because new information showed Iran was “very close to having a weapon.” But this contradicts the very recent statements from his own intelligence agencies and director of national intelligence. According to the Wall Street Journal’s reporting, officials in these agencies were not convinced by Israel’s new evidence that something dramatic had changed in Iran’s nuclear program. It also contradicts Trump’s own statements from earlier this month when he publicly discouraged Israel from attacking Iran, saying it would derail his efforts to negotiate a new nuclear deal.

It’s hard to overstate just how fast the Trump administration’s policy has shifted. Just a month ago, Trump appeared to be giving Netanyahu’s government the cold shoulder, pursuing direct diplomacy with Israel’s staunchest enemies — including Iran — and cozying up to governments in the Gulf that plainly had no appetite for a new war.

Now, Trump has not only endorsed Netanyahu’s war; he has joined it and boasted in his brief statement from the White House on Saturday that the two had worked as a team like “perhaps no team has ever worked before.” He ended his speech with “God bless Israel” along with “God bless America.”

Tonight was also a major blow to those on the right, as well as some on the left, who hoped that the Trump administration would usher in either a new era of military restraint or a shift in priorities away from the Middle East toward China. (The US has now relocated military assets from Asia for this war.)

There’s still a lot we don’t know, but it’s fair at this point to say that this is a war of Trump’s choosing.

Trump’s extraordinary gamble

In his statement from the White House on Saturday night, Trump said that the operation had been a “spectacular military success” and that the enrichment facilities had been “totally obliterated.” For the moment, we don’t have corroborating evidence of that.

Israel had mostly avoided striking these sites itself. Only the US has the powerful GBU-57 “bunker buster” bombs that can destroy Iran’s most secure nuclear sites, particularly the underground uranium enrichment facility at Fordow, and only the US has the aircraft that can carry them.

US officials told the New York Times that US bombers dropped a dozen bunker busters on Fordow on Saturday. Many experts believe the facility would be difficult to destroy and require multiple strikes, even with those bombs. Doubts about whether Fordow could be destroyed were reportedly one reason why Trump hesitated in ordering these strikes.

In his statement, Trump also implied that this was a one-off operation for now. Speaking of the pilots who dropped the bombs, Trump said, “hopefully we will no longer need their services at this capacity,” but also threatened that if Iran did not “make peace,” then “future attacks will be far greater and a lot easier.” He added: “There are many targets left.”

The hope appears to be that Iran will now be forced to cut a deal to entirely give up its nuclear program. But an Iranian regime mindful of its own legitimacy is also likely to retaliate in some form, possibly by targeting some of the roughly 40,000 US troops deployed around the Middle East.

The hope may be that these will be limited tit-for-tat strikes like those that followed the US assassination of Gen. Qassem Soleimani in 2020, though subsequent assessments have found that those attacks did more damage than was initially thought and could easily have killed far more US troops.

In any event, the Iranian regime is far more desperate now, and once the missiles start flying, it could get very easy for things to escalate out of control.

If Iran has any remaining enrichment infrastructure, either at these sites or hidden elsewhere throughout the country, the country’s leaders may now feel far less hesitation about rushing to build a bomb. There was long a view that Iran’s leaders preferred to remain a “threshold nuclear state” — working toward a bomb without actually building one. In this view, they believed that their growing capacity to build a weapon gave them leverage, while not actually trying to build one avoided US and Israeli intervention. That logic is now obsolete.

It’s also not clear that Israel simply wants nuclear concessions from the Iranian regime. While Prime Minister Benjamin Netanyahu said that new intelligence about Iran’s nuclear capabilities was the reason for starting this war, it’s been clear both from the Israeli government’s rhetoric and choice of targets that this is a war against the Islamic Republic itself, and that regime change may be the ultimate goal. Trump didn’t mention regime change in his statement, but he has now committed American military power to that Israeli war.

So far, this war has been characterized by stunning Israeli tactical successes, as well as the seeming impotence of Iran and its once vaunted network of regional proxies in its response. (Though it’s unclear how long Israel’s air defense system can keep up if Iranian strikes continue at this pace.) This may have emboldened a president who has backed off of actions like this in the past, convincing him that striking Iran’s nuclear program now would be effective and that the blowback would be manageable.

It’s quite a gamble – and this time he will have no one else to blame if it doesn’t go as planned.

---

From Vox via this RSS feed

A You’re Hot, Stay Cool sign with an AC unit and fan posted to a street light during a heat wave on 86th Street in Manhattan, New York. | Lindsey Nicholson/UCG/Universal Images Group via Getty Images

Lee Kuan Yew, the iron-willed founder of modern Singapore, was once asked what the most important invention of the 20th century was. He didn’t say penicillin, which has saved over 500 million lives, or the nuclear bomb, which has shaped geopolitics like nothing before. He didn’t even say TV!

Instead, Lee had a simple two-word answer: “Air conditioning.” Without air conditioning, Singapore, where temperatures regularly reach into the 90s with tropical humidity levels, would never have developed from a tiny city-state with a per-capita GDP that was a third of Western Europe’s in 1960 to one of the most prosperous countries in the world.

Air conditioning is as essential to the modern world as the internet itself. But like the internet, A/C gets a bad rap. Cooling already eats up 10 percent of global electricity, and demand from air conditioners is expected to triple by 2050 without tougher energy efficiency standards. Many units still use refrigerant gases that produce a planetary warming effect that is thousands of times that of a similar amount of CO2.

Air conditioning is also a physical manifestation of the energy gap between the rich who can afford it, and the poor who must sweat. It has enabled the development of energy-intensive cities in places where humans just shouldn’t live, like Phoenix. Fundamentally, A/C is seen by some as an unnecessary luxury, a prime example of a “harmful habit of consumption,” as Pope Francis once put it.

I get the point. It seems morally wrong for so many of us to use a device that contributes about 3 percent of global greenhouse gas emissions — only so we can escape the effects of that warming.

But “seems” is not the same as “is.” Air conditioning has become far more than a luxury. So on this, the second day of summer, when the East Coast is about to be enveloped by a truly suffocating wave of heat and humidity, I offer five reasons why we should be grateful for air conditioning.

It saves lives

Heat isn’t just uncomfortable. It’s dangerous, killing more Americans in a typical year than any other form of extreme weather. Access to air conditioning can mean the difference between life and death. Seven hundred and thirty-nine people died in the great Chicago heat wave of 1995, but having a working air conditioner reduced the risk of death by 80 percent. Another study looked at cities in multiple countries between 1972 and 2009 and found that more air conditioning helped reduce excess heat deaths.

As a 2021 review in the Lancet explained it, air conditioning “is set to become the most prevalent strategy worldwide for coping with hot weather and heat extremes.” And while only about 8 percent of the 2.8 billion people living in the world’s hottest regions have A/C at home, that’s an argument for closing the A/C gap — not an argument against the very real value of air conditioning.

It keeps us working

If you struggle to concentrate when the heat and humidity is high, you’re not alone. One study looked at office work and found that productivity begins to decline around 73 to 75 degrees Fahrenheit, while at 86°F, performance falls by almost 9 percent. Another study found that every 1 degree increase in average classroom temperature over a school year corresponded to a roughly 1 percent loss in students’ expected learning — but installing air conditioning eliminates about three-quarters of that effect.

As temperatures continue to increase, the importance of air conditioning in schools and businesses will only grow. A 2016 working paper finds that widespread adoption of air conditioning — especially by the most productive plants — substantially offsets the heat-induced drop in US manufacturing output, making cooling a critical adaptation tool.

It helps us sleep

The more we learn about sleep, the more important it appears to be — and keeping cool is a key part of a decent night’s sleep.

Humans fall asleep fastest around 64–68°F, while temperatures above 75°F cause vital deep sleep and REM sleep to crater. A 2024 review of more than 50 lab and field studies found that bedroom cooling increased total sleep time 15 to 20 minutes and cut the total amount of time people spent awake after falling asleep by a third.

It’s given us everything from the movies to microchips

Do you like going to the movie theater to catch a summer blockbuster? Well, you can thank air conditioning — before its invention, movie attendance always dropped during the hot summer months. It’s no coincidence that the first public air conditioner was installed in a cinema, New York’s Rivoli Theater, in 1925.

But maybe you prefer to take in your movies in the comfort of your own home? Well, producing the microchips that go into your streaming TV or smartphone requires total precision in temperature control and humidity. In short: no A/C, no microchips.

It lets millions live and travel where they want

Look, my negative feelings about living in red-hot metros like Phoenix are a matter of public record. But I am clearly in the minority: Americans love to live in hot places. Maricopa County, where Phoenix is located, added 1.2 million people between 2013 and 2023, more than any other county — and 96 percent of the new housing built to absorb those new residents comes with A/C.

What US cities like Phoenix or Houston or Atlanta have in common with Singapore and Hong Kong is that none of them would exist as anything like they are today without the widespread use of air conditioning. Before A/C, the American South was mired in poverty, far behind the rest of the country. After A/C, the South more than caught up, and the otherwise uninhabitable Southwest became a magnet for people. If you think it’s good that people can choose from a wider spectrum of places — and I do — A/C is one of the main reasons why that’s possible.

Air conditioning as it exists today is far from perfect. But it’s also necessary, especially in an ever-warming world. What we need is not less air conditioning — unless you happen to work at an office where they keep the temperature at 60°F — but better air conditioning, with more efficient units powered by cleaner electricity.

If you want to go without A/C, go right ahead (though I probably won’t be visiting your house in the summer anytime soon). But either way, it should be a choice.

A version of this story originally appeared in the Good News newsletter. Sign up here!

---

From Vox via this RSS feed

A humanoid robot shakes hands with a visitor at the Zhiyuan Robotics stand at the Shanghai New International Expo Centre in Shanghai, China, on June 18, 2025, during the first day of the Mobile World Conference. | Ying Tang/NurPhoto via Getty Images

In 2023, one popular perspective on AI went like this: Sure, it can generate lots of impressive text, but it can’t truly reason — it’s all shallow mimicry, just “stochastic parrots” squawking.

At the time, it was easy to see where this perspective was coming from. Artificial intelligence had moments of being impressive and interesting, but it also consistently failed basic tasks. Tech CEOs said they could just keep making the models bigger and better, but tech CEOs say things like that all the time, including when, behind the scenes, everything is held together with glue, duct tape, and low-wage workers.

It’s now 2025. I still hear this dismissive perspective a lot, particularly when I’m talking to academics in linguistics and philosophy. Many of the highest profile efforts to pop the AI bubble — like the recent Apple paper purporting to find that AIs can’t truly reason — linger on the claim that the models are just bullshit generators that are not getting much better and won’t get much better.

But I increasingly think that repeating those claims is doing our readers a disservice, and that the academic world is failing to step up and grapple with AI’s most important implications.

I know that’s a bold claim. So let me back it up.

“The illusion of thinking’s” illusion of relevance

The instant the Apple paper was posted online (it hasn’t yet been peer reviewed), it took off. Videos explaining it racked up millions of views. People who may not generally read much about AI heard about the Apple paper. And while the paper itself acknowledged that AI performance on “moderate difficulty” tasks was improving, many summaries of its takeaways focused on the headline claim of “a fundamental scaling limitation in the thinking capabilities of current reasoning models.”

For much of the audience, the paper confirmed something they badly wanted to believe: that generative AI doesn’t really work — and that’s something that won’t change any time soon.

The paper looks at the performance of modern, top-tier language models on “reasoning tasks” — basically, complicated puzzles. Past a certain point, that performance becomes terrible, which the authors say demonstrates the models haven’t developed true planning and problem-solving skills. “These models fail to develop generalizable problem-solving capabilities for planning tasks, with performance collapsing to zero beyond a certain complexity threshold,” as the authors write.

That was the topline conclusion many people took from the paper and the wider discussion around it. But if you dig into the details, you’ll see that this finding is not surprising, and it doesn’t actually say that much about AI.

Much of the reason why the models fail at the given problem in the paper is not because they can’t solve it, but because they can’t express their answers in the specific format the authors chose to require.

If you ask them to write a program that outputs the correct answer, they do so effortlessly. By contrast, if you ask them to provide the answer in text, line by line, they eventually reach their limits.

That seems like an interesting limitation to current AI models, but it doesn’t have a lot to do with “generalizable problem-solving capabilities” or “planning tasks.”

Imagine someone arguing that humans can’t “really” do “generalizable” multiplication because while we can calculate 2-digit multiplication problems with no problem, most of us will screw up somewhere along the way if we’re trying to do 10-digit multiplication problems in our heads. The issue isn’t that we “aren’t general reasoners.” It’s that we’re not evolved to juggle large numbers in our heads, largely because we never needed to do so.

If the reason we care about “whether AIs reason” is fundamentally philosophical, then exploring at what point problems get too long for them to solve is relevant, as a philosophical argument. But I think that most people care about what AI can and cannot do for far more practical reasons.

AI is taking your job, whether it can “truly reason” or not

I fully expect my job to be automated in the next few years. I don’t want that to happen, obviously. But I can see the writing on the wall. I regularly ask the AIs to write this newsletter — just to see where the competition is at. It’s not there yet, but it’s getting better all the time.

Employers are doing that too. Entry-level hiring in professions like law, where entry-level tasks are AI-automatable, appears to be already contracting. The job market for recent college graduates looks ugly.

The optimistic case around what’s happening goes something like this: “Sure, AI will eliminate a lot of jobs, but it’ll create even more new jobs.” That more positive transition might well happen — though I don’t want to count on it — but it would still mean a lot of people abruptly finding all of their skills and training suddenly useless, and therefore needing to rapidly develop a completely new skill set.

It’s this possibility, I think, that looms large for many people in industries like mine, which are already seeing AI replacements creep in. It’s precisely because this prospect is so scary that declarations that AIs are just “stochastic parrots” that can’t really think are so appealing. We want to hear that our jobs are safe and the AIs are a nothingburger.

But in fact, you can’t answer the question of whether AI will take your job with reference to a thought experiment, or with reference to how it performs when asked to write down all the steps of Tower of Hanoi puzzles. The way to answer the question of whether AI will take your job is to invite it to try. And, uh, here’s what I got when I asked ChatGPT to write this section of this newsletter:

Is it “truly reasoning”? Maybe not. But it doesn’t need to be to render me potentially unemployable.

“Whether or not they are simulating thinking has no bearing on whether or not the machines are capable of rearranging the world for better or worse,” Cambridge professor of AI philosophy and governance Harry Law argued in a recent piece, and I think he’s unambiguously right. If Vox hands me a pink slip, I don’t think I’ll get anywhere if I argue that I shouldn’t be replaced because o3, above, can’t solve a sufficiently complicated Towers of Hanoi puzzle — which, guess what, I can’t do either.

Critics are making themselves irrelevant when we need them most

In his piece, Law surveys the state of AI criticisms and finds it fairly grim. “Lots of recent critical writing about AI…read like extremely wishful thinking about what exactly systems can and cannot do.”

This is my experience, too. Critics are often trapped in 2023, giving accounts of what AI can and cannot do that haven’t been correct for two years. “Many [academics] dislike AI, so they don’t follow it closely,” Law argues. “They don’t follow it closely so they still think that the criticisms of 2023 hold water. They don’t. And that’s regrettable because academics have important contributions to make.”

But of course, for the employment effects of AI — and in the longer run, for the global catastrophic risk concerns they may present — what matters isn’t whether AIs can be induced to make silly mistakes, but what they can do when set up for success.

I have my own list of “easy” problems AIs still can’t solve — they’re pretty bad at chess puzzles — but I don’t think that kind of work should be sold to the public as a glimpse of the “real truth” about AI. And it definitely doesn’t debunk the really quite scary future that experts increasingly believe we’re headed toward.

A version of this story originally appeared in the Future Perfect newsletter. Sign up here!

---

From Vox via this RSS feed

A protester wearing a Trump paper mâché head stands in front of a barricade and holds a sign that reads, “Death and taxes” in New York in 2019.

Four years ago, America was on the cusp of the largest expansion of its welfare state since the 1960s.

Under Joe Biden in 2021, House Democrats passed legislation that would have established a monthly child allowance for most families, an expansion of Medicaid’s elder care services, federal child care subsidies, universal prekindergarten, and a paid family leave program, among other new social benefits.

But that bill failed — and then, so did Biden’s presidency.

Now, Republicans are on the brink of enacting the largest cut to public health insurance in American history. And the outlook for future expansions of the safety net looks dimmer than at any time in recent memory.

There are two primary reasons why progressives’ prospects for growing the welfare state have darkened.

This story was first featured in The Rebuild.

Sign up here for more stories on the lessons liberals should take away from their election defeat — and a closer look at where they should go next. From senior correspondent Eric Levitz.

First (and most straightforwardly), the Democrats are not well-positioned to win full control of the federal government anytime soon. To win a Senate majority in 2026, the party would need to win multiple states that Trump carried by double digits last year. And the 2028 map isn’t that much better. The basic problem is that Democrats have built a coalition that’s heavily concentrated on the coasts and thus, systematically underrepresented in the Senate. To win the robust congressional majorities typically necessary for enacting large social programs, Democrats would likely need to transform their party’s brand.

Second, although Democrats developed grander ambitions for social spending over the past decade, they simultaneously grew more averse to raising taxes on anyone but the super-rich. In the 2010s, when inflation and interest rates were persistently low, the party could paper over this tension with deficit spending. But Biden-era inflation revealed the limits of this strategy.

And if Congress passes President Donald Trump’s tax cut plan, then interest rates and inflationary risk are likely to remain elevated for years, while the cost of servicing America’s debts will soar. Add to this the impending exhaustion of Social Security’s trust fund, and space for new welfare programs is likely to be scant, unless Democrats find a way to enact broad-based tax increases.

Liberals could respond to all this by paring back their ambitions for the welfare state, while seeking to advance progressive goals through regulatory policy. It is perhaps not a coincidence that the two most prominent policy movements in Democratic circles today — the anti-monopoly and “abundance” crusades — are both principally concerned with reforms that require no new tax revenue (antitrust enforcement in the former case, zoning liberalization in the latter).

But expanding America’s safety net remains a moral imperative. In the long-term, Democrats must therefore strive to build the electoral power and political will necessary for raising taxes on the middle-class (or at least, on its upper reaches).

Democrats like social welfare programs. But they like low taxes on the upper middle-class even more.

Over the course of the 2010s, the Democratic leadership’s appetite for new social spending grew. Bernie Sanders’s insurgent campaigns in 2016 and 2020 put Medicare-for-All at the center of the party’s discourse, and moved its consensus on the welfare state sharply leftward. In the latter primary, even the Democrats’ most moderate contender — Joe Biden — vowed to establish a public option for health insurance and tuition-free community colleges, among other social programs.

Biden’s agenda only grew more ambitious upon taking office. No president since Lyndon B. Johnson had proposed a more sweeping expansion of social welfare than the Build Back Better Act.

And yet, while Democrats’ aspirations for social spending had become historically bold, the party’s position on taxes had grown exceptionally timid. In 2016, Hillary Clinton had promised not to raise taxes on any American family earning less than $250,000. Four years later, Biden vowed to spare all households earning less than $400,000 – despite the fact that tax rates on upper middle-class families had fallen during Trump’s first term.

Meanwhile, the Democrats’ congressional leadership was actually pushing to cut taxes on rich blue state homeowners by increasing the state and local income tax deduction.

In other words: In 2021, Democrats were promising to establish an unprecedentedly large welfare state, while keeping taxes on 98 percent of households historically low.

Officially, the party believed that it could square this circle by soaking the super-rich. After all, America’s highest-earning 1 percent had commandeered more than 20 percent of the nation’s annual income. The government could therefore extract a lot of revenue by merely shaking down the upper class.

In reality, though, Biden’s vision was also premised on the assumption that America could deficit-finance new spending with little risk of sparking inflation or high interest rates.

The Build Back Better Act did not actually raise taxes on the rich by enough to offset its social spending. Instead, Democrats leaned on budget gimmicks to “pay for” its agenda: Although the party intended the law’s new programs to be permanent, it scheduled many of them to expire after just a few years, so as to make the policies look cheaper over a decade-long budget window. Absent these arbitrary expiration dates, the bill would have added $2.8 trillion to the deficit over a decade. Even as written, the law would have increased deficits by $749 billion in its first five years.

More fundamentally, Biden’s basic fiscal objective — to establish wide-ranging social benefits through taxes on the super rich alone — only made sense in a world of low inflation.

Western Europe’s robust welfare states are all funded through broad-based taxation. This is partly because administering a large safety net requires managing economic demand. When the government expands its provision of elder care, social housing, child care, and pre-K, it increases overall demand for workers and resources in the economy. And if the supply of labor and materials doesn’t rise in line with this new demand, then inflation can ensue.

Taxes effectively “pay for” new spending by freeing up such resources. When households see their post-tax income decline, they’re often forced to make fewer discretionary purchases. Raise taxes on an upper middle-class family and it might need to postpone its dreams of a lake house. That in turn frees up labor for public programs: The fewer construction workers needed to build vacation homes, the more that will be available to build affordable housing.

But soaking the extremely rich does less to dampen demand than taxing the upper middle-class does. Even if you increase Elon Musk’s tax rate by 50 percent, he won’t actually need to reduce his consumption at all — the billionaire will still have more money than he can spend in a lifetime.

The same general principle applies to multimillionaires, albeit to a lesser extent: Raise their taxes, and they’re liable to save less money, but won’t necessarily consume fewer resources. And if they do not curb their consumption in response to a tax hike, then that tax hike will not actually free up resources.

In 2021, Democrats felt no obligation to sweat these details. For nearly a decade after the Great Recession, economic demand had been too low. Workers and materials had stood idle on the economy’s sidelines, as there wasn’t enough spending to catalyze their employment. In that context, unfunded welfare benefits can boost growth without generating inflation.

But as Democrats moved Build Back Better through Congress, the macroeconomic terrain shifted beneath their feet. Biden likely would have struggled to get his social agenda through the Senate (where Democrats held only 50 votes) even in the absence of 2022’s inflation. But that surge in prices all but guaranteed the legislation’s defeat: Suddenly, it became clear that the government could not increase economic demand without pushing up inflation and interest rates. America had returned to a world of fiscal constraints.

Unfortunately, those constraints could prove lasting, especially if Donald Trump’s tax agenda makes it into law.

Building a comprehensive welfare state is about to get harder

The most lamentable aspect of Trump’s “Big Beautiful Bill” are its cuts to healthcare and food assistance for the poor. Yet even as it takes health insurance from 10 million Americans and reduces food assistance to low-income families by about $100 a month, the legislation would add $2.4 trillion to the debt over the coming decade, according to the Congressional Budget Office.

Yet the actual cost of the GOP’s fiscal vision is even larger. To reduce their bill’s price tag, Republicans’ set some of their tax cuts to arbitrarily expire. Were these tax cuts made permanent, the bill would add roughly $5 trillion to the deficit over the next 10 years.

This is likely to render the US economy more vulnerable to inflation and high interest rates in the future.

Thus, the next Democratic government probably won’t have much freedom to deficit spend without increasing Americans’ borrowing costs or bills. Meanwhile, if that administration holds power after 2032, it will also need to find a ton of new revenue, just to maintain America’s existing welfare state.

Social Security currently pays out more in benefits than it takes in through payroll taxes. For now, the program’s dedicated trust fund fills in the gap. But in 2033, that fund will likely be exhausted, according to government projections. At that point, the government will need to find upward of $414.5 billion in new revenue, each year, to maintain existing Social Security benefits without increasing the deficit.

Given Democrats’ current stance on taxes, the imperative to keep Social Security funded would likely crowd out the rest of the party’s social welfare agenda. Indeed, merely sustaining Americans’ existing retirement benefits would almost certainly require raising taxes on households earning less than $400,000. Maintaining such benefits while also creating new welfare programs — in a context of structurally high deficits and interest rates — would plausibly entail large, broad-based tax increases, the likes of which today’s Democrats scarcely dare to contemplate.

Granted, the robots could solve all this

To be sure, it is possible that technological progress could render this entire analysis obsolete. Some analysts expect artificial intelligence to radically increase productivity over the next decade, while devaluing white-collar labor. This could slow the pace of wage and price growth, while turbo-charging income inequality.

In a world where robots can instantly perform work that presently requires millions of humans, America could plausibly finance a vast social welfare state solely through taxes on capital.

But until AI actually yields a discernible leap in productivity, I don’t think it is safe to take an impending robo-utopia as a given.

Democrats eventually need to sell Americans on higher taxes

Democrats probably can’t escape the tension between their commitments on taxation and social spending. But they can seek to mitigate it in a few different ways.

One is to scale down the party’s ambitions for the welfare state, while seeking to advance progressive economic goals through other means.

Such a retreat would be understandable. The party’s fear of raising taxes is not baseless. In a 2021 Gallup poll, only 19 percent of Americans said they would like to have more government services in exchange for higher taxes, while 50 percent said they’d prefer lower taxes in exchange for fewer services.

Meanwhile, Democrats have grown increasingly reliant on the support of upper middle-class voters. In 2024, the highest-earning 5 percent of white voters were more than 10 percentage points more Democratic than America as a whole. The lowest earning two-thirds of whites, by contrast, were more Republican than the nation writ large.

View Link

In this political environment, calling for large middle-class tax hikes could well ensure perpetual Republican rule.

In the short term then, Democrats might therefore be wise to narrow their agenda for social welfare, focusing on modest programs that can be funded exclusively with taxes on the rich.

At the same time, the party could seek to better working people’s lot through regulatory policy. You don’t need to raise middle-class taxes to expand collective bargaining rights, guarantee worker representation on corporate boards, or raise the minimum wage. And the same can be said of relaxing regulatory barriers to housing construction and energy infrastructure. (Of course, achieving any of these goals federally would require Democrats to win a robust Senate majority — one sufficiently large and progressive enough to abolish the legislative filibuster, which currently establishes a 60-vote threshold for enacting new, non-budgetary legislation.)

In the long run though, Democrats must not forfeit the pursuit of a comprehensive welfare state. America lets more of its children suffer poverty — and more of its adults go without health insurance — than similarly rich countries. These deprivations are largely attributable to our nation’s comparatively threadbare safety net. And they can only be fully eliminated through redistributive policy. A higher minimum wage will not ensure that children with unemployed parents never go hungry, or that every worker with cancer can afford treatment.

Furthermore, as technological progress threatens to rapidly disemploy large segments of the public, robust unemployment insurance is as important as ever. And as the population ages, increasing investment in eldercare will be increasingly imperative.

Democrats should seek to make incremental progress on all these fronts as soon as possible. Even if the party is only willing to tax the rich, it can still finance targeted anti-poverty spending. But absent an AI-induced productivity revolution, building a holistic welfare state will require persuading the middle-class to accept higher taxes.

How this can be done is not clear. But part of the solution is surely to demonstrate that Democratic governments can spend taxpayer funds efficiently and effectively. So long as blue areas struggle to build a single public toilet for less than $1.7 million — or a high-speed rail line in less than 17 years — it will be hard to persuade ordinary Americans to forfeit a larger chunk of their paychecks to Uncle Sam.

All this said, Democrats have plenty of time to debate the future of fiscal policy. In the immediate term, the party’s task is plain: to do everything in its power to prevent Trump’s cuts to Medicaid and food assistance from becoming law.

The path to a comprehensive welfare state won’t be easy to traverse. Better then not to begin the journey toward it by taking several steps backward.

---

From Vox via this RSS feed

Yoshua Bengio.

The science fiction author Isaac Asimov once came up with a set of laws that we humans should program into our robots. In addition to a first, second, and third law, he also introduced a “zeroth law,” which is so important that it precedes all the others: “A robot may not injure a human being or, through inaction, allow a human being to come to harm.”

This month, the computer scientist Yoshua Bengio — known as the “godfather of AI” because of his pioneering work in the field — launched a new organization called LawZero. As you can probably guess, its core mission is to make sure AI won’t harm humanity.

Even though he helped lay the foundation for today’s advanced AI, Bengio is increasingly worried about the technology over the past few years. In 2023, he signed an open letter urging AI companies to press pause on state-of-the-art AI development. Both because of AI’s present harms (like bias against marginalized groups) and AI’s future risks (like engineered bioweapons), there are very strong reasons to think that slowing down would have been a good thing.

But companies are companies. They did not slow down. In fact, they created autonomous AIs known as AI agents, which can view your computer screen, select buttons, and perform tasks — just like you can. Whereas ChatGPT needs to be prompted by a human every step of the way, an agent can accomplish multistep goals with very minimal prompting, similar to a personal assistant. Right now, those goals are simple — create a website, say — and the agents don’t work that well yet. But Bengio worries that giving AIs agency is an inherently risky move: Eventually, they could escape human control and go “rogue.”

So now, Bengio is pivoting to a backup plan. If he can’t get companies to stop trying to build AI that matches human smarts (artificial general intelligence, or AGI) or even surpasses human smarts (artificial superintelligence, or ASI), then he wants to build something that will block those AIs from harming humanity. He calls it “Scientist AI.”

Scientist AI won’t be like an AI agent — it’ll have no autonomy and no goals of its own. Instead, its main job will be to calculate the probability that some other AI’s action would cause harm — and, if the action is too risky, block it. AI companies could overlay Scientist AI onto their models to stop them from doing something dangerous, akin to how we put guardrails along highways to stop cars from veering off course.

I talked to Bengio about why he’s so disturbed by today’s AI systems, whether he regrets doing the research that led to their creation, and whether he thinks throwing yet more AI at the problem will be enough to solve it. A transcript of our unusually candid conversation, edited for length and clarity, follows.

Sigal Samuel

When people express worry about AI, they often express it as a worry about artificial general intelligence or superintelligence. Do you think that’s the wrong thing to be worrying about? Should we only worry about AGI or ASI insofar as it includes agency?

Yoshua Bengio

Yes. You could have a superintelligent AI that doesn’t “want” anything, and it’s totally not dangerous because it doesn’t have its own goals. It’s just like a very smart encyclopedia.

Sigal Samuel

Researchers have been warning for years about the risks of AI systems, especially systems with their own goals and general intelligence. Can you explain what’s making the situation increasingly scary to you now?

Yoshua Bengio

In the last six months, we’ve gotten evidence of AIs that are so misaligned that they would go against our moral instructions. They would plan and do these bad things — lying, cheating, trying to persuade us with deceptions, and — worst of all — trying to escape our control and not wanting to be shut down, and doing anything [to avoid shutdown], including blackmail. These are not an immediate danger because they’re all controlled experiments…but we don’t know how to really deal with this.

Sigal Samuel

And these bad behaviors increase the more agency the AI system has?

Yoshua Bengio

Yes. The systems we had last year, before we got into reasoning models, were much less prone to this. It’s just getting worse and worse. That makes sense because we see that their planning ability is improving exponentially. And [the AIs] need good planning to strategize about things like “How am I going to convince these people to do what I want?” or “How do I escape their control?” So if we don’t fix these problems quickly, we may end up with, initially, funny accidents, and later, not-funny accidents.

That’s motivating what we’re trying to do at LawZero. We’re trying to think about how we design AI more precisely, so that, by construction, it’s not even going to have any incentive or reason to do such things. In fact, it’s not going to want anything.

Sigal Samuel

Tell me about how Scientist AI could be used as a guardrail against the bad actions of an AI agent. I’m imagining Scientist AI as the babysitter of the agentic AI, double-checking what it’s doing.

Yoshua Bengio

So, in order to do the job of a guardrail, you don’t need to be an agent yourself. The only thing you need to do is make a good prediction. And the prediction is this: Is this action that my agent wants to do acceptable, morally speaking? Does it satisfy the safety specifications that humans have provided? Or is it going to harm somebody? And if the answer is yes, with some probability that’s not very small, then the guardrail says: No, this is a bad action. And the agent has to [try a different] action.

Sigal Samuel

But even if we build Scientist AI, the domain of “What is moral or immoral?” is famously contentious. There’s just no consensus. So how would Scientist AI learn what to classify as a bad action?

Yoshua Bengio

It’s not for any kind of AI to decide what is right or wrong. We should establish that using democracy. Law should be about trying to be clear about what is acceptable or not.

Now, of course, there could be ambiguity in the law. Hence you can get a corporate lawyer who is able to find loopholes in the law. But there’s a way around this: Scientist AI is planned so that it will see the ambiguity. It will see that there are different interpretations, say, of a particular rule. And then it can be conservative about the interpretation — as in, if any of the plausible interpretations would judge this action as really bad, then the action is rejected.

Sigal Samuel

I think a problem there would be that almost any moral choice arguably has ambiguity. We’ve got some of the most contentious moral issues — think about gun control or abortion in the US — where, even democratically, you might get a significant proportion of the population that says they’re opposed. How do you propose to deal with that?

Yoshua Bengio

I don’t. Except by having the strongest possible honesty and rationality in the answers, which, in my opinion, would already be a big gain compared to the sort of democratic discussions that are happening. One of the features of the Scientist AI, like a good human scientist, is that you can ask: Why are you saying this? And he would come up with — not “he,” sorry! — it would come up with a justification.

The AI would be involved in the dialogue to try to help us rationalize what are the pros and cons and so on. So I actually think that these sorts of machines could be turned into tools to help democratic debates. It’s a little bit more than fact-checking — it’s also like reasoning-checking.

Sigal Samuel

This idea of developing Scientist AI stems from your disillusionment with the AI we’ve been developing so far. And your research was very foundational in laying the groundwork for that kind of AI. On a personal level, do you feel some sense of inner conflict or regret about having done the research that laid that groundwork?

Yoshua Bengio

I should have thought of this 10 years ago. In fact, I could have, because I read some of the early works in AI safety. But I think there are very strong psychological defenses that I had, and that most of the AI researchers have. You want to feel good about your work, and you want to feel like you’re the good guy, not doing something that could cause in the future lots of harm and death. So we kind of look the other way.

And for myself, I was thinking: This is so far into the future! Before we get to the science-fiction-sounding things, we’re going to have AI that can help us with medicine and climate and education, and it’s going to be great. So let’s worry about these things when we get there.

But that was before ChatGPT came. When ChatGPT came, I couldn’t continue living with this internal lie, because, well, we are getting very close to human-level.

Sigal Samuel

The reason I ask this is because it struck me when reading your plan for Scientist AI that you say it’s modeled after the platonic idea of a scientist — a selfless, ideal person who’s just trying to understand the world. I thought: Are you in some way trying to build the ideal version of yourself, this “he” that you mentioned, the ideal scientist? Is it like what you wish you could have been?

Yoshua Bengio

You should do psychotherapy instead of journalism! Yeah, you’re pretty close to the mark. In a way, it’s an ideal that I have been looking toward for myself. I think that’s an ideal that scientists should be looking toward as a model. Because, for the most part in science, we need to step back from our emotions so that we avoid biases and preconceived ideas and ego.

Sigal Samuel

A couple of years ago you were one of the signatories of the letter urging AI companies to pause cutting-edge work. Obviously, the pause did not happen. For me, one of the takeaways from that moment was that we’re at a point where this is not predominantly a technological problem. It’s political. It’s really about power and who gets the power to shape the incentive structure.

We know the incentives in the AI industry are horribly misaligned. There’s massive commercial pressure to build cutting-edge AI. To do that, you need a ton of compute so you need billions of dollars, so you’re practically forced to get in bed with a Microsoft or an Amazon. How do you propose to avoid that fate?

Yoshua Bengio

That’s why we’re doing this as a nonprofit. We want to avoid the market pressure that would force us into the capability race and, instead, focus on the scientific aspects of safety.

I think we could do a lot of good without having to train frontier models ourselves. If we come up with a methodology for training AI that is convincingly safer, at least on some aspects like loss of control, and we hand it over almost for free to companies that are building AI — well, no one in these companies actually wants to see a rogue AI. It’s just that they don’t have the incentive to do the work! So I think just knowing how to fix the problem would reduce the risks considerably.

I also think that governments will hopefully take these questions more and more seriously. I know right now it doesn’t look like it, but when we start seeing more evidence of the kind we’ve seen in the last six months, but stronger and more scary, public opinion might push sufficiently that we’ll see regulation or some way to incentivize companies to behave better. It might even happen just for market reasons — like, [AI companies] could be sued. So, at some point, they might reason that they should be willing to pay some money to reduce the risks of accidents.

Sigal Samuel

I was happy to see that LawZero isn’t only talking about reducing the risks of accidents but is also talking about “protecting human joy and endeavor.” A lot of people fear that if AI gets better than them at things, well, what is the meaning of their life? How would you advise people to think about the meaning of their human life if we enter an era where machines have both agency and extreme intelligence?

Yoshua Bengio

I understand it would be easy to be discouraged and to feel powerless. But the decisions that human beings are going to make in the coming years as AI becomes more powerful — these decisions are incredibly consequential. So there’s a sense in which it’s hard to get more meaning than that! If you want to do something about it, be part of the thinking, be part of the democratic debate.

I would advise us all to remind ourselves that we have agency. And we have an amazing task in front of us: to shape the future.

---

From Vox via this RSS feed

The first big Christmas gift I remember getting was an animatronic bear named Teddy Ruxpin. Thanks to a cassette tape hidden in his belly, he could talk, his eyes and mouth moving in a famously creepy way. Later that winter, when I was sick with a fever, I hallucinated that the toy came alive and attacked me. I never saw Teddy again after that.

These days, toys can do a lot more than tell pre-recorded stories. So-called smart toys, many of which are internet-connected, are a $20 billion business, and increasingly, they’re artificially intelligent. Mattel and OpenAI announced a partnership last week to “bring the magic of AI to age-appropriate play experiences with an emphasis on innovation, privacy, and safety.” They’re planning to announce their first product later this year. It’s unclear what this might entail: maybe it’s Barbies that can gossip with you or a self-driving Hot Wheels or something we haven’t even dreamed up yet.

All of this makes me nervous as a young parent. I already knew that generative AI was invading classrooms and filling the internet with slop, but I wasn’t expecting it to take over the toy aisle so soon. After all, we’re already struggling to figure out how to manage our kids’ relationship with the technology in their lives, from screen time to the uncanny videos made to trick YouTube’s algorithm. As it seeps further into our society, a growing number of people are using AI without even realizing it. So you can’t blame me for being anxious about how children might encounter the technology in unexpected ways.

AI-powered toys are not as new as you might think. They’re not even new for Mattel. A decade ago, the toy giant released Hello Barbie, an internet-connected doll that listened to kids and used AI to respond (think Siri, not ChatGPT). It was essentially the same concept as Teddy Ruxpin except with a lot of digital vulnerabilities. Naturally, security researchers took notice and hacked Hello Barbie, revealing that bad actors could steal personal information or eavesdrop on conversations children were having with the doll. Mattel discontinued the doll in 2017. Hello Barbie later made an appearance in the Barbie movie alongside other poor toy choices like Sugar Daddy Ken and Pregnant Midge.

Despite this cautionary tale, companies keep trying to make talking AI toys a thing. One more recent example comes from the mind of Grimes, of all people. Inspired by the son she shares with Elon Musk, the musician teamed up with a company called Curio to create a stuffed rocket ship named Grok. The embodied chatbot is supposed to learn about whomever is playing with it and become a personalized companion. In real life, Grok is frustratingly dumb, according to Katie Arnold-Ratliff, a mom and writer who chronicled her son’s experience with the toy in New York magazine last year.

“What captures the hearts and minds of young children is often what they create for themselves with the inanimate artifacts.”

“When it started remembering things about my kid, and speaking back to him, he was amazed,” Arnold-Ratliff told me this week. “That awe very quickly dissipated once it was like, why are you talking about this completely unrelated thing.”

Grok is still somewhere in their house, she said, but it has been turned off for quite some time. It turns out Arnold-Ratliff’s son is more interested in inanimate objects that he can make come alive with his imagination. Sure, he’ll play Mario on his Nintendo Switch for long stretches of time, but afterward, he’ll draw his own worlds on paper. He’ll even create digital versions of new levels on Super Mario Maker but get frustrated when the software can’t keep up with his imagination.

This is a miraculous paradox when it comes to kids and certain tech-powered toys. Although an adult might think that, for instance, AI could prompt kids to think about play in new ways or become an innovative new imaginary friend, kids tend to prefer imagining on their own terms. That’s according to Naomi Aguiar, PhD, a researcher at Oregon State University who studies how children form relationships with AI chatbots.

“There’s nothing wrong with children’s imaginations. They work fine,” Aguiar said. “What captures the hearts and minds of young children is often what they create for themselves with the inanimate artifacts.”

Aguiar did concede that AI can be a powerful educational tool for kids, especially for those who don’t have access to resources or who may be on the spectrum. “If we focus on solutions to specific problems and train the models to do that, it could open up a lot of opportunities,” she told me. Putting AI in a Barbie, however, is not solving a particular problem.

None of this means that I’m allergic to the concept of tech-centric toys for kids. Quite the opposite, in fact. Ahead of the Mattel-OpenAI announcement, I’d started researching toys my kid might like that incorporated some technology — enough to make them especially interesting and engaging — but stopped short of triggering dystopian nightmares. Much to my surprise, what I found was something of a mashup between completely inanimate objects and that terrifying Teddy Ruxpin.

One of these toys is called a Toniebox, a screen-free audio player with little figurines called Tonies that you put atop the box to unlock content — namely songs, stories, and so forth. Licenses abound, so you can buy a Tonie that corresponds with pretty much any popular kids character, like Disney princesses or Paddington Bear. There are also so-called Creative Tonies that allow you to upload your own audio. For instance, you could ostensibly have a stand-in for a grandparent to enable story time, even if Grandma and Grandpa are not physically there. The whole experience is mediated with an app that the kid never needs to see.

There’s also the Yoto Player and the Yoto Mini, which are similar to the Toniebox but use cards instead of figurines and have a very low-resolution display that can show a clock or a pixelated character. Because it has that display, kids can also create custom icons to show up when they record their own content onto a card. Yoto has been beta-testing an AI-powered story generator, which is designed for parents to create custom stories for their kids.

If those audio players are geared toward story time, a company called Nex makes a video game console for playtime. It’s called Nex Playground, and kids use their movements to control it. This happens thanks to a camera equipped with machine-learning capabilities to recognize your movements and expressions. So imagine playing Wii Sports, but instead of throwing the Nintendo controller through your TV screen when you’re trying to bowl, you make the bowling motion to play the game.

Nex makes most of its games in-house, and all of the computation needed for its gameplay happens on the device itself. That means there’s no data being collected or sent to the cloud. Once you download a game, you don’t even have to be online to play it.

“We envision toys that can just grow in a way where they become a new way to interact with technology for kids and evolve into something that’s much deeper, much more meaningful for families,” David Lee, CEO of Nex, said when I asked him about the future of toys.

It will be a few more years before I have to worry about my kid’s interactions with a video game console, much less an AI-powered Barbie — and certainly not Teddy Ruxpin. But she loves her Toniebox. She talks to the figurines and lines them up alongside each other, like a little posse. I have no idea what she’s imagining them saying back. In a way, that’s the point.

A version of this story was also published in the User Friendly newsletter.Sign up hereso you don’t miss the next one!

---

From Vox via this RSS feed

Karen Read left Norfolk County Superior Court after the jury came back with a question for the judge. | Jessica Rinaldi/Boston Globe via Getty Images

Editor’s note, June 18, 2025, 4:40 pm ET: On June 18, 2025, Karen Read was found not guilty of the second-degree murder of her boyfriend John O’Keefe. She was found guilty of drunk driving. This was her second trial; to read our rundown of what was different at the retrial, click here. The story below was originally published July 3, 2024.

Every so often, a true crime case comes along that seems to be a Rorschach test — where there’s so much complicated, compelling, and contradictory evidence on all sides that it becomes easy to believe what you want to believe.

That’s arguably the best way to describe why the case of Karen Read, which deadlocked a “starkly divided” jury after an intense nine-week trial, has hypnotized and polarized the city of Boston, and increasingly the rest of the nation. Following Judge Beverly Cannone’s declaration of a mistrial, the prosecution immediately vowed to pursue a retrial; the second trial is currently scheduled to begin in late January 2025. That likely means heightened public interest and further entrenchment of the bitter camps in this case.

The prosecution alleges that Read, a successful finance analyst and adjunct professor who’s far from the “typical” murder suspect, killed her boyfriend, 46-year-old Boston police officer John O’Keefe, in the early, snowy morning hours of January 29, 2022. According to prosecutors, Read, who stood trial for second-degree murder and manslaughter, deliberately backed into O’Keefe with her SUV while she was intoxicated, then drove home, leaving him lying in the cold. O’Keefe died from blunt force trauma and hypothermia. But at trial, things were anything but clear; reports indicate the jury would have unanimously acquitted Read on charges of murder and leaving the scene, but were torn on the lesser charges, with a “soft” 9-3 split in favor of conviction for manslaughter.

Among the main pieces of evidence in the state’s favor: Read herself asking various witnesses, “Could I have hit him?” the next morning, after awakening and returning to the house to look for him.

Read, on the other hand, alleges that she’s been framed, in a spiraling conspiracy that began with a party full of witnesses lying about what happened and soon encompassed the entire Boston Police Department (BPD) and the prosecutor’s office. While from one standpoint, Read’s defense may be grasping at straws in its attempt to paint the case as a frame-up, from another, it’s the kind of police work Bostonians — and those in other cities across the country — have come to expect.

The trial has developed from what initially seemed to investigators like an open-and-shut case into a reckoning with a criminal justice system the public no longer trusts.

The murky facts of the case

Read claims that after she and O’Keefe spent until around midnight drinking at a local bar, she dropped O’Keefe off to hang out at the home of retired BPD officer Brian Albert. Several people who were either members of the BPD or affiliated with the local justice system were also at the house, but no one who was there corroborates her story. Not one of them says they saw O’Keefe enter the building; instead, they all claim they had no idea O’Keefe had even arrived at the house until his body was located outside of it in the snow the next morning.

Read, however, maintains that she watched O’Keefe go into the house before she left the scene. She alleges that O’Keefe must have been injured inside the house, that the partygoers staged the crime scene to look like a hit-and-run, and that they have lied about it ever since. Her defense argues that the injuries O’Keefe sustained resulted from a physical fight and a dog attack from a German shepherd mix Albert owned.

The defense alleges the investigation was hindered from the start by a failure to treat the Albert house as a possible crime scene and a failure to treat the witnesses as possible persons of interest; two of the primary investigative team members also have personal links to the witnesses. Further undermining the police were the bizarre methods they used, including using leaf blowers to clear away snow (and potentially evidence) at the crime scene, and using solo cups and grocery bags to collect evidence. Add to that a parade of unprofessional conduct and comments from officers, particularly from a lead investigator who was fired immediately after the mistrial, and the case has morphed into a genuine public scandal.

Throughout the investigation and the widely publicized trial, a highly vocal contingent of supporters from Boston and beyond have made Read into a true crime cause célèbre; one Facebook group supporting her has over 50,000 members. Thanks to a decade of thriving interest in true crime, trial-watchers are hyper-vigilant to potential injustices and the shortcomings of law enforcement. In the eyes of many, the case has become a referendum on corrupt cops — a subject Bostonians know all too well after decades of police misconduct ranging from cover-ups to kickbacks.

Read’s relentless conspiracy defense has been driven by incendiary star lawyer Alan Jackson, a former prosecutor turned defense attorney who’s defended both Harvey Weinstein and Kevin Spacey. O’Keefe, the victim, has been fully overshadowed by the conversation around the case, with Read’s supporters allegedly harassing his family members.

This polarization outside of the courtroom seems to have been reflected in the jury. After deliberating for less than a week, the jury tried repeatedly to declare themselves deadlocked to Judge Cannone, sending her notes to that effect. “We find ourselves deeply divided by fundamental differences in our opinions and state of mind,” a note from the day of the mistrial read, citing “deeply held convictions that each of us carry, ultimately leading to a point where consensus is unattainable.”

A deadlocked or hung jury results in a mistrial, after which the prosecution has to decide whether to retry the case. In this case, the prosecution seems intent on a retrial — despite the many, many problems with their arguments.

The prosecution’s case is a circumstantial mess

Despite presenting over 70 witnesses at trial, the prosecution’s primary evidence against Read remains largely circumstantial — meaning there’s very little direct or physical evidence backing their theory of what happened — and Read’s defense has offered up alternative theories for much of it. The timeline of the evening goes like this: After leaving a bar in Canton, Massachusetts, Read — whose blood alcohol content was likely over the legal limit — drove O’Keefe to Albert’s house. There, she either saw him safely inside or backed over him in the snow. In either event, she left the scene and drove home.

Afterward, she left him an angry voicemail declaring, “John, I fucking hate you.” Throughout the night, Read called O’Keefe and numerous other people. Early the next morning, she met up with two friends, including major prosecution witness Jennifer McCabe, who is also Albert’s sister-in-law, and drove back to Albert’s house, where they discovered O’Keefe in the snow. According to witnesses, Read repeatedly asked whether O’Keefe could have been hit by a snow plow, or whether she could have hit him. Multiple first responders also testified to hearing Read repeatedly exclaiming that she hit him, though none of them noted this in their initial reports from the scene.

This alleged repeated refrain of, “Could I have hit him?” and “I hit him,” was one the prosecution leaned on heavily as an argument for guilt. However, per witness accounts, Read also appeared to have no clear idea what had happened to O’Keefe.

What concrete evidence the prosecution did offer was hotly contested inside as well as outside the courtroom. One crucial piece: a broken taillight. An investigator who interviewed Read at her parents’ house that night reported seeing “some damage” to the right rear light of Read’s SUV; he testified that it was allegedly “cracked and a piece was missing.”

Nearly three weeks after O’Keefe’s death, the state’s lead investigator, Michael Proctor, allegedly located more fragments of a tail light at the crime scene and transported the evidence to the crime lab. Prosecutors say they found minute traces of O’Keefe’s DNA on those fragments, as well as tiny particles from the tail light on some of O’Keefe’s clothes. Yet the defense hammered home the enormous delay in locating what would have been an incredibly significant piece of evidence, as well as the unreliability of evidence suddenly unearthed by Proctor.

That’s because Proctor has multiple personal ties to both the victim and the witnesses. In texts he sent a day after the investigation began, he used a misogynistic slur to refer to Read, made it clear he thought Read’s guilt was obvious from the beginning, and even texted his sister that he hoped Read would kill herself. The defense alleges Proctor helped frame Read for the murder, deliberately damaging her tail light in order to claim the damage was caused when she hit O’Keefe with her car. In fact, a mysteriously altered video of Read’s car that purportedly showed the damage made it all the way to trial before the defense pointed out that the image was misleadingly reversed — a “mirror” video with the tail light in question inverted from its actual position.

It doesn’t help any of these optics that Proctor’s sister texted him that Albert’s wife Julie wanted to send Proctor “a gift” after the investigation was over. Proctor protested on the stand that none of these texts impacted the integrity of the police investigation, but given Monday’s mistrial, some jurors may have disagreed with him. They evidently weren’t alone: State police announced that they had relieved Proctor of duty immediately after the mistrial based on “information about serious misconduct [that] emerged in testimony.” He was ultimately fired — an extraordinarily rare move on their part.

Other facts of the case that on their surface might have supported the prosecution proved murkier on closer inspection. O’Keefe died of blunt-force trauma from a skull fracture and hypothermia, according to the medical examiner. He also had gashes on his arms that seemed unlikely to have been made from a car impact. The defense alleged these resulted from a dog attack on O’Keefe inside the house. O’Keefe also lacked many of the injuries you’d expect to see from a car impact, such as broken bones or significant bruises.

The prosecution’s witnesses are also controversial. In addition to Read’s former friend McCabe, who had ties to both O’Keefe and the Alberts, the party attendees included several people that Read’s defense argued should have been investigated by police before taking the stand. Among them: Brian Higgins, a Bureau of Alcohol, Tobacco and Firearms agent who’d been sporadically flirting with Read in the lead-up to O’Keefe’s death, and Albert’s son Colin, who as a teenager had made violent threats against other teens.

Many of the witnesses who were at the party, including McCabe and the Alberts, exchanged a litany of phone calls to one another throughout the time O’Keefe supposedly lay on the lawn, per phone records. Several later testified that the calls were “butt dials,” placed randomly and by complete mistake. As circumstances go, it’s eyebrow-raising, to say the least.

The defense’s evidence is equally unreliable

Despite all the problems with the prosecution’s case, the strongest evidence for the defense is likewise fully debatable. One of the defense’s key witnesses was a veteran snowplow driver who made multiple trips past the Alberts’ house during the time O’Keefe was allegedly wounded in the snow, yet claimed he saw no body lying anywhere on the lawn. However, snow impedes visibility, even with a bright snowplow light, and eyewitness testimony is notoriously unreliable.

Another piece of crucial evidence that many Read supporters consider a “smoking gun” for the defense has also been fully debunked by prosecution analysts. At issue: Two searches McCabe made on her cellphone at some point during the early morning hours surrounding O’Keefe’s death, including a misspelled query, “Hos long to die in cold.” The prosecution claims, and McCabe testified, that she made the searches around six in the morning when she and Read found O’Keefe lying in the snow. The official cellphone records for the search, however, show the timestamp for the search at 2:27 am.

The implication is that if McCabe searched for this then, she was part of a conspiracy to stage the scene and frame Read. But McCabe claims she merely opened that tab on her phone browser at 2:27 am, left it open, and searched the phrase hours later, after she and Read found him together. Multiple data analysts for the prosecution corroborated this interpretation of the cellphone data. Still, the defense scored a point in their favor by noting that their expert witness, who argued the search occurred at 2:27, arrived at his conclusion by using a method developed by one of the prosecution’s witnesses.

Other findings that leaned toward the defense included O’Keefe’s fitness tracker app, which showed him taking about 80 steps around the time Reed claimed she dropped him off. Yet here again, nothing about the timeline or O’Keefe’s activity is clear or corroborated by witnesses. The defense’s argument that O’Keefe was attacked inside the house, including by the family dog, is also speculative and unsupported by much evidence — which they claim is because the police never properly investigated. A medical expert testified on the stand in support of the dog bite theory, but forensic bite mark evidence is incredibly unreliable and has been more or less fully debunked, at least in cases involving humans.

So where does all of this leave us?

Corruption at the core

Ironically, outrage for the police’s conduct in this case has united Bostonians across political lines, with community activists joining the likes of local conservative pundit Howie Carr. Meanwhile, hundreds of Read supporters traveled from across the country to protest for her freedom outside the courthouse. They sometimes clashed with O’Keefe’s family and their supporters, who argue Read’s defense has turned a clear case of murder into a distorted media circus and fueled unfounded accusations of conspiracy.

The mistrial means that the narrative of the case has also expanded.

On the one hand, popular resistance to the idea that a successful, articulate white woman could commit such a brutal murder might be fueling a good deal of the support for Read — which speaks to lots of assumptions about how we view class, privilege, and who gets to be seen as a victim. We’re accustomed, after all, to seeing women like Read in the role of the victim, no matter that she is the one on trial here. When it comes to true crime, the media gives more attention and sympathy to attractive middle-class white women than it does to male victims, much less victims who are poor or people of color.

On the other hand, Boston police are famously corrupt, with a long history of bribery, fraud, and racial bias. The Read case is being compared to another possible homicide involving a horrifying allegation of police conspiracy in nearby Stoughton, but this isn’t just a regional issue. American cities are frequently beset by actual, proven police corruption. Boston isn’t unique. What does feel unique is the amount of public attention and scrutiny being given to the role of potential corruption in such a high-profile trial. It all adds up to what may well be a hopelessly irreconcilable pursuit of truth.

Update, July 22, 2024, 3:40 pm ET: This story was originally published on July 3, 2024, and has been updated with the date for Karen Read’s new trial in 2025.

---

From Vox via this RSS feed

A transgender rights supporter takes part in a rally outside of the US Supreme Court as the high court hears arguments in a case on transgender health rights on December 4, 2024, in Washington, DC. | Kevin Dietsch/Getty Images

It was obvious, if you listened to the Supreme Court’s oral argument in United States v. Skrmetti last December, that the Court would vote — most likely along party lines — to uphold state laws banning many forms of transgender health care for minors. So nothing about Chief Justice John Roberts’s majority opinion in Skrmetti is really surprising. All six of the Court’s Republicans voted to uphold these laws, and all three of the Court’s Democrats dissented.

But, as a matter of judicial craftsmanship, Roberts’s opinion is disappointing even by the standards of the Roberts Court. It draws incoherent distinctions. It relies on old and widely criticized precedents to undermine legal principles that are well established by more recent cases. At times in his opinion, Roberts seems to misread statutory language that he just quoted a paragraph or two earlier.

It appears, in other words, that the six justices in the majority started with the outcome that they wanted — bans on transgender health care for minors must be upheld — and then contorted their legal reasoning to fit that result.

Even if you share that goal, the decision in this case was unnecessary. As Justice Elena Kagan points out in a brief dissenting opinion, the issue before the Court concerned a threshold question: whether the Tennessee law at issue in this case should receive a heightened level of scrutiny from the courts before it was either upheld or discarded. The ultimate question of whether to uphold Tennessee’s law was not before the justices.

The Court’s Republicans, in other words, could have applied existing law, sent the case back down to the lower courts to apply this “heightened scrutiny,” and then ruled on the bans in a future case. Instead, Roberts’s Skrmetti opinion went further to rule on the legality of the bans, and consists of about two dozen pages of excuses for why the Court’s previous anti-discrimination decisions somehow do not apply to Tennessee’s law.

One virtue of this approach is that it minimizes the broader implications of Skrmetti. At oral arguments, several justices suggested that, in order to uphold Tennessee’s law, they might make sweeping changes to the rules governing all sex-based discrimination by the government — Roberts, for example, floated giving the government broad authority to discriminate on the basis of sex in the medical context. Roberts’s actual opinion contains some language suggesting that the general rule against sex discrimination is weaker when the government regulates medical practice, but those sections of his opinion are so difficult to parse that they fall short of the broad changes he discussed at oral argument.

Ultimately, Roberts’s Skrmetti opinion largely reveals something that close observers of this Supreme Court already know. The Court’s Republican majority is impatient. They are often so eager to reach ideological or partisan results that they hand down poorly reasoned opinions and incomprehensible legal standards.

Because the Skrmetti opinion is so incoherent, it is difficult to predict its broader implications for US anti-discrimination law. One thing that is certain, however, is that this decision is a historic loss for transgender Americans.

So what were the precise legal questions before the Court in Skrmetti?

To understand why the Skrmetti opinion is so difficult to reconcile with the Court’s previous decisions, it’s helpful to understand the precise legal questions before the Supreme Court.

The first of two questions is whether Tennessee’s ban on trans health care for minors classifies patients based on their sex assigned at birth. In United States v. Virginia (1996), the Supreme Court held that “‘all gender-based classifications today’ warrant ‘heightened scrutiny.’” “All” means that all laws that classify people based on their sex must receive additional scrutiny from the courts, not just some laws that do so.

About half of the states have laws targeting transgender health care, but the Tennessee law at issue in Skrmetti is among the strictest. It prohibits people under the age of 18 from receiving many medical treatments to treat gender dysphoria or other conditions related to their transgender status — including bans on puberty blockers and hormone therapy.

Significantly, Tennessee’s law is also quite explicit that the purpose of this law is to ensure that young people do not depart from their sex assigned at birth. The law declares that its purpose is to “encourag[e] minors to appreciate their sex” and to prevent young people from becoming “disdainful of their sex.” That is an explicit sex-based classification. Patients who Roberts refers to as “biological women” are allowed to fully embrace femininity in Tennessee. But a child who is assigned male at birth may not.

Under Virginia, in other words, Tennessee’s law — which relies on a sex-based classification — must be subject to heightened scrutiny.

To be clear, the mere fact that courts must give heightened review to Tennessee’s law does not mean that the law will necessarily be struck down. As the Court held in Craig v. Boren (1976), “to withstand constitutional challenge…classifications by gender must serve important governmental objectives and must be substantially related to achievement of those objectives.” Some laws do survive this level of scrutiny.

Roberts’s opinion raises several policy arguments for Tennessee’s law, claiming that the procedures targeted by Tennessee are “experimental,” that they “can lead to later regret,” and that they carry “risks.” A court applying heightened scrutiny could consider these arguments and whether they justify upholding the law.

But Roberts bypasses this inquiry altogether, instead denying that the Tennessee law engages in sex-based classifications at all. The law, Roberts claims, only “incorporates two classifications.” It “classifies on the basis of age” by banning certain treatments only for minors. And it “classifies on the basis of medical use” by prohibiting doctors from prescribing those treatments to address gender dysphoria or similar conditions affecting transgender people, while simultaneously permitting those treatments to address other conditions.

Roberts is correct that Tennessee’s law does draw lines based on these two classifications. But a law can do more than two things at once. And this law explicitly states that it exists to classify every child as either a boy or a girl, and then to lock them into that classification until their 18th birthday. Under Virginia, that classification demands heightened scrutiny.

The second legal question before the Court in Skrmetti was whether all laws that discriminate against transgender people are themselves subject to heightened scrutiny. Roberts, however, dodges this question by claiming that Tennessee’s law “does not classify on the basis of transgender status.” Instead, he argues, the law classifies people based on whether they have conditions such as “gender dysphoria, gender identity disorder, or gender incongruence.”

Gender dysphoria, gender identity disorder, or gender incongruence are among the defining traits that make someone transgender. Roberts might as well have argued that Jim Crow laws do not discriminate on the basis of race, but instead discriminate based on the color of a person’s skin.

To justify this distinction, Roberts points to the Court’s decision in Geduldig v. Aiello (1974), which held that discrimination against pregnant people is not a form of sex discrimination because not all women become pregnant. But, even if it is true that not all transgender people experience gender dysphoria or a similar condition, post-Geduldig decisions have long held that the government cannot evade a ban on discrimination by claiming that it is merely discriminating based on a trait that closely correlates with a particular identity.

As the Court said in Bray v. Alexandria Women’s Health Clinic (1993), “a tax on wearing yarmulkes is a tax on Jews” — even though many Jews do not wear yarmulkes.

That said, the Court’s decision not to rule on whether laws that classify on the basis of transgender status must receive heightened review is probably a blessing for transgender people, even if it is a small one. While Roberts’s reasoning on this question is muddled, his opinion leaves open the possibility that a future Court may resolve this question in favor of transgender people — although that is highly unlikely to happen unless the Court’s membership changes significantly.

Notably, Justice Amy Coney Barrett, who is close to the center of the current Court, wrote a separate concurring opinion arguing that discrimination against trans people does not trigger heightened scrutiny.

For the most part, Skrmetti is a disaster for transgender people, and especially for transgender youth. It twists the Constitution in knots to uphold Tennessee’s law. And the decision is likely to ensure that many parents of transgender children must move to blue states if they want their child to receive appropriate medical care.

---

From Vox via this RSS feed

An aerial view shows floodwaters covering farm fields and a rural road near Poplar Bluff, Missouri. In April, thunderstorms, heavy rains, high winds, and tornadoes plagued the regions for several days causing widespread damage. | Scott Olson/Getty Images

Globally, humanity is producing more food than ever, but that harvest is concentrated in just a handful of breadbaskets.

More than one-third of the world’s wheat and barley exports come from Ukraine and Russia, for example. Some of these highly productive farmlands, including major crop-growing regions in the United States, are on track to see the sharpest drops in harvests due to climate change.

That’s bad news not just for farmers, but also for everyone who eats — especially as it becomes harder and more expensive to feed a more crowded, hungrier world, according to a new study published in the journal Nature.

Under a moderate greenhouse gas emissions scenario, six key staple crops will see an 11.2 percent decline by the end of the century compared to a world without warming, even as farmers try to adapt. And the largest drops aren’t occurring in the poorer, more marginal farmlands, but in places that are already major food producers. These are regions like the US Midwest that have been blessed with good soil and ideal weather for raising staples like maize and soy.

But when that weather is less than ideal, it can drastically reduce agricultural productivity. Extreme weather has already begun to eat into harvests this year: Flooding has destroyed rice in Tajikistan, cucumbers in Spain, and bananas in Australia. Severe storms in the US this spring caused millions of dollars in damages to crops. In past years, severe heat has led to big declines in blueberries, olives, and grapes. And as the climate changes, rising average temperatures and changing rainfall patterns are poised to diminish yields, while weather events like droughts and floods reaching greater extremes could wipe out harvests more often.

“It’s not a mystery that climate change will affect our food production,” said Andrew Hultgren, an agriculture researcher at the University of Illinois Urbana-Champaign. “That’s the most weather exposed sector in the economy.”

Farmers are doing what they can — testing different crop varieties that can better withstand changes in the climate, shifting the timing of when they sow, tweaking their use of fertilizers and water, and investing in infrastructure like water reservoirs.

The question is whether these adaptations can continue to keep pace with warming. To figure this out, Hultgren and his team looked at crop and weather data from 54 countries around the world dating back to the 1940s. They specifically looked at how farmers have adapted to changes in the climate that have already occurred, focusing on maize, wheat, rice, cassava, sorghum, and soybean. Combined, these crops provide two-thirds of humanity’s calories.

In the Nature paper, Hultgren and his team reported that in general, adaptation can slow some crop losses due to climate change, but not all of them.

And the decrease in our food production could be devastating: For every degree Celsius of warming, global food production is likely to decline by 120 calories per person per day. That’s even taking into account how climate change can make growing seasons longer and how more carbon dioxide in the atmosphere can encourage plant growth. In the moderate greenhouse gas emissions scenario — leading to between 2 and 3 degrees Celsius of warming by 2100 — rising incomes and adaptations would only offset one-third of crop losses around the world.

“Looking at that 3 degrees centigrade warmer [than the year 2000] future corresponds to about a 13 percent loss in daily recommended per capita caloric consumption,” Hultgren said. “That’s like everyone giving up breakfast … about 360 calories for each person, for each day.”

The researchers also mapped out where the biggest crop declines — and increases — are likely to occur as the climate warms. As the world’s most productive farmlands get hit hard, cooler countries like Russia and Canada are on track for larger harvests. The map below shows in red where crop yields are poised to shrink and in blue where they may expand:

The results complicate the assumption that poor countries will directly bear the largest losses in food production due to climate change. The wealthy, large-scale food-growers may see the biggest dropoffs, according to the study. However, poor countries will still be affected since many crops are internationally traded commodities, and the biggest producers are exporters. A smaller harvest means higher food prices around the world. Less wealthy regions are also facing their own crop declines from disasters and climate change, though at smaller scales. All the while, the global population is rising, albeit much more slowly than in the past. It’s a recipe for more food insecurity for more people.

Rice is an exception to this trend. Its overall yields are actually likely to increase in a warmer world: Rice is a versatile crop and unlike the other staples, it benefits from higher nighttime temperatures. “Rice turns out to be the most flexibly adapted crop and largely through adaptations protected from large losses under even a high warming future,” Hultgren said. That’s a boon for regions like South and Southeast Asia.

Decreasing the available calories isn’t the only way climate change is altering food, however. The nutrition content can change with shifts in rainfall and temperature too, though Hultgren and his colleagues didn’t account for this in their study. Scientists have previously documented how higher levels of carbon dioxide can cause crops like rice to have lower levels of iron, zinc, and B vitamins. So the food we will be eating in the future may be more scarce and less nutritious as well.

And while climate change can impair our food supply, the way we make food in turn harms the climate. About one-third of humanity’s greenhouse gas emissions stem from food production, just under half of that from meat and dairy. That’s why food production has to be a major front in how we adapt to climate change, and reduce rising temperatures overall.

---

From Vox via this RSS feed