2026-04-22 20:30:00
It’s Earth Week, and these days, it’s become hard to know just what to do with this holiday. Is it a reminder to start composting, ditch fast fashion, or donate to climate causes? Or does nothing we do as individuals really matter, and it’s on governments and corporations alone to fix our climate and ecological crises?
The latter idea — that our own individual actions won’t help to heal the planet — has become almost gospel in the modern environmental movement. And it’s largely right. But there are a few actions that individuals can take that actually do make a substantial difference in turning our current environmental trajectory around. And some of the most impactful ones might not be what you’d expect.
The environmental nonprofit Project Drawdown analyzed the top 20 actions that households can take to minimize their carbon footprint. It found that reducing food waste and eating a “plant-rich” diet — one that’s lower in meat and dairy — came out tied with each other for the No. 1 spot of most impactful changes. Putting solar panels on your roof ranked third, lagging far behind. (A number of other environmental analyses have put plant-rich diets as top contenders for environmental lifestyle changes, too.)
Sign up here to explore the big, complicated problems the world faces and the most efficient ways to solve them. Sent twice a week.
And fortunately, food selection is a low-overhead choice we get to make many times each day, which makes it more flexible than other high-impact actions, like buying an electric vehicle or making energy-efficient home renovations.
Yet surveys show that people greatly underestimate meat and dairy’s enormous impact on the environment — and not just its fueling of climate change, but also its massive contribution to deforestation, water pollution, and other problems.
It’s understandable why. As I wrote a few years ago, “a power plant emitting plumes of black smoke screams pollution, while a pasture of cattle, chickens, or pigs along the highway looks natural and quaint — even eco-friendly.” So here are eight charts that break down how truly massive an impact our food choices have when you add them all up.
First, meat and dairy production are really inefficient ways to produce food.
Meat companies have to feed animals a lot of calories to produce just one calorie of edible meat, dairy, or eggs. Bruce Friedrich, president of the Good Food Institute and author of the book Meat, likens this efficiency to tossing eight plates of pasta into the trash for every one plate we eat (in the case of chicken — and it’s far higher for beef).
In this way, meat production itself is a form of food waste.
To grow all these crops, like corn and soy — and graze cattle and sheep — farmers and meat companies have turned much of planet Earth into one big animal farm, occupying more than one-third of habitable land. It’s the top cause of global deforestation and habitat loss, and thus a leading threat to wildlife.
It uses up even more land in the continental US: 41 percent.
To look at it another way, here’s the amount of land required to produce various diets:
But land isn’t the only essential resource that animal agriculture gobbles up. It’s also the largest user of freshwater.
Animal agriculture doesn’t just use up a lot of water; it also heavily pollutes waterways like rivers and streams. In the US, it’s arguably the biggest source of water pollution. The pollution primarily stems from two sources: manure from the world’s hundreds of billions of farmed animals, and the fertilizer used to grow their feed crops.
Finally, there’s climate change.
Globally, meat and dairy production is one of the leading drivers of climate change, accounting for 14.5 percent to 19 percent of global greenhouse gas emissions (including from fertilizer production, manure, deforestation, and, yes, cows’ methane-rich burps). Meanwhile, plant-based protein sources, like Impossible burgers, beans, and tofu, have much lower carbon footprints than beef, pork, and chicken.
Aside from hard data, compelling evidence on the necessity of plant-rich diets also comes from scientists themselves. In a 2021 survey of more than 200 agricultural and environmental scientists, most ranked “reducing meat and dairy consumption” as the most effective way to reduce agricultural climate emissions.
And if you’re worried that your individual actions don’t actually matter, consider this table below. It’s from leading agricultural economists Jayson Lusk and Bailey Norwood of Oklahoma State University, who crunched the numbers and determined that when consumers buy less meat, eggs, and milk, it truly does reduce production of them (more on that here).
What’s more, embracing a plant-rich diet is also much better for your health and, of course, helps to reduce the number of animals reared on cruel factory farms.
In the long run, like every other polluting industry, significant change will also have to come from governments and corporations. But so far, the meat industry has been highly effective at beating back environmental regulations. It’s hard to see how that’ll change unless more people demand it. We can start — one meal at a time — with what we eat. And we’ve got you covered with resources to get started: Check out Meat/Less, Vox’s practical guide to eating less meat and more plant-based foods.
2026-04-22 20:00:00
Every year, federal and state wildlife agencies in the US breed millions of fish and release them into the wild, all for fishers to catch. They do this because many ecosystems no longer support the thriving fish populations they once did, due to dams, pollution, and rising water temperatures.
But there’s a catch. In many cases, the fish that are being released aren’t native to these ecosystems. In Connecticut, where this video was filmed, the Department of Energy and Environmental Protection (DEEP) breeds and releases rainbow trout — from the West Coast — and brown trout, from Europe, Asia, and Northern Africa. Of course, states aren’t just throwing fish into the water at random. There are many preventative measures and monitoring systems in place to protect local ecosystems from the dangers of non-native species. But things can still go wrong.
Introducing non-native fish can wreak havoc on an ecosystem. So why are state wildlife agencies doing it by the million?
To see some fish stocking in action and learn more about the process, Vox producer Nate Krieger went on a fish stocking run with employees of Connecticut DEEP. In just under an hour, they released 675 live trout into the Mianus River for the recreation of local fishermen.
This video examines the strange paradox that incentivizes states to do something that, at least in some cases, hurts the very ecosystem they’re trying to conserve. It explores the complicated question of fish stocking in the US, examines the possible dangers to ecosystems, and shines a light on potential conservation benefits. Because recreational fishing encourages people to spend time outside, which builds a relationship with nature and the environment that could be invaluable for our future.
Read more about non-native fish stocking:
This video is part of a series supported by Animal Charity Evaluators, which received a grant from EarthShare.
2026-04-22 20:00:00
Yesteryear, the buzzy new debut novel by Caro Claire Burke, has the kind of premise it’s hard to look away from: a tradwife influencer named Natalie — a Harvard dropout who married rich at 20 — wakes up in 1855. Gone are her tastefully discrete appliances, her prized collection of luxury sweaters, her team of nannies and farm workers. In their place: an outhouse, stained homespun prairie dresses, and hours of back-breaking labor spent washing a single load of laundry with homemade lye soap.
Natalie, confronted with this brave old world, does a lot of crying. Things get especially rough for her after she tries to escape, stumbles into a bear trap, badly injures her leg, and then has to cope with 19th-century pioneer medicine. The medicinal ointment “smells like bacon grease,” and there’s no anesthetic for the stitches, so that, Natalie tells us, “it feels like my body has depleted a month’s worth of energy from the mere translation of so many nerve signals screaming EMERGENCY to my brain.”
There’s a sort of satisfaction to witnessing Natalie’s distress. You find yourself wanting to say, “How’s all that trad working for you now?” and then maybe sneer a little.
At long last, one of those perniciously appealing traditional housewife influencers — the type who’s always posting videos of herself baking bread in a sun-drenched kitchen while her adorable children romp next to her — has been forced to put her money where her mouth is. Surely now, you think, she’ll have to admit that the modern era has some things going for it.
Yesteryear is a book animated by this kind of rage, by a palpable fury at the archetype of the tradwife. That’s what makes the premise so irresistible — irresistible enough to have garnered breathless review coverage, for Anne Hathaway to sign on to produce and star in the movie after a vicious four-studio bidding war. I myself read Yesteryear in one long rush, unable to put it down.
But where the book begins to falter is when it tries to suggest that tradwives are just as angry with themselves as feminists are.
In Yesteryear, Natalie knows her content is rage bait. She refers to her followers as “the Angry Women,” noting smugly that “self-proclaimed progressive women” are “chemically addicted to hating women like me.” When, on a trip to Target, she encounters Vanessa, a high school friend who has since renounced her devout upbringing, Natalie lingers with almost erotic pleasure on how much the person must envy and despise her. “Go ahead,” she thinks gleefully. “Give yourself a migraine thinking about me.”
Natalie isn’t wrong that a lot of the attention tradwives receive ranges from critical to furious. “Is tradwife content dangerous, or just stupid?” asked a viral Cut essay in 2023. Another essay in 2020 described the sexism at its core as “the gateway to white supremacy.” In a 2024 profile of Hannah Neeleman, the influencer known as “Ballerina Farm,” who is the most prominent of the tradwives, the New York Times summarized the discourse: “Is she, as her fans would have it, a woman who has made the commendable decision to stay home, raise the kids and support the family farm? Or is she, as her detractors would argue, someone who uses social media to push for a return to traditional gender roles while glossing over the privileges that allowed her to have such a lifestyle in the first place?”
To people who consider themselves progressive, who are by and large the presumed audience for Yesteryear, tradwives aren’t women who “choose their choice”; they threaten the gains of 20th-century feminism. They try to sell women on the lie that they would be happier without birth control or educations or careers, tending endless beautiful children in a spotless, beautiful kitchen. And it’s true that a large swath of their followers are there both for the pleasure of their gorgeous pastoral lives and to be furious at them for their political propaganda.
Natalie describes the appeal of her content by analogizing it to the rancid, craveable flavor of black truffles. “People aren’t so different from pigs, apparently,” she says. “Once they learn a rotten thing can be eaten, they will eat it, and they will become addicted to it.” She believes there is a “rot” of unhappiness on her farm that comes through in her content — her own exhaustion at the drudgery of her chores, the palpable fakery of her artificial paradise — “and everyone rushed towards me with their forks.”
For most critics of tradwife content, the “rot” Natalie is describing here is the anti-feminist proselytizing, the romanticization of a bleak way of life that left many women trapped. The rot Burke is portraying in Yesteryear, however, is just straightforward influencer hypocrisy.
Influencing at its most basic form is sales, and like any overworked saleswoman, Natalie lies about her product: herself, and her allegedly pure lifestyle. She secretly douses the family’s “organic” farm in pesticides, because she knows they’ll never turn a profit otherwise. Her pastoral-chic line of Dutch ovens is made in Taiwan and drop-shipped. She has nothing but contempt for Vanessa, whom she greets warmly while internally calling her a “pick me” and a “cunt” for having named her daughter Zoe.
But Natalie’s hypocrisy goes deeper than that. We learn that she despises her dimwitted husband Caleb, whom she felt pressured to marry young and begin having kids with as soon as the wedding was over, thanks to the culture of her unnamed evangelical sect; he cannot achieve a full erection during sex, leaving her to impregnate herself with a sauce baster. Being alone with her children triggers panic attacks. Early followers tell her that her smile looks too strained, so now she compulsively practices fake smiling at all times, and has trouble dropping it when the occasion calls for solemnity.
Still, she tells herself that all the wives and mothers she knows are happier than the career women she sees bemoaning their inability to have it all. Once she has her first child and finds herself bored and miserable, she decides the stay-at-home moms she knows must be lying about their happiness. With no work history or job prospects and an ever-mounting brood of children to provide for, she can find no outlet for her intellect and creativity outside of the project of turning her life into online content.
Natalie has an intimate understanding of why her followers love to resent her, because she loves to resent modern women. She tracks her liberal college roommate Reena on social media for the sheer pleasure of hating her and her life choices, an act that mirrors career women hate-following tradwives. “She looked like a stereotype of a modern woman,” Natalie gloats over a video of Reena announcing she’s been laid off from her consulting gig, “poreless and lip-lined and shrill.”
Throughout the novel, characters create an imaginary woman out of scraps of social media content, just so that they can get mad at her. Natalie does it with Reena, and Natalie’s followers do it to her. In Burke’s telling, we do this because we are all unhappy with our own lives and want to lash out. Which is a little strange, because what is the novel Yesteryear if not the process of creating an imaginary woman out of scraps of social media content, just so that we can get mad at her?
Yesteryear has a gripping, thriller-like pacing, which it owes mostly to the delicious mystery of what exactly happened to Natalie to send her to 1855.
Did she time-travel, à la Outlander? Is she on some sort of hidden camera reality TV show? Is she being tested by God (Natalie’s favorite option)? At one point, she finds a secret cabin in the past with a sign out front that says “The Manosphere,” and I got giddy with delight over the idea that Burke was positing a world where all those podcasters decided to start building virtual realities to send uppity women for reprogramming, like an updated Stepford Wives.
Along the way, Natalie is punished by the world of 1855. There’s the bear trap, of course, and the fact that one of the first things Caleb does when we meet him in the past is slap his wife across the face so hard she blacks out. Also, the food “looks, frankly, like shit,” made with such sparse, stingy ingredients that even Natalie’s famous sourdough loaves don’t turn out right. (“The worst possible thing to happen,” she fumes.)
The real reason for Natalie’s time travel, when it comes, is deflating. Without spoiling too much, Burke’s conclusion suggests that the sadistic anger that pulses through Yesteryear — the desire to see Natalie brought down a peg, humiliated, forced to admit that what she says she wants is not what anyone who had the choice would actually want — is a feeling that Natalie shares. She wants to see herself punished as much as the reader does. She punishes herself enough to furnish the whole plot of the book.
There’s an easy smugness to this conclusion that, in retrospect, makes the project of Yesteryear less satisfying than it at first promised to be. It relies on the seductive but unlikely idea that if tradwives were really honest with themselves, they’d admit that they agree with feminists on what the problems with their lives are. It posits that Natalie, too, wants to ask, “How’s all that trad working out for you now?”
I don’t think we have to pretend that being Ballerina Farm really is as idyllic as it looks on Instagram in order for us to grant tradwives the courtesy of taking them at their word about their fundamental beliefs. Their lives might not be all that happy, but it strikes me as unlikely that tradwives secretly believe that this is because the message they are preaching is false and will make other people’s lives worse. Nor do I believe tradwives really think that they are doing something wrong, something rotten, by making the content that they do.
Even Natalie, for all her silent rage, never imagines that the liberal women she hates don’t genuinely believe in equality. It’s as though the strongest comeuppance Burke can imagine for this woman who makes us so angry is to deny that she believes the things she appears, through all her words and actions, to sincerely believe. Yesteryear punishes the tradwife by making her into someone less than substantial — and so in the end, this bingeable, buzzy novel fails to entirely satisfy.
There is something fundamentally dishonest in building an imaginary woman in order to hate her, and not even letting her hold her own principles. I suppose it’s fun to imagine a world in which a tradwife turns out to secretly love cursing and pills, where social media is not just exaggerated but an out-and-out lie, and where she punishes herself to save all the rest of us the trouble. But that’s no less of a fantasy than a bucolic farm where the bread is always perfect and the children never cry.
2026-04-22 19:30:00
2026年4月10日，有人在笔记本电脑屏幕上查看了Anthropic的“Project Glasswing”网站。| Gabby Jones/Bloomberg via Getty Images
AI的能力有多强？强大到足以让这家领先的AI公司Anthropic在本月早些时候宣布，其最新AI模型Claude Mythos Preview仅限于向少数企业发布，以应对安全风险——至少目前如此。Anthropic表示，Mythos Preview是为通用用途设计的，但在测试过程中，他们发现该模型在识别各类软件系统中的安全漏洞方面非常高效，这可能带来巨大的安全隐患。因此，Anthropic目前通过名为“Project Glasswing”的项目，仅向少数几家大型科技公司和银行分享该模型，目的是让他们有机会修补现有系统中的安全漏洞，并提前防范可能的网络攻击。
为了更好地了解Claude Mythos Preview的特性及其对网络安全的潜在威胁，Today, Explained的主持人Sean Rameswaram近日采访了The Verge的高级AI记者Hayden Field。以下是他们对话的节选（已进行删减和润色）：
“Claude Mythos是什么？Mythos是Anthropic最新开发的通用型AI模型。但在开发过程中，他们发现这个模型具备一些意想不到的特殊能力，尤其是在网络安全领域。它能够发现几乎所有操作系统中的高风险漏洞，这对黑客来说是非常危险的。如果有人拥有这个模型，他们就能获得所有关键系统中存在的漏洞和安全隐患的详细蓝图，从而找到破坏系统或窃取数据的方法。因此，Anthropic决定不向公众发布该模型，而是选择了一些负责关键基础设施的组织，以便这些组织可以利用该模型修补自身系统中的漏洞。”
目前，许多使用Claude Mythos的公司已经广为人知，包括Nvidia、JP Morgan Chase、Google等，可能还有数十家其他负责关键软件基础设施的公司。
“这个模型是如何运作的？”Hayden Field解释道：“由于它是一个通用型模型，使用方式可能与其他AI模型类似。你只需提示它查找系统中的漏洞，例如，如果你是Google Chrome，你可能会提示它检查浏览器中某些特定部分是否存在漏洞。模型会列出所有高风险的漏洞，然后你自行修补。黑客也会以同样的方式使用它，如果落入坏人之手，他们可能会利用这些信息进行恶意活动。因此，关键在于谁在使用这个模型以及他们的动机。”
“我们是否知道像Google和Nvidia这样的公司真的在使用这项技术？”他回答：“是的。Anthropic之所以发布这个模型，部分原因是为了让这些公司反馈其工作原理以及如何修补漏洞。这是一种信息共享的方式，他们希望这些公司能够测试该模型的效果，并向Anthropic报告结果。”
“Anthropic是如何选择与哪些组织分享这项技术的？”Hayden Field问道：“他们主要寻找的是网络安全防御者，即那些许多人都依赖的公司，如果这些公司遭到攻击，后果将非常严重。例如，JP Morgan Chase就是一个典型例子。此外，Anthropic还向政府提供了这项技术。”
“Anthropic的竞争对手是否也有类似工具？他们是否在开发类似技术？”Hayden Field继续提问：“据称，OpenAI也在开发类似工具。Anthropic自己也表示，他们并不认为自己会长期在这一领域领先。他们认为，全球各地的实验室可能在未来三到六个月甚至一年内发布类似技术。因此，他们现在发布Mythos，是为了让企业和银行提前做好准备，以应对未来可能由类似技术引发的网络攻击。”
“如果这种技术如此危险，存在这么多潜在风险，是否有人在讨论是否应该完全不发布这类工具，而是将其内部化？”Hayden Field问：“这是一个非常好的问题。我很高兴你提出了，因为很少有人会问，是否应该将某些AI系统发布或用于特定用途。目前，我们看到的是‘一刀切’式的AI应用，但很多时候AI并不是解决问题的正确方法。然而，对于Mythos这样的工具，人们普遍认为它现在是必要的。因为AI已经在帮助网络攻击者提升攻击能力，而且这种趋势在过去一年中愈发明显。专家们似乎都同意，我们需要用AI来对抗AI驱动的网络攻击。这就像中世纪的城堡，当战争来临，人们会加固城墙，提高防御。这些专家也告诉我，他们知道这种技术即将出现，因此建议企业现在就加强防御，以做好充分准备。”
How powerful is AI? Enough that Anthropic, a leading AI company, announced earlier this month that its latest AI model, Claude Mythos Preview, would be available only to a limited number of businesses due to security concerns — at least for now.
Claude Mythos Preview was designed for general use, Anthropic says, but during testing, the company found it extremely effective at identifying vulnerabilities in the security systems of all types of software, creating potentially massive security concerns.
So far, Anthropic is sharing the Mythos Preview model with a handful of major tech companies and banks through a program called Project Glasswing, intended to give them an opportunity to shore up any existing security vulnerabilities and get ahead of potential hacking attempts that the model could identify.
To get a better sense of what Claude Mythos Preview represents and the potential threat it brings to online security, Today, Explained co-host Sean Rameswaram spoke with Hayden Field, senior AI reporter at The Verge.
Below is an excerpt of their conversation, edited for length and clarity. You can hear the full episode wherever you get podcasts — including Apple Podcasts, Pandora, and Spotify.
What is Claude Mythos?
Mythos is [Anthropic’s] newest AI model that they designed to be a general-purpose AI model like any other. But what they realized when they were working on it was that it had these special skills that they didn’t really anticipate. It was really good at cybersecurity. It found high-stakes vulnerabilities in virtually every operating system.
That’s pretty bad if you are using that as a hacker. And to have a blueprint for a list of every big gap and insecurity and vulnerability on all these really, really high-profile systems, you’re going to be having a list of everything you could do to take those systems down or exploit data.
They realized that they better not release this to the general public because it could fall into the wrong hands. And they instead handpicked a select few organizations that are responsible for critical infrastructure to release it to so they could plug those gaps in their systems instead.
You’ve heard of many of the companies that currently have and are using Claude Mythos: Nvidia, JP Morgan Chase, Google, apparently a few dozen more that build or maintain critical software infrastructure. How does it actually work?
Since they built it as a general-purpose model, it probably works like any other model in that you’re using it and prompting it to flag all the vulnerabilities in your system.
Maybe you’re Google Chrome, and you’re looking for specific, niche parts of the browser that you think may have some vulnerabilities. You’re basically prompting the model to flag all these really high-profile gaps to you and your security, and then you’re taking that and plugging it up on your own.
A hacker would actually use it in the same way. If it fell into the wrong hands, they’d be like, “Yeah, tell me all the vulnerabilities here.” And then they’re going to take it off the platform and use that for something nefarious. So it’s basically about who is prompting the system and what their motives are.
It’s as easy as saying, “Hey, Claude, tell me how this banking system might be vulnerable.” And then Claude thinks about it for a minute, and it spits out a bunch of answers.
Essentially, yes.
And do we know that the Googles and Nvidias of the world are actually using this technology?
Yes. Part of the reason that Anthropic released this is they wanted these organizations to report back on exactly how Mythos worked and what it did to plug up the vulnerabilities and the gaps in their system. It’s an information-sharing thing.
They’re letting these companies use it to test out how well it does to plug up all these high-profile gaps, and then they have to report back to Anthropic about how it worked.
How is Anthropic choosing who to share this technology with?
I actually asked them that. They’re essentially looking for cyber defenders or companies that a lot of people depend on, and that downstream it would be a huge issue if they got hacked in any way, shape, or form.
JP Morgan Chase is a great example. Anthropic has also offered this technology to the government.
Do Anthropic’s competitors have similar tools? Are they presumably working on similar tools?
OpenAI is apparently working on a similar tool. Anthropic itself has said this isn’t something that they deem they’ll be in the lead on for too long. They think labs anywhere in the world may release this technology in the next three months, six months, 12 months.
It seems like, sometime in the next 12 months, this is going to be out there. And so that’s why they wanted to release Mythos now, so that companies and banks could get ahead of all the hacks that may be coming down the line, when similar types of technology are released to the general public, maybe months from now.
If this is so dangerous and there’s so many potential risks, is anyone having a conversation about just not releasing tools like this and just sort of shutting it down, keeping it internal?
That is a really great question. I’m so glad you asked, because not enough people ask whether an AI system should actually be released or used for certain things. Right now, we’re seeing a lot of one-size-fits-all, throw-it-at-everything type of integration. And a lot of times AI is not the answer for things.
With this, though, people tend to agree that it is something that’s needed right now. AI is already out there helping cyberattackers really step up their attacks. And we’ve been seeing that intensify over the past year. People seem to agree that you need AI to fight AI cyberattacks, essentially.
It’s kind of like medieval fortresses, where you’re adding extra stones and building up the walls at the fortress higher because a war is coming. That’s the sense I get when I talk to these experts about this. They know it’s coming. It’s just, ‘Try to shore up your defenses now so that you’re best prepared.’
2026-04-22 19:00:00
前劳工部长洛里·奇维-德雷默（Lori Chavez-DeRemer）于2026年3月12日离开白宫东厅的妇女历史月活动。她因内部调查而辞职，调查涉及其不当行为，包括在公务差旅中指示工作人员购买她的长相思白葡萄酒（sauvignon blanc）、在办公室藏酒、鼓励年轻女员工关注其父亲和丈夫，与安保人员发生婚外情，并安排公务旅行探亲访友。尽管这些细节较为特殊，但此类高层官员因争议或与总统冲突而离职的现象并非首次。
特朗普在任期内多次提名高级官员，但随后因争议或与总统的矛盾而离职。过去八周内，已有三位内阁成员辞职或被解职。据布鲁金斯学会统计，自去年1月起，约三分之一的特朗普“核心团队”成员已离开白宫，其中8人被调任其他职位，22人主动辞职或被施压离职。虽然这一离职率（内阁20%，核心团队32%）相比其首个任期的92%大幅下降，但特朗普的团队仍比其他总统更为动荡，通常只有10%的行政人员会在第一年离职。
特朗普第二任期的稳定改善并非源于更优秀的资历，而是因为更强调无条件忠诚。路透社在2024年报道中指出，忠诚是首要条件。这种对个人忠诚的重视减少了高层政策分歧和人际矛盾，但即使是特朗普的亲密朋友，如前司法部长帕姆·邦迪（Pam Bondi）和商务部部长霍华德·卢特尼克（Howard Lutnick），也可能被替换。例如，邦迪与特朗普相识十余年，而卢特尼克则被传即将离职。此外，FBI局长卡什·帕特尔（Kash Patel）也因出版多本颂扬特朗普的儿童书籍而可能面临被替换的风险。对此，笔者将为这些离职者斟上一杯长相思白葡萄酒。
This story appeared in Today, Explained, a daily newsletter that helps you understand the most compelling news and stories of the day. Subscribe here.
I will remember former Labor Secretary Lori Chavez-DeRemer less for her tenure than for her casual deployment of the diminutive term “sauvi B” — which no person over 27 should use, especially during the workday.
Chavez-DeRemer resigned Monday amid an internal investigation into her conduct. In addition to instructing staff to buy her bottles of sauvignon blanc on work trips, Chavez-DeRemer allegedly stashed liquor in her office, encouraged young female staffers to “pay attention” to her father and husband, had an affair with a member of her security detail, and arranged work travel to visit family and friends.
It’s a pretty wild story, all told. And while the specifics are unusual, the broader pattern is not. Across both of his terms, President Donald Trump has repeatedly nominated high-level officials who have later flamed out amid controversies or clashes with the president himself. In just the past eight weeks, three Cabinet-level officials have resigned or been forced out of the administration.
Who’s in, who’s out. Trump previously ousted Homeland Security Secretary Kristi Noem, the erstwhile face of his mass deportation campaign, and Attorney General Pam Bondi, who served as one of Trump’s defense attorneys during his first impeachment trial in 2020. Noem fell out of the president’s favor over a $220 million border security ad campaign that prominently featured her (among other more consequential missteps). Bondi was fired over her handling of the Epstein files, which alienated a large segment of Trump’s core base, and her failure to prosecute his political enemies.
They’ve been playing musical chairs outside the Cabinet, as well. According to the Brookings Institute, which has analyzed White House turnover going back to the Reagan administration, roughly a third of Trump’s “A Team” — the staffers who occupy the highest positions within the Executive Office of the President — have left the White House since January last year. While eight of those people were promoted into other positions, 22 resigned or were pressured to do so.
Trump’s turnover record. That turnover rate — 20% in the Cabinet and 32% among top executive staff — actually represents a marked improvement from Trump’s first term.
Turnover on the “A team” then was a whopping 92%, according to Brookings, and 14 Cabinet members left over that same period. (Remember Rick Perry? Betsy DeVos? Rex Tillerson? To say nothing of the various former Cabinet officials who have since reinvented themselves as Trump critics and pundits.) Trump still oversees a pretty volatile staff relative to other presidents, however; on average, just 10% of executive staff turn over in a president’s first year.
Trump’s loyalty test. It’s also worth considering why stability improved in Trump’s second term. The president didn’t choose better-qualified or credentialed staff necessarily; he chose more deferential ones. As Reuters put it in a story about Trump’s early search for staff in 2024, “One quality is absolutely paramount: unquestioning loyalty.”
That emphasis on personal allegiance has apparently helped reduce policy disagreements and interpersonal squabbles in the top tiers of the administration. But it hasn’t necessarily protected even Trump’s personal friends and associates from getting the axe. Bondi had known Trump for over a decade, for instance. And another rumored departure — Commerce Secretary Howard Lutnick — is Trump’s longtime friend. FBI Director Kash Patel, who has published several fawning children’s books about Trump, may also be on the chopping block next.
I, for one, will be pouring out a glass of sauvi B for them.
2026-04-22 18:30:00
2025年7月7日，海地太子港的“圣母永援教堂”举行了一名警察的葬礼仪式。| Guerinault Louis/Anadolu via Getty Images
假设你是一名正在纽约度假的外国公民，而你的祖国突然爆发内战，政治异见者和无辜平民被大量杀害。与此同时，你的美国旅游签证即将到期，返回祖国可能面临死刑。1990年联邦法律为面临此类困境的外国公民提供了人道主义庇护，允许美国国土安全部（DHS）在本国发生武装冲突或自然灾害等危机时，为符合条件的非公民提供“临时保护地位”（TPS）。TPS持有人可在美国合法工作，但需注册且不得有严重犯罪记录或与毒品走私、恐怖主义有关联。
特朗普政府对TPS持敌对态度，其上任后通过行政命令将TPS设计国缩减至13个，包括也门、索马里、埃塞俄比亚、海地、缅甸、南苏丹、叙利亚、委内瑞拉、洪都拉斯、尼加拉瓜、尼泊尔、喀麦隆和阿富汗。此举引发两起最高法院案件：Mullin v. Doe（涉及叙利亚）和Trump v. Miot（涉及海地）。原告认为特朗普政府未遵循TPS法律中的程序要求，例如未充分咨询相关部门，且以“国家安全”为由终止TPS，实则出于政治动机。
特朗普政府的法律依据是联邦法条，禁止法院审查其关于TPS的决定。但原告指出，该条款仅限制对“指定”或“终止”TPS的审查，而非所有程序。他们援引1991年McNary v. Haitian Refugee Center案的判例，认为法院仍可审查程序合规性。尽管如此，最高法院此前已通过“影子法庭”驳回了部分支持TPS的下级法院裁决，且特朗普政府认为联邦法严格限制法院干预TPS政策。
若原告成功主张程序合法性，可能迫使特朗普政府重新审查TPS终止程序。但即便如此，TPS受益者仍可能面临被驱逐的命运，因为特朗普任期尚余两年，且新任DHS部长Markwayne Mullin可修正程序错误。尽管如此，最高法院若裁定反对特朗普，仍可能为海地和叙利亚的TPS持有人争取数月的庇护期。然而，无论法院如何裁决，TPS受益者的未来仍充满不确定性，法律赋予特朗普政府的广泛裁量权可能使他们难以获得长期保障。
Imagine that you are a foreign national vacationing in New York when a civil war breaks out in your home country. Political dissidents, as well as bystanders who are unfortunate enough to get in the way of the warring factions, are being killed by the thousands. Meanwhile, the tourist visa allowing you to remain in the United States will expire soon, and returning home could mean a death sentence.
A 1990 federal law offers humanitarian relief to many foreign nationals who face this kind of dilemma. Under the law, the Department of Homeland Security (DHS) may offer “temporary protected status” to noncitizens who are already present in the United States during an “armed conflict” in their home nation, or if a natural disaster or some other catastrophe has made their home country unsafe. (Prior to 1990, foreign nationals in these circumstances could sometimes remain in the US under a program called “extended voluntary departure.” The 1990 law formalized the process that determines who may stay.)
As the program’s name suggests, temporary protected status (TPS) is temporary. DHS is supposed to periodically review the list of countries whose nationals may seek this status, and to remove countries from the list once the humanitarian crisis abates. TPS holders must register, and they are ineligible if they have a felony conviction, more than one misdemeanor conviction, or if they have ties to drug trafficking or terrorism. People with TPS status may work in the United States during their temporary residence.
The Trump administration, as part of its harsh overarching approach to immigration, is hostile to the TPS program. On his first day back in office, President Donald Trump issued an executive order with the hyperbolic title “Protecting the American People Against Invasion.” Among other things, Trump ordered his Cabinet to ensure that TPS designations “are appropriately limited in scope and made for only so long as may be necessary to fulfill the textual requirements of that statute.”
Since then, Trump’s administration terminated TPS designations for all 13 countries whose designations were due for a review. In some cases, it did so before the review was supposed to occur, and before the country’s previous designation had expired. (The full list of 13 countries includes Yemen, Somalia, Ethiopia, Haiti, Burma, South Sudan, Syria, Venezuela, Honduras, Nicaragua, Nepal, Cameroon, and Afghanistan.)
And that brings us to Mullin v. Doe and Trump v. Miot, two Supreme Court cases asking whether Trump’s apparent decision to cancel the TPS program is lawful. Doe concerns the Trump administration’s decision to strip TPS designation from Syria, a country that was recently in a civil war and that ousted its president in 2024. Miot involves Haitian nationals who fear returning to a country without a stable government, and where much of the country is controlled by criminal gangs. Both cases will be argued on April 29.
Realistically, both cases are likely to end badly for Syrian and Haitian nationals (and for other TPS beneficiaries). The Supreme Court has already blocked other lower court orders protecting TPS holders on its “shadow docket,” cases that the Court decides on an expedited basis, and the Trump administration is correct that federal law strictly limits the judiciary’s power to interfere with its decisions regarding TPS policy. (Although, as Linda Greenhouse points out, the Court has not yet blocked the lower court decisions benefiting Haitians and Syrians, so that is a point in the Doe and Miot plaintiffs’ favor.)
As the lawyers representing Syrian nationals write in their brief, federal law “forecloses challenges asserting that TPS must be extended because a country remains unsafe.” Nevertheless, they also argue that the Trump administration did not comply with “procedural mandates” that are written into federal immigration law, and that these mandates may be enforced by the courts.
Even if the Supreme Court agrees that these procedural mandates may be enforced, however, that will merely delay a reckoning over TPS. If the justices rule that Trump or his subordinates must jump through certain procedural hoops before they strip TPS protections from citizens of a particular country, the Trump administration can always just jump through those hoops.
Still, a procedural delay isn’t nothing. In the best-case scenario for Syrians and Haitians who rely on TPS, such a delay could allow them to run down the clock on the Trump administration, in the hopes that the next president will be less hostile toward immigrants and other foreign nationals in the United States. And, even if they can’t secure such a long delay, every day that TPS is in effect is a day when they won’t be deported to a place where they could be killed.
The Trump administration’s brief relies heavily on a federal law which provides that “there is no judicial review of any determination of the [Secretary of Homeland Security] with respect to the designation, or termination or extension of a designation” that a particular nation’s citizens are eligible for TPS. Trump’s lawyers argue that this statute prohibits the courts from hearing Doe and Miot altogether.
The plaintiffs, meanwhile, argue that this bar on judicial review is less total that it initially seems. Their primary argument is that the word “determination” appears multiple times in the TPS statute, and that it means only the DHS secretary’s conclusion that a particular country is or is not safe enough to allow citizens of that nation who are in the United States to return is forbidden from judicial review. Thus, they argue that other provisions of the TPS statute — including provisions that require the Trump administration to comply with certain procedures before it removes anyone’s eligibility for TPS — may be enforced by federal courts.
The plaintiffs also point to a handful of previous Supreme Court decisions, including McNary v. Haitian Refugee Center (1991), which read similarly worded bars to judicial review narrowly.
Although this argument persuaded many lower courts, it is far from clear that it will prevail in the Supreme Court. Again, the Court has already blocked lower court decisions that sought to extend the TPS program against the Trump administration’s wishes on its shadow docket, though some of those decisions were joined only by the Court’s Republican majority.
That said, the Doe plaintiffs also raise a policy argument that might convince some of the Republican justices that a total bar on lawsuits challenging the federal government’s TPS decisions would be undesirable. If no TPS-related decision by the Trump administration may be reviewed by the courts, that should also mean that a future administration could falsely claim that, say, Mexico has become so dangerous that its nationals should be eligible for TPS, and that it could do so “explicitly to accomplish mass legalization.”
So let’s assume that the TPS plaintiffs do prevail on this jurisdictional argument, and win the right to argue that the Trump administration must comply with certain procedures before it takes away anyone’s TPS status. What happens then? Even in the best reasonable case for these plaintiffs, they probably only gain a short-lived reprieve.
Assuming that the Doe and Miot plaintiffs do convince the Supreme Court to consider their procedural arguments, those plaintiffs raise several objections to the process this administration used to slash the TPS program.
The TPS statute, for example, requires the DHS secretary to consult “with appropriate agencies of the Government” before it determines whether a country should be removed from the list of nations whose citizens may claim TPS status. But, according to the plaintiffs, the Trump administration’s consultation process was extremely truncated, and seems to consist of a single email exchange between a DHS official and a State Department official, where State said in a one-paragraph email that it has “no foreign policy concerns” about rolling back TPS.
The plaintiffs also fault then-Secretary Kristi Noem for claiming that America’s “national interest” required rolling back TPS, even regarding countries that remain dangerous. While the TPS statute does require the DHS secretary to consider “the national interest of the United States” in its initial decision to designate a particular country for TPS, once a nation is on the list, the secretary is not supposed to consider this factor.
As the Miot plaintiffs argue, “it makes sense that Congress would give the Secretary discretion when making an initial designation but limit her discretion to terminate an existing designation” because, once someone has been given TPS status, they develop a “reliance interest.” It’s one thing for the United States to tell someone that it will not provide them with shelter in the first place. It’s far crueler to give them shelter, allow them to spend years building a life in the United States, then abruptly order them to resettle to another country or face deportation to a war zone.
Finally, the plaintiffs argue that Noem did not “base her decision on ‘a good-faith and objective review of country conditions,’” and that the real reason why she terminated so many people’s eligibility for TPS status was political. When Noem first announced that she was reducing the TPS program, she said that she was “getting direction” from Trump, and that “he is pausing this program.”
Rooting her decision in political conditions, the plaintiffs argue, violates the TPS statute, which required her to only consider whether “the conditions in the foreign state” still justify maintaining TPS status — that is, whether the country remains too dangerous to allow its citizens to return there.
There is one precedent where the Supreme Court sided with immigrants against the Trump administration, due to a procedural error by Trump and his subordinates. In Department of Homeland Security v. Regents of the University of California (2020), the Court reinstated the Deferred Action for Childhood Arrivals (DACA) program — which allows hundreds of thousands of noncitizens to live and work in the United States — thanks to a paperwork error by the Trump administration.
But Regents was also a 5-4 decision, with Chief Justice John Roberts joining the Court’s four Democrats. One of those Democrats, Justice Ruth Bader Ginsburg, died later that year and was replaced by Republican Justice Amy Coney Barrett. So it’s unclear whether Regents would come down the same way if it were decided today.
And even if the Court does reach an outcome similar to Regents in the Doe and Miot cases, now-DHS Secretary Markwayne Mullin could cure Noem’s procedural errors by running the TPS terminations through whatever process the Supreme Court deems adequate. And, unlike Regents, which was handed down in the final year of the first Trump administration, Trump currently has more than two years remaining in his term. So it is unlikely that TPS beneficiaries can run out the clock.
Again, a Supreme Court decision against Trump would still be a victory for Syrians and Haitians in the United States — and potentially for all TPS beneficiaries — because it would likely mean several months of safety while Mullin cures Noem’s procedural errors. But the law in this space does give Mullin a fair amount of authority that can’t be challenged in court.
No matter what the Supreme Court does, in other words, the future is probably very grim for many foreign nationals who have taken refuge in the United States.
