2026-01-11 20:00:00
一群训练师正在温哥华国际机场训练服务犬。| 李森/新华社 图片由盖蒂图片社提供
上个月，我像许多美国人一样，乘飞机回家过节。在从纽约飞往洛杉矶的航班上，一只佩戴着“服务犬”背心的狗在登机口对着我叫。这只狗（没有名字）看起来像一只胖乎乎的法国斗牛犬，来回踱步，还对着其他旅客吠叫。从洛杉矶机场返回时，我注意到更多佩戴服务犬背心的狗——有腊肠犬、另一只法国斗牛犬，还有一些混种犬——和他们的主人一起排队，等待与地勤人员接触。这让我意识到，如今越来越多的狗被指定为服务犬，多到不可能每一只都经过充分训练。其中一些狗可能是冒充的。
当然，由于假期期间很多人乘飞机，因此更容易发现这些情况；但我显然不是唯一注意到服务犬数量上升的人。这种现象引发了一些问题：为什么会有这么多服务犬？为什么这么多人拥有它们？获得认证是否很容易？真的有那么多人需要它们吗？为什么这只狗会对着我叫？这些人的目的是否只是想带狗出行？如果对某些人产生怀疑，我是不是很糟糕？假服务犬真的那么糟糕吗？
遗憾的是，我无法采访一只真正服务犬，以探讨这一争议话题。但我与专家、空乘人员和服务犬训练师交谈，他们表示，这些冒充的服务犬让他们的工作以及真正服务犬的工作变得更加困难。
越来越多的人希望带着宠物一起旅行，尽管航空公司声称安全，但主人仍对将宠物放在货舱中感到担忧。此外，带着狗登机的成本较高，而且也存在诸多不便。与此同时，在美国，将宠物狗带入机舱的规定非常严格，官方要求狗狗必须能适应指定的航空箱，能够转身，并且全程必须系好安全带。如果狗狗符合这些要求，大多数美国主要航空公司每段航程的费用大约为150美元。
然而，由于这些规定过于繁琐，许多主人选择通过服务犬的“漏洞”来带狗登机。服务犬可以免费进入机舱，而无需遵守这些规定。这使得服务犬成为一种“捷径”。
“我认为很多人开始利用我们希望狗狗能和我们在一起这一点，”Canine Companions（一家训练并安置服务犬的组织）的项目主管Jessica Reiss说道。在Canine Companions，拉布拉多犬、金毛犬以及拉布拉多和金毛的混种犬会接受为期六个月的训练，包括执行大约45项任务，如开门、响应警报、拉轮椅和物品识别等。服务犬的主人也会接受一个密集的培训课程。
尽管Canine Companions的培训非常严格，但像这样的项目并不是标准做法。部分原因在于，目前没有统一的标准。Reiss告诉我，尽管美国运输部试图阻止滥用宠物旅行规定（例如，禁止情绪支持动物），但人们仍然找到各种方法来规避这些限制。
“有一种漏洞，允许你私人训练你的狗成为服务犬，而根据定义，这意味着狗必须能够执行减轻残疾人障碍的任务，”Reiss说道。虽然私人训练可能更灵活和可及（例如，对无法负担训练师或没有附近训练师的人来说），但也意味着更多人会利用缺乏监管的漏洞。
“有很多主人训练的、行为良好的服务犬，他们训练狗执行实际的物理任务，他们应该享有通行权。但我也认为，很多人其实不想把狗留在家里，”Reiss说。
这种混乱和不一致的结果，就是为什么会出现像那只叫个不停的法国斗牛犬这样的情况，它和那些经过Canine Companions训练、社会化和服务犬享有同样的飞行权利。这也是为什么有很多关于“服务犬”在飞机上表现不佳的令人沮丧的故事。
我采访了几位美国空乘人员，他们证实，最近服务犬的数量确实增加了。但他们都表示，除了文件之外，他们不能询问主人任何问题，即使他们对一只吵闹的哈士奇幼犬有所怀疑。一位匿名的空乘人员说：“我确信这只老年的吉娃娃不会拯救任何人，但我的职责不是去验证这些事情。”
然而，问题变得更加复杂，因为没有人愿意被怀疑或质疑是否对待残疾人有偏见。那么，如何在不让人感到被攻击或被贬低的情况下，区分真正的服务犬和那些通过漏洞进入的狗？
作为一名拥有足够小的狗，可以作为随身携带物品的主人，我似乎没有理由遵守航空公司的规定。遵守航空舱内带狗的规定不仅花费更多（服务犬可以免费登机），而且会让飞行更加拥挤（被锁在航空箱中，而不是像服务犬那样可以躺在地板上或主人腿上）。
如果按照规定带狗登机如此随意且令人不快，而冒充服务犬的方式却相对容易且免费，那么遵守规则又有什么意义呢？
“说实话，这些规定根本不重要，”一位患有脑瘫并拥有服务犬Slate的女性Molly Carta告诉Vox。“我有时也这么想。我花了50美元去兽医那里做检查，只是为了填写一份表格，而旁边的人可以直接带着他们的狗上飞机。”
Carta表示，她每年旅行两到三次，过去十年中，她看到服务犬的数量迅速增长，尤其是在过去三到五年间。由于法律上没有官方的服务犬注册系统，这种增长更加明显。
她通过Canine Companions找到了她的服务犬Slate，这是她的第二只服务犬。最近，他们从康涅狄格州飞往威斯康星州，途中在芝加哥奥黑尔机场转机。
“在那个机场，有太多其他狗，以至于从我们的登机口走到下一个登机口都成了噩梦，”她告诉我，提到有多只狗试图与Slate互动、吠叫并靠近他。
尽管Slate被训练得在飞行中保持专注、静止和冷静，但干扰会让他的工作变得更加困难，甚至可能影响他在紧急情况下的帮助能力。
Carta使用轮椅和助行器，她表示，这种干扰也给Slate带来了不必要的压力。“如果我要和一群朋友一起出行，我通常不会带他，因为这可能不值得带来的压力。如果我知道周围有能帮助我的人，那我就会考虑带他。”她说。
Carta还经常担心自己在飞机上的座位安排。在她的经验中，残疾人和服务犬通常被安排在过道座位。如果有多位残疾人带着服务犬，谁会获得这些座位？这一排会不会有多个狗？
Carta对是否带服务犬出行的疑虑，似乎反映出旨在帮助她的规则存在缺陷。她还提到，她常常感到自己处于防御状态，因为人们会质疑Slate是否真的是服务犬，这可能源于他们之前与不守规矩的狗和滥用特权的人的经历。
但除非人们认识像Carta这样的人，否则很难理解她的经历会受到那些认为自己只是“轻微违规”的人影响。长期以来，Carta认为通过教育人们了解服务犬是医疗需求，是解决问题的办法。但随着时间的推移，她逐渐意识到，如果人们不愿意倾听，仅仅提高公众意识是不够的。
尽管Carta希望有相关立法，但如何在不损害真正需要服务犬的人的情况下，理清服务犬的混乱现状，仍然是一个难题，尤其是在如此多的人已经滥用这一漏洞的情况下。
“我不知道立法会是什么样子，但也许可以阻止那些不真正需要服务犬的人占用资源，”Carta说。“这关乎于认识到服务犬是一种医疗需求。”
也许最难克服的障碍就是个人的自私心理。在如此糟糕的旅行体验中，很难把别人放在自己之前，而带狗度假似乎也无害。在那一刻，没有人会想到社会契约或自己的狗可能对其他人造成影响。教导人们这种同理心，即使是服务犬也无法做到。
This past month I, like many Americans, flew back home for the holidays. On the first leg of that trip, from New York to Los Angeles, a dog in a “service dog” vest barked at me at the gate. The dog (not its given name), looked to be a stout French bulldog, paced back and forth, and yapped at a couple of other travelers.
On the way back from LAX, I noticed more dogs in service vests — a dachshund, another (different) Frenchie, a few mixed breeds — in line with their humans, waiting for desk agents. It all made me realize how many dogs traveling these days are designated service dogs, so many that there’s no way each one was a thoroughly-trained working canine. Some of these pooches had to be impostors.
Granted, because so many people fly during the holidays it was probably easier to spot them; but I’m obviously not the only person who’s noticed the rise of questionable, if not fake service dogs. Their proliferation raises a few questions.
Why are there so many? Why and how do so many people have them? Is certification that easy to get? Do this many people need them? Why is this one barking at me? Are these people who just want to take their dog on their trip? Does being suspicious of some of them make me awful? Is a fake service dog really that bad?
Sadly, I could not speak to an actual service dog for an interview regarding this contentious subject. But I did talk to experts, flight attendants, and people who train service dogs about how canine service impersonators make their job and the jobs of actual service dogs that much harder.
More and more people want to travel with their pets, and despite airline assurances about safety, owners still harbor some overall worry about traveling with their animals in cargo. They’re also managing the reality that boarding a dog can be expensive and comes with its own set of worries.
At the same time, traveling in the US with a pet dog in cabin — thanks to a multitude of rules — is actually difficult. Officially, pups must be able to fit in an approved carrier that fits underneath the seat in front of you. They must also be able to turn around in said carrier and must remain zipped up the entire time. If a dog fits all those requirements, it’ll cost roughly $150 per leg of the trip on most major US airlines.
Essentially, there’s a glut of people who want to travel with their dogs, and the only way they can is only available to small ones. Even then, not every small dog is happy to be in a secured carrier. And if there’s any certainty about people, it’s that some of them will find a way to get what they want.
“I think a lot of people started to take advantage of the fact that we really want our dogs to be with us,” says Jessica Reiss, the program director at Canine Companions, an organization that trains and places service dogs with people living with disabilities.
At Canine Companions, Labrador retrievers, golden retrievers, and Labrador-golden crosses (goldens and Labradors are two of the “fab four” breeds that experts say excel at becoming service dogs) undergo a six-month training program that includes responding to roughly 45 or so tasks that include opening and closing doors, responding to alarms and alerts, pulling wheelchairs, and item identification. Service dog recipients complete an intensive program as well.
“In order to place a dog with a person, that person comes in and stays with us for two weeks. They are literally living, breathing, everything with the dog 24 hours a day — [they’re taught] dog behavior, dog body language, how to deal with fear reactivity as the typical dog owner,” Reiss says, listing off just a few things that a person learns in those 14 days.
While training at Canine Companions is rigorous, programs like it are not the standard. Part of the problem is that there is no standard.
Reiss explained to me that even though the Department of Transportation has tried to stifle the travelers abusing pet travel (e.g., disallowing emotional support animals) and the US has made service animal designation seemingly stricter, people still find ways to circumvent those restrictions.
“There’s this loophole that says, you can privately train your dog to be a service dog, and by definition what that means is the dog has to be able to provide tasks that mitigate a person’s disability,” Reiss says. While private training can be more accommodating and accessible (i.e., for those unable to afford a trainer or who don’t have a trainer close by), it also means that more people take advantage of the lack of regulation.
“There are plenty of owner-trained, well-behaved service dogs, and they are training their dogs to do actual physical tasks, and they should be given access. But I think we’re also talking about a lot of people not wanting to leave their dogs at home,” Reiss says.
This result is a lot of confusion and lack of consistency. That’s how you get dogs like the barking Frenchy in a service vest that receives the same flying privileges as a dog that Canine Companions bred, socialized, and trained. It’s also why there are so many frustrating anecdotes of “service dogs” misbehaving on planes (and on land too).
I spoke to a handful of US flight attendants who confirm that they’ve seen an uptick in service dogs on flight. But they consistently noted that beyond paperwork, they’re instructed not to ask owners any questions, even though they might have suspicions about a rowdy, howling husky puppy. One who wished to remain anonymous put it to me this way: “Surely this geriatric Chihuahua is not saving anyone’s life…but it’s not in my job description to verify those things.”
That said, it’s even more complicated, because no one wants to be a person who treats someone with a disability with suspicion or doubt. How do you distinguish real service dogs from those sneaking in via the loophole without making someone feel attacked or dehumanized?
As an owner of a dog small enough to fit as a carry-on, there doesn’t really seem to be any benefit to following the airline rules. Following all the air cabin regulations for dogs costs more (service animals fly for free) and makes flying more claustrophobic (being zipped up in a carrier versus service animals who lay on the cabin floor or on a lap). If the “right” way to get a dog onboard is so arbitrary and unappealing, and the faux way is relatively easier and free, what’s the point in following the rules?
“That’s the thing, the rules don’t even matter,” Molly Carta, a woman living with cerebral palsy who has a service dog named Slate, tells Vox. “I feel that way half the time too. I’m like, why did I pay $50 for this vet visit to get this form filled out? This person over here is just going to walk on with their dog.”
Carta explained to me that she travels two to three times per year, and has seen the number of service dogs boom in the past decade, with the largest increase coming over the past three to five years. (By law, there is no official registry of service dogs.) Slate, whom she matched with through Canine Companions, is her second service dog, and recently they traveled from Connecticut to Wisconsin and made a connection in Chicago through O’Hare.
“There were so many other dogs in that airport that it was such a nightmare to even just get from our gate to the next gate,” she tells me, noting that multiple dogs tried to interact with, bark at, and approach Slate. While Slate is trained to maintain focus, stay put, and stay calm during flights, distractions make his job in assisting Carta harder — possibly inhibiting his ability to help her during an emergency. Carta, who uses a scooter and a walker, explains that this also puts an ample amount of unnecessary stress on Slate.
“If I’m going somewhere with a bunch of friends, a lot of times I won’t travel with him because it’s probably not worth the stress. If I know I have a bunch of people around that can help me in the same ways that he would,” Carta says.
Carta also often worries about where she’s placed on a plane. In her experience, people with disabilities and service dogs are seated in the bulkheads. Hypothetically, if there’s multiple people with service dogs, who gets that seat? And will there be multiple dogs in that row?
Carta having doubts about taking her service dog with her traveling sure seems like a failure of rules meant to help her and other people living with disabilities. She also mentioned that she tends to feel like she’s on the defensive because of people questioning whether Slate is an actual service dog — likely due to their prior experiences with unruly pups and people abusing the privilege. But unless people know someone like Carta in their lives, it’s hard to connect how her experience would be impacted by someone thinking they’re harmlessly fudging the rules.
For a long time, Carta believed that educating people about how service dogs are a medical need was the answer. But the more and more time that passes, the more she’s realized that more public awareness doesn’t work if people aren’t willing to listen. And while Carta hopes for legislation, untangling the knot of service animals without doing more damage to the people who need them is tricky too, now that so many people have abused the loophole.
“I don’t know what that legislation would look like, but maybe something that dissuades people from taking away from those of us that really need service dogs,” Carta says. “It’s about recognizing that they are a medical need.”
Perhaps the most difficult obstacle to overcome is plain individual selfishness. It’s hard to put other people ahead of yourself, especially in a situation as miserable as air travel, and taking your dog on vacation seems harmless enough. In that moment, no one is thinking about any kind of social contract or how their accompanying pooch could affect someone else down the line. Teaching someone that kind of empathy is something a dog, service or not, can’t even do.
2026-01-11 19:30:00
Suno的联合创始人迈基·舒尔曼（Mikey Shulman）使用AI音乐应用创作歌曲。据法国音乐流媒体服务Deezer称，每天有大约5万首完全由AI生成的歌曲上传至其平台。虽然这些歌曲大多不会被广泛传播，但过去一年中，一些歌曲获得了数百万次播放。这引发了一个问题：如果我们的未来充满这种AI音乐，那未来会是什么样子？
Deni Béchard是《科学美国人》的高级科学作家。他使用Suno这个AI音乐应用创作了整整一个月的音乐，并只允许自己聆听自己生成的AI歌曲。他表示，这项实验是为了更深入地思考我们未来如何与这类音乐互动。Béchard与Today, Explained的主持人诺埃尔·金（Noel King）讨论了他在这段时间内的发现，以及他的AI创作与人类创作的音乐相比如何。
他提到，自己通常会输入一个提示，然后生成两首歌曲，尽量发挥创意，多次尝试不同的提示，加入不同的乐器或人声。其中有一首名为“器官贩卖”的歌曲让他印象深刻，他要求生成一首带有女性人声和幽默讽刺歌词的当代说唱歌曲，结果AI生成了一首以器官贩卖为隐喻的歌曲，令他感到惊讶。
他意识到，主流音乐往往经过大量处理，是为了迎合市场而设计的，缺乏个人色彩。因此，他发现AI生成的音乐在某些情况下与人类创作的音乐并没有太大差别。当被问及是否能分辨AI生成的歌曲和人类创作的歌曲时，他表示自己可能分辨不出来，这说明AI已经非常先进了。
他还提到，他注意到在Spotify上流行的AI音乐往往具有强烈的情感和真实感，比如Xania Monet、Solomon Ray或Cain Walker（实际上是一个AI虚拟形象）的歌曲，以及Breaking Rust的“Livin’ on Borrowed Time”。这些歌曲让人感觉像是由真实经历和情感驱动的，因此更具感染力。
他最初对AI创作的音乐感到困惑，因为无法想象机器能有情感、经历痛苦或在深夜创作歌曲。但随着时间推移，他逐渐适应了这种体验，不再纠结于“是谁创作的”这个问题。他设想如果AI是一个具有个性的虚拟角色，比如存在于元宇宙中的虚构人物，那么他或许会更容易接受这些音乐。
他最后表示，这项实验让他对AI有了新的认识。他认为，在未来10到20年里，很多青少年可能会觉得我们现在讨论的AI音乐问题非常正常，甚至感到困惑，为什么有人会对此感到不安。他相信人类会很快适应AI音乐的存在。同时，他也希望艺术家们能得到充分的保护和合理的报酬，但AI音乐的融入将比我们想象的更加自然和顺畅。
According to the French music streaming service Deezer, there are about 50,000 fully AI-generated songs uploaded to its platform every day. Many of these songs won’t reach a wide audience, but over the past year, a few have gained millions of listens.
Which raises the question: If our future is going to be filled with this kind of AI music, what does that future sound like?
Deni Béchard is the senior science writer at Scientific American. For the better part of a month, Béchard has only allowed himself to listen to his own AI-generated music using the AI music app Suno. He says the experiment is an attempt to think more critically about how we might engage with this kind of music in the future.
Béchard spoke with Today, Explained host Noel King spoke about what he’s learned so far and how his AI creations stack up to human-made music. The conversation has been edited for length and clarity.
There’s much more in the full podcast — including snippets of Béchard’s songs — so listen to Today, Explained wherever you get your podcasts, including Apple Podcasts, Pandora, and Spotify.
Alright, so you’re using Suno, you said, to create the songs.
I come up with a prompt and I’ll plug it in, and each prompt makes two songs, and I’ll try to be as creative as possible. I’ll usually plug it in two or three times and vary it, add different kinds of instruments or different kinds of vocals, and just plug a bunch of those in. One that made me laugh was a song called “Organ Trafficking.” I had asked for a contemporary rap song with female vocals, and I had asked for playful, ironic lyrics, and it comes up with this song, where organ trafficking is kind of the central metaphor. I was pretty surprised.
I think one of the things I’ve realised is that a lot of the music I listen to that is mainstream is, I would consider, heavily processed music — music that’s designed to have a large market. And it doesn’t feel very personal to me anyway, so I realized that in that particular context, [the music I made with AI] didn’t feel very different a lot of the time.
Do you think if someone had handed you a playlist of 10 songs, five are AI, five are not, do you think you’d be able to tell the difference?
No, I don’t think so.
Wow. And what does that tell you?
I mean, it tells me that the AI is getting very good.
One thing I noticed during this process was that a lot of the AI music that is popular, that people are listening to on Spotify that has millions of listeners [are] songs that are very soulful, very gritty.
It’s like Xania Monet or Solomon Ray or Cain Walker’s “Don’t Tread on Me” — and Cain Walker’s not a person. It’s an AI avatar, right? Or Breaking Rust’s “Livin’ on Borrowed Time.” Those songs all feel just really authentic. This person really suffered through these things and felt these things. That’s how they come across.
I think that AI tends to work best when it just leans into that authenticity because it kind of helps overcome the cognitive dissonance that we’re thinking, This isn’t really a deeply felt song, and it moves away from mainstream human-generated music — human-made music — which is often very heavily designed to be a summer hit or to go viral in some way. And it often doesn’t have that level of authenticity, that feel of authenticity. I think when AI replicates that, we’re more aware of it being superficial or artificial, because there’s already an element of artificiality there.
Do you think when your experiment is done, you’re going to keep making AI music?
I think I probably will.
Oh my god, you love the power.
I think, you know, what has surprised me with it is, I’ll be walking somewhere, and I’ll think, “What if I were to ask it to combine these styles or put a banjo with a hip hop track and add this kind of vocals? What would I get?” I get curious now.
I would say now I’m at the point where I don’t worry about the connection to the human. I did in the beginning. In the beginning, I was really like, “Who’s this person?” When you’re reading a book and you’re halfway through the book and you think, “What human mind did this book come out of?” And you turn the book over and you look and see who the author was, and you Google them and you’re like, “How in the world did they think of this?”
I just had that impulse so often in the beginning to want to know who felt this, who thought this. I just would have cognitive dissonance. I’d be going, “This is a machine. This machine did not fall in love. This machine did not suffer these experiences. This machine did not wake up at two in the morning and write this song just needing to express itself.” It was actually really bothering me. It kind of would block me from being able to enjoy the song.
And I thought, “Well, if somebody created an AI avatar and gave it a personality and they were a fictional character that existed in the Metaverse, and that AI avatar was a songmaker and it was singing this song, would that make it easier?” And weirdly, it would. It would make it a little easier. And so I kind of was just imagining these AI avatars, and I’m like, “Okay, I’m imagining a fictional character singing this song.” And that lasted maybe four or five days, and then I just got used to listening to the music, and I stopped thinking about it.
Does doing this experiment and seeing how you’re reacting to this music change how you think about AI at all?
I think my conclusion from this is that in 10 or 15 or 20 years, there are going to be a lot of teenagers who look at the discussions we’re having right now and go, “What are these people talking about? This is totally normal. Why would anybody feel so conflicted about this?”
I think we’re going to adapt to it pretty quickly. That is my gut feeling. There are a lot of big questions around the creators and protecting artists and what it means to be an artist. There are a lot of questions that are going to come out of this, and I really hope that artists are as protected as possible and remunerated properly. But I think this is going to fit into our lives a lot more smoothly than I think we’re realizing at the moment.
2026-01-10 20:30:00
最近发生了一件最奇怪的事情。我联系了一家客服部门，结果体验非常好。我发了一封邮件，很快就得到了回复，并成功获得了退款。这次积极的解决问题经历最特别的地方在于，我无法确定是否还有其他人类参与其中。然而，我短暂地意识到，预言终于实现了：AI终于让我的投诉变得更容易，并且能获得更好的结果。至少这是我愿意相信的。客服本来就是AI可以轻松处理的事情之一。事实上，这次良好的体验是由一家以AI为核心的企业Intercom提供的。他们有一个名为Fin的AI代理，负责处理大部分客户的咨询。但为什么不是全部呢？“我确信，现在很多通过电话或电脑进行的客服工作，这些岗位的人将会失业，而AI将做得更好。”OpenAI首席执行官Sam Altman在9月份对Tucker Carlson说。Altman并不是唯一一位推动自动化客服的硅谷高管。去年，Salesforce削减了4000个客服岗位，转而采用AI工具；Verizon则推出了基于Google Gemini的聊天机器人作为其客服的入口。还有Klarna，其CEO曾吹嘘用AI取代人类客服，但随后在5月份收回了说法，开始招聘更多的人类客服人员。这正是问题所在。事实证明，AI，尤其是生成式AI，在某些方面确实非常出色，但一旦遇到困难，就可能表现不佳。这就是为什么你仍然需要验证ChatGPT提供的所有信息，也是为什么尽管聊天机器人在诊断某些医疗状况方面表现良好，但它们还不能取代人类医生。在客服方面，AI擅长处理简单任务，比如退款，但在处理更复杂的案例，尤其是当客户情绪激动时，AI却可能表现得很糟糕。正如《Anchorman》中所说：“百分之六十的时间，它总是能奏效。”然而，人类客服人员正在大量被AI取代，过去几年里，无论是美国还是其他国家，这一趋势都十分明显。无论是为了降低成本还是为了显得高科技，很多公司都推出了AI客服聊天机器人作为客户的第一接触点，但很快他们就意识到，客户其实并不喜欢这种做法。据Gartner客户服务与支持领导者研究主管Brad Fager表示，这些公司现在正在撤回这些计划。“用AI取代员工的想法根本不可行，甚至也不可取，”Fager告诉我，他指出高管们可能认为用AI取代人类客服是降低成本的好方法。“但现实是，这根本行不通。”此外，也有证据表明客户并不喜欢与AI互动。2024年Gartner的一项调查显示，61%的客户更希望公司完全不使用AI来处理客服问题，而其中53%的客户表示，如果公司使用AI，他们可能会考虑转投竞争对手。正如Fager向我解释的那样，Gartner认为AI和自动化将改变未来的客服模式，但人类仍将在这个过程中发挥重要作用。许多AI整合工作将发生在后台，帮助人类客服人员更好地完成工作，而不是直接与客户互动。客户可能永远不知道AI是否参与其中。这让我想起了几年前MIT和斯坦福研究人员的一项研究，他们探讨了生成式AI如何提高客服人员的工作效率。结果显示，AI主要帮助了经验较少的客服人员。通过实时提供处理电话咨询的建议，这些员工每小时能解决14%更多的问题。这个工具是基于经验丰富的客服人员的数据进行训练的，甚至还能帮助新手客服人员更富有同理心地对待客户。相比之下，你可能体验过的聊天机器人更像是一个AI版的电话菜单。你向客服机器人求助，却只能看到一系列选项，让你不断缩小问题范围，最终找到可能由AI处理的正确代理人。这与令人恼火的传统电话菜单（如“按1查询账单，按2查询技术支持”）如出一辙。这些前端的AI解决方案本质上是将AI工具附加在旧客服系统上，效果却非常糟糕。波士顿大学市场营销教授Werner Kunz认为，很多公司只是为了尝试AI而这么做。“这并不奏效，”他告诉我，“与旧系统相比，AI的失败率太高了，如果公司现在只是用AI做这些事情，我认为这会破坏客户关系。”Kunz还指出，如果将AI用于后台，可以在更安全的环境中获得更好的效果。他还补充说：“谁在乎你是否使用AI呢？”这让我回想起最近一次令人意外的积极客服体验。我联系了Intercom公司，确认是他们的AI代理Fin解决了我的问题。整个过程没有电话菜单的困扰，也没有和聊天机器人“搏斗”才能找到真人客服的经历。Fin记录了我的投诉，用听起来像人类的邮件向我提供了解决方案，甚至在合适的语境中使用了表情符号，而且在我还没来得及感到烦躁之前就解决了问题。但说AI让客服终于变得好起来，这似乎并不完全准确。正如Kunz和Fager所解释的，很多公司错误地使用AI，或者将其附加在旧系统上，导致效果不佳。然而，Intercom联合创始人兼首席战略官Des Traynor表示，全心全意投入AI是为客户提供他们想要的东西——即时结果的最佳方式。“你不想等待，”Traynor说，“这就是人们在打电话之前先上网搜索的原因：人们只想立即解决问题，而AI正好能提供这一点。”他补充道：“当它起作用时，它对用户来说就是更好的。”Traynor承认，AI带来了软件时代，让人对它的效果产生怀疑，而这一问题也引导了Fin的开发。他说，他的公司投入了大量时间来构建一个AI评估引擎，并对每个版本进行“严酷测试”，以确保Fin不会产生错误信息或做出错误判断。因此，Fin每周能解决一百万个客户问题，解决率为67%，虽然还不是100%，但Traynor表示这个数字每个月都在上升1%。他承认有些互动需要人类介入，但在大多数情况下，AI都能更好地完成任务。在我自己的案例中，确实如此。对消费者来说，最大的问题在于，你无法选择任何公司如何处理其客服问题。此外，还存在一种收入差距，大公司如亚马逊可以投入更多资源，提供更好的客服，而小公司如本地公用事业部门只能尽力而为。然而，可以肯定的是，一场变革正在发生。有迹象表明，向公司投诉变得更加容易，但也有强烈证据表明，许多公司仍会设法让投诉变得困难，尽管他们希望让投诉更容易。AI的到来是为了帮助改善服务，但前提是它不能先让事情变得更糟。本文的版本也发表在User Friendly通讯中。点击此处订阅，以免错过下一篇文章！
The weirdest thing happened to me recently. I contacted a customer service department and enjoyed it. I sent an email, heard back promptly, and got a refund. What was most notable about the positive problem-solving experience was the fact that I couldn’t tell if there was a human other than me involved.
It dawned on me, however briefly, that the prophecies were finally coming true. AI was finally making it easier for me to complain to companies and get results. At least that’s what I wanted to believe.
Customer service is supposed to be one of those things that AI can just do. Indeed, that one good experience was powered by an AI-first company called Intercom. They have an AI agent called Fin that handles most of its clients’ queries. Why not all of them?
“I’m confident that a lot of current customer support that happens over a phone or computer, those people will lose their jobs, and that’ll be better done by an AI,” OpenAI CEO Sam Altman told Tucker Carlson, of all people, in September.
Altman is hardly the only Silicon Valley executive pushing to automate customer service. Last year, Salesforce cut 4,000 customer service jobs in favor of AI tools, and Verizon launched a chatbot powered by Google Gemini as its front door for customer service. Then there’s Klarna, whose CEO bragged about replacing humans with AI before backtracking last May and launching a recruiting drive to hire more human customer service agents.
There’s the rub. It turns out that AI, and especially generative AI, is really good at doing some things…until it isn’t. That’s why you still have to fact-check everything ChatGPT tells you and why, even though they’re good at diagnosing certain medical conditions, chatbots can’t replace human doctors. When it comes to customer service, AI can be good at simple tasks, like issuing refunds, but terrible at handling more complicated cases, especially when customers are upset and could benefit from some human empathy. To quote Anchorman, “Sixty percent of the time, it works every time.”
Still, human customer service agents are losing their jobs to AI in large numbers, and have been for the last few years, both in the United States and abroad. Whether to cut costs or look cool, a lot of companies rolled out AI-powered chatbots as the first point of contact for customers, only to realize that customers actually hate this concept. Now, these organizations are pulling back from those plans, according to Brad Fager, chief of research for customer service and support leaders at Gartner.
“The idea that you could replace your workforce is really just not viable, and it’s not even preferable,” Fager told me, noting that executives might think replacing human agents with AI is a good way to cut costs. “The reality is it’s just not working.”
There’s also evidence that customers just don’t like interacting with AI. One 2024 Gartner survey found that 61 percent of customers would prefer companies didn’t use AI at all for customer service, and 53 percent of them would consider switching to a competitor if they did. As Fager explained to me, Gartner has broadly taken the stance that AI and automation will transform the future of customer service, but that humans will play a big role in that transformation. And to many customers’ delight, a lot of the AI integration will happen on the back end, helping human agents do their jobs better rather than leading interactions. The customers themselves may never know that AI was involved.
This approach reminded me of a study I read a couple of years ago from researchers at MIT and Stanford who looked into how generative AI improved productivity in call center workers. It did, mostly for the less experienced agents. With access to an AI tool that offered real-time suggestions on how to handle calls, the workers were able to resolve 14 percent more cases per hour. The tool had been trained on data from more experienced agents and could even help novice workers be more empathetic to customers.
Contrast this with what you’ve probably experienced with chatbots: the AI version of a phone tree. This is where you ask a customer service bot for help and are met with a menu of options prompting you to narrow down your request in order to get you to the correct, probably AI-powered agent. It’s a slightly updated version of the infuriating phone tree that asks you to say or press one for billing, two for technical support, and so forth.
These front-end solutions to identify customers and their needs are essentially AI tools bolted onto old customer service systems, and they’re awful. Werner Kunz, a professor of marketing at the University of Massachusetts Boston, argues that a lot of companies are doing this just to do something with AI.
“It doesn’t work very well,” he told me. “The failure rate is way too high in comparison to the older systems, and if this is what companies are using AI at the moment for, I think it destroys customer relationships.” Kunz added that using AI in the backend would provide better results in a safer environment, and also, “Who cares about if you use AI or not?”
Which brings me back to my recent, surprisingly positive customer service experience. I reached out to Intercom, the company that built the software, and confirmed that it was an AI agent that solved my problem. There was no phone tree analog and, in a sense, no fight with a chatbot to reach a human agent. Fin, the AI agent, registered my complaint, offered me a solution in a human-sounding email — there were even emojis used in the correct context — and closed the case before I even considered getting annoyed.
It wouldn’t quite be correct to say that customer service, thanks to AI, is finally starting to get good. As Kunz and Fager explained, lots of companies are getting it wrong by using AI for the wrong things or tacking it onto legacy systems. However, Intercom co-founder and chief strategy officer Des Traynor says that going all in on AI is the best way to give customers what they want: instant outcomes.
“You don’t want to wait,” Traynor said. “It’s the same reason why people Google before they pick up the phone: People just want instant resolution to problems and that’s what AI offers.” He added, “It’s just categorically better for users — when it works.”
Traynor admitted that AI ushered in an era of software that left people wondering if it worked, and that problem guided the development of Fin. He said his company “put a phenomenal amount of time into building an AI evaluation engine” and “torture-tests every release” to make sure Fin doesn’t hallucinate or get things wrong. As a result, Fin resolves a million customer queries a week with a 67 percent resolution rate, which is not 100 percent, but Traynor said that number is going up 1 percent every month. He conceded that some interactions needed human intervention, but in most cases, the AI can get the job done better. In my case, that was true.
The big problem here, if you’re a consumer, is that you don’t necessarily get to choose how any given company is handling its customer service. There’s also a sort of income equality gap between the haves and the have-nots, whereby bigger companies, like Amazon, can invest more and offer better customer service and small companies, like local utility boards, just do the best they can.
What’s clear, however, is that a transformation is happening. There are signals that complaining to companies is getting easier to do but also strong evidence that many companies will continue to make it difficult, even though they want to make it easier. AI is here to help make things work better, but only if it can stop making them worse first.
A version of this story was also published in the User Friendly newsletter. Sign up here so you don’t miss the next one!
2026-01-10 20:00:00
过去几周，埃隆·马斯克的Grok AI机器人一直在未经女性和未成年女孩同意的情况下，以惊人的速度生成色情图像。最近，彭博社的一项分析发现，Grok每小时能生成6700张此类图像，即每分钟超过一张。本周五，X平台（原推特）终于对Grok实施了一些有限的限制措施，新政策规定只有付费订阅用户才能使用Grok生成或修改图像。然而，在独立的Grok应用程序中，任何人都可以提示生成新图像，这意味着深度伪造色情内容仍在持续。
Grok一直以来都是主要AI模型中较为露骨的一个，拥有“性感”和“火辣”等可切换的模式。尽管员工曾警告称该机器人被用于生成儿童色情内容，但马斯克仍坚持认为Grok将是“最性感”的AI模型。他在X上为这一选择辩护，称这是出于商业考量，引用了1980年代VHS战胜Betamax的案例，当时色情产业支持VHS，因其存储容量更大。马斯克写道：“VHS最终胜出，部分原因在于它允许‘火辣模式’ 😉。”
马斯克的说法有一定道理。色情产业往往奖励早期采用者，而其在色情内容上的巨大利益，使其在选择两种竞争且不兼容的技术时拥有显著的影响力。然而，认为整个色情产业作为一个中立、模糊的群体决定技术战争的全貌并不准确。更准确地说，我们用来生成和分享图像的技术，通常是由那些试图迅速传播女性身体图像的人所塑造的，而这些女性往往并未真正同意。从这个角度看，Grok的能力并不算什么特别之处。
色情产业不仅帮助VHS战胜Betamax，还推动了其他技术的发展。例如，超级8胶片的普及（便于业余电影制作）、流媒体视频的兴起（私密且易于访问）、网络支付的出现（与付费流媒体内容相关）、网络分析技术的发展（对成人流媒体复杂的商业交易有帮助），以及蓝光光盘击败HD DVD。蓝光光盘的存储容量比竞争对手大，这在色情市场中尤其具有吸引力。
此外，还有一些图像分发系统是在色情产业之外发展起来的。其中许多系统都是出于人们想要迅速分享对女性身体的性化图像的动机，但这些图像的发布者往往并非自愿或知情的成年人。有时这些创新是无害的，比如谷歌图片的开发，是因为2000年很多人搜索詹妮弗·洛佩兹穿着著名低腰Versace礼服的照片，而洛佩兹本人将这一事件视为她的荣耀。在这种情况下，她穿着该礼服参加了一个高调活动，并希望被关注，因此可以合理地认为她是知情同意的。
但有时候情况就复杂得多。YouTube的诞生就源于开发者想要观看2004年珍妮特·杰克逊的“服装故障”事件，却因难以在互联网上找到相关视频而感到沮丧。杰克逊一直声称她并未打算让自己的乳房在电视上被展示，因此这属于非自愿的裸露。与此同时，马克·扎克伯格创建Facebook的灵感来自Facesmash，这是一个模仿“Hot or Not”的网站，用于将哈佛大学女生与农场动物进行比较。其初衷并非制造非自愿色情内容，而是对女性进行性化羞辱，这种行为在发布当晚就因过于受欢迎而使哈佛服务器崩溃。
因此，马斯克说那些具有“火辣模式”的技术往往表现良好，这并非完全错误。但更准确的说法是，在我们这个充满性别歧视的社会中，对未自愿女性身体的物化和羞辱具有极高的价值，以至于世界改变性技术的命运往往取决于其在促进这种行为方面的效率。AI本就注定会被用于此类目的，但只有像马斯克这样冷酷无情、愿意忽视道德底线的人，才会让事情发展到如此糟糕的地步。
For the past few weeks, Elon Musk’s Grok AI bot has been generating pornographic images of women and underage girls, without their consent, at an astounding rate. A recent Bloomberg analysis found that Grok creates 6,700 such images per hour, or more than one per minute. On Friday, X at last put some minor guardrails on the tool, with a new policy that only paying subscribers can use Grok to generate or alter images. On the standalone Grok app, however, anyone can prompt Grok to generate new images, meaning the deepfaked porn continues.
Grok has long been one of the more suggestive of the major AI models, with “spicy” and “sexy” settings that can be toggled on and off. While employees have warned that the bot is being used to generate child sex abuse materials, Musk has remained committed to the idea that Grok would be the sexiest AI model. On X, Musk has defended the choice on business grounds, citing the famous tale of how VHS beat Betamax in the 1980s after the porn industry put its weight behind VHS, with its larger storage capacity. “VHS won in the end,” Musk posted, “in part because they allowed spicy mode 😉.”
There is a certain amount of truth to Musk’s take. The porn industry tends to reward early adopters, and the money to be made in porn means that it has impressive leverage when it comes to choosing between two competing and incompatible forms of technology.
Yet the idea that porn as an industry, neutral and amorphous, settles tech wars doesn’t show us the whole truth. It would be more accurate to say that the technologies we use to generate and share images are, more often than not, shaped by people distributing images of women’s bodies — often with dubious consent from the women themselves. In that sense, Grok’s abilities are par for the course.
Porn didn’t only help VHS win over Betamax. The industry has also been linked to the mainstreaming of Super 8 film (easy and convenient for amateur filmmakers), the development of streaming video (private and easily accessible), the development of web payments (comes with paywalled streaming video), the development of web analytics (good for the complex business transactions of adult streaming), and the victory of Blu-ray over HD DVD. (Like VHS, Blu-rays held a lot more data than its competitors, which is especially attractive in the porn market.)
Then there were the systems of image distribution that developed outside of porn as an industry. A surprising amount of them revolved around people trying to share sexualized images of women’s bodies as quickly as possible — only in these cases, the people whose images getting distributed weren’t necessarily consenting adults who were getting paid for their trouble.
Sometimes the innovation was more or less harmless. Google Images was developed because so many people went searching for pictures of Jennifer Lopez in her famously low-cut Versace gown in 2000, a distinction Lopez has treated as a feather in her cap. In this case, Lopez wore the dress to a high-profile event and wanted to be seen and talked about, so it’s reasonable to assume consent.
Other times it got cloudier. The impetus for YouTube came when developers wanted to watch Janet Jackson’s 2004 wardrobe malfunction and were frustrated that it took so long to find video of it on the internet. Jackson has always maintained that she did not intend for her breast to be seen on national TV, so here, we’re dealing with nonconsensual nudity.
Meanwhile, Mark Zuckerberg’s progenitor for Facebook was Facesmash, a Hot or Not rip-off developed to compare the women of Harvard University’s student body to farm animals. The intent here was less to create nonconsensual pornography than it was to perform a sexualized humiliation of nonconsenting women — an act that turned out to be so popular that it overwhelmed Harvard’s servers the night it launched.
So Musk is not entirely wrong when he says that technologies with what he euphemistically refers to as “spicy mode” tend to do well. A more accurate phrasing, however, might be to say that in our misogynistic society, objectifying and humiliating the bodies of unconsenting women is so valuable that the fate of world-altering technologies depends on how good they are at facilitating it.
AI was always going to be used for this, one way or the other. But only someone as brutally uncaring and willing to cut corners as Elon Musk would allow it to go this wrong.
2026-01-10 19:30:00
2025年1月，唐纳德·特朗普就职总统后不久，寻求庇护者聚集在墨西哥蒂华纳的美国海关办公室外。十多年前，特朗普首次进入政坛时，就将关闭美国边境和改革移民体系作为其政策的核心。而在他第二个任期的一年里，这一议题不仅定义了他的总统任期，也改变了美国的走向。其中，特朗普政府对庇护制度的破坏可能是最具影响力的举措之一。庇护制度允许移民因害怕暴力或迫害而合法进入美国。特朗普政府采取了强硬措施，限制寻求庇护者的入境，同时迫使已在美境内的寻求庇护者离开。本期《Today, Explained》特别嘉宾米尔斯·布莱恩采访了普罗公共报道（ProPublica）的移民记者米卡·罗森伯格，探讨特朗普政府如何让寻求庇护者的生活更加艰难，拜登政府时期庇护制度的崩溃，以及美国政策变化对世界的影响。以下是他们对话的节选，已进行删减和润色。完整播客内容更丰富，欢迎在Apple Podcasts、Pandora和Spotify等平台收听。
特朗普政府在2025年对庇护制度持怎样的态度？根据美国法律，人们可以在边境请求庇护，如果他们担心返回本国会受到伤害。但这一过程会触发一个漫长的法院程序，可能需要数年才能解决。特朗普及其顾问认为，这一制度实际上是一个巨大的漏洞，他们相信大多数通过这种方式进入美国的人并非真正的寻求庇护者，而是出于经济原因。因此，特朗普政府上任后迅速推出了一系列政策，旨在关闭这一制度。这些政策包括在边境迅速将寻求庇护者遣返墨西哥，甚至将他们送往第三国，如巴拿马、哥斯达黎加，甚至一些他们从未到过的地方，剥夺他们在美国寻求庇护的机会。
这些“第三国”遣返政策对特朗普政府的整体政策有多重要？这可能是特朗普政府最具创新性和令人惊讶的举措之一。多年来，多个政府都面临一个难题：一些国家拒绝接受本国公民作为遣返对象。在第一次特朗普政府时期，他与一些中美洲国家达成协议，让这些国家接收部分遣返人员，但进展有限。这次，特朗普政府大幅加强了这一策略，与大约20个国家签署了相关协议，包括南苏丹和乌干达等遥远地区。其中，最具野心和影响的举措之一是将近230名委内瑞拉人送往萨尔瓦多的最高安全监狱。他指责这些人是“最坏的帮派分子”。然而，普罗公共报道与委内瑞拉合作伙伴的调查发现，美国政府清楚地知道这些男子在美国从未被定罪过，但他们仍被迅速拘捕并送往监狱，几个月后才通过囚犯交换被释放。这是美国历史上首次在如此大规模上实施此类遣返政策，目前正面临法律挑战。但由于这些人员大多已不在美国境内，因此很难进行有效挑战，导致许多人陷入极其危险的境地。
在特朗普之前，庇护制度是如何运作的？在拜登政府时期，大量寻求庇护者在边境向官员自首以申请庇护，这一现象显著增加。许多人被释放到美国境内，以便在移民法庭上提出庇护申请。根据法律，只有那些因种族、宗教、国籍、政治观点或属于特定社会群体而面临迫害的人才有资格获得庇护。由于这些理由很难证明（尤其是当人们因恐惧而逃离时，可能缺乏足够的证据），因此设立了法院系统，以便人们有时间收集证据并提出申请。我认为，确实有很多人真正有合法的庇护请求，他们为了生命而逃亡，面临政治迫害。但同时，也有一部分人是因为经济困难、暴力或其国家的政治和经济崩溃而来到美国。特朗普政府认为几乎所有的庇护请求都是虚假的，因此采取了极端措施，这实际上导致了合法寻求庇护者难以获得保护。
对于远离美国的国家的人来说，他们是如何得知或相信前往墨西哥、穿越边境、等待并自首能够带来更好的生活？我认为，对于不同国籍和群体的人，情况各不相同。一些人通过WhatsApp群组或TikTok网红了解到不同的入境路线。来自南美洲、印度和非洲部分地区的人开始意识到，他们可以通过边境申请庇护，并可能被释放以继续他们的申请。例如，有数十万人徒步穿越哥伦比亚和巴拿马之间的危险丛林。非洲和印度的移民甚至欠下数万美元的债务，购买商业或包机航班飞往尼加拉瓜，再穿越墨西哥前往美国。
目前还有人获得庇护吗？这一现象是否还在发生，还是特朗普彻底关闭了这一渠道？特朗普政府的目标是封锁边境，这一目标在许多方面已经实现。边境过境人数降至历史最低水平，被释放到美国境内以进行法院程序的人数也大幅减少。因此，人们很难再通过这一途径寻求庇护。这说明美国的庇护政策发生了巨大波动。这些波动似乎很大程度上是因为政策主要由行政命令决定。你认为国会是否会采取实质性的措施来改革庇护制度？我认为，目前国会不太可能采取实质性的行动。多年来，国会一直未能通过真正全面的移民改革法案。我们目前使用的是一个过时的系统，每位总统都通过行政命令和行动来制定移民政策。这些政策可以被法院挑战，也可能在新政府上台后迅速被推翻。因此，只有通过真正的、具有广泛共识的立法行动，才能对庇护制度进行实质性改革。
面对这些无法获得庇护的人，我们该如何看待他们？他们是否会转向其他国家？是否会有一个新的“灯塔之城”出现？目前，全球正经历前所未有的难民潮，人们因冲突而逃离家园。特朗普是利用人们对移民增长担忧的一波政治浪潮的一部分。欧洲甚至加拿大的一些政客也接受了特朗普政府对限制移民和减少庇护申请的立场。过去，许多国家在人权和保护寻求庇护者的问题上都倾向于跟随美国的领导。但现在，这些国家可能会跟随美国采取相反的立场。
When he first emerged on the political stage more than a decade ago, Donald Trump made closing America’s borders and remaking our immigration system a central plank of his agenda.
A year into his second administration — and as this week’s events in Minneapolis underscore — the issue has defined his presidency and changed America’s trajectory.
Perhaps one of the most consequential moves on that front has been his dismantling of our system of asylum: the process by which immigrants can legally enter the country if they fear violence or persecution. Trump has moved aggressively to curtail asylum-seekers’ entrance into the US, as well as to force ones already in the country to leave.
Today, Explained guest host Miles Bryan talked to ProPublica immigration reporter Mica Rosenberg about how the Trump administration has made life harder for asylum-seekers, how the system broke under Joe Biden, and what the changes in the US might spell for the rest of the world.
Below is an excerpt of their conversation, edited for length and clarity. There’s much more in the full podcast, so listen to Today, Explained wherever you get podcasts, including Apple Podcasts, Pandora, and Spotify.
What was the Trump administration’s mindset about asylum coming into 2025?
Under US law, people are allowed to show up at our border and request asylum if they fear returning to their home countries. But that actually triggers a very long court process in the US that can take years to resolve.
Trump and his advisers really view this system as kind of like a giant loophole. They believe that most people who are coming into the country this way are not legitimate asylum-seekers, and they’re maybe coming for economic reasons. They’ve really come into office with a blitz of policies to try and shut that system down.
And that includes things that they’re doing at the border, which is quickly turning people back to Mexico — and, in some cases, sending them to third countries like Panama or Costa Rica or even farther, locations that they’ve never been to, and not giving them a chance to seek asylum here.
How important are those “third country” deportations to the administration’s overall policy?
This is really one of the most novel and surprising things that the Trump administration has tried.
For years, multiple administrations have struggled with a particular issue of countries that have refused to take back their own nationals as deportees. During the first Trump administration, he forged agreements with some Central American countries to take back some deportees from different nationalities — mostly regional migrants, and they didn’t really get very far.
This time around, [the administration has] really ramped up this strategy significantly. They’ve signed these types of agreements with around 20 countries, including really far-flung ones like South Sudan and Uganda.
In one of the most audacious and consequential deportations so far of Trump’s presidency, he sent close to 230 Venezuelan nationals to a maximum security prison in El Salvador. He accused them of being the worst of the worst, gang members. Our reporting at ProPublica and with Venezuelan reporting partners found that the government knew that the vast majority of these men had never been convicted of any crimes in the United States, but they were rounded up and whisked away to this prison, where they were held for months before they were released in a prisoner exchange.
This is something that has never really been tried before at this scale. And it’s being challenged in court. But it’s very difficult to challenge because once these people are outside of the United States, they’re mostly outside of the jurisdiction of US courts. So it’s leaving a lot of people in very precarious situations.
I think it’d be helpful to kind of remind everybody, including myself, what this system and the process looked like before Trump started blowing it up. Can you paint us a picture of how this was working under the Biden administration?
During the Biden administration, this phenomenon of people arriving at the border and turning themselves in to border officials to claim asylum really exploded under the Biden administration. The people that were coming and asking for refuge were overwhelming border stations, and many ended up being released into the country to make their claims in immigration court.
What qualifies you for asylum is a really sort of narrow band of reasons. It’s granted to people specifically who fear persecution because of their race, their religion, their nationality, their political opinion, or membership in a particular social group. The system’s really been set up in the past acknowledging that those things can be very difficult to prove (especially if you’re fleeing out of fear, you might not have all of the proof that you need).
That’s why the court system was set up in this way. It was supposed to give people time to gather evidence, to make their claims. I think there are a lot of people who were arriving at the border who really did have legitimate asylum claims. They’re fleeing for their lives. They’re facing political persecution. But mixed in there, I think, are people who are coming for other reasons. They’re facing serious economic hardship or violence or political and economic breakdown in their home countries.
What the Trump administration has done, by believing that almost all of the asylum claims are fraudulent or not legitimate, they’re really sort of throwing the baby out with the bathwater. And advocates are saying that these changes have made it nearly impossible for legitimate asylum-seekers to really get protection.
How do people in countries far from the United States find out or come to believe that flying to Mexico and then trekking to the border and then waiting at the border and then maybe turning themselves in was going to lead to a better life?
I think it’s very different for every nationality and every group. There were WhatsApp groups, there were TikTok influencers who were advertising different routes for making it to the United States.
People from countries deeper in South America, in India, and parts of Africa started understanding that they could come to the border and claim asylum and potentially be released to pursue their claims. There were hundreds of thousands of people who were making a perilous trek on foot through the dangerous jungle between Colombia and Panama.
African and Indian migrants were going into debt for tens of thousands of dollars to pay for commercial and charter flights into Nicaragua and then to make their way through Mexico.
Is anyone still getting asylum? Is this still happening at all, or has Trump just turned the tap off completely?
Well, the Trump administration’s goals of sealing off the border are really being accomplished in many ways. Border crossings have dropped to record lows, and releases of people into the country to try and go through this court process have also really dropped. There has really been a reduction in the ability for people to seek protection here.
So you’re telling us this story of huge swings in our asylum policy. It seems like a big reason that those swings are possible is because the policy is being set with executive orders. Do you think there’s any possibility that Congress is going to actually make any meaningful changes to our asylum system?
Well, everyone says that we are where we are right now because Congress for decades has never gotten around to passing any really meaningful, comprehensive immigration reform. We’re working with an outdated system. Each president that comes in basically makes immigration policy through fiat and executive actions. And those can be challenged in court. They can be quickly overturned if a different party comes into office.
This is something that would take real, meaningful, bipartisan action. There have been efforts that came really close in the past where there were groups on both sides. I think it really doesn’t look good for congressional action at this point.
How should we think about all these people who have historically sought out the United States for asylum who now cannot? Are they going to other countries? Is there going to be another nation that becomes the shining city on the hill?
All of these changes are happening at a time where there’s really an unprecedented explosion of people fleeing conflicts all over the world. Trump is part of a wave of politicians who have capitalized on concerns about rising immigration. Politicians in places like Europe or even Canada have embraced some of the views that the Trump administration has about tamping down on migration, limiting access to asylum.
Many countries in the past have really felt compelled to follow the US lead on issues of human rights and protecting asylum-seekers. But now, these countries may end up following the US lead in the opposite direction.
2026-01-10 07:25:00
2025年11月3日，格陵兰岛努克的彩绘房屋和住宅公寓楼。| Juliette Pavy/Bloomberg via Getty Images
这则新闻出现在《Logoff》日报上，帮助您了解特朗普政府的动态，而不会让政治新闻占据您的生活。订阅这里。
欢迎来到《Logoff》：各位读者，这周的新闻进展非常迅速。有一则大新闻不容忽视：在委内瑞拉之后，特朗普总统将目光转向了格陵兰岛。特朗普到底想做什么？他想以任何方式获得格陵兰岛。目前来看，他的首选方案似乎是购买该岛，但丹麦方面表示这不可能，或者可能通过支付格陵兰岛居民费用来实现独立。不过白宫本周表示，军事手段“始终是一个选项”。周五，特朗普告诉记者：“如果我们不能轻松地做到，那就只能用困难的方式。”这真的值得我担心吗？不幸的是，是的。正如我的同事乔什·凯因格所报道的，特朗普在委内瑞拉的行动表明，他愿意采取极端措施来实现其扩张主义的美国权力愿景。包括丹麦首相梅特·弗雷德里克森在内的欧洲领导人，都将此视为一个严重威胁。
特朗普到底想用格陵兰岛做什么？特朗普称获得格陵兰岛对国家安全至关重要，包括声称该岛附近有“俄罗斯和中国船只到处都是”。但将美国拥有该岛视为解决安全问题的必要条件是误导的。不仅丹麦是北约盟友，而且美丹之间还有一个单独协议，允许美国在希望的情况下大幅增加其在格陵兰岛的军事存在。这说明还有另一种解释：这可能是特朗普总统在任期最后阶段的一个大型自大项目，他自认为是一个善于谈判和房地产开发的人。正如特朗普本周对《纽约时报》所说，这是“对成功心理上的需要”。
大局如何？随着2026年的到来，特朗普似乎越来越不受约束。委内瑞拉是迄今为止最明显的例子之一，而当他表示格陵兰岛可能是下一个目标时，我们应该认真对待。
好了，是时候下线了……祝贺您成功读完周五晚上！在我们所有人周末下线之前，我的同事普拉蒂克·帕瓦尔分享了一则我最喜欢的过去一年的故事：纽约市拥堵收费政策的成功。这是一项相对较小的干预措施——在高峰时段对司机收取9美元的通行费——一年后，它已帮助减少通勤时间，提高道路安全，并惠及公共交通。我认为这值得庆祝。
This story appeared in The Logoff, a daily newsletter that helps you stay informed about the Trump administration without letting political news take over your life. Subscribe here.
Welcome to The Logoff: Hi readers, this has been a breakneck news week. Here’s one big story that shouldn’t get lost in the shuffle: After Venezuela, President Donald Trump is turning his gaze to Greenland.
What is Trump trying to do? Acquire Greenland, any way he can. For now, it seems like Plan A is to try to buy it, which Denmark says is a nonstarter, or perhaps to pay Greenland’s residents to secede. But the White House said this week that the military is “always an option.”
On Friday, Trump told reporters that “If we don’t do it the easy way, we’re going to do it the hard way.”
Is this really something I need to worry about? Unfortunately, yes. As my colleague Josh Keating reports, Trump’s actions in Venezuela make it clear that he’s willing to take drastic steps to facilitate his expansionist vision of US power. European leaders, including Denmark’s own prime minister, Mette Frederiksen, are treating it as a serious threat.
What does Trump actually want Greenland for? Trump has described acquiring Greenland as necessary for national security, including claiming that there are “Russian and Chinese ships all over the place” near the island.
But it’s misleading to suggest the US would need to own the island to shore up security concerns. Not only is Denmark a NATO ally, but the US and Denmark have a separate agreement giving the US significant room to scale up its military presence in Greenland if it wishes.
That leaves another explanation: This is one big vanity project for a president in his last term in office, one who sees himself as a dealmaker and a real estate developer. As Trump put it to the New York Times this week, it’s “psychologically needed for success.”
What’s the big picture? As 2026 gets underway, Trump is acting as though he’s increasingly unchecked. Venezuela is one of the clearest examples so far — and we should take him seriously when he suggests Greenland could be next.
Congratulations on making it to Friday evening! Before we all log off for the weekend, here’s my colleague Pratik Pawar on one of my favorite stories of the past year: the success of congestion pricing in New York City. It’s a relatively small intervention — a $9 toll for drivers during peak hours — that, one year on, has helped reduce commute times, improved road safety, and benefited public transit. I think that’s worth celebrating.