Share This Podcast

• Check out Lila’s piece – https://tinyurl.com/mr39butw
• Call Congress – 202-224-3121
• Subscribe to the What A Day Newsletter – https://tinyurl.com/3kk4nyz8
• What A Day – YouTube – https://www.youtube.com/@whatadaypodcast

Follow us on Instagram – https://www.instagram.com/crookedmedia/

TRANSCRIPT

Jane Coaston: It’s Tuesday, January 20th, I’m Jane Coaston, and this is What a Day, the show just wondering what Congress is up to. You guys having a nice time? Good January so far? Anything new? Any, say, foreign policy debacles involving our NATO allies you want to prevent? [music break] On today’s show, hooray, it’s one year since President Donald Trump took office. We’re 25% of the way there. Technically, we have fewer Trump days ahead of us than behind us, if you count his first term. And did Trump’s tariffs make us all richer? Look in your wallet or stay tuned to find out. But let’s start with artificial intelligence. If you’ve ever used Claude or ChatGPT, congratulations. You are among millions. According to Pew, 34% of U.S. adults had used ChatGPT by the middle of 2025. That’s double the number of users since 2023. And half of all Americans have used some AI large-language model, including ChatGPT, Claude, Gemini, and Copilot. That’s either fascinating or terrifying, depending on how you feel about AI. And AI development is only getting faster. Over the last few months, you may have heard about Claude Code, another product of Anthropic, the company that makes Claude. Here’s an ad from Anthropic attempting to explain Claude Code to normies. 

[clip of Anthropic ad] Claude Code is an agentic coding tool that lets you work with Claude directly in your terminal. 

Jane Coaston: Okay, I’m already lost. But here’s what you need to know. Claude Code is making coding incredibly easy, and thus making everyday life easier too. For example, you could use it to go through millions of data points in a health research study or monitor your plants’ vitals to help keep them alive, or build a bot to help you unsubscribe from every stupid email list you stupidly signed up for five years ago, even if you’re a normie. Okay, to be fair, that’s what I do with Claude Code. But the thing about Claude Code that’s really cool, or, again, absolutely terrifying is that it might be learning how to make itself better, an idea called recursive self-improvement. Researchers think that’s one critical step towards AGI, or artificial general intelligence, an AI that could problem solve and apply something it’s learned in one space to another, something that, as of right now, only people can do, but maybe not for long. Which, again, is either amazing or the doom of our civilization. Anyway, to talk more about Claude code, what it does, and what it could do in the future, I spoke to Lila Shroff. She’s an assistant editor at The Atlantic with a focus on AI. Lila, welcome back to What a Day. 

Lila Shroff: Thanks for having me back. 

Jane Coaston: You wrote that a friend tried Claude Code and said, it just does stuff. Now, I’ve used Claude because apparently I need a robot to tell me different recipes to use skinless boneless chicken for. But for people who have used Claude and have never used Claude code, what is it? And why are people so excited about it? 

Lila Shroff: Yeah, Claude code, it’s like an AI chatbot, like ChatGPT or Claude, the ones many of us are more familiar with, but it’s also technically an agentic coding tool, which are words that probably don’t mean a ton to many of use, but it, it’s sort of a more powerful chatbot that happens to run by generating code, but also happens to do a bunch of other cool stuff. It’s like a super power chat bot. 

Jane Coaston: Okay, could you give me an example? You used the example in your piece of someone using the chat bot to go through all of this person’s text messages and figure out how many times he said, like, lol. 

Lila Shroff: Yeah, so I talked to one man who wanted to create basically a Spotify wrapped for his text messages. And so using Claude, he said, hey, take a look at my texts and do some analysis. And it did a lot of things. One thing was it filtered for how he was saying, haha, versus, lmao, and he also, you know, had a slide of who he ghosted, but maybe a more practical example, I talked to another woman who said she’s looking for a new office space for her company. She was communicating with a ton of different realtors, and so she had Claude code. She asked it to go through her iMessages and compile a table with all the listings she’s been sent or sent to someone else. And I think the thing, a lot of the time you ask ChatGPT for help and it will tell you to copy-paste something or you know do this thing or that thing. And and basically, since Claude is good at coding, it’s just good at doing things on the computer. 

Jane Coaston: You installed Claude Code. After using it yourself, what surprised you the most about what this AI tool can do? 

Lila Shroff: This was one of those moments, there’s been a few moments in my life where I’ve been completely astounded by a technology. I think, you know, getting in a driverless car for the first time was sort of surreal. You hear about it and you think, yeah, sounds cool. And then you try it and there’s something about it that just, you feel like you were catapulted into the future a little bit. This was one of those moment. It was, it was to me, almost more impressive than using ChatGBT for the first time. Here’s an example. I had this massive data set that had been sitting on my desktop for months, this health data I wanted to write a story about. And I was struggling to do good analysis on it. And I asked Claude Code, hey, can you you know take a look at this data and answer these questions? It produced results. Um. The question now for me is how how valid were those results? And that is what’s interesting here, is a lot of the times you can ask it to do more powerful stuff. But if you don’t know how to build an app and it builds you an app, you have the app. But does the app actually function with good cybersecurity practices? Um. That’s harder to tell. 

Jane Coaston: Now given all of the questions I’m asking you, you can tell I have not used Claude Code. And you wrote that the app’s popularity skyrocketed in December, and everyone in tech was using Claude Code, but why do you think it hasn’t popped off for the general public? What are the barriers to entry? 

Lila Shroff: So for one, it costs money. This isn’t like ChatGPT where you can sign up for free. It’s more than your average streaming subscription. So you need to be excited enough to feel like there’s real utility there for you maybe going into it. Um. And then until now, until pretty recently, it’s been accessible through the computer terminal, which isn’t that scary to use, but if you’ve never taken a coding class, it looks a little bit like crazy hacker in a movie. Um. But they just they actually just released a new product that’s not not as scary looking for for non-technical users. And I think, you know, Anthropic, when I talked to people that worked there, they said they were surprised because they did build this as a coding tool. They built it for developers. And so seeing it all of a sudden pop off among all these non-technical people, um they themselves were were kind of surprised to see that. 

Jane Coaston: You talk about how Claude Code could turn out to be an inflection point for AI progress, especially towards artificial general intelligence, or AGI, because there are early signs of what’s called recursive self-improvement. First of all, what’s recursive self improvement? 

Lila Shroff: Basically, the idea is at some point we’re going to get to possibly AI systems that can start making themselves better. One of my co-workers, um Matteo Wong, has put this pretty nicely. You know if the GPT-5 model starts improving the GPT-6 model, which improves the GPT-7 and on and on and you get this takeoff of kind of rapid improvement. This is this is somewhat theoretical and and so you know we can reason about whether it’s a legitimate theory of how AI progress will continue. Um. But uh the the um Anthropic employee who created Claude Code said they’re starting to see Claude come up with ideas of what to build next and so for him it was kind of early sparks of this. 

Jane Coaston: Now if you could see what my face is doing, I just made the face that you said Claude is coming up with ideas and I’ve seen Terminator 2 judgment day so I am very anxious about this and I think that if anyone’s been paying attention to people being concerned about AGI, there’s either been like AGI is going to save us all or AGI’s going to kill us all because it’s going become a god and everybody loses their minds. So based on what you’ve seen, how worried about AI spinning out of control, given that you just said that someone at Anthropic is seeing Claude come up with ideas. Humans come up ideas. AI coming up with ideas is new and very scary. So how worried should we actually be? What are experts saying about this? 

Lila Shroff: You know I’ll start by just saying I haven’t seen whatever is going on inside of Anthropic, so I can’t say um what exactly they’re seeing and whether it is actually  you know AI coming up with novel ideas or just doing a good job of kind of recreating already existing human ideas. Even if if this is actually what’s going on, I think that I, I’m not polarized to the you know abundance, crazy AI future or the extreme we all need to go hide out. I think the the biggest questions I have around um around employment and automation, I think this is the first time I’ve started to think, wow, I can really see how there’s a lot of work that we do on a day-to-day basis that could be pretty radically transformed and perhaps um automated with the technology. That sense is becoming more clear, and so how this plays out on the jobs front is the biggest question in my mind. 

Jane Coaston: You report that Anthropic, the maker of Claude code, discovered Chinese state-sponsored hackers were using the tool for cyber espionage. Can you tell us more about that? 

Lila Shroff: Yeah. I mean, if me who has a very limited programming experience using Claude code can all of a sudden be a much better programmer, people with nefarious or ill intent can also be kind of leveled up. It’s, you know, we all kind of get a boost, the good guys and maybe the people–

Jane Coaston: Right. 

Lila Shroff: –with a little bit less, uh worse intentions. And so I think that kind of creates an arms race effect on the cybersecurity front where you get more powerful ways of tracking down bad actors, but then you also get bad actors doing crazier things, and so there’s a lot of open questions, I think um, there, and cybersecurity in particular is an area I also have some concerns about. 

Jane Coaston: Yeah, I think that the concerns for me, I mean we’ve seen this with other AI tools like Grok where it’s creating through users, but you know the users wouldn’t have the ability to do this without Grok, child sexual abuse material and non-consensual sexual content. What does that tell us about the risks of AI? I think you just said something really smart which is that like if you know, the white hats are going to use it to do even better white hat things and the bad people are going to use it to do even worse things, what does that mean for the challenge of regulating these technologies?

Lila Shroff: That’s a big question. I think you know there is a degree to which this is all happening really fast. The sorts of questions that get introduced by a much more powerful tool like Claude Code, um you know I think this kind of makes me think across the board. You know in education, how do you want students learning to use these tools? Um. There’s been a lot of talk about whether we should have this streamlined national AI policy or what states figure things out. I think there’s just a ton of confusion and lack of direction as to how we handle this. And so I think hopefully hopefully this is a moment that sort of prompts people to start putting their heads together and thinking more seriously. 

Jane Coaston: Lila, thank you so much for coming back to the show. 

Lila Shroff: Thanks for having me. 

Jane Coaston: That was my conversation with Lila Shroff, Assistant Editor at The Atlantic. We’ll link to her piece in the show notes. We’ll get to more of the news in a moment, but if you like the show, make sure to subscribe, leave a five-star review on Apple Podcasts, watch us on YouTube, and share with your friends. More to come after some ads. [music break]

[AD BREAK]

Jane Coaston: Here’s what else we’re following today. 

[sung] Headlines. 

[clip of Danish Prime Minister Mette Frederiksen] The meeting we had last week left me with the clear impression that the president honestly and full-heartedly wants to acquire Greenland. But we also made it crystal clear that this is a red line. 

Jane Coaston: The Danish foreign minister spoke with reporters Monday reiterating the quagmire that is Trump versus Greenland. And if he was looking for optimism today, he didn’t find any. That’s because Trump’s petty little fingers were aflutter this weekend, texting the prime minister of Norway. In the texts acquired by PBS Monday, Trump said in part, quote, “considering your country decided not to give me the Nobel Peace Prize for having stopped eight wars plus, I no feel an obligation to think purely of peace.” He’s like a six-year-old who didn’t get the trophy, quits the soccer team, and tries to take the ball home. In a statement about the exchange, the Norwegian prime minister said, quote, “I have on several occasions clearly explained to Trump what is well known, namely that it is an independent Nobel committee and not the Norwegian government that awards the prize.” So it’s like, a six year old stole the trophy from an entirely different soccer league, also he’s in charge of the military for some reason? The message comes on the heels of Trump threatening new tariffs on the European nations back in Denmark and Greenland. The EU will weigh a potential response at an emergency summit in Brussels on Thursday. The Supreme Court is set to hear arguments on Wednesday in a case involving President Trump’s attempt to fire Federal Reserve Board Governor Lisa Cook. Last year, Trump tried to fire Cook, citing accusations of mortgage fraud. Cook has denied the allegation and no charges have been made against her. She sued to keep her job, and the Supreme Court issued a brief order allowing her to stay on the board while it considers the case. So here we are. According to sources, Fed Chair Jerome Powell plans to attend the session, a show of support by the Central Bank Chair. It’s unusual, but the Fed has really been going through it lately. Just last week, Powell revealed that the Department of Justice subpoenaed the Fed and threatened a criminal indictment related to his congressional testimony over the cost of a renovation of fed buildings. 

[clip of President Donald Trump] I always say tariffs is the most beautiful word to me in the dictionary. Then I was reprimanded by the fake news. They said, what about love, religion and God? I said, I agree. Let’s put God number one. Let’s put religion number two. Love, I don’t know, we got to put that number three, I guess, right? And then it’s tariff, because tariffs are going to make us rich as hell, it’s going to bring our countries–

Jane Coaston: That was President Trump exactly a year ago today, professing his undying love for tariffs. And he hasn’t stopped since. But have tariffs made us all rich as hell, as the president promised? It seems unlikely. In fact, new research published by the Kiel Institute for the World Economy found Americans are actually footing the bill for the country’s tariffs. Researchers at the German think tank analyzed more than 25 million shipment records, spanning a total value of almost $4 trillion in U.S. imports. From that, they found that foreign exporters took on about 4% of the tariff burden, leaving 96% to be passed to U.S. buyers. The researchers found that ultimately, tariffs are bad for everyone. U. S. Companies will be confronted with shrinking margins and consumers will face higher prices in the long run. The report concluded, quote, “the claim that foreign countries pay these tariffs is a myth. The tariffs are, in the most literal sense, an own goal.” Reading us for filth. As of today, January 20th, we are officially one year into President Trump’s second term. And just in case our previous headlines didn’t make it clear, things aren’t going great. That’s why this afternoon, at 2 p.m. Local time, Americans in all 50 states are expected to walk out of their jobs, schools, and businesses to protest the country’s quote, “escalating fascist threat.” The call to action comes from the same group behind the monumental Women’s March during Trump’s first term. They’re calling it the Free America Walkout. According to the website, a free America begins the moment we refuse to cooperate. This is not a request. This is a rupture. This is protest and a promise. In the face of fascism, we will be ungovernable. So no matter what’s on your agenda today, we hope you celebrate this year’s Trump-aversary like a true American, whether it’s exercising your freedom of speech, right to assemble, or liberty to demand better from your government. Because that’s what this country was built on. Not tariffs or Greenland or demanding talk show hosts be fired for not being nice enough to the President of the United States. Actually, we don’t have to be nice to the president of the United State at all. Because this is America. And that’s the news. [music break] One more thing, immigration and customs enforcement is very busy in Minneapolis right now, as you may know. Breaking car windows, hospitalizing children with flashbangs, screaming obscenities, they’ve got lots to do. But something else they’ve been doing is arresting Native Americans. Yes, Native Americans. Last week, members of the Oglala Sioux tribe in Minneapolis sounded the alarm after multiple members of tribe were allegedly detained by ICE. Taken away with no details as to where they are, or when they’ll be released. The Twin Cities is home to at least 50,000 American Indians, and they are being targeted by ICE with allegations of harassment and abuse. And American Indians from across the country, from Utah to Arizona to Washington State, are reporting the same thing. Despite literally being Indigenous Americans, ICE is going after them. Here’s Native American actress Elaine Miles on CNN, talking about an incident in December in which she was told that her ID looked, quote, “Fake,” by ICE agents in Redmond, Washington. 

[clip of Elaine Miles] I kept telling them that it was from a federally recognized tribe in Eastern Oregon and it’s a federal ID and only enrolled members can get those um because they kept saying anybody could make ’em. 

Jane Coaston: Now, you might be wondering, why in the hell would Indigenous Americans, literally the most American of Americans, be targeted by ICE? The answers, as far as I can tell, are racism and government overreach. Last September, the Supreme Court found that, quote, “apparent race or ethnicity could be used as a relevant factor that would allow ICE to detain someone.” These stops, now called Kavanaugh stops for Justice Brett Kavanaagh, are now being used against Indigenous Americans because, to put it bluntly, ICE thinks they look foreign. But you heard Elaine Miles mention her tribal ID, which is a real form of identification and can be used as proof of citizenship. There are 574 federally recognized American Indian and Alaska Native tribes in the United States. Those tribes receive specific rights and protections, including some powers of self-government and limited sovereignty. And that gets to the government overreach part. See, because of treaties made between the government and American Indian tribes, ICE doesn’t have the same powers on Indian land it has elsewhere. But as you might guess, it would really, really like those powers. Specifically, ICE wants American Indian Tribes to sign 287-G agreements, which would create a relationship between the federal government and Indian tribes to enforce immigration laws. And according to the president of the Oglala Sioux tribe, Frank Star Comes Out, federal officials said that they wouldn’t give up information about the people they detained until the tribe signed a 287-G agreement. But Frank Star Comes Out said no. In a letter to multiple Trump administration officials, he wrote, quote, “We will not enter an agreement that would authorize or make it easier for ICE or Homeland Security to come onto our tribal homeland to arrest or detain our tribal members.” In an interview with the Washington Post, Minnesota State Representative Liish Kozlowski who is of Anishinaabe-Ojibwe descent, said that the Trump White House wants to, quote, “crush dissenters.” But they said, quote, “the thing is, they’ve never been able to crush our spirits ever.”[music break]

[AD BREAK]

Jane Coaston: That’s all for today. If you like the show, make sure you subscribe, leave a review, don’t bet on sports, and tell your friends to listen. And if you’re into reading, and not just about how we are in the midst of the NFL playoffs and the NBA is in full swing, and we just watched the College Football National Championship, and I just want to remind you that betting on sports is a terrible idea, like me. What a Day is also a nightly newsletter. Check it out and subscribe at Crooked.com/subscribe. I’m Jane Coaston. And most Americans think sports betting is ruining the integrity of sports. So just say no to betting real American currency on sports. Or politics. Or pretty much anything. [music break] What a Day is a production of Crooked Media. It’s recorded and mixed by Desmond Taylor. Our associate producers are Emily Fohr and Chris Allport. Our producer is Caitlin Plummer. Our video editor is Joseph Dutra. Our video producer is Johanna Case. We had production help today from Ethan Oberman, Greg Walters and Matt Berg. Our senior producer is Erica Morrison and our senior vice president of news and politics is Adriene Hill. Our theme music is by Kyle Murdock and Jordan Cantor. We had help today from the Associated Press. Our production staff is proudly unionized with the Writers Guild of America East. [music break]